ML17158C152
| ML17158C152 | |
| Person / Time | |
|---|---|
| Site: | Susquehanna  |
| Issue date: | 05/14/1997 |
| From: | Poslusny C NRC |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| NUDOCS 9705150193 | |
| Download: ML17158C152 (26) | |
Text
May 14, 1997 NOTE TO:
Document Control Desk FROM:
Chet Poslusny
SUBJECT:
NON-PROPRIETY DOCUMENTS FROM PENNSYLVANIA POWER
- 5. LIGHT CO.
Docket No: 50-388 Enclosed are three non-proprietary versions of enclosures to documents previously sent to the Commission and already on the docket.
They are enclosure to PLA 4605 dated 4/9/97, enclosure to PLA 4595 dated 3/27/97, and enclosure to PLA 4611 dated 4/16/97.
Please place them on this docket and provide a copy to the PDR.
If you have any questions please call me on 415-1402.
Thanks.
>~000~
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII 9705i50i93 9705l4 PDR ADQCK 05000388 P
N-P~Pai&-~ "f NCLOSURE B TO PLA-4595 Page1of14 g>jp) 3/~g/97 PP&L Response to NRC Request for Additional Information on EMF-97-010, Rev. 1 Question 1:
Provide details on the calculation of transient BCPRs incorporating the flow dependence.
Explain why the adjustment of hCPR described in EMF-97-010, Rev.
1 is acceptable as opposed to re-running the transients and directly calculating a bCPR which accounts for the flow dependence.
Discuss the impact on anticipated operational occurrences and MCPR related accidents.
~Res onse1:
A fundamental element of the Current Licensing Basis for the Susquehanna Units is that 99.9%
of the fuel rods are expected to avoid boiling transition during normal operation or Anticipated Operational Occurrences (AOOs).
As discussed in the response to Question 2, the MCPR Safety Limit represents a set of conditions at which 99.9% of the fuel rods are expected to avoid boiling transition.
The transient analyses calculate conservative values of the change in Critical Power Ratio (dCPR) for limiting AOOs.
The limiting dCPRs from the analyses are added to the calculated MCPR Safety Limitto produce the MCPR Operating Limits as functions of core power and core flow.
PP8L has performed reload licensing analyses for several reloads on each unit to calculate dCPRs for the limiting AOOs using NRC approved methods (References 1 and 2).
The analyses performed for each reload include (Reference 2):
- 1. Loss of Feedwater Heating
- 3. Generator Load Rejection Without Bypass
- 4. Turbine Trip Without Bypass
- 5. Rod Withdrawal Error
- 6. Recirculation Flow Controller Failure
- 7. Rotated Bundle
- 8. Mislocated Bundle The ANFB correlation applied to ATRIUM-10 fuel has been shown to have a dependence on bundle flow in its prediction of critical power.
Since some AOOs exhibit bundle flow changes, a
method of calculating dCPR is required such that the hCPR is sufficient to ensure that the MCPR Safety Limit is not exceeded during the event.
This methodology is outlined in Section 6.4 of EMF-97-010, Rev. 1. Details of the process are discussed below.
t=NCLOSURE B TO PLA-4595 Page 2 of 14 PP8L Response to NRC Request for Additional Information on EMF-97-010, Rev.
1 Calculation of Ad'usted ACPR The transient methodology to calculate an adjusted hCPR described in Section 6.4 of Reference 3 can be implemented in a number of ways.
One approach to implement this methodology is to incorporate the flow bias in calculated MCPR directly into the computer codes used to calculate hCPR and run the analyses.
The resulting calculated dCPRs would thus directly contain the adjustment.
The process for this approach is the following. First, the event is analyzed and the CPR is calculated as a function of time. The calculated CPR at each time will be adjusted to account for the flow dependence.
The adjustment is performed using the ECPR values from Table 6.2 of Reference 3, except that values of ECPR less than 1.0 (i.e., the correlation is conservative) are treated as equal to 1.0.
This approach is consistent with the usage of the flow dependence for calculating MCPR Safety Limit, described in the response to Question 2.
