ML20058P218
| ML20058P218 | |
| Person / Time | |
|---|---|
| Site: | Framatome ANP Richland |
| Issue date: | 12/10/1993 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20058P104 | List: |
| References | |
| NUDOCS 9312230156 | |
| Download: ML20058P218 (1) | |
Text
m
..j
.jp#"Y R
g, -
.3 E'-
. UNITED STATES
~l
?'
4i NUCLEAR REGULATORY COMMISSION
...../
WASHINGTON, D.C. 2055b0001 December 10, 1993 Mr. R. A. Copeland Manager, Reload. Licensing Siemens NJclear Power Corporation j
2101 Horn Rapids Road P.O. Box 130 Richland, WA 99352-0130
Dear Mr. Copeland:
SUBJECT:
ACCEPTANCE FOR REFERENCING OF LICENSING: TOPICAL REPORT EMF-92-081, o
" STATISTICAL SETPOINT/ TRANSIENT METHODOLOGY FOR WESTINGHOUSE TYPE REACTORS" (TAC NO. M85061)
The staff has reviewed the topical report submitted by Siemens Nuclear Power..
Corporation by letter dated May 29, 1992. The report is acceptable for referencing in license applications to the extent specified-and under the limitations stated.in the enclosed report and U.S. Nuclear Regulatory-Commission (NRC) evaluation. The evaluation defines the basis for acceptance of the report.
The staff will not repeat its review of the mai.ters described.in'the report.
'I and found acceptable when the report appears as a reference in license:
applications, except to assure that-the material presented applies to the.
specific plant involved. NRC acceptance applies only to-the matters' described' in the report.
In accordance'with procedures established in NUREG-0390 the.
NRC requests that Siemens Nuclear.. Power Corporation ~ publish accepted versions 1
of the_ report, proprietary and non-proprietary, within 3 months of receipt of.
j this letter. The accepted versions shall incorporate this-letter and the enclosed evaluation between the title page and.the abstract and an -A (designating accepted) following the report identification symbol.
t i
If the NRC's criteria or regulations change so that its conclusion.that the
- i report is acceptable is invalidated, Siemens Nuclear Power Corporation and/or
?
the applicant referencing the-topical-report will be expected to revise and 1
resubmit its respective documentation,.or submit justification ~for the a
continued applicability of the topical: report without revision of ~ the.
respective documentation.
Sincer i
4k M
l g
Asho C. Thadanbtor:
Divi ion of-Systems Safety and Analysis Offi~ce of Nuclear Reactor Regulation-
Enclosure:
i EMF-92-081 Evaluation j
+
9312230156' 931210 "
i PDR REVGP NRCCRCR:
CORRESPONDENCE PDR l
i r
I l*
j,("*%,
l
, e n
r-8 UNITED STATES
~
i NUCLEAR REGULATORY COMMISSION g
,/
wAssinoTou. o.c. rosss-cooi ENCLOSURE SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION l'. ELATING TO TOPICAL REPORT EMF-92-081(P)
" STATISTICAL SETPOINT/ TRANSIENT METHODOLOGY FOR WESTINGHOUSE' TYPE REACTORS" '
SIEMENS NVCLEAR POWER CORPORATION 4
1.
INTRODUCTION In a letter of May 29, 1992, from R. A. Copeland to' T. E. Murley (NRC),
Siemens Nuclear Power Corporation (SNPC) submitted topical report EMF-92--
t 081(P),. " Statistical Setpoint/ Transient Methodology for Westinghouse Type Reactors," for NRC review.. The report describes the SNPC methodology for performing statistical transient analyses, including the statistical development of the trip protective systems, utilized in Westinghouse-type-pressurized water reactors.
The NRC staff was supported in this review by.its consultant, Brookhaven National LaForatory.. The staff has adopted the findings recommended in the consultant's technical evaluation report (TER) which is attached, i
2.
EVALUATION The attached TER provides the evaluation.
3.
CONCLUSIONS The staff has reviewed the SNPC topical report EHF-92-081(P). and the.
supporting doctmentation submitted in response lto staff requestsLfor j
additional information. On the basis of this review, the staff-. concludes that i
EMF-92-081(P) is acceptable for-referencing-in licensing actions' by SNPC with l
respect to the statistical setpoint methodology for Westinghouse reactors, subject to the limitations stated in Section:4.0 of the attached TER.
