ML20072L591
| ML20072L591 | |
| Person / Time | |
|---|---|
| Issue date: | 03/31/1983 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| References | |
| NUREG-0966, NUREG-966, NUDOCS 8303310361 | |
| Download: ML20072L591 (147) | |
Text
.
NUREG-0966 l-Safety Evaluation Report
' related to the D2/D3 Steam Generator Design Modification U.S. Nuclear Regulatory Commission Office of Nuclear Reactor Regulation March 1983
,p" "*%,
- k
)
7 L
RBR3nRee'S 096b R PDR
,. -,, ~.
, - - -. ~
~ --
+,
-.. -.. -~
.n
,U, e
+6 '
');
m.
. y s
1 y,-
~
l
,e l,
' NOTICE
.j Availability of Reference Materiela Cited in NRC Publications t
.~
Most documents cited in NRC publications will be a'voilable from one of the following sources:
i a
4
[
f
, 1. The NRC Public Document Room,1717 H Street, N.W..
~
-Washington, DC 20556 1
g'
+
4.
2.iThe NRC/GPO Sales Program, U.S.' Nuclear Regulatory Commissions j
p 4
Washington, DC 20555
.t
?
i:
~
]
lf _
- 3.~ The National Technical Information. Service, Springfield, VA 22161 d
'Although the listing thaf follows represents the majority of documents cited in NRCI Eblicatims,'
, -l it is not Intsnded to be exhaustive.
n 1
- Referenced documents available for inspection and copying for a fee from the NRC Public Docui
(
ment Room include NRC correspondence sad internel N RC memorands; NRC Office of Inspection '
y and Enforcement bulletins, circulars, information notices,. inspection and. investigation: notices; Licensee Event Reports; vendor reports and correspondence; Commission papers; and applicant and 1
- licensee docaments and correspondence.
3 The following documents la the NUREG series are available for purchese from the NRC/GPO Sales Program: formal NRC staff and contractor reports, NRC-sponsored conference proceedings, and '
'NRC booklets and brochures. Also available are Regulatory Guidcs, NRC regulations in the Code of Federal Regulations, and Nuclear Regulatory Commission issuances Documents available from the National Technical-Inf'ormation Service include NUREG series i
- reports and technical reports prepared by other federal agencies and reports prepared by the Atomic :
Energy Ccmmission, forerunner agency to the Nuclear Regu!4 tory Commission.
De uments available from public and.special technical libraries include._all open literature items; i
D such as books, lobrnal and periodical articles, and transactions. Federal Register notices,' federal and
[
state leghfation, and congressional reports can usually be obtained from'these libraries.-
i i'
' Documents such as theses, dissertations, foreign reports' and translations, and non NRC conference.
h proceedings are available for purchase from the organization sponsoring the publication cited.
l l
Single copies of NRC draft reports are available free upon written request to the' Division of Tech-nical. Information and Document Control U.S. Nucleer' Regulatory Commission, Washingtonf DC -
l i
20555.
5 Copies of Industry codes and standards used in a substantive manner in the NRC regulatory process.
{
! H
-are maintained at the NRC Library,7920 Norfolk Avenue, Bethesda,' Maryland, and are available L ie
-there for reference use by the public. Codes and standards are usually copyrighted and may be -
purchased from the originating organization or, if they are' American National Standards, from the-American National Standards Institute,1430 Broadway, New York, NY 10018.
t J
$6.50 GPO Printed cor / price:
0
~
Y h,
k t
,f,
a i
M
- var v-y yy wg el9
-r irg y
y 4 y-sp y
4
-v M y,
,e ee'a'
-w-"n 9 *r t = m- - m = - -pr - e e e =e4 e ~
- t -e rw
-w e-*-e*
d'
~e9
'rP v m-Jh
e NeM e-*'Ce
Safety Evaluation Report related to the D2/D3 Steam Generator Design Modification l
Y t
l U.S. Nuclear Regulatory Commission Office of Nuclear Reactor Regulation March 1983 fa a'**%,
l l
l l
l i
l A
4 i
i ABSTRACT This Safety Evaluation Report (SER) related to the D2/D3 steam generator design modification has been prepared by the Office of Nuclear Reactor Regulation of the U.S. Nuclear Regulatory Commission. The purpose of this SER is to issue the staff's evaluation of the acceptability of the design modification for both installation and full power operation in the D2/D3 steam generators based on the Design Review Panel Report of Janua,7y 1983.
1 i
L i-L l
l l
l l
i h
i NUREG-0966 iii
s CONTENTS
~l 4
1.0 INTRODUCTION
AND
SUMMARY
^1-1, 2.0' DESCRIPTION OF MODIFICATION.................,...................
2 3.0 EVALUATION.......................................................
3-1 3.1 Thermal Hydraulics and Model-Testing........................
3-1 i.
3.2 Structural Mechanics........................................
3-4 i-3.3 Materials, Welding, and Nondestruct!ve Examination...........
3-6 3.4 Tooling.and Installation....................................
3-6 3.5 Radiological Considerations /ALARA...........................
3-6 3.6 Quality Assurance.............................,.............
3-7 3.7 Chemistry.and Cleanliness Control...........................
3-7 3.8 Inservice Inspection and Testing............................
3-7
4.0 REFERENCES
4-1 APPENDIX A PRINCIPAL CONT RIBUTORS...................................
A-1 APPENDIX B DRP REP 0RT.................................................
B-1 APPENDIX C CHRON0 LOGY.................................................
C-1 J
i B
f 1
- e a
i 1
1
1.0 INTRODUCTION
AND
SUMMARY
History On October 21, 1981, a steam generator tube leak occurred at Ringhais Unit 3 (Varberg,. Sweden), a tb'ee-loop Westinghouse plant with Modei D3 steam genera-tors, resulting in plant shutdown.
From the resulting invebcigation, the occur-i rence of some type of accelerated wear mechenism avolving interaction between the steam generator tubes and the tube support plates was occurring.
Eddy cur-l
. rent testing (ECT) was performed on all three steam generators. The ECT results indicated that preferential wear was' occurring.in the outer three rows of tubes in the preheater section (Rows 47, 48, and 49).
A task force was established by Westinghouse to identify ar.d correct the cause of-the problem. To accomplish this objective, the task force gathered infor-mation relative to.the problem such as ECT, data from operating plants, pulled tube data, tube vibration data, information from analytical models, and infor-
. esting-mation from a series of air and water scale-model test facilities.
W house determined that this type of accelerated tube wear is characteristic of r
only the preheat section of its Model D and E steam generators.
Several conceptual modifications to reduce the tube vibration and the resultant wear were developed by Westinghouse during the above investigations.
From these, Westinghouse proposes to install an internal manifold in the Models D2 and D3 (D2/D3) steam generators, and is continuing to evaluate solutions for the 04, DS, and E steam generators. Therefore, this report. discusses only the rodification proposed for the D2/D3 steam. generators.
Design Review Panel During the first several months after the Model D steam generator problem.was identified, Nuclear Regulatory Commission (NRC) staff worked with Westinghouse and the utilities listed below on an individual basis.
In an effort to con-serve NRC staff resources, the concept of a third party design review of the proposed modification was initiated. This review was intended to lessen the need for a detailed technical review by the NRC, and the third party's report was proposed to serve as the basis for the NRC Safety Evaluation Report.
Ten-nessee-Valley Authorit.y (TVA), South Carolina. Electric and Gas Corpany (SCE&G),
and Duke Power Company agreed to pool resources to form a Design Leview Panel (DRP) to perform a-third party design review which would examine all aspects of the final Westinghouse modification design for the Model D2/D3 steam generators.
The following table lists all of the U.S. plants with D2/D3 steam generators.
Plant Type No. of S/G Owner McGuire 1 02 4
Duke Power Company McGuire 2 D3 4
Duke Power Company Catawba 1 D3 4
Duke Power Company Summer D3 3
SCE&G Watts Bar 1 D3 4
TVA Watts Bar 2 D3 4
TVA 4
NUREG-096E l-1 e
w
e
,. - - - - - -.,. -., + - - -. - -.. - - - - -.
+-----.-.-,.m
.---e.--.-i.
= -,
r
*e--
e-
-et-~v
Three chairmen, one from each utility, were:in. charge of the DRP.
InchJding
~
the chairmen, the DRP consisted of 20 utility employees and 2 consettants with.
a total of 432 years of experience.- The objective of the DRP was to determine the acceptability of the final modification design selected by Westinghouse for installation in the D2/D3 steam generators and then submit a report to the NRC with.their conclusions.
The review performed by the DRP covered many'different.
areas. - These include thermal hydraulics, model testing, radiological exposures, structural mechanics, stress analysis, quality assurance, inservice inspection, materials, tooling, welding, chemistry control, and insta11atit Throughout the' review process, by meetings and telephone conversations with the DRP by.the NRC staff, the staff observed that the DRP-reviewed appropriate areas,' asked.
pertinent questions, conducted ~an appr 1riate depth of review and overall that the expertise of the DRP and their approach to the review of the modification was appropriate and acceptable. A chronology of events concerning staff involve -
ment in the D2/D3 steam generators and DRP review is provided as Appendix C.
J The DRP report of January 1983 (Ref. 1, propr!etary and nonproprietary versions) relies extensively on technical data summarized.in the Westinghouse report,
" Westinghouse Preheat Steam Generator D2/D3 Design Modification Evaluation
' Package,". December 1982 and addenda (Ref. 2, proprietary and nonproprietary versions); however, these data were made available to the DRP throughout the review process prior to December 1982.
Attached as Appendix B, which was re-written by the DRP for reader ease of understanding, is a nonproprietary version of the DRP report submitted March 2, 1983 without the appendices.
(Pages omit-ted from the March 2, 1983 submittal were submitted on March 10, 1983.) The DRP has concluded that the modification to the D2/03 steam generator preheater section is not an unreviewed safety question, can be installed, and the steam generators can be operated safely.
4 Sumary The staff has reviewed the DRP report and finds the DRP report acceptable with some exceptions and comments in the areas of thermal hydraulics, structural i
mechanics, radiological considerations, quality assurance, and inservice inspection and testing.
Thi? report discusses these exceptions and comments that.along with the DRP repurt (Appendix B) form the NRC staff's safety evaluation of the D2/03 steam generator modification.
As a result of this review, the staff has established the following additional requirements:
(a) Radiological Considerations Each utility shall perform a comparative utility specific radiological l-assessment prior to initiating SG modification and, upon completion of the modifications, shall perform a summary radiological assessment in accord-ance with C.3.c of Regulatory Guide 8.8.
+
(b) Quality Assurance Each utility shall apply its ID CFR 50 Appendix B, QA program to the SG i
modifications.
l NUREG-0966 1-2
(c) -Feedwater System i
Any feedwater system modifications that are contemplated as a result of i
the SG modifications shall be evaluated in accordance with NUREG-0800, j
Section 3.6.2, " Determination of Ruptures Location' and Dynamic Effects Associated with the Postulated Rupture of Piping." The evaluation shall include pipe break location and jet thrust analyses.
(d)
Inservice Inspection and Testing For the initial two U.S. plants which install the modification, the first ECT inspection conducted after operation with the modification installed (as evaluated in Section 5.8, Appendix B) shall be expanded.
The plant's first inservico~ inspection, as defined in the plant Technical Specifica-tions, i.e., 3% random sample inspection, shall be added to the DRP rec-ommended preheater section ECT inspections.
The surveillance program shall be expanded so that the plant's loose parts monitoring system is utilized
~
to monitor the manifold's structural integrity.
With respect to eddy current testing (ECT) inspections, Technical. Specifications shall be re-vised to include a minimum of 240 tubes 'in the preheater section in addi-tion to the 3%' random sample inspection.
The manifold shall be visually inspected whenever the eddy current testing (ECT) is performed.
Visual.
examination of the manifold by fiberoptic boroscope techniques shall be-done in accordance with B&PV Code Section XI-IWA-2211.
The staff finds that the modification of the D2/03 steam generators is acceptable and that the modified steam generators can be operated at 100% of the?r design capacity. Finally, the staff finds that cperation with the modification installed will not be inimical to the health and safety of the public.
1 I
l I
r l
NUREG-0966 1-3
l l
2.0 : DESCRIPTION OF MODIFICATION
- The.D2.and D3 types-of' preheat steam generators are nearly identical in design.
A schematic of the Model D3 steam generator.is provided as Figure 1.
The details of the D2/D3 preheat steam generator design modification, including major compo-nents, are shown in Figure 2.-
In the existing design, as shown in Figure 1.0.4 of Reference 2, flow entering the steam generator passes thrcach a four-venturi reverse. flow limiter and is deflected by an' impingement plate before entering
~
' the tube bur.dle.
It has been demonstrated that this combination of features leads to severe turbulence in the inlet plenum,;the major; factor contributing to wear of the steam generator tubes.
The design _mooification for the D2/D3 preheat tieam generators to alleviate these severe turbulence conditions consists of replacing the flow impingement plateLassembly with an internal manifold. assembly, and, replacing the four hole reverst flow limiter with a 19 hole reverse flow limiter. The new configuration is shown in Figure 2.
The modification is designed to isolate'the~ tube _ bundle from the highly turbulent feedwater inlet flow and to. distribute the feedwater
.more uniformly over the full area of the bundle at the inlet plenLa region.
j' The flow splitter divides the flow into several channels as it enters the mani-
-fold.
The larger number of holes in the reverse flow limiter results in smaller l
jets providing a more fully;rieveloped flow entering.the manifold. -The momentum force on the manifold assembly is reduced without compromising the function of the reverse flow ilmiter.
The manifold box consists of two perforated plates, the inlet and exit plates.
The minimum hole size of the inlet plate was dictated by the need to ensure that' E
build up of contaminants will not plug the holes during service. The : kinetic-energy of-the flow jets exiting the inlet plate is dissipated by impinging on t
the ligaments of the downstream exit plate.
The flow enters the preheater inlet plenum region through the toles in the exit plate with a relatively uniform flow velocity.
No attachment welds are made to the pressue boundary portion of the steam generator.
Section 5.7 of Reference 2 contains additional details con -
cerning the installed configuration and assembly of this design modification.
Additional details concerning the other components are contained in Section 5.0 of Reference 2.
i i
NUREG-0966 2-1
~
l e
Stearn Outlet To Turbine V
A
/
i Secondary Moisture m,x f[
anway Separator I
1:vi: 1 Primary Moisture J
Separator j
-',J Nozzle h
9(
=
l i
/
a>
m Anti-Vibration Bars T b Bundle
) [ ue Tube Support Plates b
j ig e
i t
A A
Feedwater inlet
- Preheater s
_i,.
Y J
Flow Distribution Baffle
! D #3
' *l yTube Sheet
- y i
Ii I
l Coolant Inist Manway
- See Figure 2 for modification detail.
f l
Figure 1.
Schematic Model 03 Steam Generator l
l l
i j
NUREG-0966 2-2 l
1
BING I
INTERNAL M ANIFOLD 1
y RAPPER W
(
' MANIFOLD
/ SUPPORT SLEEVE g
FLOW SPLITTER
~
B AFFLE' PLATE NO. 6
/
~
- ssi j
jTHERMAL SLEEVE
/ REVERSE FLOW
'f
/
LIMITER VENTURI j
- .. W
, ?-
}
/
1 O
./ y
'0
~
- j. /
,, - y
\\
~~',,,
\\
CYLINDER SUPPORT
/
/{
FEEDWATER EXIT ENTRANCE INLET NOZZLE PLATE Pl. ATE t! AFFLE PL ATE NO.5 Figure 2.
Modification of D2/03 Preheater Inlet NUREG-0966 2-3
3.0 EVALUATION 3.1 Thermal Hydraulics and Model Testing The basic causes of the excessive tube vibration as perceived by the staff and its consultants are relative 1; large tube-to-tube support plate clearances for the spans used. These design features increase the probability that a tube m.,y " float" in a tube support plate (TSP) such that the TSP is ef fectively inactive, that is, it dc9s not act as a st.pport (TSP-inactive).
If a tube
" bridges" or " floats" in a support plate, it will have an effectively longer span, lower natural frequency, and will inherently be more susceptible to fleid excitation forces.
Based on the results of an extensive investigation involving field data, model tests and computer modeling, Westinghouse has concluded that high-level turbulence and nonuniform flow in the inlet region with the D2/03 impingement plate is responsible for the majority of the tube wear. Westinghouse further concluded that localized regions of very high wear, such as those found near the edges of the impingement plate, are explained by a fluidelastic mechanism, where hydrodynamic forces are an important contributing factor. The mechanism is characterized by a critical flow velocity below which the vibration amplitudes are small and above which the amplitudes increase rapidly.
The vibrations occurring at a natural frequency of the tubes in the fluid are defined as fluidelsstic instability.
The tubes may vibrate in steady orbital patterns or they may precess.
Such-a motion results in accelerated tube wear at the tube / tube support interface.
The DRP report concluded, and the staff concurs, that the proposed Westinghouse design modification successfully achieves the design objectives of reducing inlet turbulence and attaining nearly uniform flow at the inlet.
Tests performed in the 0.417-scale model and in the Swedish State Power Board
(
(SSPB) tests have shown the tube vibration response to be reduced substantially with the internal manifold when compared with the response with the impingement plate (base case).
For the SSPB facility, it had been a staff concern that flow over spans away from the inlet is not adequately considered relative to its effect on the fluid drag force as well as its contribution to excitation of tube motion.
This con-l cern was alleviated, in part, by results from the 0.417-scale model test program.
L Tests show that agreement between vibrational responses in single pass flow and in multipass flow is very close. Westinghouse attributes this to the lower crossflow velocities over the spans away from the inlet.
In addition, informa-tion obtained from Westinghouse indicates that the " baffle plate" leakage in the field units is higher than leakage in the 0.417-scale model.
These results demonstrate that the 0.417-scale model test is conservative in the sense that crossflow over spans away from the inlet span is greater in the model than in the field units.
It also explains the lower crossflow velocities, over spans away from the inlet, to be expected in the field uaits.
Based on the above, staff concerns relative to flow over spans away from the inlet have been ade-quately addressed.
NUREG-0966 3-1
Tube support provided by the baffles.is of fundamental importance because it Laffects'the potential for a tche to experience damaging vibrations.
It had been a staff _ concern that not all possible support configurations,~and possibly not even the worst configurations, were being simulated.in the 0.417-scale j '
tests by the selective urdercutting of-the tubes at baffle locations.
- However, this concern also was. alleviated by information provided by: Westinghouse rela-tive to the SSPB tests.
In this case, discussion of the_SSPS tests revealed.
the following:
(1) tube-to-baffle plate diametral clearances were on the high side, relative to the design clearance and were thus ennservative (note that j
the as-built field unit clearances were at the maximum tolerance); (2) the-1 r.
support plates were initially adjusted to give tube vibration t>esponse spectra approximately replicating field. data from operating steam generators (the con-figuration giving the most active tube vibration conditions was used as the baseline for testing); and (3) a support condition " sensitivity" study was per-formed with a total of 50 different tube. support configurations tested.
- Thus, the staff concludes that the question of accounting for possible variations in support plate effectiveness has been adequately addressed in the SSPB tests.
The staff concludes from the 0.417-scale tests that (a) the scitation is pri-marily a result of the inlet flow, (b) the effects of crossflow over spans away from the inlet are negligible, and (c) the worst support configuration was simulated in the SSPB tests.
The results of the SSPB tests provide convincing M
evidence that the internal manifold modification achieves a substantial reduc-tion in tube vibration response.
In particular, results show that for the most active tubes the response at 100% power with the manifold is comparable to e
response at'40 to 50% power for the base case (with impingement plate).
l The tube vibrations measured in the 0.417-scale model, the SSPB model and'the L
operating plants, when scaling' factors were considered, were determined to have~
the same frequencies and comparable. amplitudes.' The turbulence forces measured l
in the 2/3-scale model, when scaled, were the same as the SSPB-full-scale turbulence force measuresant. The fluid velocities measured in the 2/3-scale model and SSPB full-scale model were consistent with the tube vibration behavior measured in the 0.417-scale model and the SSPB full-scale model.
Also, the relative improvements in flow distribution and the tube vibration of the mani-i fold r. edification compared with the base case was. found to be consistent enong the scale models.
Therefore, the staff concluded that the. scale models can be relied on for evaluating major aspects of tube vibration performance of the I
manifold modifications.
Field data from the operating steam generators have indicated that for-power levels of 40 to 50% or less, tube wear is low and acceptable from the standpoint of plant operations without the modifications.
NRC had approved operation of the McGuire plant, with as-built steam generators, at 50% power level or less.
l Therefore, a realistic acceptance criterion for the manifold modification is to require vibration response' reduction at full power with the manifold to levels of response measured in operating plant steam generators (with impinge-p ment plates) at 40 to 50%.
The Westinghouse design modification satisfies this j
criterion based on results of the SSPB tests, and supporting test results from l
the 0.417-scale model tests.
f Westinghouse places considerable emphasis on its nonlinear, analytical model.
Among other things, it bases, in large part, its conclusion that turbulence is.
NUREG-0966 3-2 9p,.