For example, if the unadjusted CPR at time t is 1.15, and the ECPR corresponding to the bundle flow at time t is 1.05, the adjusted CPR is calculated as:
CPR3djUS(pd
= 1.15 / 1.05
= 1.095 The minimum adjusted CPR during the event is subtracted from the time zero adjusted value to produce the "adjusted dCPR" described in Section 6.4 of Reference
- 3. PP8L intends to implement this more direct approach to calculate the adjusted hCPR at a later date.
Different conservative approaches are also available using data from existing (unadjusted) transient analyses.
These "off-line" approaches are based on the fact that the critical power correlation does not affect the transient physical response of the hot bundle (flows, heat fluxes, qualities, pressures, etc.).
The critical power correlation is, thus, applied after the physical response of the hot bundle is calculated.
Using previously calculated hot bundle responses (bundle flow, heat fluxes, etc.), the time dependent change in hot bundle flow rate at the position of minimum Critical Heat Flux Ratio (CHFR) is determined.
Using the flow dependence of ECPR, a CPR adjustment based on the flow decrease vs. time is calculated.
Since SPC's transient analysis codes have the ability to calculate CPR at each time step, the hCPR is determined in a similar manner as in the direct approach discussed above.
Specifically, the calculated CPR at each time will be adjusted (i. e., divided) by the ECPR corresponding to the bundle flow at that time. The minimum adjusted CPR during the event is subtracted from the adjusted time zero value to produce the "adjusted bCPR".
Since PP8 L's transient analysis codes do not currently have the ability to calculate CPR at each time step, a more conservative off-line approach is used.
This conservative off-line approach is based on a conservative determination of the time of minimum CPR.
NCLOSURE B TO PLA-4595 Page 3 of 14 PP&L Response to NRC Request for Additional Information on EMF-97-010, Rev. 1 Once a scram occurs during a limiting transient, the bundle heat flux decreases rapidly, which tends to make the CPR increase rapidly. At some time after a scram, the tendency of the CPR to increase due to decreasing heat flux will dominate the tendency for CPR to decrease due to the adjustment produced by the flow decrease producing a minimum value for the event.
Based on the heat flux and bundle flow behavior for the event, this time of minimum CPR can be conservatively determined. A conservatism in PP8L's off-line approach is that the increase in CPR after the time of the original minimum due to the decreasing heat flux is not credited (see Figure 6.3 of EMF-97-010, Rev. 1).
The calculated CPR at time zero and at the time of minimum are adjusted (i. e., divided) by the ECPRs corresponding to the flow rates at their respective times.
The adjusted minimum CPR at the conservatively determined time of minimum CPR is subtracted from the adjusted time zero value to produce the "adjusted d CPR".
PP8L will be using this conservative off-line approach to calculate adjusted bCPRs for establishing MCPROLs in the initial Core Operating Limits Report (COLR) for U2C9.
As indicated earlier, PP8L intends to modify its transient analysis codes and utilize the direct approach described above (i.e., flow dependent adjustment to CPR calculated directly during an event).
The adjusted bCPRs calculated with this direct approach will subsequently be used to determine revised MCPROLs and revise the U2C9 COLR. PP8L will inform the NRC of the implementation of this direct approach prior to the issuance date of the revised U2C9 COLR.
Both the direct approach and the "off-line" approaches described above accomplish the calculation of adjusted bCPR as described in Section 6.4 of Reference 3 and result in dCPRs which account for the flow dependence exhibited by ANFB.
Since the effect of the Additive Constant uncertainty is accounted for in the Safety Limit methodology, no additional conservatism in the d,CPR methodology needs to be introduced because of the small increase in uncertainty at low flow conditions (see response to Question 5).
The Additive Constant uncertainties are comparable for co-resident ATRIUM-10 and 9x9-2 fuel types.
For 9x9-2 fuel the Additive Constant uncertainty over the entire flow range is','or ATRIUM-10fuel the Additive Constant uncertainty is for flows Mlb/hr.
For the few ATRIUM-10 rods with the Additive Constant uncertainty As discussed in the response to, Question 5, an increase in the low flow Additive Constant uncertainty from results in an increase of < 0.01 in the MCPR Safety Limit. Therefore, the effect of this small variation in low flow'uncertainty will have a negligible impact on the final MCPROLs.