In a separate action, the HTP DNB correlation has been approve'd by the NRCland,-
1 therefore, may be used with the SNPC statistical setpoint~ methodology i
. described in EMF-92-081(P).
-l
}
I o
TECIINICAL EVALUATION REPORT TECHNICAL EVALUATION OF THE SNPC STATISTICAL SETPOINT/ TRANSIENT METHODOLOGY TOPICAL REPORT EMF-92-081(P)
J. F. Carew November 20,1993 Prepared for the Office of Nuclear Regulatory Regulation U.S. Nuclear Regulatory Commission Washington, DC 20555 NRC FIN L-2589-3, Task-1
. Reactor Analysis Group Applied Technologies Division Brookhaven National 1.aboratory Upton, Long Island, New York 11973 W
TECHNICAL EVALUATION REPORT Topical Report
Title:
SNPC Statistical Setpoint/ Transient Methodology Topical Report Number:
EMF-92-081(P)
Report Issue Date:
May 1992 Originating Organization:
Siemens Nuclear Power Corporation
1.0 INTRODUCTION
By letter dated May 29,1992 (Reference-1), the Siemens Nuclear Power Corporation (SNPC) has submitted the Statistical Setpoint/ Transient Methodology Topical Report - EMF 081(P) for NRC review and approval. The topical report provides the methodology that SNPC intends to use in determining the trip protective system settings for Westinghouse Oy) type reactors. A detailed description of the methodology and the treatment of setpoint parameter uncertainties together with a sample calculation are included in the Topical Report. The primary difference relative to the presently accepted SNPC methods for Westinghouse reactors is that in the proposed methodology the calculation and measurement uncertainties are treated statistically rather than deterministically. The statistical approach used is similar in many respects to the SNPC statistical setpoint methodology that has been approved for application to Combustion Engineering (CE) plants (Reference-2). The SNPC methodology makes use of response-surfaces, together with Monte Carlo sampling techniques and Second-Order Error Propagation -
l (SOERP) methods, to determine appropriate setpoint uncertainty tolerances. The construction and testing of the response surfaces is based, in part, on the approved GSUAM - Generic
)
Statistical Uncertainty Analysis Methodology (References 3 and 4). The setpoint methods are intended for the determination of the Overpower-AT (OPAT) and Overtemperature-AT (OTAT) serpoints which provide 95/95 protection for the fuel melt and DNBR limits, respectively. The methodology will also be used for analyzing transient events and determining or verifying the reactor protective system trips.
The review of the Topical Report focused on the conservatism included in the determination of the OPAT and OTAT setpoint parameters and the statistical method used for determining the 95/95 uncertainty limits. The SNPC methodology is summarized in the following Section-2, and the technical evaluation of the important issues raised during this review is presented in Section-3. The technical position is given in Section-4.
2.0
SUMMARY
OF THE TOPICAL REPORT 2.1 Ovemower-AT Reactor Trio The Overpower-AT reactor trip provides 95/95 protection from fuel melt during operational transients and Anticipated Operational Occurrences (AOOs). The form of the OPAT trip is the same as used by Westinghouse (Reference-5). The trip includes a constant term with coefficient K, and a term proportional to the core-average temperature with coefficient K to 4
6 account for the effects of coolant density and heat capacity on the relationship between AT and the core-average temperature. An additional core-average temperature term with coefficient K5 is included to account for transient effects such as piping and thermal delays. Compensation for the time dependence of the measured AT and core-average temperature is accounted for in the 2
OPAT trip equation using the precalculated sets of time-constants (rg, 7, 7 ) and (r6 7 ).
2 3
7 respectively.
The steady-state coefficients K and K are determined so that for the design peaking 4
6 factor F, the core power level at which fuel centerline melt occurs is avoided. The power q
peaking measurement uncertainties and the engineering factor are included in the K -coefficient 4
using a Monte Carlo procedure and the remaining uncertainties are subt acted directly from the value of K.
4 The calculation of the K and K OPAT temperature coefficients is based on the design 4
6 basis peaking-F. In order to account for power distributions more severe than the design g
peaking, the OPAT limit is calculated for a range of axial power distributions. Ajoint probability distribution is then determined in order to allow for uncertainty in both the power level and the axial flux difference-AI. The power distribution adjustment to the OPAT limit -
the F(AI) reset function -is determined using this joint probability distribution and by requiring that the fuel-melt power level be avoided with 95% probability and 95% confidence.