+*.wg--
9-.+,
p.w--
,-w y
e-yw.p,-.y, g.pw-.97
- i. tag
,.,9.__
p y9,.-,.
.p_.
g
.rp.w_e,..w.-
e i
i responsible for the majority of tube wear on results of the nonlinear model Lusing,megsured turbulence forces as input. The model is also used to pradict u
wear over the design life of the. steam generator using wear coefficients established from laboratory _ tests. NRC t.oncerns regarding the nonlinear mod 21-ing include the following:
4 (1) Measurement of turbulence forces-(the input to the model) by itself is very difficult; in addition, there i; the need to extrapolate from the 2/3-sca'e model to the full-scale unit.
1 (2)- The model results are shown to'be very sensitive to support plate offset-and tubesheet rotation; as a consequence, the model is tuned to get results t
that. match field data.
Nevertheless, the model is useful!for gaining insight and guidance as to pre-dicting wear; however, the staff does not accept it as a reliable wear predic-tion model. The DRP also_ recognized the uncertainties associated with the Westinghouse wear medel.and wear prediction method and the need to add addi-tional conservatism to account for the uncertaintics.
.The DRP lists aspects of the design modification that could have a net negative effect leading to accepting the Westinghouse results with some caution.
The NRC staff and its consultant agree with the listing and note that the_ list directly reflects concerns expressed by the staff and its' consultant during the review of the Westinghouse program. The following items (not included-in the DRP list appearing on page 5.1-2 in Appendix B) are additional staff concerns which further adds to the negative aspecte of the modification design:
(1) One of the thermal-hydraulic design requirements is the relation of the_
manifold natural frequency to the relevant fluid excitation frequencies, p
It is not clear to the staff whether this objective was achieved.
(2) Another design objective is to limit the' gap flow velocity to less than a specific magnitude at full feed flow.
This objective does not appear to have been satisfied.
(3) In view of the SSPB test data and the fact that the characteristics of.the j
tube array have not been altered, the possibility exists that fluidelastic instab_ility is an excitation mechanism in some regions of a modified steam generator and should be recognized as such. Westinghouse hns acknowledged the possibility of fluidelastic instability. The nonlinear model used by Westinghouse cannot model the instability.
If the fluidelastic instability is still present after modification, the wear prediction based on the non-I
' linear model response to turbulence excitation will not give a reliable estimate of wear attributed to instability.
Accamulated wear data from operating plants will provide important information on the extent of fluid-elastic instability, if any, in modified steam generators.
The~ staff posi-tion on this issue is somewhat at variance with the DRP, which states that b'
after the modification fluidelastic instability will be absent.
(4) The staff suspected the existence of a possible overall flutter-like L
dynamic instability of the entire manifcid structure behaving as a canti-lever in a strong axial flow field. With frequencies reduced as noted NUREG-0966 3-3 La,
.m,
_, _ - -.. _ _.. _. ~ _ _,.
earlier, such a mechanism could produce a large first-mode-frequency vibra-tion.
Data from the SSPB and other model tests, however, do not point to the existence of such a dynamic instability.
Based on the data available to date, this concern has greatly diminished but it has' not been conclusively eliminated.
The not effect of the negative aspects discussed above and in the DRP evalua-tion is to create some uncertainty about how effective the internal design modification will be in preventing tube wear in the inlet region.
The staff believes that some tube wear, probably at greatly reduced levels, will con-tinue to occur even with the modification in place. Tube wear predictions remain uncertain as a consequence.
However, by providing periodic inspections on the initial two plants on a shortened interval as discussed in the DRP report, this concern has been adequately addressed.
Additionally, the staff has the following observations:
(1) The problem is extremely complex and is dependent on the ability of a tube to float within a tube support plate (TSP) under operating condi-tions, namely, with thermal distortions of plates and tubesheets, and with the inherent hydrodynamic forces in effect.
These conditions are very difficult to simulate in any model test.
If a tube is floating within a TSP, the potential for fluidelastic instability exists even with the design modificatio'n.
(2) Although individually the scale model's and the SSP 8 te'st' data may be deficient in certain respects, collectively the data can be relied on to evaluate the vibration performance of the manifolci modification.
- 3. 2 Structural Mechanics
(
l As a result of their review, the DRP concludes in Section 5.2.4 of Appendix B that the Westinghouse design modification is structurally adequate for all applicable loads, including those loads resulting from pipe rupture events with l
the exception that the manifold bolts allowable stresses could be exceeded dur-ing the design forward flushing transients.
Recommended actions to address this structural inadequacy (identified in Section 5.2.1.4, Appendix B) are described in Section 6.0 of Appendix C.
The staff conducted a detailed review of the specific areas of the structure that were highly stressed or where the caiculated fatigue usage factors were determined to be greater than 1.0.
These areas are identified in Tables 9.1.8-9.1.12, 9.1.11.25, and 9.2.12 of Reference 2.
Supplemental criteria used to justify that these areas meet their function for the intended service were provided in Reference 2.
These criteria were also reviewed by the staff.
Subsequent to the DRP Evaluation Report, each of these areas was re-evaluated l
by Westinghouse with the specific temperature limit on the forward flushing procedure to optimize system performance and component fatigue consideration relative to forward flushing.
A weld associated with the splitter ring, which is not addressed in the DRP evaluation, shows little sensitivity of the fatigue l
NUREG-0966 3-4
~ - -
i usage factor to changes in-fonvard flushing parameters.
This behavior of the weld results from a different heat transfer coefficient of Qis weld ar.d its location.
The primary contributor to fatigue usage at this weld is the plant load follow transient (100% power -.15% power - 100% power), assumed to occur once a day for the life of the plant. A more realistic assessment of the expected pl6nt load follow transi_ent (one,every otner day as opposed to once a day) reduces the fatigue usage factor to within an acceptable range. A crack initiation / crack propagation evaluation for this weld shows that the weld will meet its intended function under the deff t;ed transient conditions for the life
.of the plant.
For the remainder of the preceding component areas, forward f Nshing pa 7 meters-
.do have an effect on the calculated fatigve usage factor.
The results of the
're-evaluation for these areas show that, for a minimum-temperature of 200*F for water entering the steam generator, the fatigue usage is less than 1.0 for any purge flow rate in the range of 14% to 3%.
Based on this re-evaluation, the conclusion remains that the modification components will meet their function for the intended service.
At the staff's request, an assumed failure of the manifold attachment wela dur-ing a postulated feedwater pipe break event was avaluated by Westinghouse.
The result of the impingement of the manifold against the tubes was analyzed.
In this analysis, the first-row tubes are replaced by an equivalent tube.
One of the staff's concerns was that although the displacement of the equivalent tube will be identical to displacement of the actual tube, the strain distribution in the equivalent tube will be different from that in en actual tube.
As a result of a staff analysis, an assessment of the strain in the equivalent tube was concluded to be acceptable.
_'Another staff concern was whether or not a clean break / release of the manifold represents the worse case compared with the situation in which one side of the weld breaks first so that the impact is not the equivalent of s flat plate impacting all 38 tubes, but rather the edge of a plate impacting a few tubes.
Although the total impulse may be less, the load is taken by a smaller number of tubes. As a result of subsequent discussions with Westinghouse, the staff concluded.that the clean break / release represents the only realistic mode of impact.
The staff, therefore, concurs with the DRP conclusion that the first row of l
tubes is capable of absorbing the energy of the impacting manifold without rupture of the tubes during this postulated scenario.
Feedwater system modifications to eliminate the forward flushing transient, which result in overstressed conditions on some of the modification components discussed earlier, would involve rerouting of piping.
They may also involve changes in operating temperatures in the piping resulting in recategorization of the affected piping from high energy to moderate energy or vice versa.
The DRP evaluation does not address 'this aspect of the modification.
The staff will require that all affected piping be reevaluated on a plant-specific basis
~
in accordance with Standard Review Plan Section 3.6.2, " Determination of Ruptures Locations and Dynamic Effects Associated with the Postulated Rupture of Piping" (NUREG-0800, Ref. 3).
The reevaluation should include pipe break locations and jet thrust analyses.
l NUREG-0966 3-5 L.
3.3 Materials, Welding, and Nondestructive Examination Materials selection for the manifold and its fabrication techniques, and non-destructive' examinations (NDE) to be conducted during construction of the mani-fold assembly are discussed in Section 5.3 of Appendix B.
As a result of its review of both Reference 2 and Re DRP evaluation, the staff has concluded that the materials, welding, and NDE aspects of the Jesign, construction, assembly, and installation of-the proposed modification have been adequately addressed and are acceptable.'
3.4 Tooling and Installation Tooling and installation aspects of the proposed modification are discussed in Sections 4.1, 4.2, and 5.4 of-Appendix B.
The staff has concluded that tooling and installation have been satisfactorily addressed and reviewed by the DRP.
3.5 Radiological Considerations The staff reviewed Sections 4.4, 5.5, and 6.0 of Appendix B report concerning radiological considerations.
The staff finds the proposed measures consistent with 10 CFR 20.1(c) and Regulatory Guide 8.8 and, therefore, agrees with the
~
DRP conclusions with the following comments.
Utility-Specific Considerations Each facility should perform a comparative utility-specific radiological assessment of the proposed modification for their operating units prior to task.
initiation to determine the applicability of proposed worker radiological pro-tective measures and ALARA considerations, and determine how best to integrate this program into their own facility radiation protection program. Where sig-nific6nt differences exist in any of the radiole,gical parameters considered by Westinghouse (e.g., equipment, dose rates, radiation sources, doses, t mining),
these should.be evaluated and compensating radiation protection /ALARA actions described.
On completion of the modification, each facility should perform a summary radiological assessment of the task, as is recommended in C.3.c of Regulatory Guide 8.8, to enable the staff to evaluate the radiological results of the modification and determine if additional or different radiological con-trols need to be considerd. The summary should include the following:
(1) The collective occupational dose estimate updated weekly.
If the updated estimate exceeds the person-rem estimate by more than 10%, the licensee shall provide a revised estimate, including the reasons for such changes, l
to the NRC within 15 days of determination, l-l (2) A final report shall be provided to the NRC within 60 days after completion i
of the repair.
This report will include:
j (a) a summary of the occupational dose received by major task, and (b) a comparison of estimated doses with the doses actually received.
l NUREG-0966 3-6
3.6 ' Quality Assuance The staff reviewed Section 5.6 of Appendix B concerning1 quality assurance (QA).
The staff agrees with the DRP conclusions regarding QA given in Section 5.6.3 of their-report with the following comments:
(1) The audits and surveillances recommended by the DRP have been conducted by Westinghouse. They provide additional assurance'of the DRP's conclusions.
(2) The lead technical member of the DRP's QA review team is to verifysimple-mentation of the QA record plan at a later date as implementation of the plan will require several months of effort.
4,
(3)- Each utility should apply its 10 CFR 50 Appendix B QA program to the modi-fication of its' steam generators to verify that Westinghouse and other organizations involved in the work are_doing the work in accordance with the pertinent requirements of 10 CFR 50 Appendix B.
3.7 Chemistry and Cleanliness Control e
Chemistry and cleanliness aspects of implementing the proposed modification are discussed in Section 5.7 of Appendix B.
The staff has concluded that all aspects of chemistry and clear.liness control have been satisfactorily addressed and reviewed by DRP.
3.8 Inservice Inspection and Testing The program for inspecting the installed manifold. surveillance to be conducted during operation, and postoperation inservice inspection are discussed in Sec-tion 5.8 of Appendix B. ~ Based or_its. review-of Westinghouse's proposal as reviewed and evaluated by DRP, the staff has concluded that the testing and inspection program for the initial twc plants installing the modification was developed in a canner that will provide sufficient verification of post-modification performance. The staff concludes that this program should.be applied to the initial two U.S. plants installing the modification.
Further-more, the first eddy current test (ECT) inspection to be conducted after opera.
tion.with the modification installed sEL11 be expanded.
The plant's first inservice inspection, as defined in the plant Technical. Specifications, i.e.,
p 3% random sample inspection, chall be added to the DRP recommended preheater section ECT inspections.
In addition, the overall surveillance program shall be
-expanded so that during operation, the plant's loose parts monitoring system is used to monitor the manifold's structural integrity.
The staff further requires that plants with installed modifications shall revise their Technical Specifications tc-include inspection of a minimum of 240 tubes in the preheater sections in addition to the 3% random sample inspection as i
required by current Technical Specificatioris. Visual inspections of-the mani-fold shall:also be conducted whenever a scheduled ECT is performed for monitor-ing the post-medification-performance. The visual examination of the manifold by fiberoptic boroscope techniques shall be done in accordance with ASME Boiler and Pressure Vessel Code Section XI, IWA-2211 Visual Examination VT-1, which describes procedures for conducting remote visual examination.
i i
NUREG-0966 3-7
4 i
Tube Vibration Moaitoring Program l
The internal manifold was evaluated in both semiscale and full-scale tests and shown to reduce tube vibration response to acceptable levels; namely, levels corresponding to 40% power with the original design.
To verify that the mani-I fold achieves this reduction in an operating steam generator, a number of tubes must be instrumented with accelerometers mounted internally. Westinghouse and Duke Power Company have recommended that four tubes be so instrumented for McGuire Unit 1.
la particular, the selected tubes include two window tubes, two non-window tubes, a tube on the periphery of the bundle that is exposed to j
" skimming" flow, and a central tube. The staff has reviewed this recommenda-tion and concurs with it.
As discussed earlier, because relatively large tube-to-TSP. hole diametral clearances still remain, the staff believes that the potential for a tube to float within a TSP exists.
Furthermore, calculations based on the assumption of uniform flow show that at high power levels fluidelastic instability is possible if a tube can vibrate in a TSP-inactive mode.
Therefore, an additional objective of the accelerometer measurements will be to determine (1) if any of the tubes are vibrating in a TSP-inactive mode, and (2) if a threshold power l
level exists cove which acceleroneter measurements indicative of fluidelastic instability cccur.
In the original design two tubes at McGuire Unit 1 were previously instrumented with biaxial accelerometers and data were obtained under operating conditions.
Therefore, the staff considers it important that these original instruments be left in place for the post modification vibration monitoring program and that i
to the extent possible the same. type accelerometers and mounting techniques be used for the additional instrumentation, thereby allowing direct comparisons of tube response before and after modification.
With regard to axial positioning for the additional accelerometers, as noted above it is o' considerable interest to determine if a TSP is inactive.
In such a case, un accelerometer located at the particular TSP elevation would provide useful information. However, it is impossible to determine a priori for which tubes, and at what TSP location, such a condition might exist.
I Frequency response analysis from a mid-span accelerometer will provide informa-tion from which it can be determined that the tube is vibrating in a support inactive mode, and further excitation will be the strongest over the-inlet span (i.e., the span between TSPs five and six).
For these reasons, the additional accelerometers will be located at mid span locations between TSPs five and six.
The staff concurs in this approach.
I NUREG-0966 3-8
l
4.0 REFERENCES
(1) Letter froin S. K. Blackley (Design Review Panel) to H. R. Denton (NRC),
dated January 19, 3983, subject: " Utility Design Review Panel Evaluation Report, D2/D3 Steam benerator Design Modification, January 1983."
(2) Letter from E. P. Rahe (Westinghouse) to D. G. Eisenhut (NRC), dated December 23, 1982, subje:t: Westinghouse Preheat Steam Generator D2/03 Design Modification Eval.satSn Package.
(Proprietary report enclosed.
Proprietary addenda were submitted on Dec. 29, 1982; Jan. 2, 1983; and Jan. 14, 1983.)
Letter from E. P. Rahe (Westinghouse) to D. G. Eisenhut (NRC), nonpro-prietary version of report submitted February 2, 1983 (includes addenda).
(3)
U.S. Nuclear Regulatory Commission, NUREG-0800, "Si.andard Review Plan for the Review of Safety Analysis Reports for Nuclear Power Plants,"
July 1981 (formerly NUREG-75/087).
NUREG-0966 4-1
APPENDIX A PRINCIPAL CONTRIBUTORS This Safety Evaluation Report is a product of the NRC staff and a consultant.
The following were the principal contributors to this report.
Name Title Jon B. Hopkins Project Manager Jai Raj Rajan Mechanical Engineer Louis Frank Senior Materials Engineer Richard J. Serbu Health Physicist Conrad McCracken Chemical Technology Section Leader John G. Spraul Quality Assurance Engineer M. W. Wambsganss Consultant (Argonne National Laboratory)
I NUREG-0966 A-1
APPENDIX B UTILITY DESIGN REVIEW PANEL EVALVATIO,4 REPORT 02/03 STEAM GENERATOR DESIGN MODIFICATION JANUARY, 1983 DUKE POWER COMPANY TENNESSEE VALLEY AUTHORITY SOUTH CAROLINA FLECTRIC & GAS COMPANY l
l NUREG-0966 B-1 i
UTILITY DESIGN REVIEW PANEL-EVALUATION REPORT 02/03 STEAM GENERATOR DESIGN HODIFICATION JANUARY, 1983 DUKE F3WER COMPANY 11NNESSEE VALLEY AUTHORITY SOUTH CAROLINA ELECTRIC & GAS COMPANY TABLE OF CONTENTS
- 1. 0 INTRODUCTION 1.1 History and Need for Preheater Modification
- 1. 2 General Description of Model 02/03 Steam Generator 1.3 De:cription of Review Process 2.0 DEdiGN CRITERIA AND REQUIREMENTS 2.1 General Criteria 2.2 Materials Criteria l
2.3 Thermal Hydraulic Design Requirements
- 2. 4 Mechanical Design Requirements l
2.5 Transient Analysis
- 2. 6 Evaluation of Design Criteria i
I 3.0 OESCRIPTION OF MODIFICATION l
l 3.1 Description of Physical Design 3.2 fluid Flow Aspects of Design 3.3 Materials t
i l
1 4.0 DESCRIPTiUN OF INSTA.LATION 4.1 General Description of Modification Installation 4.2 Rigging and Tooling 4.3 Welding /NDE Requirements 4.4 Radiological Censiderations 5.0 EVALUATiDN 5.1 Thermal Hydraulics and Model Testing 5.2 Structural Mechanics 5.3 Materials, Welding and NDE 5.4 Tooling and Installation 5.5 Radiological Considerations /ALARA 5.6 Quality Assurance 5.7 Chemistry and Cleanliness Control 5.8 Inservice inspection and Testing 6.0
SUMMARY
Appendix A - Design Review Plan Appendix B - Design Review Panel Members Appendix C - Design Review Panel Meeting Minutes - June, 1982 Appendix D - Design Review Panel Meeting Minutes - September, 1982 Appendix E - Summary of NRC Comments II
- 1
s
- 1. 0 INTRODUCTION i
1.1 HISTORY AND N6ED FOR PREHEATER MODIFICATION On October 21, 1991, Ringhals Unit 3, a three loop Westinghouse plant with Model 03 split flow steam generators, was shut down due to a steam generator tube leak of 2.5 gpm. Upon investigation, tube R49 C55 (cold leg) had worn a small through wall hole at support plate 3.
The unit had 5
operated for about 2200 hours0.0255 days <br />0.611 hours <br />0.00364 weeks <br />8.371e-4 months <br /> at power levels above 75% at the time of the leak.
It was apparent that some type of accelerated wear mechanism involving interaction between the support plates and tubes was occurring.
~-
i Eddy current testing (ECT) performed on all three steam generators indicated that prefarential wear was-occurring in the outer three rows of tubes in the preheater section (Rows 47, 48, and 49).
Several tubes were i
removed from 'the affacted steam generator to better characterize the wear pnenomena.
Almaraz Unit 1 and McGuire Unit 1 both have Model 0 split flow steam generators and were operating at the time of the Ringhals 3 occurrence.
Both units were shut down and ECT inspections were performed. Almaraz 1 had been at power levels of 75% or greater for about 1500 hours0.0174 days <br />0.417 hours <br />0.00248 weeks <br />5.7075e-4 months <br /> and had i
similar indications in its Model D3 steam generators to those observed at Ringhals 3.
McGuire 1, which was in the startup test program and had not been above 50% power, showed no signs of tube wear in its Model 02 steam generators. Operation of all three of these units continued over the next year w;th restrictions on power level.
2 1.1-1 2
i 8
Westinghouse responded to the problem by establishing a task force to identify and correct the cause of the problem.1This involved gathering information relative to ECT measurements-from operating plants, pulled tube data, behavioral information from analytical models'and information from a series of air and water scale model test facilities. Monitoring instrumentation was installed external to and inside several s was generators to detect and measure tube vibration. The information from all these sources was then used for the following purposes:
1.
Map fluid velocities and turbulence in the preheater inlet plenum, 2.
Determine. resultant motion of the steam generator tubes, and 3.
Correlate feedwater flow to tube motion and wear.
Integral with the above investigations, Westinchouse-developed several conceptual modifications to. reduce tube vibration and the resultant wear.
