ML20236S058
| ML20236S058 | |
| Person / Time | |
|---|---|
| Issue date: | 07/17/1998 |
| From: | NRC (Affiliation Not Assigned) |
| To: | |
| Shared Package | |
| ML20236S037 | List: |
| References | |
| NUDOCS 9807240105 | |
| Download: ML20236S058 (14) | |
Text
- _ _ _ _ _ _ _
l r%
$t UNITED STATES y"
g j
NUCLEAR REGULATORY COMMISSION o
't WASHINGTON, D.C. 20555-4001
%,.....p ENCLOSURE SAFETY EVALUATION REPORT BY THE OFFICE OF NUCLEAR REACTOR REGULATION RELATING TO TOPICAL REPORTS EMF-1997(P). REVISION 0. "ANFB-10 CRITICAL POWER CORRELATION." AND EMF-1997(P) SUPPLEMENT 1. REVISION 0. "ANFB-10 CRITICAL POWER CORRELATION: HIGH LOCAL PEAKING RESULTS SIEMENS POWER CORPORATION 1 BACKGROUND EMF-1997(P) and its supplement (EMF-1997(P) Supplement 1, Revision 0), describe the methodology behind the application of the ANFB-10 correlation to the Siemens Power Corporation's (SPC's) ATRIUM-10 fuel design, Reference 1. EMF-1997(P) and its supplement provide test data taken specifically at the Siemens test facility at Karlstein, Germany, in development of the application of the ANFB-10 correlation to the ATRIUM-10 fuel design, and to the determination of the associated correlation, " Additive Constants."
The ANFB-10 correlation is a modification of the original ANFB correlation, described in Reference 2. However, the definitions of the associated parameters (dependent and independent) as described in Reference 2, are not changed for the application of the new ANFB-10 to the SPC ATRIUM-10 fuel design. The technical analysis of this ANFB-10 correlation and its exclusive application to the ATRIUM-10 fuel appears below.
The additive constants are determined in accordance with the NRC approved procedure described in References 1 and 2. The uncertainties associated with these additive constants are then used in the approved SPC safety limit methodology for boiling-water reactor (BWR) fuel i
designs. The approved methodology is used to ensure that less that 0.1 percent of the fuel rods j
are.in boiling transitiun during steady-state operation and during anticipated operational occurrences.
h i
1 l
l
}
l 9807240135 9eo717 PDR TOPRP EMVEXXN C
.PDR 3
L l..
2 2 TECHNICAL EVALUATION I
1 I
The ANFB-10 correlation is a new correlation designed and developed to address the critical power behavior of the SPC ATRIUM-10 fuel design The ATRIUM-10 fuel design has part-i-
length fuel rods that were not adequately captured by the base ANFB correlation.
I The ANFB-10 correlation is very similar to and is a modification to the base ANFB critical power i
correlation, Reference 2 which predicts planar average critical heat flux for assemblies with all full-length fuel rods. To develop the ANFB-10 correlation, the base correlation was modified to account for the non-homogeneous effects caused by the presence of the part-length fuel rods in the ATRIUM-10 fuel design. The dependence of the ANFB-10 correlation on the ATRIUM-10 bundle design is included by adjusting the values of the additive constants to match the critical power test data. SPC has performed a series of critical power tests and determined the l
ATRIUM-10 design-specific additive constants. The ANFB-10 correlation, when used with these additive constants, reproduces the measured critical power to within the correlation's standard l
deviation.
The ANFB-10 correlation is an empirically derived expression that is a complex function of the input parameters: local coolant enthalpy, mass flow, and pressure. These input parameters cover the ranges of pressure, mass velocity, and inlet cooling, consistent with expected operating and accident conditions. The correlation is based on local coolant conditions predicted from uniform and non-uniform axial power distribution test data. The correlation also includes correction factors to account for geometry characteristics of the' fuel.
t l
Low-flow and high-flow behavior of the correlation are captured by refining the parameters in the
[
' correlation equations, Reference 2. These parameters address the impacts of the variations in j-
. the local enthalpy from the planar average enthalpy. One of these pararneters is F-effective, which characterizes the fuel rod local behavior, such as enthalpy rise. The additive constants account for the fuel bundle geometry and spacer effects on the critical power behavior of the bundle, Reference 2.
k L
l 1
1
3
\\
2.1 ANFB-10 Data Base and Test Strateay
{
l The ANFB-10 data base consists of datataken at the SPC test facility at Karlstein, Germany.