Generation of MCPR 0 eratin Limits The MCPR Operating Limits are determined by adding the bCPRs from the transient analyses to the MCPR Safety Limit. The MCPR Operating Limits for the Susquehanna Steam Electric Station Units are determined as a function of core flow and core power.
The Operating Limit at a given power/flow condition is the greater of the flow dependent and power dependent MCPR Operating Limits at the relevant conditions.
CLOSURE B TO PLA-4595 Page 4 of 14 PP8 L Response to NRC Request for Additional Information on EMF-97-010, Rev. 1 For the new methodology, this process is essentially unchanged except that the MCPR Safety Limit is a function of core flow.
For a transient analysis at a specific core power/core flow condition, the MCPR Safety Limit for that core flow is added to the adjusted dCPR for that transient.
Take the example of a limiting GLRWOB analyzed at 102% power / 87 Mlb/hr. Ifthe MCPR Safety Limit at this flowis, and the adjusted dCPR is, the MCPR Operating Limit would be, which would be entered on the flow dependent MCPR Operating Limit curve in the Core Operating Limits Report (COLR). The cycle specific transient analyses are utilized in this way to produce the power and flow dependent MCPR Operating Limits. Therefore, the MCPR Operating Limits are derived as a function of core power and core flow, as before.
The flow dependent Safety Limit and adjusted dCPRs will result in significantly higher MCPR Operating Limits for U2C9.
Using the old (unadjusted) methodology for the example given
- above, the MCPR Safety Limit would be, and the unadjusted bCPR would be approximately,'resulting in an Operating Limitof Accidents The only MCPR related accident is the Single Loop Pump Seizure.
To analyze this event, the bCPR is adjusted to account for the change in flow (Section 6.4 of Reference 3).
This significantly increases the calculated dCPR since there is a significant flow decrease for this event.
The next step is to calculate the number of fuel pins in boiling transition (assumed to fail) using the Safety Limit methodology discussed in Section 6.3 of Reference 3 and described in the response to Question 2.
For this calculation, the flow dependence of ECPR and the increased uncertainties at low flows and increased uncertainty for local peaking factors greater than are used.
Thus, the flow dependence in ANFB for ATRIUM-10 fuel is directly incorporated into the calculated number of pins in boiling transition and, hence, the calculated offsite dose.
The results will be used to establish MCPR operating limits for single loop operation which assure that the offsite dose is a small fraction of 10CFR100 limits, as required.
Question 2:
Describe the details of the new methodology for computing the flow dependent MCPR Safety Limit. How is the process different from the standard methodology (non-flow dependent limit)'
Describe how the flow bias is incorporated into the calculation.
Res onse2:
Excessive overheating of the fuel rod cladding can result in cladding damage and the release of fission products.
In order to protect the cladding against overheating due to boiling transition, the Thermal Power, High Pressure and High Flow Safety Limits (Sections 2.1.2 and 3.4.1.1.2 of the Susquehanna SES Unit 2 Technical Specifications) were established.
CLOSURE B TO PLA-4595 Page 5 of 14 PPRL Response to NRC Request for Additional Information on EMF-97-010, Rev. 1 NUREG-0800, Standard Review Plan Section 4.4, specifies an acceptable, conservative approach to define this Safety Limit.
Specifically, a Minimum Critical Power Ratio (MCPR) value is specified such that at least 99.9% of the fuel rods are expected to avoid boiling transition during normal operation or anticipated operational occurrences.
Boiling transition is predicted using a correlation based on test data (i.e., a Critical Power Correlation). The Safety Limit MCPR calculation accounts for various uncertainties such as feedwater flow, feedwater temperature,
- pressure, power distribution uncertainties, and uncertainty in the Critical Power Correlation.
Currently, the ANFB Critical Power Correlation is used to predict boiling transition for SPC fuel at Susquehanna.
The proposed cycle speciTic Safety Limit MCPR values (two-loop and single-loop) were calculated using SPC's NRC approved licensing methods, as modified by EMF-97-010, Rev. 1.
The new methodology addresses an observed flow dependence in the ANFB critical power predictions for ATRIUM'-10fuel as well as an increased correlation uncertainty for high local peaking factor rods.