Many designs were conceived; several reached the conceptual design stage and ultimately one design was selected as the " reference" design. During the preliminary testing cf this reference design, an " alternate" design was developed. This alternate design was the internal manifold which is the final design proposed for installation in the Model b2/03 steam generator.
w 1.1-2 i
' a
- 1. 2 GENERAL DESCRIPTION OF MODEL 02/03 STEAN GENERATOR Heat generated in the Wastinghouse' pressurized water reactor (PWR) is removed from the~ core by the primary coolant water which is transported to the steam generators by reactor coolant pumps.
Each' primary coalant circulation loop in the W stinghouse PWR design has.one primary coolant e
circulatior: pump and one vertically mounted U-tube steam generator. -The steam generators are designed with the following integral sections:
a preheater section, an evaporator section and a steam drum section (see Figure 1.0.1 of the Westinghouse report). The steam drum section is the upper part of the steam generator assembly which contains the moisture separators. The evaporator section is an inverted U-tube heat exchanger containing 4674 (3/4-inch diameter) inconel tubes through which primary coolant water is circulated f*om the reactor fuel to transfer heat from the primary coolant to water in the secondary side of the steam generator which causes the secondary side water to boil. The primary coolant water l
enters the hot leg section of the U-tuces at approximate'ly 618 F and flows through the' discharge or cold leg side of tha U-tubes exiting the steam generator at approximately 558'F. The primary coolant flows from the steam generato= through the reactor coolant pump and is returned to the reactor vessel where it is reheated.
Feedwater is pumped into the secondary or shell side of the~ steam gene-rators where it boils and generates steam to drive the turbine generator.
In order to enhance heat transfer to the incoming feedwater, the Model D series steam generators incorporate a series of baffle plates in the steam generator around a portion of the cold leg which forms the pre-1.2-1 1
--a
. heater section. Feedwater flowing into the steam generator first passes-through a venturi insert in the main feed nozzle which serves as a backflow restrictor to limit the rate of blowdown fron'the steam gener-
'h
~
ator in the event of a main feedwater line break. Upon entering the preheat section, feedwater is diverted by a flat circular impingement platt which has a slightly larger diameter than tha inside of the inlet nozzle. Feedwater tt.en moves through the flow paths in the support and baffle plates in the preheater.
- 1. 2.1-FEEDL'ATER BACKFLOW RESTRICTOR Located in the inlet of the feedwater nozzles of both the D2 and D3 steam generators as presantly designed is-a backflow restrictor made of inconel.
The purpose of this device is to restrict the rate of flow of feedwater from the secondary side of the steem generator in the event of a feed-water line break. The backflow restrictor assembly, which fills the inside diameter of the thermal sleeve in the feedwater inlet nozzle, has four even'y spaced holes, each of which has a radius on the inlet side and a throat diameter of.approxima;<1y 3-3/16-inches with a diverging conical discharge side. The thickness of the flow restrictor at the location of the holes is approxitately 10-1/4-inches. The backflow restrictor is attached to the thermal sleeve in the feedwater nozzle by an intermediate carbon steel sleeve to which the backflow restrictor is i
welded after the intermediate sleeve has been wetaed to the thermal sleeve. The dissimilar-metal weld (i.e., inconel-to-carbon steel) is at the intermediate sleeve-to-backflow restrictor weld joint.
1.2-2
,-m-a-
.u
-, - +
MM l
1.2.2 IMPINGEMENT PLATE ASSEMBLY The impingement plate sssembly for the Model 02/03 is located inside of b
tt.e preheat section of the steam generator concentric to the feedwater nozzle and provides a barrier between incoming feedwater and the U-tubes in the preheat section. The impingement plate assembly is made up of two 3/4-inch thick circular carbon steel plates which are 17-inches in diameter. These plates are concentric and parallel to each other 1-inch apart and are %parated by two 1-inch square horizontal bars which are attached tb the inner plate by fillet welds prior to assembly with the outer plate. The outer plate is attached to the 1-inch square bars by two 5/8-inch diameter plug weld; for each bar. When assembled in the ster.m generator, the ends of the 1-inch square bars are welded to two 3/4-inch diameter vertical bars which are attached to preheater plates 5 and 6 and are located outside of the perimeter of the U-tubes.
The outer impingement plate is attached to the thermal sleeve in the fee &ater nozzle by four carbon steel bars which are 2-inches wide, 1/2-inch thick, and 6-inches long. The four bars are welded perpendicu-lar to the outer impingement plate at 90 degree intervals, and the other end of the 6-inch length is welded to the inside diameter of the feed-water inlet nozzle thermal sleeve.
1.2.3 PREHEATER The preheater section of the steam generator is made up of a series of eleven semicircular baffle and tube support plates thtcugh which the cold 1.2-3
leg tubes of the steam. generator are routed. 'The plates are spaced in j
the first 140-inches above the tubesheet.
The'five plates located below the feedwater nozzle-are spaced from 7-1/4 to 19-1/4-inches ' apart; plates 5 and 6, at the feedwater nozzle, are located 21-1/4-inches apart; and spacing of the platos above the feedwater. nozzle range from 10-1/4 to 21-1/4-inches apart.
Feedwater flows into the preimat section of the steam generator between plates 5 and 6 the bottom plate is number.one (03 only). The incoming feedwater flows horizontally across the cold leg tubes, and a percentage goes down through flow. slots in plate 5 and reverses direction by 180 degrees. The baffle ~and tube support plates cause three additional 180 degree direction reversals to 0:: cur in the water flowing downward in the preheat section before the water. discharges into the boiling section around tne hot leg tubes.
The-principal function of the two bottom baffle plates, is to control the direction of feedwater flow.
Clearance holes drilled in these plates for the 3/4-inch outside diameter steam generator tubes are slightly larger than the tubes. The remaining percentage of the feedwater which flows upward in the preheat section makes 3 changes of horizontal direction of flow by 180 degrees as it flows around the tui:e support plates, and the last direction chznge is'90 degrees-from horizontal' to flow upward. The two top support plates have flow holes centered between-tube columns and rows. Holes for tubes in the support plates are slightly L
larger than the tube outside diameter.
1.2-4
I
1.3 DESCRIPTION
OF REVIEW PROCESS _
1.3.1 DESIGN REVIEW PANEL (DRP)
During the first several months after the Model D Steam Generator problem was identified, NRC staff worked with Westinghouse and the several utilities on an individual basis. Only one unit (McGuire-1) with Model 0 steam generat?.rs was licensed in the United States. However, a second unit (Summer) was nearing completion of constructior and 4 additional units (McGuire-2, Watts Bar 1-2, Catawba-1) were scheduled.
In an effort to conserve resources and to minimize the time for NRC review and approval of the proposed design changes, the concept of a third party design review of the proposed modification was initiated.
Such review was intended to lessen the need for a detaiied technical review by the NRC and its final report was proposed to serve as the basis for the NRC l
Safety Evaluation Report.
Responding to the identified needs for a thorough and coordinated review and for conservation of resources, Tennessee Valley Authority, South Carolina Electric and Gas Co., and Duke Power Company agreed to pool resources to form a Design Review Panel to exanine all aspects of the final Westinghouse design for the Model 02/03 modificaion. A review plan was developed to serve as guidance for the Design Review Panel (DRP).
A copy of this plan is included in Appendix A.
l Members of the DRP were selected from the three utilities' staffs.
Dr.
l F. L. Eisenger was retained as a consultant by TVA to serve as a panel l
l 1.3-1 l
member and to assist in the review effort.
Dr. Eisenger has extensive experience in multiple areas of steam generator design and operation. A listing of the members of the DRP and a summary of their background and qualifications is shown in Appendix B.
1.3.2 SCOPE OF REVIEW The plan of the DRP was to review the final modification design selected by Westinghouse for installation in the steam generators, and to deter-mine whether or not it was acceptable, and to report on conclusions reached.
In performing the review, many different areas have been examined in detail. These include - thermal hydraulics, model testing, ALARA, structural mechanics, stress analysis, quality assurance, inservice inspection, materials, tooling, welding, chemistry control, and installa-tion. All of these areas have been addressed by Westinghouse, and the DRP review conclusions in each of these areas are contained in the report.
This DRP report provides an overview of the work done by the panel in datormining the acceptability of the modification. The panel had the benefit of numerous technical meetings with WestinghoJse, including two full Design Review Panel meetings, multiple teiephone conference calls, and correspondence.
In addition, DRP technical members have had access to Westinghouse test data and results and to original calculations and
{
design details. Table 1.3-1 summarizes the nectings ant other activities 1.3-2
attended by various members of the DRP.
Copies of the minutes of the two full Design Review Panel meetings ars' contained as Appendices C and D.
During the course of the DRP review, many technical issues were examined i
in detail.
The more important issues and their resolution are discussed in the appropriate sections of this report. These issues were resolved through both informal and formal means.
Written and verbal questions were given to Westinghouse.
Formal responses to the questions were provided and are noted in the files of the DRP.
In all cases, outstand-I ing items and issues were resolved to the satisfaction of the DRP.
The NRC Staff was in attendance at the two full Design Review Panel meetings.
At the conclusion of each of these meetings, staff members had an opportunity to make comments. Apoendix E is a summarization of the comments made by the NRC in the DRP meetings.
All of these items were aduessed during the review and resolved to the satisfaction of the DRP.
The DRP relied to a great extent on the material contained in a Westing-house document entitled " Westinghouse Preheat Steam Generator - 02/03 Design Modification - Evaluation Package - December, 1982." A one volume version was provided to all ORP members and NRC observers in the June ORP meeting.
In the September ORP meeting, Westinghouse distributed an updated two volume version of the report which superseded the one volume version.
(Three copies of this September report were transmitted to NRC by letter dated September 30, 1982 from Mr. E. P. Rahe, Westinghouse to Mr. Darrell G. Eisenhut, NRC/0NRR.) Numerous rsfarences are made to this document in the ORP report.
(References to the Wstinghouse Design 1.3-3 l
\\
Report or Westinghouse report refer to the above mentioned document.) In addition, an updated version of the Westinghouse report was transmitted to NRC by letter NS-EPR-2696 dated December 26, 1982 from Mr E..P. Raha of Westinghouse to Mr. Darrell G. Eisenhut,'NRC/0NRR. This update was supplemented by additional information tran mitted to NRC by letter NS-EPR-2699 dated December 29,'1982 from Mr. E. P. Rahe of Westinghouse to Mr.'Darrell G. Eisenhut, NRC/0NRR.
l
\\
l 1.3-4
i j
TABLE 1.3-1 Model 02/03 Steam Generator Preheater Modification Design Review Panel Activities N
}
G h h
A N Ae N W N ) W : $ ! /
=
l LBeaa Srfa6 Y
Y M
1 Y
S._kran_hn X
X X
X X
X X
X
-1
. _I Y
Y Y
.X C. r--
7 a M
X X
X X
Y Y
Y M
Y G._rm har X
X X
._ I X
X Y
1 Y
M. [mias.
Sffs __
X X
X J
._M F. Entrarse M
M X
X iY Y
Y
_X Y
I. Ma=a M
X X
X X
X X
_X Y
1 Y
, _.___ X.
C Assat1_.hmE X
Y L btass M
M Y
X Y
M X
Y ~
LMm M__
X Y
LM-sfrs M
y M
M v
v W, jirfanammat har M
M I
M V
1 Y
L Pma u SffsG Y
Y X
Y M
M V
lL.bnene SilIG X
X X
X X.
Y V
[. Asese
&g_
X X
X Y
Y M'
IL_StagfRg_ hag X
X X
M Y
1' Y
__ _. _ X R. M M
I M
M 1
Y X
X C._Ilms
.Bar _
X M
M 1
X X
LWests M
X X
X Y
M
=
1 y
X v
&_linymEA _S(fgi X
X M
Y.
I M
M Y
r Y
y LNutamn M.
M Y
_1 Y
Y
.X 4
B._Wittian _ Bat r
M v
v v
Sie
_E X
X r
v IbiEt AL NETHES WITH W MAE IED RN
- Veslis $@ 6f/ - fM PiTTSMNGl MEA We ESS NOTED oHMISE.
2.0 DESIGN CRITERIA AND REQUIREMENTS 2.1 GENERAL CRITERIA To ensure that the modification would be compatible v5 the steam-
- generator, Westinghouse established design criteria and requirements that have been app?ied to'tbe manifold and the flow ifmiter. The general criteria are as follows:
1.
The safety classification of the components shall be established in accordance with ANSI N18.2-1973 and addendum ANSI N18.2a-1975.
2.
Material, design, coastruction, analysis and installation of the modification shall use the ASME Code Sections II, III, V, IX and XI as guidelines.
l l~
The DRP has reviewed the application of these general criteria to the modification design. The DRP has determined that the modification has been appropriately classified with respect to safety as Safety Class 2 for the flow limiter and as Non-Nuclear Safety for the manifold assembly.
The materials, design, construction, analysis and installation of the manifold have appropriately utilized the applicable ASME Code sections L
and subsections as guidelines, d
2.1-1
- 2. 2 MATERIALS CRITERIA The materials chosen for the modification have been selected to be compatible with the steam generator materials and with the steam generator service condi-tions~in which they will function. The' design review panel has determined tnat the materials chosen for the modification meet these criteria. Materials considerations are discussed further in Sections 3.3 and 5.3 of this report.
I 1
1 i
l 2.2-1
2.3 THERMAL-HYOPAULIC DESIGN REQUIREMENTS The major thermal-hydraulic design requirements / objectives for the preheater modification specified by Westinghouse are listed below:
1.
The nominal full power.feedwater flow condition to be used in design and
. analysis are:
feedwater flowrate - 4.1 x 10s pounds /hr, feedwater temper-ature - 440*F, secondary side pressure - 985 psig.
2.
.The design of the reverse flow limiter shall be such that at the intersec-tion of the wrapper and nozzle thermal sleeve the maximum local velocity is less than a specified minimum value under normal flow conditions.
3.
The design objective for the internal manifold flow distributor is to-
. distribute the feedwater over the tube bundle inlet as uniformly as possible.
4.
The design objective of the internal manifold flow distributor is to limit the average velocity in the gaps between tubes to less than a specified l
minimum value at full feed flow.
5.
The internal manifold flow distributor shall limit the crossflow velocity in the preheater at 100 percent feedwater flow.
l l
?.3-1 i
6.
Theinternalmanifoldflowdistributorshallinotbesubjecttoflow induced vibration.
It.shall be a design goal that the manifold ~ natural frequency be at least twice the relevant fluid excitation frequencies.
7.
Analysis shall be made of the potential for sludge collection. This analysis must show that deposition is i:ot harmful or else a method of.
correction for: sludge deposition must be designed and developed.
2.3-2
2.4 MECHANICAL DESIGN REQUIREMENTS-The major mechanical design requirements / objectives for.the preheater modifica-tion specified by Westinghouse are listed below.
' 1.
The modified reverse flow limiter shall produce uniformly distributed exit flow. The flow area of the reverse flow limiter shal? be equal to or less than 0.22 square feet. The reverse flow limiter may not be attached by welding to any portion of the steam generator shell or nozzle pressure-boundary.
2.
The internal manifold shall be designed to distribute the feedwater uniformly across the face of the inlet area of the tube bundle. The internal manifold shall be supported from the nozzle thermal liner.
3.
The modification components shall be designed to operate with the water chemistry conditions.specified by Westinghouse to operating plants.
4.
--The design of the modification shall not cause significant excitation due to steady-state pressure oscillations.
Stresses introduced fr.to struc-tural members shall be less than the fatigue endurance stress (at 1010 I
cycles) for the applicable caterial.
S.
For purposes of the ASME Code Calculations only, a 40 year design shall be considered.
,\\:
2.4-1
6.
Due to the modification, the reduction in tube wall thickness must be demonstrated to be less than 40 percent of the original wall thickness over the design period of the steam generator.
a r
l 2.4-2 l
1 2.5 TRANSIENT ANALYSIS The effect of normal operating conditions and transients on the modification was analyzed. Normal operating conditions are specified in Section 2.2-1.
Transient conditions which were considered are listed in Tables 2.2-1 through 2.2-5.
The t;ansien.t conditions for which the feedwater inlet modifications must be evaluated can be divided into water hammer pressure transients and therm.41/ flow transients.
Although both thermal and flow transients may be associated with the same event they can be analyzed separately because the. transient loads do not occur simultaneously.
Steady state pressure pulses / forcing functions can be generated on the secondary-side of the steam generator as a result of valve closure / opening, pump start /.
stop, hydraulic resonance, pump oscillations or steam bubble collapse. The design of the-steam generator modification considered the effects of steady state pressure pulses.
l 2.5-1
F
-Table 2.5-1 Normal Operating Conditions
' 1.
Plant Heatup/Cooldown 2.
Plant loading at 5 percent of full power / minute'.
^
3.
Plant unloading at 5 percent of full power / minute 4.
Small step load increase 5.
Small step load decrease 6.
Large step load increase with Steam Dump 7.
Feedwater Cycling at Hot Shutdown.
8.
Steady State Fluctuation 9
Plant Loading and Unloading N; tween 0 and 15 percent power [Fe vard Flushing Purging]
10.
Loop Out of Service hormal Loop Shutdown Normal Loop Startup 11.
Boron Concentration Equalization
- 12. Turbine Roll Test 13.
Reactor Conlant Pump Startup/ Shutdown 14.
RCS Venting 15.
Feedwater Heaters Out of Sarvice l
)
p i
l l
., -. - -, - - - -..,. -,. =.. -, -.. - - -, ~, - -,.. _. - - - -.
.,-n
V Table 12. 5-2 Upset Condition Transients 1.
Loss of Load 2.
Loss of Power 3.
Partial Lnss of Flow 4.
~ Reactor Trip From Full Power-A.
No inadvertent cooldown B.
With cooldown and no steam injector C.
With cooldown and steam injection 5.
Inadvertent'RCS Depressurization 6.
Inadvertent Startup of Inactive Loop 7.
Control Rod Drop 8.
Operating Basis Earthquake (OBE) 9.
Excessive Feedwater Flow 10.
Inadvertent Safety Injection
- 11. Bypass Line Tempering Valve Failure 12.
Excessive Bypass Feedwater
- 13. Check Valve Closure
- 14. Upset Bubble Collapse
\\
f;
f Table-2.5-3
_ Emergency Condition Transients 4
1.
Small Loss of Coolant Accident 2.
Small Steam Rreak 3.
Complete Loss of Flow Note: The main feedwater nozzle is not subjected to any emergency transients.
Therefore, emergency condition transients'are not considered in the fatigue usage calculations. However, the effects of the transient were considered in the structural evaluation of the modification.
i s
Table 2. 5-4 Faulted Condition T ansients j
1.
Reactor Coolant Pipe Break (Large LOCA) 2.
Large Steam Line Break 3.
Feedwater Line Break / check Valve Closure 4.
Reactor Coolant Pump Locked Rotor 6.
Control Rod Ejection 7.
Simultaneous Feedline/ Steam Line Break 8.
Tube Rupture 9.
Faulted Bubble Collapse Note: Faulted condition transients are'not considered in the fatigue usage calculations. However, the effects-of the transients were considered in the structural evaluation of the modification.
l t
(
(
1 I
,.. ~. _ _ _ _ _ _ _ _ _. - -.
Table 2. 5-5 Test Condition Transients 1.
Primary Side Hydrostatic Test 2.
Secondary-Side' Hydrostatic Test 3.
Tube Leakage Test -
s Secondary Side Pressure 200 psi 400 psi g
600 psi 840 psi
' 4.
Primary Side Leakage Test-5.
Secondary Side Leakage Test (Manway Closure)
Note: Test condition transients are not considered in the fatigue usage calculations. However, no changes to the original analysis was neces-sary due to the feedwater inlet modification.
Test conditions do not result in loads being applied to the de' sign modification.
l
i 2.6 EVALUATION OF DESIGN CRITERIA The original goal was to design a modification which met all the original daign criteria for the steam generator. Westinghouse was able to achieve this goal with one exception - forecast tube wear.
Conservative analyses by Westing-house predict that a small number of the tubes'could experience wear during the 40 year design life of the unit. Given that the modification does distribute the feedwater flow evenly over the tube bundle, further reduction in tube wear due to redistribution of flow is not feasible. This wear is predicted to occur over a long enough time interval that the inservice inspection can be relied on to indicate the need for corrective action. Thus, this deviation from the design criteria is judged to be acceptable. This is discussed in more detail in Section 5.1.
All other criteria are deemed appropriate for the design of the modification.
)
2.6-1
i
3.0 DESCRIPTION
OF MODIFICATION
3.1 DESCRIPTION
OF PHYSICAL DESIGN 1he 02/D3 preheat steam generator design modification including major components, is shown in Figures 5.0.1 and 5.0.3 of the Westinghouse Design Report.
For comparison, the existing design is shown in Figure 1.0.4 of the Westinghouse report. The major components and items of the Westinghouse design modification consist of the internal manifold (6 individual boxes); the manifold bolts, nuts, studs and lucking cups; the manifold support sleeve; the flow splitter, the reverse flow limiter support cylinder and the reverse flow limiter.
Except for the six internal manifold boxes which are bolted together using the manifold bolts, studs, nuts, and locking cups; all other components are either welded to adjacent modification components or to the thermal (nozzle) liner.