The test setup consists of electrically heated bundles that are physically the same as the ATRIUM-10 fuel assembly. The tests are designed to reproduce the local conditions typically oresent in a boiling-water reactor (BWR) fuel asrembly, und support the full range of applicability for the ATRIUM-10 correlation.
Three different test programs were developed to accumulate a data base representative of the appropriate statistical requirements for the ATRIUM-10 fuel design. The tests selected and the number of points required were dictated by the requirements of the statistical design of experiment (SDE), Reference 3. This approach ensures that an adequate number of tests is performed and that sufficient data are gathered to perform an appropriate simulation of the behavior of the ATRIUM-10 fuel design.
Both steady-state and transient tests were performed as part of the validation of the ANFB-10 l
correlation. In each case, the tests were designed to include test runs with peaked rods located adjacent to the internal water channel.
The database comprises more than 900 data points taken in a large number of tests performed at the SPC test facility. The database consists of upskew, downskew, and cosine axial power shapes. The database includes peripheral rods, corner rods, rods on the interior of an assembly, and rods adjacent to the water canister (channel), a feature unique to the ATP M design fuel.
The local power peaking pattems were selected to determine the effects of the upskew axial power profiles as compared to the cosine power profiles in several regions of the test bundle.
l Local power peaking data were also collected at the corners, and peripheral rows, as well as around the internal water canister to ensure complete understanding of the fuel critical heat flux (CHF) behavior, particularly in these regions.
4 The internal water canister is a major and unique characteristic of the SPC's ATRIUM fuel design. It replaces a 3X3 matrix of fuel rods. The rectangular canister is designed so that the subchannels around it are regular in size, typical of those addressed by the original ANFB correlation. The test matrix of the ATRIUM-10 fuel at the SPC test facility, was modified to include tests to confirm the behavior of the fuel surrounding the internal water canister. Analysis of test data demonstrated that the ATRIUM-10 fuel did not show any abnormal behavior around the intemal water canister.
2.2 ANFB-10 Data Base and Test Strateav l
The correlation parameter F-effective (F rr) accounts for the local peaking factor effect on the e
bundle critical power. Ferr is constructed in two parts. One part depends solely on the peaking factors of the rod of interest and its immediate neighbors (Ferro), the other part, termed the
" additive constant," accounts for other local effects, such as spacer and bundle geometry effects. These spacer and bundle geometry effects influence the critical power behavior of the entire bundle. Therefore, an offset term is applied to each rod in the bundle, subject to the rod's position in the bundle. This offset term is called the " additive constant." The additive constant can be considered as a flow /enthalpy redistribution characteristic of a particular lattice / spacer design. So the additive constants are unique to a particular fuel design. They are explicitly determined for each lattice / spacer design configuration and are utilized in design calculations for the ATRIUM-10 fuel bundle, Reference 4.
When validating the ability vf the correlation to predict steady-state as well as transient upskew and downskew axial power shapes, only the cosine test data were used in the determination of the additive constants. When these additive constants were used to predict transient behavior thee results were good, thus validating the use of the additive constants in steady-state and transient calculations. The additive constants are experimentally determined from a large data bank representative of the power profile expected during the operational range of the
(
ATRIUM-10 fuel design.
I l
i-5 3 STATISTICAL ASPECTS OF THE ANFB-10 CORRELATION The statistical aspects of the ANFB-10 correlation consist of applying appropriate statistical techniques, References 5 and 6, to the ANFB-10 database. These techniques involve the evaluation.of the 1) distribution characteristics,2) figures showing critical power ratios (CPR) with respect to each characteristic within the correlation,3) descriptive statistics for subgroups of data,4) descriptive statistics for additive constants, 5) additive constants uncertainty, and
- 6) conservatism of the ANFB-10 critical power correlation.
The correlation study examined the CPR in three series of tests. A total of 18 tests were performed: 12 tests were conducted with a chopped-cosine shaped axial power profile,2 tests with a downskew power profile, and 4 tests with an upskew power profile. Each test was performed on a different assembly, and each assembly contained 91 rods, Reference 1. Each I
test was repeated many times (" runs"), such that the entire range of inlet sub-cooling, pressure j
and mass flow were covered. Eight hundred fifty-seven runs were used in the development of SPC's ANFC-10 correlation. As is standard practice, the remaining data points (in excess of 300), were used for validating the ANFB-10 correlation.