2.2 Overtemperature-AT Trip The Overtemperature-AT trip provides the required protection from DNB and hot-leg saturation during normal operation and Anticipated Operational Occurrences (AOOs). The form of the Overtemperature-AT trip is the same as that used by Westinghouse (Reference-5) and includes a constant term K, a term proportional to the core-average coolant temperature with i
coefficient K, and a term proportional to the pressurizer pressure with coefficient K. The time 2
3 dependence of the measured AT and the core-average temperature are accounted for in the 3
1
I OTAT limit using the predetermined sets of time constants (ri,r2,r3) and (r4,7,7 ), respectively.
5 6 The OTAT steady-state coefficients K, K and K are determined to provide DNB and hot-leg 1
2 3
saturation protection at the 95/95 probability / confidence-level. The DNB uncertainty margin is calculated using a response surface for AT. A similar analysis is performed for the hot-leg saturation AT setpoint. The response surfaces are determined using the SNP approved GSUAM procedures. The statistical uncertainty penalty is determined using a Monte Carlo approach, rather than the SOERP method, and includes the response surface parameter uncertainty as well as the trip input temperature and pressure measurement uncertainties.
In order to provide DNB protection for axial power shapes more severe than the design axial distribution, an F(AI) reset trip adjustment is included in the OTAT trip function. The first step in determining the reset function is to identify the DNB limiting axial power distribution for each AI interval. A AT response surface is then generated for this axial distribution at the conditions determined to be most sensitive to the uncertainties. The resulting AT distribution is then used to determine the joint probability distribution. The F(AI) adjustment is calculated -
from this joint probability function.
1 23 Neutronics Analysis q
Both the OPAT limit for fuel melt and the OTAT limit for DNB protection depend on the core axial power distribution. This dependence is included in the trip functions via the F(AI) reset function. The relationship between the local peaking (F and Fg) and Al are determined n
using a three-dimensional neutronics model. The axial flux difference and power distribution are calculated as a function of core operating conditions.
4 u
3.0
SUMMARY
OF THE TECHNICAL EVALUATION The Topical Repon EMF-92-081(P) provides a detailed description of the statistical methodology that SNPC intends to use for the setpoint determination for Westinghouse type -
reactors. The review focused on the applicability and conservatism of the methods used for calculating the trip parameters (K -K ), and the validity of the statistical approach employed in 3 6 determining the allowance for design and measurement uncertainties.
Several important technical issues were identified during the initial review which required additional information and clarification from SNPC. This information was requested in Reference-6 and was provided in the SNPC response included in References 7-9.
This evaluation is based on the description and examples presented in the topical report and the supporting information provided in References 7-9. The evaluation of the major issues raised during this review are summarized in the following.
3.1 Setooint Methodology 3.1.1 Transient Effects The time-dependent effects associated with instrument delay and piping lag are accounted for by the K -transient coefficient in the OPAT trip setpoint equation. These effects are 5
considered to be hardware related and, in Response-17 of Reference-7, SNPC has indicated that this parameter will not be modified with the application of the EMF-92-081(P) Methodology.
The OPAT trip is designed to protect against slowly evolving transients in which there is no significant fuel temperature overshoot. In Response-3 of Reference-7, SNPC has indicated that 5
)
l 1
it will verify the adequacy of this term and the protection of the fuel centerline temperature from fuel melt by performing event specific transient analyses.
The time-dependent processing and compensation of the OPAT and OTAT trip input i
signals is performed using the seven time constants 73 - r7 These plant-specific equipment-related constants will be provided from information supplied by the reactor vendor. These time constants will not be recalculated as part of the SNPC setpoint methodology, however, the adequacy of the values will be confirmed by SNPC in plant-specific transient analyses (Response-19, Reference-7).
3.1.2 Response Surface Methods The statistical uncertainty analyses for the DNB and hot-leg saturation limit lines and the statistical transient analyses are performed using response surfaces generated with the SNPC_
GSUAM methodology. The response surfaces are typically low order polynomials in the i
independent variables and the specific functional forms are selected to. provide the required
-)
.J accuracy. Specifically, the selection is based on the fitting statistics, residual plots and the overall goodness of the fit. In Response-9, SNPC has indicated that the response surface fitting error is determined and included in the Monte Carlo uncertainty analysis.