No attachment welds are made to the pressure boundary portion of the steam generator, Figures 5.1-2 and 5.7-1, and Section 5.7 of the Westinghouse report contain additional details concerning the installed configuration and assembly of this design modification.
The reverse flow lim' iter, the flow rplitter and the interncl manifold are further dis-cussed below. Additional details concerning the cther components are contained in Section 5.0 of the Westinghouse report.
)
The reverse flow limiter is similar to the existing flow limiter except
(
it now has 19 individual venturi holes instead of 4.
The total throat area of the new 19-hole venturi is 0.22 square foot, which is the same as the existing 4-hole venturi. The reverse flow limiter serves the func-3.1-1
tion of limiting the reverse flow from a steam generator following a postulated feedwater line break accident. The new flow limiter will have an outside diameter of approximately 13 inches, with a total length of approximately 83s inches. The support cylinder, to which the flow limiter attaches is designed such that in the event of failure of the attachment weld, the flow limiter would be caught downstream by the smaller diameter portion of the support cylinder.
The flow splitter and splitter plate are an integral unit. The splitter consists of six radial vanes equally spaced at 60 degree intervals and welded to an outer ring, a conter hub, and the flow splitter plate. The flow splitter will have an outside diameter of approximately 14 inches and a length of approximately 13% inches.~~ Proper engagement between the manifold and the flow splitter is ensured by two alignment pins and a 1/4" counterbore in the entrance plate of the manifold to match the ciameter of the splitter plate.
The internal manifold consists of six manifold boxes.
Each manifold box is a welded construction with two perforated plates and a sector of a circular collar.
Each box consists of an exit plate, ertrance plate, flange plate, front plate, end plate (corner boxes only), and two wr-tical rib plates.
Figure 5.4.1 of the Westinghouse report shows addi-'
tional details. The exit plate is drilled with a uniform pattern of hnles.
The entrance plate is also drilled with a uniform pattern of holes with diameters as indicated in Figure 5.4.2 of the Westinghouse report except that the holes are offset from the hole pattern in the exit plate.
3.1-2
Additional details, thicknesses,'subcomponent dimensions'and other pertinent information are provided in Section 5.4 of_the Westir.ghouse
. Design Report.
4 t
i i
i l
I 3.1-3 l-
3.2 FLUID FLOW ASPECTS OF DESIGN As compared to the previous impingement plate performance, the manifold-provides a more uniform flow distribution over the face of the tube bundle. This results in the following improvements:
- )I 4?g Qdijf.f..
x u.3 a.
Lower peak flow velocities.
f.'Z :0*
'i S ;t, r er b.
Elimination of reverse flows which existed in the midcle area.
c.
Lower fluctuating velocity components.
d.
Lower vibration resulting from Tbove velocity improvements.
x:
M
. 9,x p..' --
- t '
fr.].
The.new 19-hole flow restiictor also helps to improve the performance of
"' i the manifold by providing more evenly distributed flow into the manifold.
3.2-1
3.3
' MATERIALS
-3.3.1 GENERAL MATERIALS CONSIDERATION The primary factor _ involved in selecting materials for the various parts of the manifold assembly are listed below:
-~
j 1.
Strength and fatigue resistance to withstand transient loadings.
2.
Erosion resistance to withstand high velocity feedwater flows.
3.
Corrosion resistance te withstand environmental effects; e.g.,
pitting, stress corrosion cracking.
+
.e Table 3.3-1 is a summary of the components in the manifold arsembly and' the materials used to construct each component. -This table also includes.
the basic design function of each element of the modification.
l The primary material used in the manifold is Inconel Alloy 600. This material is used in the manifold boxes, flow splitter, reverse flow
~
limiter and various support sleeves and cylinders.
A second material used for major components is carbon steel, SA 106 Grade f
C.
It is used in those applications in the modification where interface with the nozzle therma! liner is required.
l l
l 3.3-1
The studs,' bolts and nuts.used to fasten the manifold boxes together are critical to overall ' integrity'of the assehbly.
Type 410 stainless steel was chosen for this application. -This material exhibits superior resist-l i
anceito general corrosion and 'to-stress corrosion cracking.
ts' C*R S
~
1
. i t
'9(
The ferritic stainless steel was purchased to a specification that required room temperature yield strength within specified ranges.
~
- The fatigue resistance of the fasteners was enhanced by.using rolled threads and undercutting the nut. This icproved fatigue performance was conservatively verified using the procedures in ASME Boiler and Pressure Vessel Code Section III Appendix II.
"~
Other features of the fasteners have been considered which will enhance box alignment, provide for easier installation and maximize load carrying
,a v..
. capability. These are discussed in detail in the Westinghouse Report, m
i Section 5, F.
4 e
3.3-2 1
- -]
i L
TABLE 3.3-1 D2/D3 DESIGN MDDIFICATION MATERIAL COMPOSITION ITEM DESCRIPTION DESIGN FUNCTION MATERIAL CONF 05I M REVERSE FLOW PREVENT RAPID S/G LLOWDOWN INCONEL 600 LIMITER REDUCE FLOW NON-UNIFORMITY SUPPORT REVERSE, FLOW LIMITER SA-10 SUPPORT CYLINDER OBTAIN GREATER' FLEXIBILITY TO GRADE C RESPOND TO THERMAL TRANSIENTS INCONEL S00 REMOVE FLOW LIMITER ATTACHMENT FROM PRESSURE. BOUNDARY j
FORM A WELD DAM (IF NEEDED)
SA-10 BACKING RING FOR SUPPORT CYLINDER TO GRADE C THERMAL LINER WELD INTERNAL DISTRIBUTE FLOW OVER TUBE INCOMEL 600 MANIFOLD BUNDLE UNIFORMLY B0XES REDUCE TUBE VfBRATION SUPPORT MANIFOLD SUPPORT REMOVE MANIFOLD ATTACHMENT INCONEL 600 SLEEVE WELD FROM VICINITY OF WRAPPER /
SA-106-LINER WELD GRADE C FLOW DISTRIBUTE FLOW TO BOXE5 INCOnEL 600 SPLITTER SA-106 GRADE C STUDS / BOLTS HOLD MANIFOLD BOXE5 TOGETHER SA 193-B6 (410 SS)-
NUTS HOLD MANIFOLD BOXE5 TOGETHER-SA 194 GR. 6 (410 SS)
LOCKING CUPS
-RtIAIN BOLTS & NUTS ~
INCOMEL 600 4
'., 4 -
~
4.0 DESCRIPTION
OF INSTALLATION 4.1
~ GENERAL DESCRIPTION OF M001FICATION INSTALLATION This section' describes the sequence in which the preheater modification
.4i i
components are installed and the manner in which they are attached. The installed configuration of the 02/03 Design Modification is shown in Figure 5.0.1 of the Westinghouse' report.
i The modification components have been designed to provide a minimum diametral clearance between the component support surfaces and the I
nozzle thermal sleeve for ease of installation. Tooling is provided to handle and position the components. All.of the installation field welds are similar metal welds performed using the ASME Boiler and I
Pressure Vessel Code,Section III, Subsection NG and Section IX as guides.
The six manifold box assemblies are iridividually-inserted in a prescribed j
i sequence through the inlet nozzle into the preheater inlet plenum.
The
.,]f six manifold boxes are then fastened together with bolts and tapered studs while being supported inside the inlet plenum on baffle plate 5.
The bolts and studs are torqued to give a specified axial preload 6000 lbs. and preload tensile stress 42 Ksi.
The manifold support sleeve is inserted into the inlet nozzle thermal sleeve, then positioned,and tack welded in place to the thermal sleeve.
4.1-1
L y
I 4
I
-l s
The internal manifold is raised and its suppcet collar is inserted into
}
the thermal sleeve agaicot the sanifold support sleeve.
The manifold is located, radially, by the necked-down end'of the support collar that fits into'the counterbore of the support sleeve.
After centering the manifold 4
and assuring proper clearances between baffle plate 5 and 6, the wrapper e
and the tube bundle, the manifold is tack welded to the support slaave.
The welus are completed and inspected after performing a trial fit of the flow splitter. The manifold bolts are then untorqued and retorqued and the locking cups are crimped at a minimum of 2 places 180* apart.
The flow splitter is then inserted into the manifold. Alignment pins are provided in_ the flow splitter for orientation of the flow splitter
~
relative to the manifold. The installation weld is completed after the seated position of the flow splitter is verified.
s The next step in the installation of _ the modification is to install the reverse' flow limiter support cylinder. The support cylinder is located from the flow splitter vanes. The ' distance from the flow splitter vanes
_e to the reverse flow limiter venturi exit surface is controlled to make l
's the field installation of the modification consistent, hydraulically, with the component positions raintained during development testing.
After the welding operations are completed between the support cylinder and thermal-sleeve, the reverse flow limiter venturi is positioned inside J
the support cylinder. The welding operations are subsequently completed between the support cylinder and reverse flow limiter venturi. This step completes the installation of the modification within the feedwater nozzle.
4.1-2 l
4.2 RIGGING AND TOOLING Special tooling was developed for installation of the modification in accordance with the' design drawings and documents. These tools'were r
designed to; accommodate ease of installation, ALARA considerations, and
- to. minimize plant.down time. -Where possible, commercially available
-equipment was modified to perform special tasks such as automatic welding and reverse flow ifaiter removal. Several operations. required. develop-ment of "first of a kind" tooling. Tiiese operations include manifold insertion and bolt crimping.
A tool qualifica?. ion program was implemented to verify field readiness of the complete tooling system. A modification crew training program is' also in place.
Personnel will have actual hands-on experience using the tools before going to the field.
4.2.1 TOOLING Tooling and processes have been developed for removal of the base case components and for modification installation. All tooling has been extensively tested on full scale mockups for verification of task per-formance and tooling reliability.and suitability. The modification tooling program has been separated into two distinct groups of activi-ties,.one being the process of disassembly and the other, installation of new components.
In addition, should it be necessary, the modification components'could be removed after installation with this tooling package.
4.2-1
o 4.2.2 MAJOR TOOLING TASKS A.
Flow Limiter Assembly Removal h
1 A key objective in selecting trie tooling design was to develop a -
. lightweight and compact tooling system that would be well suited for use in the restricted nozzle area. An internally mounted lathe was selected for this task. The lathe is powered by three 0.9 HP air motors each delivering 150 ft-lbs stall torque.at 90 psi.
The unit is built to be lightweight and easy to set up. Four locator pads incorporated into the unit bring it into square and on center with the nozzle. Once the unit has been positioned inside the nozzle, all the machining operations can be controlled remotely by controlling the radial feed rate on a built-in digital counter.
An automatic machining tool designed for facing, boring, leveling and cut-off was selected as a back-up to the internally mounted lathe.
Its primary purpose will be nozzle I.D. boring and machining operations which may be required after weld cutting and weld build-up/ropair procedures. The system incorporates a radially adjustable tool bit holder mounted on a slide which is activated axially by means of an automatic mechanism. The unit is externally clamped on the feedwater nozzle, thus, providing high visibility and access by the operators. The unit'is powered by a hydraulic power supply which provides a working pressure of up to 3000 psi. The system has 4.2-2 Ja
-the capability of being controlled remotely, thus limiting radiation exposure.
A Flow Limiter Handling Fixture was designed for removal of the Elow Limiter from the nozzle once the attachment weld is removed. This tool incorporates a cam-lock feature which engages the I.D. of two flow holes. This tool also includes two jacking studs which can be used to apply a pulling force if-required during removal.
B.
Impingement Plate Assembly. Removal A metal disintegrator machining (MDM) process'is used for impinge-ment plate removal. The process can best'be described'as spark erosion, but it differs from electrical discharge machining (EDM) in several ways.
The process employs water as a coolant fluid, rather than an oil l
based dielectric; fluid. Furthermore, it uses a low voltage power supply in. lieu of the high frequency power supply used by the EDM process. Finally, it must use a vibratina electrode synchronzied I
with the positive peak of the power transformer wave.
The MDM cutting tool uses a very fine grain grachite electrode which f
prcvides a high metal removal rate with low electrode wear.
Differ-ent shaped electrodes are used for each cut. The MDM system con-a sists of the following subassemblies:
l 1
1 4.2-3 l-
1.
XY Tool Carrier' provides support and guidance for the disin-tegrating head. The desired position.is entered by a thumb -
wneel for each axis, X ar.d Y.
2.
Power Supply
. includes 20 KVA water cooled transformers, controls and. power coolant circuits.
-,c
., y 3.
Coolant Handling System - consists of a 60 gallon tank with a 5 clicron filter section, a catch pan and curtain,' suction pump and centrifugal pump which delivers clean water to the elec-trode. Using this system, almost all the MDM coolant fluid will be collected and processed, thus minimizing the loss of contaminated water in the-steam generator.
4.
Control Panel - The MDM system is controlled remotely by a panel with indicator lights, electrical instruments, switches -
and gauges to inform the operator of the process parameters and movements of the tool.
C.
Manifold Installation and Assembly Special handling fixtures have been' designed for manifold installa-tion and assembly. The manifold is installed in segments and then f
bolted together. After manually torquing these bolts, a special crimping tool is used to lock the bolts'in place. Once the manifold is assembled, it is mated. to the bimetallic ring and the entire assembly is then properly aligned in the nozzle using special 4.2-4
I tooling for squaring and centering the manifold assembly in the nozzle.
D.
Welding of Modification Components 2
An automatic gas-t8ngsten arc welding system (GTAW) has been de-signed and qualified for production welding. The unit, mounted inside the feedwater nozzle, uses cross-seam oscillation, arc voltage control and pulsed current which can be synchronized with oscillation.
Although it is possible to preprogram all the welding parameters on a programme,, the process must be monitored by an operator who can alter the process parameters as necessary to ensure a satisfactory weld.
Because visual access to the weld is blocked by the welding head, fiber optics are used to monitor the welding operation from outside the working area. The welding equipment also includes a power source and a control pendant which allows the operator to remotely
~
control all welding functions.
As a backup, manual stick electrode welding can be performed when ALARA considerations permit.
k-4.2-5
6 4.2.3 AUXILIARY TOOLIbG A.
Tent, Hoist, Suction and Filter System A platform enclosure tent will be secured on the working area. A s
hoist capable of lifting any tool or component during installation is also provided.
A suction and filter system will be used to clean up both the thermal sleeve and plenum area.
B.
Measuring and Inspecting Equipment t
A special digital readout measuring tool has been designed for support plate distance gaugingc--Special self-centering gauges are used for internal bore measuring.
t C.
Communication and TV Systems All operations will be remotely monitored on a TV and audio system once a particular machine has been set up.
Visual inspections will be assisced by closed circuit television.
4.2.4 MOCKUPS AND TRAINING Full scale mocku's of the steam generator feedwater nozzle and preheater p
i section adjacent to the feedwater nozzle have been constructed.
Th;, are used for the following purposes:
4.2-6
1.
Develop and verify detailed welding procedures, parameters sheets and sequencing information.
- 2.
Check out the tooling prior to field implementation.
' e Tn.
3.
Demonstrate the performance of production tooling.
4.
Train personnel for field work under restricted conditions.
The training program which is monitored by Westinghouse Quality Assurance pers'annel, includes both classroom and hands-on training. Trainees must demonstrate physical ability to do the job and pass a written examina-tion.
l l
J 4.2-7
4.3 WELDING AND NDE REQUIREMENTS 4.3.1 WELDING AND NDE CODES All in:tallation field welds are similar metal welds performed using the l
ASME Boiler and Pressure Vessel Code,Section III, Subsection NG and Section IX as guidds.
In addition, all welding procedures and welders are qualified in accord-ance with Section IX plus additional requircents invoked by the desigrer.
4.3.2 WELDING PROCESS SELECTION A remote controlled automatic gas tungsten arc welding system (GTAW) has been designed and qualified for production welding with the operator utilizing fiber optics and/or closed circuit TV to view the weld area.
Shielded metal arc welding may be performed as an alternative when ALARA 5
considerations permit.
4.3.3 NDE REQUIREMENTS All NDE will be performed in accordance with requirements and acceptance standards contained in ASME Soiler and P nssure Vessel Code,Section III, Subsection NG-5000 for the NDE method used.
Inconel to inconel field i
welds are examined by the liquid penetrant method.
Carbon steel to car-I bon steel field welds are examined by the magnetic particle or liquid penetrant process.
1 4.3-1 U:
4.3.4 WELD CONFIGUAATION Details of weld configuration and location are' depicted in Figure 5.7-1 of the Westinghouse report. Associated NOE. techniques are also' depicted-
'in this figure.
r b
i I
l l
4.3-2
i 4.4 RADIOLOGICAL CONSIDERATION The design modification to the preheat section of the Westinghouse D2/D3 steam generators presents radiological considerations that are in general less.lfaiting than those normally associated with steam _ generator work.
.The primary reason is that this work is restricted to the secondary side of the generators and consequently does not. involve the intense beta and gamma fields associated with primary side modifications.
In addition, the modificatien will be made on plants with limited operating history and thus limited primary and secondary side contamination.
The primary source of radiation exposure will come-from radioactive corrosion products deposited on-the inside f.e., primary side of the steam generator U-tubes.
Exposures due to contamination and inhalation are, in general, likely to be insignificant.'
Contamination on the secondary side of these generators, most of which are relatively new, is expected to be negligible. Westinghouse will, however, address in procedures the steps necessary to perform the work should local ventilation of the steam generators be required due to unexpected airborne contamination problems.
Westinghouse has used the CORA computer code to determine surface activi-ties and the KAP-VI shielding code to determine expected radiation fields in the work area around the.feedwater nozzle. Westinghouse estimates of l
1 rem /hr'at the surface of the tube bundle falling off to 0.015 rem /hr on the work platform are consistent with levels measured at Duke Power 4.4-1
'l
- Company's McGuire 1 nuclear generating unit. These levels are considered reliable for the purpose of estimating dose rates likely to be encoun-tered.during.the modification.
Based on their extensive background in making steam generator repairs, Westinghouse has prepared detailed man-hour / exposure rate studies by subtask. The resulting total dose of approximately 25 person-res per steam generator appears to be a reasonable estimate of expected dose for:
the modification. This estimate is based on actual timing data derived.
in mock-up training. Tasks that show a significant increase in estimated dose will be reviewed wit $ the intent'. of improving procedures in order to decrease the subtask-dose. The dose commitment is estir.ated to be divided such that 16 person-rem per generator will accrue due to actual internal generator modifications and 9 person-rem will accrue 'due to ancillary preparation and restoration work.
I Westinghouse estimates that post installation eddy-current and fiber l-optic inspection will contribute approximately 9-person-res per generator per inspection. One additional eddy current examination after several 6 months operation is anticipated. Because of the high dose commitments associated with eddy current testing, it is important to avoid an accelerated t
testing program, for the purpose of increasing a data base, unless nuclear safety directly mandates such testing.
(
'The quantities of potentially contaminated waste generated in this modification are small compared to those encountered in normal cutage work. There has been an effort to exclude oil and to rely on water' base
\\
4.4-2 i
e
lubricants. The modification should not adversely impact the' normal was e handling facilities found at' reactors nor cause the need for any t
~
- increased capabilities.'
4 1
d 4.4-3 l
5.0 EVALUATION 5.1 THERMAL HYORAULIC AND MODEL TESTING 5.
1.1 DESCRIPTION
OF TEST MODELS In the process of this study, several test arrangements were used by.
Westinghou,se, These included water table and air flow models for initial design concept verification, followed by 1/6, 5/12, and 2/3 scale models in Westinghouse's laboratories and a full scale model at the Swedish State Power Board facility in Alvkarleby, Sweden, a.
' The 1/6 scale model was used foY flow analysis (flow velocities 'and fluctuating flow velocities) using the method of Laser-Deppler-Anemometry (LDA).
b.
The 5/12 scale (0.417 scale) model is a linearly scaled version of.
all the prototype dimensions, including tuoe-to support baffle clearances. This model is capable of simulating a single ficv pass at the inlet as well as the actual multipass condition.
It was used primarily for tube. vibration response studies. The tubes were instrumented internally with strain gauges and accelerometers.
Various tube support conditions were achieved by undercutting the tested tubes at the support plates. An undercut of up to 14 mils diametral clearance was used.
5.1-1
From the scaling ~1aws, Lthe model _ response from turbulent excitation ij.
will' equal the prototype values at flow velocities of about 80 percent prototype velocities. The fluidelastic critical' velocities are sin 11arly predicted at 80 percent of prototype velocities, l
assuming 1.he same damping. The ratio of model and prototype tube frequencies-is 2.2.
The model has been used extensively-for evaluation of several types-of manifolds, for tube response studies with single pass and multi -
pass flow conditions, simulation of worn tubes and for demonstrations of fluidelastic stability.
The tube response studies indicate a substantial reduction of vibration levels from the baseline case to the manifold.
The multipass versus single pass responses were virtually identical for the tube support configurati'ons tested.
Westinghouse' reported no fluidelastic vibration _at any_of.the " normal" support conditions. However, the tubes did respond to fluidelastic type vibration in both the baseline and the manifold cases when the l
tubes were undercut to an extreme full scale diametral gap.
c.