The multiplicity of runs within each test was required in order to involve various contributions of j
input factors (inlet flow, inlet subcooling, and pressure). For most of the runs, these factors were selected at random, following standard statistical procedures, References 5. For dryout testing, additional runs were made following a 2-level,3-factor factorial design to ensure that the entire range of interest (including " corner to corner") was represented, Reference 3.
Review of SPC calculations shows that the average CPR appears to be very near 1.0. That ratio is retained without any apparent trend across inlet mass velocity (Mlb/hr-ft), enthalpy (Btu /lbm), pressure (psia), the best estimate of the Ferr, and the axial offset. The overall CPR mean for the 857 runs was calculated to be 0.998. Several measures of variability were presented in the submittal in association with the overall mean. The most conservative measure of variability is the total (unpartitioned) variance calculated from the 857 runs, yielding a variance of 0.375921x10~3 and a standard deviation of 0.193887x10'.
I l
l
6 To evaluate the quality of the correlation, the staff independently calculated a CPR 95/95 upper tolerance limit, Reference 7, for each test, for each profile, and for the entire set of 857 runs.
)
SPC was subsequently asked to submit an independent set of tolerance limit calculations for comparison purposes. Apart from rounding errors and conservative table interpolations, SPC's calculations were in total agreement with NRC's calculations. The highest CPR upper 95/95 tolerance limit does not exceed 1.05 for any test or any grouping of tests. For the 857 runs, treated as one group, the upper tolerance limit is calculated as 1.032. This lin,it is interpreted to mean that one is 95 percent sure that at least 95 percent of the population of runs yield a CPR
. value no higher than 1.032. SPC's calculations also show that for any test or grouping of tests, the percentage of runs that fall below their associated tolerance limits is at least 96.8 percent.
The submittal contained charts and tables reflecting CPR behavior across different mass velocities (Mlb/hr-ff) for individual tests. Although some tests showed higher CPR values associated with high mass velocity, the reverse was true for other tests, and no dependency
. between CPR and mass velocity was apparent.
Another objective of SPC's study that involves statistical consideration was the determination of the additive constant for ATRIUM-10. The additive constant is a statistical adjustment to the Ferg to account for the effect of the rod's geometric position within the assembly. This adjustment has two components: a calculated additive constant and a measure of uncertainty associated with the calculation. In the development of the additive constants, SPC used only the cosine profile data. As is standard practice, the measure of the associated uncertainty was calculated from the entire database, containing cosine, upskew, and downskaw test data, The main contributors to this uncertainty are two sources of v6riability: "within test variability" i
and "between test variability." The within test variability was given as a weighted average, in which the weighting factors were the number of runs per test. The between test variability was l
given as a weighted average of the difference between the Frr for a rod in a test bundle and the 3
l l
average Fer, for the test bundle. Weighting factors using the number of rods observed in boiling transitions and the number of tests were incorporated in the statistical process to determine the total standard deviation of the additive constante The square root of the sum of the squares (the two sources of variability) gave the measure of variability associated with the calculation of h.-
i i
7 the additive constant. In-depth review of the statistical section of the submittal, led the staff to concur with the statistica! methods used and the results obtained by the vendor.
4 ANFB-10 CORRELATION BEHAVIOR The ANFB-10 correlation was tested to ensure smooth functions and no significant discontinuities in its behavior over the entire range of operability of the fuel. Flow, enthalpy, and pressure-dependent functions within the correlation, such as the " Tong Factor" correction for the ATRIUM-10 fuel, were investigated for their behavior over the entire applicable range of the fuel.
A number of tests were conducted to determine the sensitivity of the major functions within the ANFB-10 correlation to flow, inlet -subcooling, pressure variation, Ferg, and axial power shape.
Review of the data, figures, and tables, indicated that the ANFB-10 correlation behaves well over the applicable range of the fuel.
5 ANFB-10 CORRELATION VALIDATION SPC performed several tests to validate the behavior of the ANFB-10 correlation in steady-state and transient events. The validation database consisted of more than 300 steady-state data points that were not included in the correlation database. The data were collected from tests conducted on an ATRIUM-10P assembly that contained more part-length fuel rods than are usually found in a typical ATRIUM-10 assembly. These tests were conducted to demonstrate the ability of the ANFB-10 correlation to capture the effects of the part-length rods, as well as the correlation agreement with the data. The predicted ANFB-10 correlation critical power vr. sus the measured critical power for these tests showed very good agreement. Two sets of transient tests were performed as part of the validation process. Both tests were designed to examine l
peak rods around the intemal water canister. The first test had rods with a chopped cosine-shaped axial power profile and the second had rods with an upskew cial power shape. Another purpose of these tests was to validate the concept that the additive constants can be derived from steady-state cosine tests and applied to other axial shapes under transient conditions.