The relation of predicted critical power to bundle flow, the flow dependent correlation uncertainties, and the increased uncertainty for high local peaking factor rods are derived directly from ATRIUM~-10 critical heat flux test data and are used as input to the Safety Limitanalyses.
To address the flow dependence, safety limit calculations are performed at various core flows to generate a core flow dependent MCPR Safety Limit. The Safety Limit MCPRs (two-loop and single-loop) are defined as functions of core flow and assure that at least 99.9% of the fuel rods are expected to avoid boiling transition during normal operation or anticipated operational occurrences.
Details of the process are provided below.
The computer code is used to apply the SPC MCPR Safety Limit methodology described in ANF-524 Revision 2. The version of is used to apply the revised MCPR Safety Limit methodology described in EMF-97-010 Revision 1. The MCPR safety limit is determined by a statistical convolution of the uncertainties associated with the calculation of thermal margin. A Monte Carlo procedure simulates a variety of reactor states around a base state, where the reactor states for the various trials are determined by varying the nominal reactor conditions according to the magnitude of their uncertainties.
The calculation procedure can be divided into five parts; 1)
Initialization (Establish Nominal Conditions) 2)
Outer Loop (Monte Carlo Trials) 3)
Inner Loop (Fuel Assembly Calculations) 4)
Rod Loop (Rods in Boiling Transition Calculation) 5)
Evaluation of Results
0 NCLOSURE B TO PLAP595 Page 6 of 14 PP&L Response to NRC Request for Additional Information on EMF-97-010, Rev. 1 In the first part of the calculational procedure (Initialization),
. This establishes the nominal conditions for the analysis. A basic principle in the initialization is that the MCPR calculation performed to establish the nominal conditions is consistent with how MCPR calculations are performed during core monitoring with POWERPLEX-II. This means the F-effective for the limiting MCPR assembly is calculated using local peaking factors which do not account for channel bow.
Similarly, in the revised methodology for ATRIUM-10 fuel, the flow bias in ANFB is not accounted for. No changes are introduced in the first part of the calculational procedure by the revised methodology.
In the second part of the calculational procedure (Outer loop), the core wide conditions for the Monte Carlo trial are calculated.
The nominal values for are perturbed based on their corresponding uncertainties to calculate the perturbed values used in the Monte Carlo trial. No changes are introduced in the second part of the calculational procedure by the revised methodology.
In the third part of the calculational procedure (Inner loop), the assembly conditions for the Monte Carlo trial are calculated.
The nominal values for are perturbed based on their corresponding uncertainties to calculate the perturbed values which will be used in the Monte Carlo trial. The F-effectives are then calculated from the perturbed local peaking factors and additive constants.
These F-effectives are used in the fourth part of the calculational procedure to determine if a rod is in boiling transition.
In the fourth part of the calculational procedure (Rod Loop), the number of rods in boiling transition are calculated. The flow dependent ECPR (ratio of critical power predicted by ANFB divided by the critical power measured in the test) is implemented in the fourth part of the calculational procedure.
The result of each Monte Carlo trial is a calculated number of fuel pins in boiling transition.
Following the desired number of trials, the results are evaluated.
In the fifth part of the calculational procedure, the evaluation of results, the number and fraction of rods in the core which are predicted to be in boiling transition with 95% confidence are determined.
No changes are introduced in the fifth part of the calculational procedure by the revised methodology.
Question 3:
Provide the technical bases for the uncertainties for local peaking factors )
provide the technical bases for the flow dependent uncertainties used.
Also
NCLOSURE B TO PLAR595 Page 7 of 14 PP8 L Response to NRC Request for Additional Information on EMF-97-010, Rev.
1 The additional Additive Constant uncertainty for local peaking factors As seen in Table 6.1 of Reference 3, three other rods experienced dry out during two tests at different local peakings.
In each test case, the difference in the Additive Constants derived from individual tests was the same as or less than the Additive Constant uncertainty derived from the entire cosine test data base.
The flow dependent uncertainties are based on the Additive Constant uncertainties determined during the cosine series of tests, STS-17.
When the Additive Constant uncertainties are calculated in the cosine tests the effects from all peaking patterns are considered.
This includes low bundle flow tests, system pressure and the effects of combining peaking patterns.