The 2/3 scale model simulated the feedwater inlet region between I
plates 5 and 6: This model is used for studies of flov velocities 5.1-2
-.. - -... - ~,
s
1 l
l and fluid forces actirg upon the tubes.
Special instrumentation consisting of velocity probes.and load sensing cells were used for flow velocity measurements and measurement of static and. dynamic forces acting upon the tubes.
Pressure losses in the baseline case and the internal manifold design were also studied in this model.
The model operates at a pressure of 55 to 60 lb./in.2 with a flow rate of 4239 gal./ min. simulating the 100 percent rated velocity condition with the water temperature at 100*F. Detailed velocity.
saps and tube loading spectra used in the vibration and wear evalua-tions were obtained from this model, d.
The SSPB full scale model duplicated the inlet pass of the model 03 preheater and included a set of-full length tubes and supports representing the full straight length of tubes.. Upper and lower flow passes were.not simulated; and therefore, the model was re-stricted to single pass flow effects. Adjustment to support plates was possible to simulate steam generator operating conditions and achieve replication of tube vibratica during operation based on comparison with field measurements.
Tube vibration response, tube-to-support baffle impact forces and flow velocity measurements were taken in this model at different flow levels, including an overload conditicn of 102 parcent of full flow.
The scaling laws indicate that the tube frequency of the model is 96 percent of the prototype, and equal vibration displacements will be l
5.1-3 l
- achieved at 95 to 99 percent of the protctype velocity.
Ninety three percent of the prototype fluid elastic critical velocity -
should produce fluidelastic instability in the redel, assuming equal
. dailping..
'In the baseline configuration, tube support plate adjustments were made until a certain number of tubes responded in a similar way as in the operating Ringhais 3 and Almaraz 1 steam generators. With
- the same support conditions, two manifold designs, the prototype and
' the production manifold were tested.
4 A sensitivity study was' performed to simulate a range of tube support positions, moving support-plates 3 and 8 from the initial:
position in conjunction with repositioning of the impact gauges at support plate 5.
(Support plates 5 and 6 have'a fixed position-and are not moveable.) Fifteen different cases were tested, yielding a 'otal of 50 tube-support configurations.
t The flow velocity measurements were in good agreement with the results obtained from the 2/3 scale model. Tube responses at the equivalent full flow (100 percent) with the manifold were about equal to the tube responses-at the 40 percent flow condition in the l
baseline case. The wide variety of model testing gave a good insight into the inlet flow characteristics and the structural
{
behavior'of the tube bundle.
.j 5.1-4 l
l L
1 5.1. 2 -
FLOW VELOCITIES AND 2LUID FORCES AT TUBE BUNOLE INLET
~
The baseline design with-the impingement plate produced an uneven feedwater flow at the entrance of the tubs bunale. Uneven flow distribution occurred in the inlet pass (between support olates 5 and 6).
The flow distribution was uneven also in the sense that reverse flow with negative velocities occurred in areas around the impingement plate. The local velccities coached a maximum positive value behind the impingement plate and a maximum negative velocity at the edge of the ispirgement plate.
In addition to the mean flow, pulsating (turbulent) flow occurred, having a significant fltctuating component-at the tube bundle entrance.
The internal manifold flow distributor significantly changed the distribu-tion and the character of the incoming feedwater flow. A more uniform distribution'was achieved, both in the vertical :.nd in the side-to-side directions, with all the flow being positive (inward into the tube bundle). Pressure pulsations or velocity fluctuations were also signifi-
~
cantly reduced. The average velocity varied across the inlet to the tube bundle.
f 1
5.1-5
l 5
Individual tubes with the manifold design'could experience higher velocities averaged over the inlet span. The fluctuating velocity component at the tube bundle entrance has been reduced.
A redistribution of the forces occurred with the manifold. 'In the center of the' tube bundle, in front of'the manifold, a more_ uniform flow distribution was achieved. On both sides of the tube bundle, in the region ~outside of the manifold, a reduction in the forces resulted.
c The tube-to-support plate -fepact forces have been substantially reduced.
o A reduction of thess forces up to a ratio of about 20 to 1 occurs from the baseline to the manifold case.
5.1.3 SYSTEM RELATED THERMAL HYDRAULICS The manifold is essentially designed for tne same generic steady state and transient thermal hydraulic conditions as the steam generators. The manifold design conditions are the same as steam gene ator secondary side design conditions. The steady. state design conditions for the 02/03 steam generator are a maximum secondary side pressure and temperature of 1185 psig and 600*F, respectively. Nominal full power temperature,:
pressure and feedwater flow are.440*F, 985 psig, and 4.1 x 108 pounds / hour, respectively. For transient conditions, specific NSSS design transients for particular 02/D3 plants are generally bounded by the generic NSSS (Model 0) design transients when that plant possesses a Westinghouse feedwater control system, the plant satisfies the Westir.ghouse steam generator interface criteria, and the plant possesses either a feedwater l
5.1-6
bypass system or a temperature pegging system. The Duke, South Carolina Electric and Gas Company, and TVA plants meet these Westinghouse require-ments.
The design transients which Westinghouse considered are grouped into categories of slow thermal and flow transients, and relatively fast pressure pulse (water-hammer) transients. Of the various thermal tran-sients investigatcd, eight were selected by Westinghouse as bounding for the manifold desiga and used as input to the stress analysis. These transients are noted with an asterisk in Table 8.1-1 in the Westinghouse report. The bounding thermal transients for the different types of events and operating conditions considered are reactor trip with cooldown and safety injection, large stap decrease (95 percent step), plant loading (5 percent per minute), two banks of feedwater heaters taken out of service, excessive feedwater flow due to inadvertent opening of a feedwater control valve, plant unloading (5 percent per minimum), exces-sive bypass feedwater flow, and forward flushing transient.
The waterhammer pressure transients which were considered by Westinghcuse are due to unit loading switchover (from bypass to main feedwater nozzles),
unit unloading switchover, feedwater isolation from 100 percent flow, excessive feedwater flow, check valve closure during an upset condition, feedline break-check valve closure, and bubble collapse waterhammer (rapid condensation of steam bthbles when mixed with cold water).
Except for the bubble collapse transient, the severity of the loads associated with the above pressure transients according to Westirighouse, is a function of the initial flow, flow overshoot and the rate of change of 5.1-7
flow caused by feedwater valves opening or closing. The most-severe loads due to pressu"e transients are caused by the fecdline break-check valve slam event, which is a faulted condition. This feedline break ~
check valve slam transient bounds all of the pressure transients, includ-I g the main steam line break event.
Specific-loads and additional, details concerning these transients'are provided in Section 8.0 of the Westinghouse Report. The Design Review Panel concurs with Westinghause that these steady state and transient conditions are bounding and conser-vative, and thus, are acceptable to use for input to the stress analysis.
~
5.1.4 TUBE VIBRATION TESTING The 0.417 scale model and the SSPB full scale model were used for tube vibration testing.
. i Vibratory acceleration-and strain gauge data were collected on tape and utilizing the Nicolet 660 analyzer, RMS acceleration and power spectral density (PSD) plots.were produced. A.nethod of criculating inferred tube
~
displacement from 0.417 model strain gauge vibration data was used. -A double integration routine was used for calculating displacement from acceleration data, again using the Nicolet Analyzer.
The response spectra were obtained in two perpendicular directions and at several locations along the tube axis of the instrumented tubes.
Tube support conditions were varied to obtain similar vibration response i
spectra to those measured in the operating steam generators for the baseline case.
These were compared with the spectra obtained for the 5.1-8
t internal manifold-cases. The frequency content of the viaration spectra j
played a particularly important role in these comparisons..In'tha baseline cases, the typical frequency content of the maximus' responses.-
2 were in high and low frequency ranges. Wh'le the-lower frequency range -
' signifies a " loosely" held tube with only a few firm supports, the high frequency response represents the tube " fully" supported at all support baffles.
The tests with the manifold show a number of tube responses still having the low frequency content. Most of them,7however, show the " fully" supported tube condition with the high frequency content. The vibration levels, however, are generally much lower.
Low frequency response of 25 Hz and lower were obtainec in the SSPB tests with the manifold but these.
were said to be due to pump pulsation and/or associated with the structural-modes of the steel supporting structure. A review of the vibration data for the cases with the manifold would indicate two types of. tube motion.
1 p
5.1. 5 TUBE VIBRATION EXCITATION-MECHANISMS i
The dominant vibration excitation mechanism in the original D2/03 is consi,dered to be turbulence or turbulent t,uffetina. This conclusion was.
reacne,a on the basis of studying the response spectra of the measurements taken at the operating plants and also by matching the wear pattern of tunes pulled from operating steam generators with the predicted vibratory motion of a nonlinear tube model excited by turbulence excitations. This-was done as follows:
from the 2/3 scale ~model tests, an equivalent forcing function acting upon the tube clong its length between plates 5-l~
5.1 -...
i l
and 6 in the inlet pass was established. Using a nonliaear tube model (see Subsection 5.1.7), tube motions were predicted at several-points along the tubes. Since these motions were consistenc with the wear patterns of pulled tubes,' conclusions were drawn as to the type of excitation the tubes receive.
The majority of the data from inservice experience is well described on-the basis of turbulant excitation with reasonably.well-bounded tube wear rates. However, some of.the data.with large wear rates outside of the
" turbulence" range are considered to be possibly caused by fluidelastic e
instabilities. Thus'a second mechanism, affecting only some tubes in the baseline case, seems to have played a role.
A review of the test data for the cases with the internal manifold does not indicate the existence of the fluidelastic mechanism, et least not within the normal range of plant operating conditions which'have been tasted in the 0.417 scale and the SSPB full scale models.
In the absence of the fluidelastic mechanism, the tube responses can be determined on the basis of turbulent excitation.
5.1.6 INDEPENDENT FLOW AND VIBRATION ANALYSIS i
In parallel with Westinghousa's testing and design efforts, the thermal hydraulics-and vibration group of the Design Review Panel proceeded with a theoretical flow and vibration analysis of the steam generator pre-heater section.
The purpose of this analysis was to gai1 a deeper and
-independent insight into the problem.
The idea was to investigate the 5.1-10
.. ~
entire preheat region and determine the flow velocity distribution and the vibration characteristics of the tubes.
A three-dimensional (finite difference) computer flow analysis has been performed and flow velocities throughout the tube bundle have been datermined. Using representative flow velocity distributions for tubes in the inlet area and around the flow restrictors, a parametric study of a large number of combinations of support conditions with one, two, and three missing supports has been performed, using a linear multispan' tube vibration computer model. The analyses provided the following results:
a.
The tubes in the first two rows of the preheater inlet are most susceptible to vibration, b.
There are no other' critical areas inside the tube bundle that are susceptible to damaging vibration.
c.
The external passes will experience a maximum cross-flow velocity
- much lower than that~of the inlet pass.
d.
The " fully" supported tubes (at each support baffle) demonstrate'a high natural frequency.
)
'e.
Tubes with missing supports have lower frequencies.
5.1-11
f.
The study also predicted that drastic reduction of natural frequencies will bring the tubes into the range of fluidelastic instability at or below the full load condition.
g.
There is an effect of the external cross passes upon the fluidelastic instability of the tubes. The reduction of the fluidelastic critical velocities due to the flow through the external cross passes (passes in addition to the inlet pass) is typically in the 20 to 30 percent range, depending upon the actual tube support conditions.
t 5.1.7 COMPUTER TUBE VIBRATION MODEL A three-dimensional model of a singri tube from the tubesheet up to the U-bend has been developed with the Westinghouse WECAN computer program.
This model is multispan and utilizes annular gap elements at the support plate locations simulating the tube-to-support clearances and permitting orbital tube motion. Support plate offsets can also be modeled.
The model is loaded at discrete points within the' inlet span between support plates 5 and 6 by a force time-history excitation derived from test measurements on the 2/3 scale model.
Theoutputisinthefobofdisplacementtime-historiesfromwhich 4
response spectra can be generated for analysis of frequency compositions.
The generated displacement time-histories and contact forces serve as a f
basis for calculating the integrated tube-to plate contact forces times 5.1-12 l-i i
+ - ~ -, - - -
travel distance per unit time to determine wear rates used in tube wear evaluations.
5.1. 3 TU8E WEAR PREDICTION METH005 The principal design objective of the preheater inlet modification is to limit tube wear.
Westinghouse's method for predicting tube wear consists of establishing a limiting tube wear volume based on the minimum allowable tube wall thickness for tubes in service and comparing it with an estimated wear volume. Westinghouse has developed an analytical model to estimate the wear for any specific tube and tube support plate location in 'the C2/03 preheater for a given period of time based on a maximum bounded wear rate for that tube and support plate location.
The limiting tube wear volume for a single tube wear scar selected as a basis for evaluation of the modification performance corresponds to the 40 percent tube wall reduction limit (see Figure 4.2.3.1A and Section 7.2.1 of the Westinghouse Report) which, according to Westinghouse, is a conservative limit of the structural capability of the tube and bounds the allowable wall reductions permitted by the utilities' technical specifications. The tube-wear, volume-wear depth relat.ionship is based on data obtained from seven steam generator tubes (each containing multiple wear scar sites) which were removed from four different steam generators at Ringhals 3 and Almaraz 1.
Field eddy current test results were also reviewed, correlated, and compared with the above data. Westinghouse's 5.1-13
i approach with regard to limiting tube wear volumes and tube wall reductions is reasonable and acceptable.
With the single tube scar wear volume limit established, an analytical wear model was developed by Westinghouse to estimate wear volume for specific tube and support locations. Westinghouse's wear equ3 tion in which the single tube scar wear volume is the product of tube wear coefficient (K), the tube wear distribution factor (D), the equivalent
[
full power operating period (T), and the impact-sliding work rate for the total tube (WR), is based on an Archard's wear equation, wnere wear volume is the product of the wear coefficient, force, and distance.
In order to use the Archard equation, two modifications Dy Westinghouse are required. First, Westinghouse deve hped a relationship for work rate performed at a particular tube-support location. The work rate (WR) for the entire tube is then obtained by summing of the individual work rates for each tube-support location. The total work (W) performed on an entire tube for an operating time (T) is determined by taking the product of the tntal work rate (WR) and operating time (T). The total wear volume (VT) for an entire tube is the product of total work (W) and the tube wear coefficient (K).
Second, in order for Westinghouse to obtain the total maximum wear at a particular tube-support location, Westinghouse developed a relationship based on field eddy current and pulled tube data from the Ringhals 3 and Almaraz 1 units.
(Shown in Figure 7.2.13 of the Westinghouse Heport.)
l In order to incorporate this information into the Westinghouse wear equations, a tube wear distribution factor (D) was conceived and defined.
l 5.1-14 l
This calculated wear volume is compared to the limiting single scar wear volume to determine whether or not the tube may require plugging after a specified equivalent full power operating period (T). These types of comparisons have been made by Westinghouse and an example of one is shown in Figure 7.2.15 of the Westinghouse Report.
The wear equations and basic wear method concept-developed by Westing-house are, in our opinion, well-conceived, reasonable, and an acceptable approach and method for modeling and bounding tube wear. A description of the specific inputs used in the Westinghouse wear equations and the adequacy of these inputs is discussed below.
The tube wear coefficient (K) is an empirical term used originally with
-12 Archard's wear equations, and having units of in.3/lb.-in. x 10 Depending upon the type of tube motics at the tube-support interface, one of two coefficients (K) are used by Westinghouse. These wear coefficient values are derived from Westinghouse and historical wear test data.
Comparison of one of the coefficients with coefficients conservatively calculated using initial tube-to+ plate gaps and using field data is favorable (See example in Section 6.6.3 of the Westinghouse Design Report A). Westinghouse also has stated that these coefficients determined from field data are conservative, since the experimental value of K is greater than wear test data from tests that more closely duplicated the manifold type tube motion. Westinghouse also states convincingly for a second type of tube motion, the wear coefficient chosen is conservative.
We concur that using the specified wear coefficients is app'ropriate.
5.1-15
The tube wear distribution factor (D) is also an empirical tarm which was
. derived and developed by Westinghouse so the maximum single scar wear j
volume can be determined when total tube scar we n volume is known.
The tube wear distribution factor (D) is defined above, and as previously mentioned, is based on the upperbound curve of field tube pull and ECT data as shown in Figure 7.2.13.
In Section 7.2.3.2 of the Westinghouse Report, Westinghouse indicates that the upperbound curve encompassed the dominant trend of the data; however, a number of data points including some of which Westinghouse felt represented wear due to fluid elastic excitation at high power levels (>75%), were above this upperbound curve.
These data points, according to Westinghouse, were excluded because they did not represent the general tube population; wear patterns were incon-sistent with the general data bank;1'luidelastic excitation is not a plausible excitation mechanism in a modified steam generator; and abnormal tube-support plate geometries, gaps, and other "as-built" conditions may l
have been present. We agree with Westinghouse's basic statement that the upperbound curve bounds most of the relevant data or encompassed the dominant trend of the data. However, there is an element,of uncertaini~ty l
when using this Westinghouse upperbound curve. An absolute upperbound curve, which bounds all the data (and worn tubes) shown in Figure 7.2.13, 1
regardless of the cause of wear or tube-support configurations, would minimize the uncertainty to a more acceptable level. This absolute upperbound curve would extend from the existing Westinghouse curve in Figure 7.2.13 of the Westinghouse Report through the worn tube data point i
at approximately 140 x 10 and 240 x 10'4 on the maximum single scar
-4 wear volume and total scar wear volume scales, respectively. This absolute upperbound curve adequately bounds the data. The use of the 5.1-16
i tube wear distribution factor (D) when based on the above absolute upperbound curve is adequate. The projected life of a tube (worn either coroletely through the wall or 40 percent of the wall) when using this absolute upperbound curve as compared to the Westinghouse curve is reduced by approximately 36 percent.
-The equivalent full power operating period (T) is used to input length of operating service into the wear model. This concept has been previously accepted by NRC for use on the McGuire 1 unit. This parameter is conserva-tively estimated since total work rate and wear rate increase with respect to power exponentially.
(See Figure 3.2.4.9 of the Westinghouse Report).
The work rates for the' total tube (WR) are calculated from the Westinghouse nonlinear flow induced vibration (FIV) model, using the WECAN computer program. Four force histories (corresponding to the four test force gauge measurements) are applied simultaneously (in the two support plate plane directions) as input to the FIV model for a real time duration of 4
approximately 0.3 seconds. The FIV model then provides, as output,- time histories of tube motion, linear response spectra, tube contact forces, l
workrate, and other miscellaneous information.
The FIV model output was verified by Westinghouse in basically three different steps.
First, Westinghouse was able to match the analytical model tube motion with wear scar locations and characteristics obtained from pulled tubes from Ringhals 3.
Second, the FIV model results for displacement and accelerometer spectra, when compared with field data l
5.1-17 l
L
_u..
a from an instrumented tube at Almaraz 1 compared closely and favorably.
Third, the FIV model was able to produce the pattern of wear, and approx-inately the same work rates as determined from data in the field for selected tubes throughout the bundle.
A comparison of the work rates predicted by the analytical model and those obtained from field data is shown in Figure 7.2.9 in the Westinghouse report. Most of the data is bounded by the turbulent forcing functions used in the analytical model. For those work rates which are not bounded, fluidelastic excitation and, perhaps, physical steam generator abnormalities are the explanation provided by Westinghouse for their deviation.
Westinghouse has performed some limited investigations on the effects of a fluidelastic forcing function on the base case and the manifold design.
Since Westinghouse did not believe that fluidelastic forcing functions could not be accurately calculated, they approximated the forcing function and selected three different fluidelastic forcing function magnitudes for input into the analytical model for the base case.
For the manifold design, mathematical studies showed that fluidelastic forcing functions would be reduced by at least a factor of five and the maximum work rate expected would be less than the maximum work rate j
expected for turbulent induced conditions. Although the absolute values presented by Westinghouse may be only approximate, we concur with Westing-house that the maximum turbulent induced work rate for manifold design is adequate to bound the wear for fluidelastic mechanisms.
I 5.1-18
Westinghouse hs performed several sensitivity studies with regard to tube-to-ttbe support clearances, initial-physical configurations and geometries, gap sizes, and other similar items adequately.
The Westinghouse model and analysis, because of its good correlation with field data and SSP 8 test data, was applied to the manifold design. The results of this analysis for seven selected tubes with reference support conditions are shown in Table 7.1.2.4-2 of the Westinghouse Report. A comparison of work rates as a function of position in the bundle for the manifold and the base cases from both the analytical model and full-scale tests are shown in Figure 7.2.11 of the Westinghouse Report.
For the manifold, both the analytical model and the full scale tests indicated that much of the tube motion had been-changed. Also, work rates for the manifold for a given bundle location in Row 49 are at least one order of magnitude less, and the maximum work rate is two orders of magnitudes less than for the' base case.
Similarly, the maxieum wear rate for the manifold design is also two orders of magnitude less than the base case.