The methods used to construct the response surface are generally conservative. However, the SNPC methodology response surface selection criteria does not in general result in the base-point having the maximum statistical uncertainty. In Response-8 of Reference-7 and Response of Reference-9, however, SNPC demonstrates that for the case of the DNB and hot-leg 1
saturation lines and the Chapter-4 DNB transients this selection criteria does result in the o
maximum statistical uncertainty and is therefore acceptable for these applications. However, the 6
validity of this criteria is based on the specific selection of uncertainty variables and the assumed uncertainty estimates. If additional variables are added to the DNB and/or hot-leg saturation response surfaces or the uncertainty estimates change, this criteria should be reevaluated.
The proposed response surface selection criteria is not applied to the Chapter-4 low flow and pressurization transients. In Response-2 of Reference-8, SNPC has indicated that for these transients the analyses will be performed at the operating conditions which result in the most conservative transient setpoints.
3.1.3 Axial Shane Reset Function - F(AD The axial shape reset function F(AI) provides the reduction in the AT trips to account for axial power distributions more severe than the design peaking. In Response-12 of Reference-7, SNPC has indicated that the limiting conditions of operation have been included in the
\\
i determination of F(AI). Limiting core power level and axial power shape combinations are i
selected. Severe axial xenon distributions are used to determine the axial power distribution as a function of fuel rod burnup, power level and control rod insertion.
-(
The axial power distributions used to determine the DNB OTAT F(AI) penalty for axial
)
power shape depend on the core power level. The most DNB limiting axial power shape is selected and provides a bounding F(AI) penalty function.
3.1.4 DNB Ouality Limit The DNB correlations are typically only applicable below a specific upper quality limit.
j 1
The SNPC HTP DNB correlation is limited to qurdities below the HTP quality limit while the i
application of the W-3 correlation is limited to qualities less than 15%. While the hot-leg saturation limit line will typically limit the exit quality to less than the HTP limit, this may not 7
h be the case-for other DNB correlations.
However, SNPC has indicated in Response-2
- (Reference-7) that the XCOBRA-IIIC code used to calculate the DNB OTAT limit line provides a specific edit to insure that the DNB correlation is within the applicable quality limit.
3.1.5 Steam Generator Safety Valve Limit Line The Steam Generator Safety Valve (SGSV) limit line provides the lower limit on the DNB i
AT setpoint at high core average temperatures. If the steam generator tube plugging levels are l
unchanged SNPC will use the vendor SGSV limit line. However, if tube plugging levels increase, SNPC should adjust the SGSV line as described in Response-16 of Reference-7.
3.2 Application of the Methodolocy 3.2.1 Fuel ard Plant Desiens The setpoint methodology makes specific assumptions concerning the plant configuration and core design. The proposed methods are applicable to Westinghouse plants which have OPAT and OTAT protective system trips of the form given in Sections 2.1 and 3.1, respectively, of the topical report. The methodology is applicable to fuel designs for which SPNC has NRC-approved methods for evaluating DNB and fuel melt.
?
In Response-1 of Reference-7, SNPC notes three potential simplifications of the t
application of the statistical methodology. First, in plant-specific applications certain variables may be treated conservatively by taking them at their deterministic limits rather than treating them statistically. Second, certain licensing applications will only require the verification of a prior vendors setpoints. In this case, the setpoint coefficients K will not be recalculated but i
rather an independent verification of the setpoints will be performed. Finally, for certain plants 8
i
[
l L
i de total uncertainty allowance associated with the trip channel may be provided by the reactor 1
1
(
vendor. In this case, the individual trip channel component uncertainties will be replaced by a single statistical total uncertainty allowance.
3.2.2 Setooint and Event Anplications The setpoint/ transient analysis methodology is applicable to the setpoint determination for specific reactor trips. The steady-state methods of Chapters 2 and 3 are applicable to the determination of the OPAT and OTAT trip setpoints. The transient analysis methods of Chapter-4 are intended for the analysis oflimiting DNB and pressurization events, and the determination of the OPAT, OTAT and other specified trip setpoints. The transient analysis methods of i
Chapter-4 may be used to either determine or verify these setpoints (Response-18, Reference-7).
3.2.3 Code and Methods Aporovals The proposed setpoint methodology employs several SNPC core performance codes /_
correlations including RODEX2, XCOBRA-IIIC, XTG, and the XNB and HTP DNB correlations. In Response-11 of Reference-7, SNPC has indicated that, except for the HTP DNB correlation, all of these methods have been approved. The HTP DNB correlation is presently under review and it n=t receive NRC approval before it may be used with the statistical i
setpoint methodology.