When the upskew dry out test, STS-29.1, was binned on flow the Additive Constant uncertainties for each individual bin decreased.
Therefore, the more conservative Additive Constant uncertainty originally derived based on all the cosine tests, is used in Safety Limit analyses with the ANFB flow dependence.
Question 4:
Describe how the flow dependent MCPR Safety Limit affects plant monitoring and protection systems.
Specific items to consider are the process
In addition, consider the impact on the Susquehanna Steam Electric Station power/flow map. Are the bases for any of these items affected?
~Res onse 4:
The implementation of a flow dependent MCPR Safety Limit (MCPRSL) was evaluated to determine the potential effect on the operation or bases of SSES Technical Specification plant monitoring and protection systems.
In addition, a review of the SSES Core Monitoring System and SSES Power/Flow map was also performed.
The result of these evaluations is that the implementation of a flow dependent MCPRSL does not affect the operation or bases of any SSES Technical Specification plant monitoring or protection system, the SSES Core Monitoring Software System (I. e., process computer), or the SSES Power/Flow map.
NCLOSURE B TO PLA-4595 Page 8 of14 PPBL Response to NRC Request for Additional information on EMF-97-010, Rev.
1 The following systems are discussed in detail below;
~
Rod Block Monitor (RBM) System,
~
Control Rod Program Controls (Rod Sequence Control System (RSCS) and Rod Worth Minimizer (RWM)),
~
Control Rod System,
~
APRM Flow Biased Scram and Rod Block System, and
~
Core Monitoring Software System (CMSS).
The effect of a flow dependent MCPRSL on the SSES Power/Flow map was also evaluated and is discussed below.
Rod Block Monitor S stem The basis for the RBM system is to prevent fuel damage in the event of an erroneous rod withdrawal from locations of high local power during reactor power operation.
Control rod drift events at SSES have demonstrated that the RBM system may not, in certain situations, initiate a rod block for a control rod which drifts out of the core. In response to these events, PP8L committed to performing the cycle specific Control Rod Withdrawal Error (CRWE) Analysis without taking credit for the mitigation of the event by the RBM system. Therefore, to prevent fuel damage in the event of an erroneous rod withdrawal the CRWE analysis is performed assuming that the error control rod is withdrawn from the full in to the full out position. The calculated dCPR for the CRWE analysis is added to the MCPRSL to establish a MCPR Operating Limit (MCPROL) which maintains the required margin of safety.
The bCPR for the U2C9 CRWE analysis is adjusted for the flow dependence of ANFB as discussed in the response to Question 41 above.
Although the RBM system is not directly credited in any licensing analyses, it is currently in use at SSES functioning as a defense in depth mechanism to mitigate the consequences of erroneous control rod withdrawals.
The RBM does not function as a system which is utilized to maintain the required margin to safety limits or prevent fuel damage.
Therefore, the implementation of a flow dependent MCPRSL does not affect the operation or basis for the RBM system.
Control Rod Pro ram Controls The RSCS and the RWM enforce the established control rod withdrawal and insertion sequences to assure that the maximum insequence individual control rod or control rod notch reactivity worth does not insert an amount of positive reactivity which could result in a peak fuel enthalpy greater than 280 cal/gm in the event of a control rod drop accident (CRDA). The RSCS and RWM logic functions at a reactor power less than or equal to the low power setpoint of 20% of rated thermal power.
Therefore, the basis for the RSCS and the RWM is to mitigate the consequences of a CRDA initiated from 20% of rated thermal power or below. The analysis of the CRDA does not utilize the ANFB critical power correlation.
NCLOSURE B TO PLA-4595 Page 9 of 14 PP&L Response to NRC Request for Additional Information on EMF-97-010, Rev. 1 The RSCS and the RWM do not function as the mitigating system to protect against fuel failure due to boiling transition and, as such, are not affected by the implementation of a flow dependent MCPRSL.
Control Rod S stem The control rod system is defined as the control rods and the support systems (e.g., hydraulic, electrical) necessary to enable the control rod system to perform its intended function.
The bases for the control rod system are to ensure that adequate core Shutdown Margin (SDM) is maintained and to bring the reactor subcritical at a rate fast enough to prevent the violation of the MCPRSL during an Anticipated Operational Occurrence (AOO).