We concur with Westinghouse that the manifold design has significantly reduced the impact forces, the work rates, and the wear rates for the affected tubes in the Westinghouse 02/03 preheat steam generators. We also note that because of the urgency in which the Westinghouse modifica-tion program was implemented that not every lead or question could be pursued to a complete and absolute conclusion or answered with absolute certainty. There are some elements of uncertainties in the Westinghouse 5.1-19 j
program; however, we beleive that these uncertainties have been minimized to an acceptable level by being conservative in approach, by utilizing I
l other aspects of the program to compensate, or by other means. Those major uncertainties, which have not been discussed above, are discus;ed below.
In order to conservatively and safely predict when the first steam gener.
ator tube will require plugging at 40 percent tube wall reduction, addi-tional conservatism has been added to the Westinghouse wear model tu ac-count for uncertainties. As noted above, the tube wear distribution fac-tor (D) was adjusted so that the projected life of a tube (to 40 percent wall reduction) was decreased by approximately 36 percent. The tube wear l
scar volume relationship to tube wat1 thicimess reduction was also reviewed.
The uncertainty associated with this relationship can be further minimind by reducing the wear scar valume acceptance limit by 15%. This additional conservatism would further reduce the projected life of the tube another 21 percent for a total life reduction of 57% when compared to Westing-house's'esthods.
Since we have accepted Westinghouse's basic approach.
and methodology as being valid,' the only other term that has any uncer-tainty associated with it is the work rate.
Using the above conservative inputs and Westinghouse's maximum calculated work rate, the estimated projected life of the tube subjected to the greatest wear is specified for a 40 percent tube wall reduction.
If a two year inspection interval for all tubes of interest is selected and the above' conservative inputs are used again, the wear rate to obtain 40 percent wall reduction is Y
0.47 inch-lb./sec., and to obtain 65 percent wall reduction is 0.66 inch-lb./sec. These wear rates are 2.66 greater than those used by 5.1-20
WesH,ghouse.
For a 1.5 year inspection interval, the results are 0.62 inch-lb./sec. and 0.88 inch-lb./sec., respectively, with the conservatism factor increased to 3.55. For a one year inspection, the results are 0.94 inch-lb./sec. and 1.32 inch-Ib./sec., respectively, with the conservs-tism factor increased to 5.329. For six months, these numbers are 1.875, 2.642, and 10.65, respectively. As shown above, we can introduce as much conservatism as needed by shortening the inspection interval.. Note that for the six month period, the wear rate required to cause a 40 percent wall reduction is approaching the wear rate observed for the 100 percent power base case in the field. Clearly, these magnit'Jdes of wear rates are hign since the manifold wear rates will be similar to 40 percent base case wear rates. The results of this evaluation show that Westinghouse Panifold design is safe and that tube wear will be predictable enough to allow for normal inspection intervals and plugging ~of tubes if necessary.
For Model 0 steam generator tubing, 65 percent wall reduction - the safety limit - is the maximum amount of wall loss the tube can sustain and maintain integrity under the most severe accident conditions as determined by analysis and test.
Forty percent wall reduction is the limit for some plants (others use 50 percent) for determining when steam generator tubes must be plugged.
This limit takes into consideration an eddy current uncertainty of 10 percent and allows an additional 15 percent (5 percent for 50 percent plugging limit) wall reduction for wear during the next inspection period. For worn tubes, a 24 month maximum inspection period is required 5.1-21 j
i by the NRC.
If wear rates are high enough such that a 24 month inspection interval is inadequate, a shorter inspection can be used. Westinghouse is proposing that an inservice inspection program be established for modified plants which includes eddy current testing of the tubes expected to have the maximum response after an initial operating interval of six to twelve months. We feel this Westinghouse program is very conservative. Since Duke's McGuire 1 and other plants have been operating for six months or more at 40 percent or greater levels with no significant tube wear, and it has shown that the manifold design for 100 percent power levels has reduced the fluctuating forces to approximately the same level or less than the base case forces at 40 percent power level, we are confident-that the manifold design will be safe to operate.
Inspection intervals for later units and for future inspections of operating units will be based on plant operating wear data as it is accumulated.
As mentioned above, the inspection intervals will be set so that tube wall reduction safety limits are not exceeded.
5.1. 9 CONCLUSIONS Analytical and test results show that the modification of the flow inlet conditions by the new reverse flow limiter and internal manifold flow distributor ~ill substantially reduce the adverse conditions of high I
w local flow velocities, reverse flows and high fluctuating flows. As a result of this modification, the turbulence induced forces will be i
significantly reduced and a more uniform drag force distribution will be achieved.
5.1-22 l
I
=.
The new flow velocity distribution in the inlet region results, however, in increu ed average flow velocities, in the middle portion..The average flow velocities on the sides of the inlet region can be expected to be reeuced by about one-third.- The forces introduced by the higher mean i
flow velocities will have a beneficial effect of trying to push tiie tubes against the supports,:but at the same time the nigher velocities will make the tubes more. susceptible to vibration if. firm supports are not achieved due to the clearances and support plate positions. On the sides of the inlet region, the positive effect of the reduced crossflow velocities is effset by a reduction in the force.
A substantial reduction (generally three to five-fold) in vibration levels is predicted.
In conjunction with this, a substantial reduction in the tube-to-support plate impact forces-(up to about 20 to 1) is also predicted..These reductions are due primarily to less violent, more unifow flow conditions. The low frequency tube vibration component so typically present in the baseline case still appears in many cases with i
the manifold, but the work (and wear) rates associated with these tubes have been significantly reduced in magnitude.
l A wide variety of tests in the full scale and the 0.417 scale arrange-ments have been performed with the main objectives to determine the effect of varyina tube sucoort conditions on tube response and fluid-elastic vibration and the effect of the external cross pass flow upon tube vibratory responses. The main results obtained were as follows:
(1) the effect of the external passes is negligible; (2) no fluidelastic vibration occurred under " credible" tube support conditions simulating 5.1-23
the "as-built" and operating conditions; and (3) no excessive vibration resulted, although tubes acted as unsupported at some support plates, as indicated by the presence of low frequency components (40 to 50 Hz and lower).
Under extreme adverse support conditions where the tubes were supported at plates 5 and 8 only, with no contact possible at other support points, fluidelastic vibration has been demonstrated both in the baseline case l
(with impingement plate) and with the manifold. The resulting-fluid-elastic vibration levels with the manifold were shown to be within the bounding limits of those obtained from turbulent excitation.
Theoretical analysis using linear models predicted the possibility of fluidelastic vibration in the tubes at the inlet with the manifold design.
It was this analysis which led to the extensive testing of tLbe support configurations, searching for the fluidelastic vibration. No sucfi vibration has been demonstrated to occur within the expected range of support conditions. Additional effects, not incorporated in the linear model, such as friction, etc., have apparently played a role in preventing the fluidelastic vibration from occurring.
Considerable emphasis is being placed on the nonlinear single tube vibration model used for the prediction of vibration levels, types of vibrating motions and consequent calculations of tube wear rates and tube life predictions. The wear rates predicted with this model showed good-correlation with actual tube wear experienced in service.
5.1-24
]
6
. The-wear prediction model conservatively bocids the majority of the in-service wear scars. The model has a built-in feature that predicts a i
decreasing rate of progression of wear with time.
Back calculations using wear scars of in-service' tubes have proven the validity of the model for the correlated tubes.
I
- The results obtained by the manifold modification, as outlined above are
- all favorable. Following is a list of aspects which could conceivably have a not negative effect and thus must lead us to accept the results
.obtained with some caution.
1.
The-effect of the external crosspasses could not have been fully tested (fully compensated for) in the full scale model. The drag force and the dynamic force effect in these passes may. conceivably influence the tube vibratory behavior resulting in greater vibration levels.
2.
There is a considerable increase in average crossflow velocities in the center portion of the tube bundle with the manifold.
If actual i
support conditions were to be outside of the range of tested support conditions, fluidelastic vibration could result.
3.
The drag forces on the sides of the tube bundle will be lower, possibly leading to some tubes having fewer effective support points and higher vibration levels.
5.1-25
4.
The vibration results of the 0.417 scale model should te taken with caution since.no direct scaling of the tube-to-support baffle clearance effect is possible.
5.
Despite the good predictions in a number of cases, the nonlinear.
model needs improvements and further validation. Among the most obvious ones would be the introduction of friction forces at the supports, velocity related damping and a force input distributed
-along the entire tube length (all crossflow effects).
6.
The wear prediction model, especially the feature of the decreasing rate of progression of wear with time needs further. validation.
7.
The primary causes of this vibration /weer problem have been identi-fied as inlet flow turbulence and support plate clearance. The modification addresses only improvements in the inlet flow condi-tions and not in the tube support clearance conditions.
Based on the above, it is concluded that tube wear rate cannot be pre-cisely predicted for the actual steam generators. However, this uncer-tainty may be adequately addressed by providing for periodi: inspections and analysis of tube wear, thus precluding tube leaks.
Such a program is recommended and has been required by Westinghouse.
The proposed modification by the new flow limiter and internal manifold design will greatly improve the vibration-associated problems.
It will substantially reduce the vibration levels, the tube-to-support plate 5.1-26
impact forces and the corresponding tube wear rates and thus markedly increase one tune life spans. The vibration-associated wear, however, will not be fully eliminated. Tubes will still vibrate at mucn lower levels and some may experience wear. This could lead to need for come tubes to be taken out of service within the predicted life-span of the steam generator.
M9 wever, since the predicted wear rates of the modified design are acceptably low and inservice inspections will be carried out, the prob-ability of primary to secondary side leaks will be virtually eliminated.
Thus, there is no safety concern with this modification. Therefore, this modification meets the acceptance criteria and is acceptable.
N' 5.1-27
5.0 STRUCTURAL MECHANICS The preheater modification does not. form any part'of tne pressure boundary for the primary or secondary system; however, it is subjected to various' mechanical and thermal loadings.
Extensive calculations have been:
provided by Westinghouse to show that the modification parts'.can withstand all applicable loads without detrimental effect on pressure boundary components. The ORP reviewed these calculations and-the summary report in sufficient detail to conclude that the structural integrity had been adequately addressed. The following section summarizes the results of this review.
5.2.1 MODIFICATION STRUCTURAL CONSIDERATIONS 5.2.1.1 Stress Criteria ASME Boiler and Pressure Vessel Code Subsection NB and NG and Appendix F were used as guides in determining the stress criteria for the manifold components.
In addition, plastic-dynamic analysis and crack propagation analysis were used to supplement this criteria.
I i
5.2.1.2 Design Loads i
The loading conditions for which structural analysis was performed are described in Section 9.0 of the Westinghouse Report. The types of loading conditions included in the analysis are thermal transients, 5.2-1 l-l-
d 1
A thsreal stratification, thermal striping, steady-state pressure oscilla-tions, water hammer pressure effects, seiseic loads and feedwater line
~
pipe break..The main steam 'line break's loading was determined by.
Westinghouse to be less severe than the feedwater line break.and was, therefore, not used in the modification analysis.
4
- 5. 2.1. 3 Analysis Methods i
Most of the analysis was performed using finite element methods. Different-r models were used depending on symmetry of structure and load ' cases.
It-was stated by Westinghouse that model' sizes were sufficient to cssure that all components within all load paths for a given loauing were adequately included.
Likewise, Westinghouse assured that the analysis of all critical welds included the effects of all component loading utilizing the subject weld as a load path. Where elastic analysis criteria could not be' met, an elastic / plastic evaluation' was performed.
The thermal striping analysis' consisted of calculating the through-thickness temperature distribution and the surface' stresses on a 1" thick plate and 5
r a 0.2" thick plate for the water temperature fluctuations given by thermal hydraulic analysis. These were considered to be representative and conservative generic cases for the material thickness in the modifica-tion componerts.
It was reported that this produced ar, alternating stress which is below the endurance limit stress of 12,600 psi for Inconel 600. This was considered as an additional stress to be combined with other stresses resulting from the thermal stratification cases.
In general, the i
5.2-2
L thermal striping stress did not occur at the location of maximum stress due tol stratification and considering the low magnitude of the striping stress, S
.it resulted in an insignificant contribution to fatigue usage.
The analysis for steady state pressure fluctuation resulted in the-development of curves of allowable peak-to peak pressure oscillations 4
versus frequency. These curves were developed for critical modification components most subject to this loading. These components are the manifold structure, the manifold bolts, the flow splitter vancs, the -
entrance plate, eNit plate, internal manifold attachment weld, flow limiter attachment weld and flow splitter attachment weld. These curves are based on limiting the oscillating pressure stresses.at any frequency to the endurance limit for the material.
.At the request of the DRP, Westinghouse developed a technique to determine the effect of spectral combinations to get probable peak values for steady. state pressure fluctuation. Based on SSPB test data, and calculated transfer functions, Westinghouse has shown that these peak stresses are also below the endurance limit.
In addition, they have been included in-the accumulated fatigue usage calculations. Through the evaluation of accelerometer and pressure transducer test data, Westinghouse has. demon-strated to the panel that there is essentially no. correlation between tube vibration and nozzle inlet presrure oscillations. However, there is some correlation between inlet pressure oscillations and manifold vibrations.
Inasmuch as feedwater inlet nozzle pressure oscillations would be expected 5.2-3
p I
to be dependent on plant specific feedwater system components such as pumps, i
valves and piping layout, the DRP recommends that these pressure oscillations be initially monitored in operating plants throughout the design operating-flow ranges. This data,should then be evaluated for effect on fatigue usage of_ manifold components. This testing and. evaluation is not urgent.in nature, but should be done at the earliest convenient time.
y, An assumed failure of the manifold attachment weld during a postulated i
~
feedwater pipe break event was evaluated and the result of the impingement of the manifold against the tubes was analyzed. The ORP has reviewed the detailed Westinghouse evaluation and has'also made an-independent evaluation 4 -
of this postulated event. The DRP concluded that the first row of tubes are capable of absorbing the energy of the impacting manifold without
-rupture of the tubes during this postulated scenario.
The calculations for_ natural frequency.of the manifold assembly were questioned by the DRP, and the ORP recommended that the analysis be revised by extending the analysis model to the feedwater nozzle. This J
was done by' Westinghouse.
Inasmuch as the manifold / thermal liner assembly receives support from the feedwater nozzle and the thermal wrapper, the
~ demonstrated effect of this change in the model was minimal.
Thermal stratification analyses for all components have been performed by Westinghouse. Three flow stratification cases were evaluated to investigate the. impact of the relative elevation location of the cold-to-hot interface-5.2-4
. m..
.~ ~,
on the stresses and interaction loads between mar.ifold boxes. As a result of these analyses, design modifications were made to reduce interaction bolt loads to within allowable values.
The DRP believes that referenced flow stratification produces the most severe loading likely in the normal / upset category of events. The stresses resulting from these cases have been treated in the same manner as the thermal transient cases and are included in the evaluation for code compliance.
Westinghouse was asked to provide detailed drawings of the thermal wrapper /shell interface so the support of the wrapper could be evaluated by the DRP. The drawings were provided and showed that the wrapper is integrally joined with the tube support plates and the thermal liner.
When combined with the tubes, this mass is considered to be a dynamic restraint for possible vibration of the manifold / thermal liner model.
Thus, the DRP agrees that the assumed model boundary conditions are adequately described and treated.
5.2.1.4 Manifold Assembly Analysis Results Analysis results on the reverse flow limiter model have shown that all stresses meet ASME code limits used as guides. Since maximum primary plus secondary stresses in the ligament of the reverse flow limiter and in the thermal liner to feedwater nozzle weld exceed 3 S,, further
)
evaluation by the simplified elastic plastic analysis in accordance with NB-3200 was performed by Westinghouse. This evaluation showed that the weld and flow limiter both meet the applicable criteria, including usage factors less than 1.
The finite element modeling of the weld between the 5.2-5
feedwate'r nozzle and the thermal liner was reviewed by the ORP. Manu-facturing records indicate that the modeling of the joint is in agreement with the installed configuration for all units. This will be verified by Westinghouse.with ultrasonic examination of the nozzle when accessible during the modification.
In the faulted event of feedwatcr line break / check valve slam, the structural integrity of the reverse flow limiter.to support cylinder weld was verified by assuring that the plastic' strain in the weld does not exceed the strain limit of the material.
Analysis cf the manifold assembly was used to determine fastener loads, boundary conditions for detailed models and overall manifold response to loading conditions. Analysis of several detailed models of the several box components was also performed. The results of these analyses demon-strated that code requirements are satisfied for normal and upset loads.
For the faulted case of feedline break / check valve slam, local regions of.
I the manifold box and the bolts do not meet the requirements of non-mandatory Appendix F.
The evaluation of this case is discussed in Section 9.1.6.6 i
of the Westinghouse Report.
The DRP agrees that due to the nature of the.
load and constraint provided by the structural configuration, this event 1
will not result in a failure of tne manifold that could damage the primary pressure boundary.
In addition, Westinghouse has provided a simplified dynamic elastic plastic analysis showing that the strain in
(
the inelastic regions is less than the minimum elongation for the material.
5.2-6
l
' ~.........
LO Mar.ifold flange to thermal liner weld analysis shows that code requirements for primary and secondary stresses are satisfied for normal, upset ano faulted cases. However, in a portion of the sleeve to liner weld, a fatigua usage factor grer.ter than 1.0 was calculated. Based on the evalu dion discussed in Westinghouse Raport, Section 9.1.8-10, the DPP concludes that the sleeve to liner weld will meet its intended function.
The manifold boxes are connected by bolts and tapered studs. These fasteners were evaluated for tension, shear and bending from the manifold interaction, for applied preload and for loads caused by differential expansion between the bolts and the fastened parts. These analyses considered the relative stiffness of the various parts. Threaded fastener stresses were shown to satisfy code requirements for normal and upset loads with one exception.
Analysis indicates that the bolt loading caused by the low flow cold water forward flushing event using the system design values for flow rate and water temperature as given in Section 9,0 of the Westinghouse report results in overstressing of the fasteners.
The evaluation of the fasteners for faulted loads was reviewed along with the manifold analysis.
Flow splitter and flow splitter plate analyses were done in two parts.
One model was analyzed te determine the overall behavior of the assembly
)
and stresses in the splitter vanes, support ring and associated welds. A separate analysis was used for the splitter plate to allow a more detailed analysis of the perforated plate.
It was determined that for some thermal load cases, the splitter plate or alignment pins contact the 5.2-7
= -
manifold. The resulting forces have been considered-in the analysis of.
both components.
The resulting stresses satisfy code allowables with the exception of the fatigue usage factor for the splitter to ifner attachnert weld. This weld-is-discussed in the Westinghouse Report, Section 9.2.1.3.
Based on the conservatism in the transients and analysis and the crack growth evaluation of this region, the DRP concludes that this evaluation is acceptable.
In addition to the evaluation of stresses and code allowables, the results of the analysis were used to ovaluate the effects of postulated cracks in highly stressed regions. This evaluation used the procedure of Section XI of the ASME Boiler and Pressure Vessel Code.
Initial flaws were assumed and crack growth for the plant lifetime was calculated. All crack growth calculations indicated that the modification would function satisfactorily ev a with the postulated cracks.
The effects of. residual stresses due to weld shrinkage on crack growth were reviewed and determined to be insignificant.
The critical crack size for the fasteners was determined for stresses equal to the code allowable. The critical depth for a crack completely l-around the circumference was found at expected temperatures and at a conservatively low temperature of 32*F.
l 5.2-8
l The limiting load capacity of critical welds was determined assuming a circumferential crack partially through the weld thickness.
Results showed a factor of at least 3 betwean the applied faulted loads and the calculated limit loads.
It is concluded that large initial flaws will not result in failure of the modification parts.
Westinghouse was requested to conduct further investigation into the possibility of stress corrosion cracking between the thermal liner and the close fitting components within the liner. The DRP reviewed the results and concluded that the material selection and other design considerations were adequate to avoid this type failure.
- Further, general, galvanic, stress and crevice corrosion and erosion resistance design considerations were addressed.
5.2.2 STRUCTURAL CHANGES DURING DESIGN The ORP comments and expressed concerns were taken under advisement by Westinghouse as the design progressed.
Significant changes which evolved during this period includes the following:
The manifold attachment weld was moved farther upstream into the nozzle to give more flexibility to the structure and reduce thermal loads on the weld l
1 5.2-9
l A cylinder was added to relocate the above weld. The cylinder was made binetallic to eliminate the previous binetallic field weld.
All field welds will now be mono-metallic.
The reverse flow limiter support cylinder attachment weld was moved farther downstream from the thermal liner attachment weld to minimize the effects of the shrinkage of the modification weld on the existing thermal liner attachment weld.
The' design contour of the reverse flow limiter cylinder attachment weld was revised from convex to a flat fillet to minimize structural discontinuity.
The finite element models used for the Ylow splitter analysis were revised to include the vanes.
5.
2.4 CONCLUSION
S It is the opinion of the DRP that the Westinghouse design modification is structurally adequate for all applicable loads including those loads resulting from pipe rupture events with the exception that the fastener allowable stresses could be exceeded during the design forward flushing transients.
Recommended actions to address this structural inadequacy 5.2-10 I
ei
L (identified in Section 5.2.1.4) are described in Section 6.0.