Tests to simulate the load rejection with no bypass (LRNB) events were performed by varying power, pressure, and flow decay. Power forcing functions were programmed to produce l-
\\
i~
8 l
transient heat flux on the surface of the rod typical of an LRNB event. Parameters monitored during the tests were power, inlet flow, system pressure, inlet temperatures, and cladding temperatures.
The transient thermal-hydraulic code, XCOBRA-T, References 8 and 9, was used to predict the test results using the ANFB-10 steady-state critical power correlation. XCOBRA-T calculated the fluid conditions at a specified time step. The critical heat flux (CHF) was calculated at each axial position and time step, then compared to the corresponding measured rod heat flux at the surface of the rod. The ratio of the calculated critical heat flux to the measured rod heat flux is defined to be the critical heat flux ratio (CHFR). When this ratio is unity, it is referred to as the minimum critical heat flux ratio (MCHFR) and it signifies " boiling transition" in a transient event.
Comparison of measured and calculated time-to-boiling transitions, for cosine and upskew transient tests, showed that the XCOBRA-T calculated time-to-boiling transition values were conservative when compared to actual boiling transition time. This validation confirmed the use of the steady-state ANFB-10 correlation and the associated additive constants in evaluating transient events.
5.1 COMPARISON OF ANFB AND ANFB-10 CRITICAL POWER CORRELATION FOR THE ATRIUM-10 FUEL DESIGN Critical power data can be represented by the experimental critical power ratio (ECPR). ECPR I
is the ra6o of the critical power predicted by the correlation for the test conditions to the critical power measured in the test. The original ANFB critical power correlation, Reference 2, was shown to be non-conservative when predicting critical power for upskew axial power shapes.
Critical power versus mass flow indicated that for ATRIUM-10 fuel, the original ANFB correlation l
has a flow dependence that is a function of mass flow. The lower the mass flow, the further from j
unity the ECPR became. In essence, ANFB has a flow-dependent bias and over predicts the critical power for upskew axial profiles at low mass flow. Review of the upskew tests from both
)
the ANFB and ANFB-10 correlations shows substantialimprovement of the ANFB-10 correlation i
over the ANFB correlation. Comparison of the data showed that the upskew bias that existed in the ANFB correlation has been substantially removed. The staff agrees with the SPC's
)
conclusion that no bias of concern remained when the ANFB-10 correlation was used to l
I
9 evaluate the selected upskew data. Additionally, when all the upskew test data were included, no observable trend existed.
6 ANFB-10 HIGH LOCAL PEAKING FACTORS During a recent inspection of SPC, Reference 10, the NRC staff expressed concern over a lack of sufficient dryout test data for bundle local peaking factors greater than 1.3, included in the development of the original ANFB correlation. To address this issue, SPC conducted a series of tests (encompassing many runs) to demonstrate the ability of the ANFB-10 correlation to predict fuel behavior beyond the 1.3 peaking factor. Specifically, the ECPRs as calculated by the ANFB-10 correlation for test data with peaking factors up to 1.3, were compared with the ECPR
\\
as calculated by the ANFB-10 correlation for test data with peaking factors beyond the 1.3.
I Three tests were conducted: two with peaking factors below 1.3 and one with a peaking factor higher than 1.3. The latter test was conducted to validated the fact that the niean ECPR and its associated variance for data with peaking factors higher than 1.3 are consistent with the ECPRs and their associated variance for data with peaking factors below 1.3. The objective was to show that the ECPR variance is independent of the local peaking factors, and that no additional variance is necessary for fuel rods with local peaking factors higher than 1.3.
6.1 Test Description and Results of Evaluation The validation test was performed at the SPC test facility described above. The test was done to demonstrate that the dryout behavior of the ATRIUM-10 assembly with high local peaking was adequately predicted by the ANFB-10 correlation with additive constants calculated as described above, and that no additional additive constant uncertainties were required for fuel rods with high local peaking factors.
L The validation test was performed on a greater than 1.3 peak-to-average chopped-cosine axial power profile with the local peaking pattern typical of a limiting fuel rod. The data acquisition process utilized during the validation test was the same as the process used for all the other data acquired.