The ability of the control rod system to ensure adequate core SDM is related to the overall capability to insert control rods into the reactor core.
The capability to insert control rods into the reactor core is dependent upon mechanical and thermal hydraulic control rod, system related parameters such as system pressure and flow and is unaffected by the analytical results of the MCPRSL analysis.
The rate at which the control rods are inserted directly affects the rate at which the reactor is brought subcritical thus preventing the violation of the MCPRSL. The rate at which the control rods insert is dependent upon mechanical and thermal hydraulic control rod system related parameters such as system pressure and flow and is unaffected by the analytical results of the MCPRSL analysis.
Therefore, implementation of a flow dependent MCPRSL does not affect the operation or bases of the control rod system.
APRM Flow Biased Scram and Rod Block S stem Although the APRM Flow Biased Scram and Rod Block System is not directly credited in any licensing analyses they are currently in use at SSES functioning as defense in depth mechanisms to mitigate the consequences of plant transients. The setpoints for this system are defined in the SSES Technical Specifications and limit plant operation to the region covered by the cycle specific plant licensing analyses.
The setpoints define an operating region which is unaffected by the cycle specific licensing analyses.
This operating region is an input to the cycle specific licensing analyses establishing the initial conditions for some licensing events.
The APRM flow biased scram and rod block system does not function as the primary system which mitigates the consequences of an event, maintains the required margin to safety limits, or prevents fuel damage.
Therefore, the APRM flow biased scram and rod block system is unaffected by the implementation of a flow dependent MCPRSL.
NCLOSURE B TO PLAP595 Page 10 of 14 PP8L Response to NRC Request for Additional Information on EMF-97-010, Rev.
1 Core Monitorin Software S stem The primary function of the Core Monitoring Software System (CMSS) is to monitor the performance of the plant such that plant operation is consistent with applicable licensing analyses.
To ensure operation consistent with applicable licensing analyses cycle specific operating limits are developed and input to the CMSS.
The performance of the plant is measured against these cycle speciTic operating limits via calculations performed by the CMSS.
Therefore, the CMSS and the cycle specific operating limits collectively ensure that the plant operation is consistent with applicable licensing analyses.
PP8L utilizes the SPC POWERPLEX-II (PPX-II) CMSS to monitor plant performance during power operation.
The cycle specific input for the PPX -II CMSS consists of data which describes the neutronic and thermal hydraulic characteristics of the fuel assemblies and reactor core.
In addition, the cycle specific input includes the applicable fuel type specific fuel mechanical design limits (Linear Heat Generation Rates),
LOCA limit (Average Planar Linear Heat Generation Rate),
and the cycle specific MCPROLs.
The MCPROLs are developed based on the combination of the cycle specific hCPRs calculated by PP8L and the MCPRSL calculated by SPC.
For U2C9, the MCPROLs will be developed based on the combination of the cycle specific adjusted dCPRs calculated by PP8L (as discussed in response
¹1 above) and the core flow dependent MCPRSL calculated by SPC (as discussed in response ¹2 above).
The resultant MCPROLs will be incorporated into the PPX-II CMSS as part of the cycle specific input.
Although the PPX-II CMSS will be calculating CPR values which do not account for the flow dependent bias of ANFB, comparison of these values to the MCPROLs contained in the Core Operating Limits Report (COLR) and input to PPX-II is a correct and consistent use of the values.
Susquehanna SES control room operator training stresses maintaining plant operation within the limits specified in the COLR.
An operator action which is based solely on the calculated CPR values from the CMSS without consideration of the proximity of these values to the applicable MCPROLs would be a deviation from current operating practices and a violation of current plant procedures.
Therefore, use of the non-flow adjusted CPRs from the PPX-II CMSS for other than the intended purpose (i.e., comparison to the MCPROLs) is not considered to be a realistic scenario at Susquehanna SES.
The operation and performance of the PPX-II CMSS is unaffected by the implementation of a core flow dependent MCPRSL because the CMSS utilizes MCPROLs to assess current core performance and does not directly use MCPRSLs.
Although the U2C9 MCPROLs will reflect a core flow dependent MCPRSL, this dependence will be transparent to the PPX-II CMSS and will manifest itself in the core power and core flow dependent MCPROLs.