In addition, the locking devices for the threaded fasteners act also as positive capture devices for the fastener parts, thereby preventing inadvertent loose parts inside the steam generator, e
5.2-11
i 5.3 MATERIALS, WELDING, AND NDE 5.3.1 DISCUSSION Early in the design phases emphasis on material selection and material-related problems was deferred until the final design alternative to resolve vibration problems had been identified.
Subsequently, the panel had several questions related to material properties, stress corrosion
. cracking, corrosion, fatigue, heat treatment, welding procedure and sequence, residual weld stress, NOE technique, etc. The DRP expressed concerns about the interaction of weld stresses due to proximity of welds, stress corrosion cracking in confined spaces and in proposed binetallic welds as well as the adequacy of' assumptions and finite element modeling for certain welds, the details of which are discussed in Section 5.2.1 and 5.2.3 of this report. Westingnouse revised some analysis acdels, provided a detailed evaluation of some assumptions, changed the shape of a proposed weld, and incorporated a bimetallic manifold support sleeve to relocate a bimetallic weld.
The Westinghouse report. evidences the fact that the efforts in problem
~
solv'ng emanated from a'" design for welding" concept rather than a " weld the design # concept.
Fracture mechanics has been utilized, residual weld stresses have been evaluated and the impact of weld shrinkage has been assessed in combination with working stresses to determine weld adequacy.
To evaluate materials, welding, and NDE training activities, DRP members visited Westinghouse Pittsburgh design facilities as well as the Forest J
5.3-1
Hills and Advanced Reactor Division sites. Training activities and weld tooling applications at the Westinghouse Tampa facilities in Tamps, Florida were observed and found to support adequately the tasks identi -
fied in Sectio'n.4.2.4 of this report.
5.3.2 MATERIALS The materials used in the modification are inconel, stainless steel and carbon steel (Table 3.3-1) :These materials have been well proven in the steam generator environment..Th-is is true with respect to both chemistry and flow considerations. However, a concern for these materials is in areas where the inconal and carbon steel are connected through a binetal-lic weld. This aspect is separate from the basic materials issue and is discussed in Section 5.2 of the Westinghouse report.
The fasteners are a more difficult problem. Adequate strength must be provided in order to withstand transient loads while maintaining accept-able general corrosion and stress corrosion cracking resistance.
In addition, fatigue strength must be acceptable. This combination has been provided through the use of Type 410 ferritic stainless steel. Section 5.0 of the Westinghouse report deals with these considerations in greater depth.
(
5.3.3 WELDING l
Weld design, fabrication and NDE requirements were selected from ASME Section III Subsection N8, NG and Section IX and are judged to be appropriate for 5.3-2
~
this modification. All welds are to be performed in accordance with-welding procedures and by welding personnel qualified to the requirements of ASME B'iler and Pressure Vessel Code Section IX. All field w1 ding ^
o is expected to be performed using the automatic gas tungsten are welding process, controlled remotely and equipped with fiber optics' and/or' closed circuit TV for viewing the welding arc. Manual gas tungsten arc or shielded metal' arc process will be used as a backup if needed due to site unique conditions which might arise. Welding machine operators are re-quired to be ASME qualified to use the manual gas tungsten arc process in addition to the machine operator qualification requirements of Section IX.
Weld design and fabrication are to be performed to the requirements or Section III Subsection NG. The installation and welding seque.nces are described in Section 4.1 of this report. Weld tooling is, describe 2 in Section 4.2 and the training of welding personnel is described in Sec-tions 4.2 and 4.3.
't_
5.3.4 NON-DESTRUCTIVE EXAMINATION (NDE)
All non-destructive examination wi.ll be performed to the requirements and
-acceptance standards contained in ASME Boiler and Pressure Vessel Code Section III NG-5000 for the NDE method being used. Magnetic particle (MT) or liquid penetrant (PT) method will be used for ferritic materials and liquid penetrant method will be used for non-ferritic materials.
Attempts were made to perform volumetric examination of welds by radi-l ography.
It was determined that meaningful results could not be obtained 5.3-3
in all cases. ~To simulate volumetric' examination, NT and PT will be performed progressively, i.e., root' pass, intermediate pass (es) when
'needed and. final pass.
5.
3.5 CONCLUSION
Materials,. welding, and NDE are an integral part of the design process.
It is concluded that adequate weld tooling, training, and contingency methods and techniques.have been considered and are available to resolve unexpected problems.
It is: concluded further that the materials, weld-ing, and non-destructive examination aspects of implementing the proposed modifications have been addressed adequately, and are satisfactory.
~9 5.3-4
. ~.
5.' 4 TOOLING AND INSTALLATION 5.1. 4 DISCUSSION A detailed evaluation was performed of the adequacy of the preposed modification installation process and associated special tooling. This evaluation included numerous meetings with Westinghouse representatives including drawing and procedure reviews along with hands-on experience with the various tools and equipment during mock-up testing.
The evaluation criteria utilized in this review included the folicwing:
1.
Modification installation requirements 2.
Efficiency and reliability of the installation process and tooling 3.
Personnel safety 4.
Potential for inadvertent steam generator component damage 5.
Mock-up adequacy 6.
Tooling qualification 7.
Procedure adequacy 8.
Personnel training 9.
Logistic support requirements The design requirements for the preheater modification installation F
(i.e., fit-ups, alignments, and inspections) nave been addressed in the tooling design, and compliance with these requirements can be reasonably I
assured during the installation process.
5.4-1
i Each of'the toolir.g designs were reviewed for process efficiency and equipment reliability. Each of.these goals have been substantially met during the installation development program.-
Appropriate precautionary steps have been included in the modification
. procedures to prevent unnecessary personnel hazards.
In addition, the tooling design incorporates features which minimize th'e potential for inadvertent damage to the steam ganerator. Specifically, consideration has been given to design features which will prevent the loss ~of loose parts in the steam generator, mechanical deformation of steam generator components and thc i.7troduction of chemical contaminants by use of unqualified materials.
All tooling operations were fully developed and tested utilizing proto-typic mock-ups of the feedwater nozzle / partial preheater sectic, of the 02/03 steam generator. The mock-up testing was performed as a means of tooling design verification, equipment qualification and training of Westinghouse personnel. These mock-ups are also being utilized for optimizing the installation procedures and training of site support
~
personnel. The methodology of this mock-up testing program was examined and judged to be extremely effective in providing mean'ingful input to the tooling and procedures development process.
5.
4.2 CONCLUSION
f In summary, it is concluded that the tooling and procedures are adequate.
to accomplish the proper installation of the preheat ~er modification as designed.
5.4-2 i
5.5 RADIOLOGICAL CONSIDERATIONS /ALARA 5.5.1 DISCUSSION The design modification and its installation procedures as presented in the Westinghouse design report have been compared point by point to the
. recommendations of Regulatory Guide 8.8.
It is clear that ALARA conc'epts have been' incorporated at appropriate stages of the design evolution.
The most obvious indication of this commitment was the requirement that the modification be accomplished with remote tooling' requiring minimum entry by personnel into the 1 res/hr field proximal to the tube bundle.
The final installation relies on remote tooling for positioning, machin-ing and welding.
Several other ALARA features should be mentioned. Westinghouse conducted an extensive program of full scale mock-ep training and equipment qualifi-cation.
Installation crews will have extensive hands-on experience with all phases of the installation before attempting the work in_ a radiation area. This training included practice'in protective clothing and respir-ators in the unlikely case the latter would be required.
In addition to remote tooling, remote monitoring systems, TV cameras with and without fiber optics systems will be used extensively to reduce the time spent by workers in the high radiation areas.
Post task ALARA reviews will be used to improve. job performance on subsequent comparable tasks.
.5.5-1
5.
5.2 CONCLUSION
This repair can be accomplished within the limits given in 10 CFR 20 and Westinghouse has satisfactorily incorporated the ALARA guidance found in Regulatory Guide 8.8.
Accordingly, it'is concluded that adequate consid-eration has been given to ALARA matters and the modification is accept-able from a radiological standpoint.
i l
i L
l l
5.5-2
5.6 QUALITY ASSURANCE 5.6.1
SUMMARY
Westinghouse presented to the Design Review Panel (ORP)'that their standard quality. assurance program for design control'and design verifica-
~ tion was applicable to the 02/03 steam generator modification program.
They emphasized that the standard programs were-defined in Topical Report WCAP 8370, Revision 9A, Amendment 1 (which was NRC approved) and imple-mented via WCAP 9550, along with various Westinghouse Division Manuals.
i Within the Westinghouse standard program the means of performing design-verification is defined as design review supplemented by' alternate analysis and/or testing programs as required. This approach is consistent with.QA~
criteria for verification and a formal design review program is in exist ence at Westinghouse.
A distinction must be made, however, between design control-techniques (and supporting QA programs) and design' verification. Whereas design' verification is the independent confirmation that a design is technically adequate once it is designed, design control techniques and supporting QA-programs are those controls in place during the evolution of the design.
Design control, therefore, assures that the in process design work is controlled by such actions as interface. reviews, document control of inputs and outputs, computer program verification, etc.'
I 5.6-1
Because of the distinction between design verification and design control, the QA discipline reviewer of the ORP approached the design verification task in the QA area as one requiring verification that the design control and supporting QA programs were in place and functioning. This approach differed from the Westinghouse approach to QA relative to design review (verification) in that Westinghouse participation emphasized QA items associated with the final manifold design, such as inspectability, rather than actions that assure the in process design work is controlled, such as interface reviews, document control of inputs and outputs, computer program verification, etc. Formal participation in the Westinghouse design reviews on the part of QA was by way of the Westinghouse Design Review Committee Secretary who organizationally belonged in the Design Integrity and Product Assurance (DI and PA) department (QA). Westinghouse also assigned a QA project manager to the design task force to assure that the existing QA programs were in use during the design process since the task force was assembled from groups who may or may not have been familiar with the QA programs. The QA project manager had no Westinghouse Design Review Committee function.s for verification purposes.
5.6.2 SCOPE OF REVIEW The verification activities performed by the DRP, were associated with how well standard QA programs functioned with respect to design control and to what extent the Westinghouse QA organization was involved in the standard programs as they directly related to the manifold design.
In addition, the Westinghouse QA participation in the verification (Design Review) was evaluated. The DRP evaluation of the Westinghouse programs was based on 5.6-2
m comparisons to criteria of Quality Assurance Regulations as judged appropriate for manifold design and design verification. QA criteria.of 10CFR50, Appendix.B, and requirements of Regulatory Guides 1.8,- 1.28, 1.33, 1.37, 1.38, 1.39, 1.58, 1.64, 1.74, 1.88,'1.116, 1.123, 1.144, and 1.146 were considered during the evaluation process.
i The items deemed important to Design control and verification of the
- manifold and results of the review of the items by the QA discipline of the DRP are discussed in the following subsections.
5.6.2.1 Field Data-Field inputs consisting of eddy current examination results-and acceler-ometer data were utilized by members of the Westinghouse design task force.
to analyze the problem and aid in establishing performance parameters needing to ~ be met by the manifold. The procedural controls associated with these tasks, the qualification of equipment and personnel, the calibration control of equipment, the controls associated with Westing-house procurement of testing services, and the Westinghouse QA surveil-lance.and audits associated with these functions were examined.
The results of the' examination revealed that methods were used for taking data from 4 or 5 plants. All of this data was not acquired under direct t
contractual auspices of Westinghouse. Eddy. current testing (ECT) at one plant was performed under contract from the plant owner to a testing agency who used their personnel and equipment and also analyzed the data.
Another plant hah the utility personnel perform the ECT under Wastinghouse 5.6-3
direction usino a' Westinghouse procedure and had-the same testing agency
~
as the first. plant perform the analysis. Another olant utilized a differ--
ent contractor under contract from the utility for total scope of services and provided data only to Westinghouse. And yet another plant had work.
done by the common testing agency under contract to Westinghouse directly.
In all cases, the same type of equipment was used and the testing method was the same with'the testing frequency exceptions which were accounted for during evaluation.
The most Westinghouse involvement with ECT data acquisition was by steam
' generator task force engineering personnel who visited all.the plants providing data,-established the relationship of the absolute method of ECT to the wear scars on tubes removed from plants, and utilized and compiled the data.
No' Westinghouse Quality:Asturance surveillance or audit activi-I ties were conducted with respect to work implementation'and field data acquisition.
For the'ECT testing agency used by Westinghouse procurement controls were used and audits were performed by qualified Westinghouse QA personnel that attested'to.the agency's program (procedures, calibration,'
training, etc.) acceptability.
The accelerometer instrumentation was provided by Westinghouse as " pro-cured and calibrated." The steam generator task force engineering per-sonnel were involved with supply, placement, acquisition of' data and utilization of results, while the utilities with the instrumentation provided the service.
5.6-4
.--_.-~..,,--,,_,_..--...,_.....m..
m-, _ _ _.,
...__,.,___.3
_, -,.. _ -,,,. ~.,
,y
l.
l' The field ECT data used during the design process appears to be credible.
This credibility is based on the facts that the data was acquired from
~
reliable subcontractors with the same type equipment that was within a calibration program; the relationship of absolute ECT measurement to wear scar. volume was clearly established; the results were readily compar-able to setup readings and to each other; and Westinghouse engineering-personnel closely monitored the process. Also, the fact that accelerometer instrumentation was installed in operating plants with assumed acceptable controls with Westinghouse engineering personnel involvement, tend to support the credibility of the data supplied from these instruments. The l
standard Westinghouse QA Program controls in-this area provided minimal QA organization involvement that could be related directly to the manifold design effort.
5.6.2.2 Qualification, Indoctrination and Training The steam generator task force consisted of personnel whose qualifications
.were a matter of record and commensurate with their functions.
- Further, indoctrination and training in procedures and programs (including QA programs) was conducted for task force members and docuentation of the training was available. QA personnel qualifications for personnel per-forming audits complied with the standard QA program in place at Westing-house which meets regulatory requirements.
In all, no problems with qualifications, indoctrination, or training existed for personnel directly involved in the manifold design or design verification.
5.6-5
l l
5.6.2.3 Computer Procram Verificatica and Control Various computer programs were employed by the manifold designers to support the technical acceptability of the' design..The Westinghouse standard _QA program provides for independent verification of computer programs and configuration control of computer programs to assure only authorized changes are made. The internal QA audit program provides for routine checks of_ design control.which includes computer program verifica-
]
tion and control. The means by which specific programs were verified and controlled _via the configuration control system was not examined in detail.
Emphasis was placed on whether computer programs used for mani-fold design were verified and controlled and whether Westinghouse QA personnel verified this by audit. A select number of progress modified in the design review meetings were confirmed to be in configuration control.
1 The Westinghouse internal QA audits of design control that addressed computer ' program verification and cohfiguration control were also exam-ined. While exam'ining this area, it was noted that the internal audits were programmatic in nature as required by the Westinghouse standard QA program. The-audits confirmed that the process was acceptable; but did not confirm that specific computer programs were included in the design process.
Based on this fact, and the fact that the design was a task force effort where nonfamiliarity with standard program requirements might exist, a recommendation was made for Westinghouse to perform a narrow scope, manifold specific audit to provide further assurance that all programs used were verified and controlled. Westinghouse agreed to schedule and perform this audit.
It is concluded with the accomplishment 5.6-6 l-
of this audit, that ample. computer program verification and coritrols
~ existed and.were.in place to support design of the manifold and that Westinghouse QA effort existed to confirm these controls.
5.6.2.4 Model Testing Various model testing associated with.the design of the manifold was performed. Westinghouse planned t5e various scaled down testing for'the purpose of scoping the problem. The Swedish State Power Board (SSPB) full scale test was planned to to confirm (verify) functioning of the design.
The SSPB test was required to have QA program controls while other scaled-down tests performed at various Westinghouse facilities or subcontractors were not required to have QA proqram controls. Much testing had already been performed when the DRP began its review.
Beginning with the SSPB full scale test, QA controls associated with testing were evaluated to encompass items such as:
. provisions for procedures for conduct of testing,
. control and calibration of test instrumentation and equipment,
. procurement controls' associated with calibration and/or testing ser-vices, as approp'riate,
. procurement control associated with model manifolds being tested, 9.
. dimensional verifications of the model and model manifolds,
. qualification of personnel involved in testing 5.6-7
verification by Westinghouse QA of test' implementation adequacy by
~
review, audit and surveillance techniques.
These items were considered appropriate to test control and were pursued both at the testing facility in Sweden and Westinghouse headquarters in Pittsburgh.-
Problems related to SSPB testing were identified in several areas. While at the test facility, a DRP member discovered that some procedural discre-
. pancies existed and that calibration status of certain instrumentation was
' questionable. Westinghouse took actions to correct these discrepancies in l
conjunction with performing an audit at the facility.
i A local calibration service was contracted to supplement the Westinghouse calibrations to assure instrumentation was within calibration.
This subcontractor utilized standards that were traceable to nationally recog-nized standards Elthough not the. United States National Bureau of Standards.
Westinghouse QA participated in the effort to correct calibration problems via audit at the test facility.
Subsequent evaluations by the DRP in Pittsburgh disclosed that no QA followup at the SSPB test facility was planned to assure that tests were run in compliance with test procedures, and that no test procedure reviews were planned by QA to assure compliance to test prospectus. Also, no i
formal review of the Swedish calibration subcontractor was planned.
Recommendations were provided to Westinghouse in these areas and Westing-house committed to revisit the test facility for surveillance to assure 1-5.6-8 i
,,_...-+,+=p-,-
,,--mw a
m.
_,,,. +,...
,pe
- 9.,. +,,,. -..,
,-v,,
m, 99,p-m.+-
m
l l
procedures'were, implemented and at the same time tsview the procedures for compliance. - These actions should provide further assurance of controlled testing of the_ manifold. Alternate means of verifying acceptability of instrument calibration were-presented to preclude formal-QA evaluation of-the-subcontractor.
It is concluded that this approach provided adequate assurance th'e instruments were within calibration.
Concurrent with full scale testing, discussions were held with Westing-house concerning other scaled down model testing conducted by various Westinghouse organizations and/or other vendors. Westinghouse concluded that some of this testing was required for design and verification of the manifold design. Accordingly, design controls and associated QA program controls should have been applied, irt particular, to the.417 and 2/3 scale models and to. wear test facilities.
In general,'those problem scoping tests that were subcontracted.were performed via procurement controls that confirmed the vendor has an acceptable test program which included elements of procedural and calibra-tion control.
Evidence of procurement document reviews, prospectus reviews and audits on the part of Westinghouse QA were available, although no plans for in process surveillance and procedure review for acceptabil-ity were planned or. performed (similar to SSPB). _The test.proolem scoping performed by Westinghouse organizations generally had no QA program i
controls applied.
As.a result of evaluating model testing and observing the test facilities, it was recommended to Westinghouse that they continue pursuit of QA 5.6-9
program controls on testing (other than full scale) that they had begun by considering formal identification of those tests needing QA program controls and a retrofit (rerun) using appropriate controls such as proce-dure review for compliance, calibration centrol and surveillance of.
implementation. As an alternate, it was suggested that for testing alreadyaccomplishedadocumentedanalysisbeperformedkojustifythe adequacy of testing based on items such as:
. model to model correlation,
. expected / predicted results,
. correlation of results to operating plant data,
. consistency of results (test after test),
. repeatability of base case results.
Westinghouse reviewed-the model testing programs to provide a master list of those tests being utilized to substantiate the design or verify the design and provided the necessary QA program controls including QA review and surveillance to provide further assurance that controlled testing took place and results were credible.
Most of the program implementation took place on newer tests and those easily rerun.
The alternate approach was utilized to document acceptability of some testing. Overall, the DRP reviewer concluded Westinghouse actions appropriate.
(
Controls associated with production of model test manifolds were evalu-ated. Manifolds were manufactured at a Westinghouse vendor under procure-ment controls required by the standard Westinghouse QA programs.
Verifi-cation of supplier qualifications a,nd capabilities was accomplished by QA 5.6-10
l audit. Although criginal plans did not include QA surveillance at-the manufacturing facility or independent dimensional checks, these activities were subsequently performed by Westinghouse at the recommendation of the DRP.
5.6.2.5 Westinahouse Desian Review' Committee Activities Evaluation of the Westinghouse Design Review Committee activities was
~. performed to the extent that the process was. determined to be program-natically acceptable ~and functioning. Reports of meetings were detailed and it appeared that technical subjects were addressed and open items
- pursued and resolved or carried for future examination..In certain cases-alternate calculations suppcrted theTeview.
~The Westinghouse QA participation on the committee was evaluated and a representative of DI and PA (QA) served'as secretary.
Based on the documentation of the proceedings, the QA items considered involved only-manufacture and installation. This approach differed from that taken by-the DRP. Westinghouse apparently relies on activities conforming to its program to the extent that conformance to the design control p. w as neea-not be pursued for eacn oesign. This approach is acceptable.
Q 5.6.2.6
. Manufacture of Manifolds k
The results of the design effort to be utilized in the manufacture of manifolds were evaluated. These results consisted of drawings and specifi-cations for the manufacturing facility.