10 Although only one validation test was conducted, the many runs that were performed resulted in an adequate database. The database consisted of data collected on previous tests on the same rod at different peaking factors. The validation data were collected on the same fuel rod located in the same location as before, which was on the outer row of a 3x3 corner. This location was chosen because the critical power for the 3x3 corner is, in general, higher than the critical power in the 4x4 or the 3x4 corners of the assembly. The location was also chosen because it would serve to provide the third data point for the same rod at the same location. In addition, the neutronic design of the bundle is such that a peripheral row rod is usually limiting, j
l l
6.2 Determination of the ECPR for the Hioher Peakina Factor The value of the Fm for the higher peaking factor was determined by using the local peaking pattern for the validation test in combination with the ANFB-10 additive constants. This in turn, was used to evaluate the data with the ANFB-10 critical power correlation and to determine the ECPRs for the validation test data. The ECPR values for the validation test data, along with equivalent values from previously conducted tests on the same rod with lower peaking factors,
.was then plotted and tabulated. Vendor-submitted overlay plots and tabulated data show the similarity of the data for the three tests: two with a local peaking factor of 1.3 or less, and one with a local peaking factor greater than 1.3. Graphs of critical power as a function of increasing rod local peaking for the three tests showed a linear behavior in the more conservative direction, thus demonstrating the correlation's trend to be consistent on rods with varying local peaking
. factors. - The results of the validation test indicated that the higher the local peaking, the more conservative the calculated critical power ratio became. The standard deviation remained about the same as that for an individual test. The three points were very consistent in that they form a straight line, demonstrating the point that the local peaking factor calculation (algorithm) remained consistem and well behaved for local peaking factors below and above 1.3.
Examination of the data gave no indication that the trend should not continue. Indeed, statistical analysis of the validation test data demonstrated that the variance of the ECPR does not depend
- on maximum local peaking value. The difference in the predicted additive constant for the local peaking factors below and above 1.3 was well within the additive constant uncertainty for all rod positions. The staff concurs with this conclusion. The staff also concurs with the vendor's conclusion that the analysis of the validation data has demonstrated that the ECPR uncertainty
n O
D 11 is not affected by local peaking and, therefore, there is no need for augmenting the additive constant uncertainty caused by local peaking greater than 1.3, up to 1.5. The value of 1.5 local peaking factoris a minor extrapolation of the experimently determined local peaking value of 1.45, Reference 1.
Although local peaking factors may exceed 1.5 in controlled bundles, these bundles by definition are not limiting bundles, consequently, they do not factor in the calculation of the Minimum Critical Power Ratio (MCPR) safety limit. If however, in the process of calculating the MCPR safety limit, the local peaking factor of 1.5 may be exceeded, an additional additive constant uncertainty of 0.018 will be imposed on a rod by rod basis. The additional uncertainty of 0.018 is
. calculated as an upper 95 percent confidence limit (obtained from the chi-squared distribution) of the variability component of the additive constant for the three tests, Reference 12. These i
conditions have been agreed upon by Siemens Power Corporation.
7 NONCONFORMANCE ISSUES The submittals, as documented in Reference 1, are SPC's corrective action response to Part (1) of Nonconformance 99900081/97-01 as stated in Attachment 11 of SPC's letter to the NRC,
. dated February 24,1998. The Nonconformance stated that:
(a) As used to determine the safety limit critical power ratio (SLCPR) and the operating limit minimum critical power ratio (OLMCPR) for the ATRIUM-10 fuel assemblies for Susquehanna Unit 2 Cycle 9, the ANFB correlation included a nonconservative flow bias and, therefore, was outside the NRC-approved SPC methodology.
(b) The application of the ANFB correlation to the ATRIUM-10 fuel assemblies for Susquehanna Unit 2 Cycle 9 showed that the local peaking factor (F) was t
[
9reater than 1.3, which placed it outside the NRC-approved range for the ANF8 r
correlation of less than 1.3.
d e
12 In response to this Nonconformance, SPC issued an interim methodology using the ANFB correlation with a flow-dependent MCPR safety limit. NRC approval was obtained for this interim
{
methodology for application to the first cycle of the ATRIUM-10 fuelin the Susquehanna Unit 2 plant, Reference 11. This interim methodology also addressed NRC concerns related to the extrapolation of the ANFB predictions for local peaking higher than 1.3. The licensee (Susquehanna) submitted appropriate Technical Specifications changes to address safe
]
operation of the units with this interim methodology.