3.2.4 Allowance for Uncertainties The EMF-92-081(P) methodology performs a statistical determination of the 95/95
~
probability / confidence-level uncertainty allowance for the Reactor Protection System trip _
setpoints.
This analysis requires plant-specific uncertainty estimates for the important measurement uncertainty components. In Response-6 of Reference-7, SNPC has indicated that
'l 9
1
i the uncertainty values for the average coolant temperature, core power level, pressurizer pressure, bistable, channel linearity and reproductibility, and axial flux difference measurement will be provided by the reactor vendor. The plant-specific power peaking uncertainties used will be based on the Technical Specification values.
t I
i t-b i
i r
4
?
i
-i i
t 10 D
4.0 TECHNICAL POSITION The Siemens Nuclear Power Corporation Statistical Setpoint/ Transient Methodology Tropical Report EMF 92-081(P) and supporting documentation provided in References 7 and 8 have been reviewed in detail. Based on this review, it is concluded that the SNPC methodology is acceptable for determining the Reactor Protective System setpoints for Westinghouse type reactors subject to the conditions stated in Section-3 of this evaluation and summarized in the following.
- 1) DNB and Hot-lec Saturation Response Surfaces The validity of the maximum arithmetic difference criteria for determining the most conservative DNB and hot-leg saturation response surfaces is based on the specific selection of uncertainty variables and their estimated uncertainties.
If ad6itional uncertainty variables are added to the DNB and/or hot-leg saturation response surfaces or the uncertainty estimates change, this criteria should.be reeva[uated (Section-3.1.2).
- 2) Low Flow and Pressurization Transient Response Surfaces The response surface base. point selection criteria does not determine the maximum statistical uncertainty for the low flow and pressurization transients. For these transients the base-point should be selected at the operating conditions which j
result in the most conservative transient setpoints (Section-3.1.2).
11
- 3) Steam Generator Tube Plugging The steam generator safety valve line provides the lower limit on the DNB AT
_j setpoint at high core average temperatures. If steam generator tube plugging levels increase the SGSV line should be adjusted (Secdon-3.1.5).
i
approval before it may be used with the statistical setpoint methodology (Section-3.2.3).
)
1 1
12 i
REFERENCES l
1.
" EMF-92-081(P) Statistical Setpoint/ Transient Methodology for Westinghouse Type Reactors, Siemens Nuclear Power Corporation, May 1992," Letter, R.A. Copeland (SNPC) to Thomas E. Murley (NRC), May 29,1992.
2.
" ENC Setpoint Methodology for CE Reactors: Statistical Setpoint Methodology," XN-NF-507(P) (A), Supplements 1 and 2, Exxon Nuclear Company, Richland, WA 99352, i
September,1986.
3.
" Generic Statistical Uncertainty Analysis Methodology," XN-NF-81-22(P) (A), Exxon Nuclear Company, Richland, WA 99352, November 1983.
j i
4.
" Expanded Generic Statistical Uncertainty Analysis Methodology," XN-NF-507(P) (A),
Supplement 1, Appendix A, Exxon Nuclear Company, Richland, WA 99352, September,
'I 1986.
j 5.
" Design Bases for the Thermal Overpower AT and Thermal Overtemperature AT Trip
.I Functions," WCAP-8745(P), Westinghouse Electric Corporation, Pittsburgh, PA 15230, March 1977.
13
i 1
1 6.
" Request for Additional Information EMF-92-081(P), Statistical Setpoint/ Transient Methodology for Westinghouse Type Reactors," Letter, R.C. Jones (USNRC) to R.A.
Copelend (SPC), July 16,1993.
7.
" Responses to NRC Questions on SPC Statistical Transient Methodology," Letter, R.A.
Copeland (SPC) to R.C. Jones (NRC), August 13,1993.
i 8.
" Responses on the SPC Statistical Transient Methodology," Letter, R.A. Capdand (SPC) to L. Kopp (NRC), September 24,1993.
9.
" Response to Question on EMF-92-081(P)," Letter, R.A. Copeland (SPC) to L. Kopp.
(NRC), October 5,1993.
i i
p 14 i
1 9
I William T.
Russell 7
'DYSTRIBUTION
- Centralt Files PDR SRXB R/F AThadani RJones
. LPhillips JConran (MNBB-3701)
LTremper (MNBB-4503)
LSapp (12H-5)
EPoteat (12H-7)
LKopp LKopp R/F