The QA involvement in the specifi-1 5.6-11
______-___ K __-
l cation review, approval and issuance was in accordance with the Westing--
house standard program and'is considered acceptable.
A tour.of the manufacturing facility (Westinghouse-Electro-Mechanical Divi-sion) was performed. This facility has a manufacturing quality assurance program capable of assuring an acceptable manifold. A description of techniques used for control and inspection were provided the DRP and the conclusion is'that there is a high probability the manifolds will be manufactured acceptably.
5.6.2.7 Installation of Manifolds Throughout the design and design verification (review) process,' attention was given to the installation of the manifolds. The tooling and service organizations within Westinghouse prepared a program to accomplish these tasks concurrent with the design review. The Westinghouse QA organization monitoring of tpis program was-investigated to determine its acceptability to applicable regulatory criteria.
Both Westinghouse (Tampa) and Westing-house (Pittsb gh) activities associated with the installation program' were examined.
i The Westinghouse standard programs and QA organization involvement in this area encompassed enough to provide control and were implemented commen-surate with the status of work activities. The QA involvement consisted of reviews'of functional specifications for tooling used as the basis for tooling design reviews, tooling qualifications, and tooling procurement.
Westinghouse QA also reviewed the qualification test procedures prepared 5.6-12 1
l
1 by the design engineers for utilization in the acck-up program at Tampa to prove functionality.-
l
-l J
The installation procedure results from these qualification efforts. -In-addition, QA monitored step-by-step mock-up activities to assure accept-ability of the process.
Similiarly, non-destructive examination tech-niques and procedures to perform the examinations were reviewed and monitored by Westinghouse QA.
Training of craftsmen was audited by Westinghouse QA and. training was i
provided to Westinghouse QA personnel who will staff the on-site installa-I tion. Training programs ~are well documented.
Calibration of tooling, gages, NDE equipment, etc., is within the Westing-house calibration program and internal QA audits were performed and are planned to. verify continued acceptable implementation of the-program.
The Westinghouse effort in this area is considered adequate with good QA organization participation. A recommendation was provided Westinghouse to verify that the mock-up nozzles used for, training be verified as dimen-sionally representing an actual steam generator. Westinghouse responded.
to the recommendation which should provide further assurance of adequate installation planning.
r 5.6-13
l
-5.5.2.8 Records
^
Records of the design and design verification effort at Westinghouse were in use and dispersed among the task force at the time the DRP review was conducted. -Westinghouse has documented a plan whereby a responsible individual will interface with participants and compile the necessary records package for permanent retention.
It is concluded that if the planned activities-are executed, records relative to the manifold design and verification will be adequate.
5.6.2.9 Audits Westinghouse utilized standard program audits of design control supple-mented by a steam generator task force specific audit to monitor the manifold design. The audits that were routinely performed complied with the standard program in that they were performed by qualified auditors and' were followed and closed in an acceptable manner. Throughout the ORP review additional audits and surveillances relating directly to the manifold design and verification were recommended and incorporated by Westinghouse. Taking all Westinghouse audit and supplemental reviews' into account, an adequate internal audit program relating to the manifold exists.
5.
6.3 CONCLUSION
S Based on the details of the DRP QA discipline review, it is concluded that:
5.6-14 i
l I
Design activities were monitored to the extent that the Westinghouse QA effort reinforces conclusions drawn in the technical areas, Further assurance is provided that tests and analysis were performed in a controlled fashion based on Westinghouse QA involvement, and There is a high probability that manufacturing and installation of manifolds will experience negligible problems based on Westinghouse QA involvement.
t I
5.6-15
_-______________-_________a
I l
l L
l 5.7 CHEMISTRY AND CLEANLINESS CONTROL 5.7.1 CHEMISTRY The steam generators to be modified are in three categories:
(a) those oper-ated at power, (b) those in wet layup but not operated, and (c) those in dry stcrage. During installation of the Westinghouse preheater modification, a major task is to minimize corrosion of the steam generator internals while creating no hazards for those personnel carrying out the modification or working in the containment.
Among the various methods and techniques considered were an inert steam gener-ator atmosphere, partial wet layup, hot arr flow, dehumidified air circulation and hot dry out. Each of these were discounted primarily for the following reasons:
inert atmosphere personnel hazard; partial wet layup - hydrazine and ammonia vapor hazard to personnel; hot air flow through - ineffective in drying out the steam generator and physical problems involving inflow and venting; dehumidified air circulation - same as previous except for venting.
l The major defenses against corrosion'which do not impact personnel are the-1 i
cagnetite film which is much more passive than the base carbon steels and.
einimizing exposure of the steam generator internals to the containment at-mosphere.
L Steam generators that have operated at power have heavy magnetite films on the h
interior carbon steel surfaces.
Similar films of 6 mils or more on fossil boilers have been shown to protect carbon steel in a humid atmosphere from further corrosion for over one year. Considering that the proposed modifica-5.7-1 L
tion'can be completed in 40 to 60 days from initial' pipe cut to final pipe weld, no corrosion degradation ~of the steam generator should occur.
Steam generators in wet layup also have protective magnetite films on the interior carbon steel surfaces. Although these films are not as heavy as
" operated at power" steam generators, they should be adequate to protect the
' generator for the limited time of exposure to the atmosphere.
1 Steam generators, which are currently in dry storage under an inert atmosphere,-
should be opened and the inert gas replaced with air.
It is recognized that a film of hematite may form during the' modification period, however, the film should be thin and readily converted to protective magnetite during subsequant hot functional testing. Adequate prctection of the generators may also be achieved by placing them in wet layup or by reestablishing an inert atmosphere following modification.
5.7.2 CLEANLINESS CONTROL The cleanliness preservation was investigated with respect to the manufacturing process and the modification installation'. The manufacturing process was reviewed regarding cutting oils, degreasing solvents, marking materials, packing siaterials, shipping containers and storage requirements. All r,aterials introduced into the steam generator were evaluated as to the chemical composi-tion, especially in respect to their halogen sulfur and low melting metal l
content.
5.7-2 l
E L1 I
iL ii Westinghouse process specifications which adequately define cleanliness criteria l
l have been approved for these applications.
l 5.
7.3 CONCLUSION
-It is concluded that the chemical and cleanliness aspects of implementing the proposed modification have been adequately considered. There were no safety questions and none are expected from the proposed chemical and cleanliness reouirements. Adequate contingency methods and techniques have been considered and are available to resolve any unexpected events.
5.7-3
5.8 INSERVICE INSPECTION /NO TES' TING
- 5.
8.1 INTRODUCTION
l The surveillance program described in Sections 5.8.2 and 5.8.3 addresses three aspects of the modification installation:
1.
Long term structural integrity of the modification:
2.
Effect of the modification on other steam generator components.
3.
Performance of the nodification with respect to wear.
To assess the long term integrity of the manifold and the effects of the modification on the steam generator components, Westinghouse has proposed a visual inspection program.
To assess the performance of the modification, with respect to tube wear, Westinghouse nas proposed an eddy current test program.
Both of these programs will be applied to all plants to be modified.
In addition to these programs, Westinghouse has proposed a program to monitor accelerometers in tubes in the first two plants to receive the modification.
l These programs are discussed in Section 7.3 of the Westinghouse Report.
5.8.2 VISUAL INSPECTION 5.8.2.1 Inspection Scope The visual inspection is ganeral in nature. As much of the assembly as is a
accessible will be inspected with particular emphasis on any critical areas.
f 5.8-1 l
The general inspection will cover three major areas:
)
1.
Mechanical inspection 2.
Weld inspection 3.
Sludge / deposit inspection The components and welds which will be inspected are listed in Table 5.8-1. This table lists each specific component and weld to be inspected along with identification of specific inspection items.
The scope of the internal manifold assembly visual inspection will include all accessible areas. With respect to applied mechanical loads, feedwater flow loads and adverse chemical build'up, no critical areas have been identified. While the inspection scope identified in Table 5.8-1 is somewhat detailed, the examination will be general in nature. The primary goal of the inspection is to ascertain the overall integrity of the-assembly after a periou ef power operation.
In addition, this' inspection will allow the identification of impending problems.
For example,6the crud / sludge inspection will identify areas of potential corrosive attack well before significant damage can occur.
Based on the results of the examination, remedial-programs will be developed to address any problem
(
areas.
5.8.2.2 Access and Equipment Access to the feedwater inlet manifold can be gained through a radiography inspection port.
This small opening allows inspection of portions of 5.8-2
I the manifold and manifold inlet assembly, including the flow limiter, flow splitter, manifold boxes and associated welds.
The primary inspection tool will be a fiberoptic boroscope guided to the i
areas of interest by hand.
Still photographs and videotape of appropriate portions of the inspection will be used for a permanent record.
5.8.2.3 Schedule A complete visual inspection will be made subsequent to modification installation and prior to initial power operation on all units.
A second complete inspection will be performed concurrent with the first shutdown for eddy current inspection after the modification.
Followup inspections will be at the discretion of the individual utility.
l 5.8.3 EDDY CURRENT TESTING (ECT) 5.8.3.1 Eddy Current Test Scope The primary area of interest in the 02/03 steam generators are Rows 45 through 49, all columns (approximately 297 tubes).
The tubes in these rows will be inspected as part of the preheater modiff-cation verification program. Table 5.8-2 is a list of the tubes to be examined. The cold leg side of the tubes will be inspected from support i
plate 1 through 10 for the Model D3 (2 through 10 for Model D2).
5.8-3
The ECT program is the primary means of assessing in plant performance of the modification. Comparisons between the initial examination results and results after a period of power operation will be used to determine whether the modification is effective in eliminating tube wear.
In addition, this data will be used to determine any follow-up examination required. The ECT program recommended by the ORP is summarized in Table 5.8-3.
The ECT program includes those tubes which are most likely to be affected by the modification. The Westinghouse ECT program in Section 7.3 of the Westinghouse Report identifies additional tubes beyond the first five rows to be tested.
In the judgement of the ORP testing of these addi-tional tubes is not necessary to verify acceptable performance of the modifi-cation.
The DRP considers testing of these additional tubes to be an option to be decided by individual licensees Because the modification is expected to reduce the rate of tube wcar to zero or to an acceptably small value, only the tubes directly affected by wear need to be included in the inspec-tion program.
5.8.3.2 Equipment and Techniques The equipment and tcchniques used to perform the ECT will depend on the equipment and techniques utilized in previous inspections. The equipment or technique used should produce results compatible with previous inspection results and should have a demonstrated capability to detect small volume wear scars.
In addition, careful consideration should be given to the selection of a calibration standard.
It has been shown that this can influence the accuracy of the inspection.
5.8-4
E '
5.8.3.3 Schedule The above ECT inspection will be conducted at the time of installation of the modification. This inspection will serve as a baseline for future inspections and will be conducted on all plants making the modification.
The next ECT inspection of the first plant to operate in the modified condition will be performed after approximately six effective full power months of operation. After this inspection, the inspection interval will be determined in accordance with the requirements of Regulatory Guide 1.121 based on the measured wear rate.
For units other than the initial operating unit, the allowable inspection interval should be based on the results of the inspection of the first operating unit.
5.
8.4 CONCLUSION
S The DRP has reviewed the testing and inspection program proposed by Westinghouse in Section 7.3 of the Westinghouse Report. While it is a comprehensive program, it goes beyond what is required to demonstrate acceptable performance of the modification.
Individual licensees may elect to perform the entire Westinghouse program, but it is the DRPs judgement that certain portions of this program should be considered optional.
The testing and inspection program discussed in this ORP report is considered to be what is required to verify acceptable modifica' tion perfcrmance.
f 5.8-5
The vibration verification testing proposed by Westinghouse for the first two plants is a reasonable program which will be implemented.
It is expected that as a minimum this program will be followed on McGuire 1 and Almaraz 2.
Number and location of accelerometers will be determined by Westinghouse and the plant ownors, h
)
5.8-6
TABLE 5.8-1 (1 of 2)
Visual Inpsection Plan Summary
- Component Scope of Inspection Manifold Box Mating Surface Mating Surface Gaps Mating Surface Fretting Near Erosion j
Fastener Locking Clips engaged and intact l
Bolt head to manifold gap Manifold Box General Inspection for Loose Parts Manifold Support Assembly General inspection for Loose parts, Flow Splitter Erosion, Fretting Near, Corrosion Support Cylinder and Cracking Flow limiter Thermal Sleeve i
Flow limiter to support cylinder weld Service induced cracking Splitter to liner weld Service induced cracking Manifold support cylinder to liner Service induced cracking
. weld Support cylinder dissimilar weld Service induced cracking i
Support cylinder to manifold weld.
Service induced cracking i
j Splitter vane to hub weld Service induced cracking Support cylinder to liner weld Service induced cracking
- In each case, only those portions of the particular component which are accessible is considered to be included in this inspection.
t
- s a
TABLE 5.8-1 (2 of 2)
Visual Inpsection Plan Summary i
Component Scope of Inspection j
Splitter vane to cylinder weld Service induced cracking I
Flow limiter - Leading Edge Sludge / deposit buildup.
Flow splitter - Leading Edge Sludge / deposit buildup.
Entrance plate 1/4" holes Sludge / deposit buildup.
Top of Horizontal Flow Splitter -
Sludge / deposit buildup.
Entrance Plate Joints Top of Horizontal Entrance Plate Sludge / deposit buildup.
Ribs Top of Horizontal Surfaces in Mani-Sludge / deposit buildup.
fold boxes.
Box External General integrity of assembly dimensional relationship between box and tubes maintained.
4 d
1
TABLE 5.8-2 Tubes to be Inspected for Preheater Modification Verification Preheater Inlet Region Row Column 49 31 through 84 48 29 throv3b 96 28 through 87 47 46 26 through 89 45 25 through 90 I
l l-l
TABLE 5.8-3 Eddy Current Test Program Location Time Sequence Reason Preheater inlet - 5 rows 1.
After modification prior Baseline for preheater modification (45-49) (cold leg only) to startup performance verification.
2.
After power operation Establish the tube wear rate of the period (See Section 5.8.2.3) modified steam generators for input to the Inservice Inspection Program requirements of Regulatory Guide 1.121.
l
I l
)
6.0 SIMERY The preheater modification developed by Westinghouse for installation in Model 02/03 steam generators provides a substantial improvement in preheater hydraulic performance and reduction in tube vibration to an acceptable level. This improvement was accomplished by creating a more uniform, less turbulent flow field entering the tube bundle. The proposed modification replaces the existing impingement plate assembly with an internal manifold and replaces the 4 hole reverse flow limiter with a 19 hole limiter. The design of the modification was reviewed extensively by the utility Design Review Panel in parallel with the normal Westinghouse design review process. The DRP review is documented in Sections 1.0 through 5.0 of the report.
1 Westinghouse performed a safety evaluation of the preheater modification which is set forth in Section 11.0 of the Westinghouse report. The conclusion from this evaluation is that modification of the 02/03 steam generators does not represent an unreviewed safety question. The panel concurs in this conclusion.
l Two major areas of interest to the DRP were the effectiveness of the modification in reducing tube vibration and the assurance of structural I
adequacy of the modification. The DRP concluded that the modification will provide for reduced tube vibration by minimizing turbulence and buffeting in the preheater section. This determination was reached after a thorough review of test models and testing results as well as analytical models and analysis results.
It is concluded that the modification will 6.0-1
result in substantially improved flow conditions in the preheater. It is noted, however, that some tube wear, in excess of the design objactive of zero wear, may occur over the projected 40 year life span of the steam generator. This is judged to be acceptable based upon proper provisions for tube inspection and remedial action, as required.
The ORP performed a detailed review of the stress analysis performed, which included examination and evaluation of analytical methods and models. The DRP concluded that the socification did meet the established structural design criteria, with one exception.
In this case, the low flow / cold water forward flushing transient was found by analysis to produce loads which resulted in inadequate stress margins in bolting.
Responsive to this finding, Westinghouse established operating limits on purge flow rates and temperature which would assure stress margins were preserved in bolting. These limits establish that for fluid assumed to be at 32*F, minimum purge flow rate must be 2.5%.
For lower flow rates entering the steam generator, fluid temperature of 200*F minimum must be assured. The Design Review Panel believes these limiting conditions are achievable with adequate plant instrumentation and specific operating procedures. Consistant with the maintenance of these operating parameters, the proposed modification is adequate.
In its consideration of functional achievability, the Design Review Panel notes there are other options available which alleviate the need for such precise operations control and still provide assurance design margins are maintained. These options include changes in plant systems design to preclude need for forward flow.feedline warming techniques.
The panel is 6.0-2
_ =.
~_.
cognizant of unique feedwater designs in virtually every facility having l
02/03 steam generators, and it is expectea that individual utilities will consider these options.
The DRP reviewed the materials selected for use in the modification, welding criteria and processes, NOE requirements and installation methods.
The DRP concluded that Westinghouse has adequately addressed each of these areas.
In addition, it is concluded that adequate consideration has been given to ALARA principles in the proposed installation methods.
Chemistry and cleanliness control during the installation phase was reviewed with the conclusion that this area has been adequately considered.
Quality Assurance measures relative -tm design, testing, analysis, manu-facturing and installation activities were reviewed.
It is concluded that these measures were sufficient to assure the adequacy of the modifi-cation.
The inservice inspection and testing program recommended by Westinghouse for the first two plants to be modified (McGuire 1 and Almaraz 2) was reviewed.
It is concluded that this program will provide sufficient verification of modification performance.
Based upon its review, the Design Review Panel concludes that the proposed r
j modification can be installed in the D2/03 steim generators and that they can be operated safely at 100% of their design capacity.
I 6.0-3
APPENDIX C CHRONOLOGY Ringhals tube leak October 21, 1981 NRC meeting with Duke Power Company November 20, 1981 and Westinghouse NRC meeting with Westinghouse February 17, 1982 Design Review Panel (DRP) formed May 12, 1982 NRC meeting with DRP Chairmen May 19, 1982 Initial DRP meeting (NRC in attendance)
June 30-July 1, 1982 NRC meeting with DRP Chairmen August 25, 1982 DRP meeting (NRC in attendance)
September 29-30, 1982 DRP review complete January 14, 1983 DRP report submitted to NRC January 19, 1983 (proprietary report submitted Jan. 17,1983)
NRC meeting with DRP January 21, 1983 i
NUREG-0966 C-1
,p-,
-y-y 9
,----,m
,-9 r
- 8'#
U.S. NUCLEAO EEZULATORY COMMISSION yy BIBLIOGRAPHIC DATA SHEET NUREG-0966
- 4. TITLE AND SUBTtTLE (Add Volume No., of apprmriatel
- 2. (Leave b/mk)
S2fsty Evaluation Report related to the D2/D3 Steam G:n=rstnr DeRign Modification
- 3. RECIPIENT'S ACCESSION NO.
l l
- 7. AUTHORIS)
- 5. DATE REPORT COMPLETED M ON TH l YEAR March 1983
- 9. PERFORMING ORGANIZATION NAME AND MAILING ADDRESS (Include 2,p Codel DATE REPORT ISSUED Divi; ion of Licensing March 1983 Offica or Nuclear Reactor Regulation
- 6. (Leave blank /
U.S. Nuclear Regulatory Commission W uhington, D. C.
20555 s.ft,,,,b,,nki
- 12. SPONSORING ORGANIZATION NAME AND MAILING ADDRESS (/nclude tip Codel p
- 11. CONTRACT NO.
- 13. TYPE OF REPORT PE RIOD COVE REo (/nclussve dates /
- 15. SUPPLEMENTARY NOTES
- 14. (Leave D/ mal
- 16. ABSTR ACT Q00 words or less)
Thie Safety Evaluation Report (SER) related to the D2/D3 steam generator design modifica-tion has been prepared by the Office of Nuclear Reactor Regulation of the U.S. Nuclear R:gulatory Commission. The purpose of this SER is to issue the staff's evaluation of the acceptability of the design modification for both installation and full-power operaticu in the D2/D3 steam generators based on the Design Review Panel Report of January 1983.
l l
- 17. KEY WORDS AND DOCUMENT AN ALYSIS 17a. DESCRIPTORS 17tr IDENTIFIE RS/OPEN-ENDE D TERMS l
19 SE CURITY CL ASS (Tn,s eeporrl 21 NO 0F PAGES 18 AV AIL ABILITY STATEMENT Unclassified Unlimited 20 SECURITY CLASS (Thes pagel 22 PRICE Unclassified NHC F OHV 335 67 77)
1, c
r
~
~
, i; UNITED STATES..
- ast class ea.t.
. NUCLEAR REGULATORY COMMISSION
~
Po"stact a sats pa.o -
.unnec
-. WASHINGTON, D.C. 20M6 ~
.4 easm a c,.
~
.Meast see g '
i s
' opriciat susnefss.
.,4 PENALTY FOR PhfVATE USE. 4300 W
- e.,
-c w
7 1,s
.g
+
i s
i
['
120555078677 1 AN
~
Lc h8C A5F Olv CF 1100 PCLICY E PUP PGT Bh-POR AUREG
~
CC 20555 5
NGTLh m
m
.A Y
+"
J k
s 1
N Y
M i
i I
t r
~
..,. -.,-..