]
The NRC staff concludes that with the submittal of EMF-1997 and its Supplement, the vendor has shown that the application of the ANFB-10 correlation will provide a rigorous treatme,'t of the ATRIUM-10 performance and will eliminate the flow-dependent MCPR safety limit. The submittals also demonstrated that the correlation does not require any alteration or special consideration for application to high local peaking, up to a local peaking factor of 1.5. Thus, with the submittal of EMF-1997 and its Supplement, all problems identified in the inspecion report (Nonconformance 99900081/97-01) related to the dryout methodology for ATRIUM-10 fuel have been addressed.
8 CONCLUSION The staff has reviewed the analyses in Topical Reports EMF-1997(P), Revision 0, "ANFB-10 Critical Power Correlation Application," and its Supplement, EMF-1997 (P) Supplement 1, Revisior 0, "ANFB-10 Critical Power Correlation: High Local Peaking Results," and concludes that on the basis of its findir,gs (presented above), Topical Reports EMF-1997(P) and its Supplement, are acceptable for licensing applications, as per SPC's agreement, subject to the following conditions:
- 1. The ANFB-10 correlations (as described in these submittals, Reference 1) are applicable to Siemens Power Corporation ATRIUM-10 fuel with a design local peaking factor up to 1.5.
j 13
- 2. If however, in the process of calculating the MCPR safety limit, the local peaking l
factor of 1.5 is exceeded, an additional uncertainty of 0.018 will be imposed on a rod by rod basis.
- 3. ANFB-10 correlation range of applicability:
{
Pressure (psia) 571 to 1415 Inlet Mass Velocity (Mlb/hr-ff) 0.115 to 1.5 Inlet Subcooling (Btu /lbm) 5 to 149 l
f
e e
14 8 REFERENCES 1.
Letter from H.D. CL,;9t to the U.S. Nuclear Regulatory Commission, submitting Topical Report EMF-1997(P), Revision 0, October 30,1997, and Letter from H.D. Curet, submitting Topical Report EMF-1997(P) Supplement 1, Revision 0, January 29,1998.
2.
Letter from R.A. Copeland, Transmittal of (A) version of ANF-1125(P) to the U.S. Nuclear Regulatory Commission, April 27,1990.
3.-
"StatisticV Methods for Nuclear Material Management, " NUREG/CR-4604, PNL-5849, December i988.
4.
Letter from J.F. Malley to the U.S. Nuclear Regulatory Commission, " Request for Additional Information (RAl) to Topical Report EMF-1997(P), ANFB-10 Critical Power Correlation,"
Revision 0, May 7,1998.
5.'
G.J. Hahn, and S.S. Shapiro, " Statistical Models in Engineering," Wiley,1967.
6.
H.W. Lilliefors, *On the Kolmogorov Test for Normality with Mean and Va,iance Unknown,"
Journal of American Statistical Association, Vol. 62, June 1967.
7.
R.E. Odeh and D.B. Owen, " Tables for Normal Tolerance Limits," Table 1, Marcel Dekker, ink,1990.
8.
"XCOBRA-T: A Computer Code for BWR Transient Thermal-Hydraulic Core Analysis," XN-NF-84-105(P)(A), Volume 1 and Volume 1, Supplements 1 and 2, Exxon Nuclear Company, Richland, WA 99352, February 1987.
j l
9.
"XCOBRA-T: A Computer Code for BWR Transient Thermal-Hydraulic Core Analysis, Void
'j Fraction Model Comparison to Experimental Data," XN-NF-84-105(P)(A), Volume 1, Supplement 4, Advanced Nuclear Fuels Corporation, Richland, WA 99352. February 1987.
- 10. Letter, Samuel J. Collins to David G. McLees. " Demand for Inforrnation and Notice of Nonconformance (Inspection Report 99900081/97-01)," EA97-495, datt i October 27,1997.
- 11. Letter from Chester Poslusny, Senior Project Manager, NRC, to Mr. Robert G. Byram, Pennsylvania Power and Light Company, May 7,1997.
- 12. Letter from J.F. Malley to the U.S. Nuclear Regulatory Coma.ission, " Additional Information to the Tonical Report EMF-1997(P), " ANFB-10 Critical Power Correlation," Revision 0, June 8, W98.
l l
l
.