PP8L has utilized core power and core flow dependent MCPROLs for a number of cycles on both SSES Units, therefore, core power and core flow dependent MCPROLs will not be a unique occurrence for U2C9.
P
~
~"
1
NCLOSURE B TO PLA-4595 Page 11 of 14 PP8L Response to NRC Request for Additional Information on EMF-97-010, Rev. 1 SSES Power/Flow Ma General Electric developed a generic Power/Flow map which defined the expected operating domain of the BWR 4 reactor.
This generic Power/Flow map was validated and modified to reflect actual SSES performance using data obtained during initial startup testing of the Units and, as additional data is obtained, is periodically re-verified or modified. In addition, the SSES Power/Flow map contains representations of the SSES Technical Specification system setpoints discussed above (i. e., the APRM Flow Biased Scram and Rod Block and the RBM Systems).
The SSES Power/Flow map defines the allowable operating domain for the SSES Units.
This Power/Flow map is used as an input to the cycle specific licensing analyses to determine initial conditions for analysis of cycle specific AOOs. The MCPROLs determined as a result of these cycle specific licensing analyses support SSES operation over the range of power/flow conditions defined on the SSES Power/Flow map.
For U2C9, similar to past cycles, the operating domain defined by the SSES Power/Flow map will be used as input to the cycle specific licensing analyses.
The resultant U2C9 MCPROLs will allow operation over the range of power/flow conditions defined on the SSES Power/Flow map although the ability of the plant to operate over the allowed range of power/flow conditions may be affected due to the these MCPROLs as a result of the conservatisms incorporated to address the flow dependence of ANFB.
Therefore, the implementation of a flow dependent MCPRSL does not affect the SSES Power/Flow map because the domain established by this map is based on measured plant data or SSES Technical Specifications setpoints which themselves are unaffected as discussed above.
Question 5:
Explain in detail the process used to extend the ATRIUM-10 upskew axial power distribution data to the lower flow range (bundle flows less than
) ~
~Res onse 5; During the cosine series of dry out tests, the ATRIUM-10 bundle was tested over an entire range of expected bundle flows. For the upskew testing, the range of flows tested was smaller.
To cover the range of flows of interest for the safety limit, the upskew test data was scaled to extend its flow range.
When the experimenter looks at the raw test data from a dry out test, he expects to see two things based on experience.
The first is that for a given flow, the plot of bundle critical power vs. inlet subcooling should be approximately a straight line.
The second is that, if the bundle inlet subcooling is held constant, the plot of critical power versus bundle flow should be a parabola.
Both of these trends are seen in the ATRIUM'-10data.
Figure 5.1 of Reference 3 shows the plot of inlet subcooling versus flow for STS-17.4, and Figure 4.3 of Reference 3
shows a plot of bundle flowversus critical power for a constant subcooling.
NCLOSURE B TO PLA-4595 Page 12 of 14 PP&L Response to NRC Request for Additional Information on-EMF-97-010, Rev. 1 These two trends led to the method of scaling the data.
The plots of bundle critical power versus inlet subcooling for STS-17.8 (the cosine test with the complete range of flow) were first compared with the same plot for STS-29.1 (the upskew test).
In order to determine the low flow dependence of the ECPR (predicted / measured critical power), ANFB critical power predictions were made using the Additive Constants described in Reference 3.
These values are combined for each flow with the cosine test uncertainties for an increased Additive Constant uncertainty of,
' for use in the MCPR Safety Limit calculations for low bundle flows.
Figure 1 is a plot that compares the cosine and upskew data points at an inlet subcooling of 20 Btu/lb.
As a reasonableness check on the flow scalings a simple least squares power flit was performed on the known upskew flows.
The Barnett correlation, an industry standard correlation, was used to provide additional assurance that the scaling method gave credible results.
Figure 2 is a plot of the results.
- First, the comparison of predicted critical power using the Barnett correlation was made to known data. As can be seen from the plot, the Barnett correlation underpredicts critical power for each high flow point. Conversely, for the low flow scaled data, the Barnett correlation predicts higher critical powers than the scaled data, indicating that the scaling process is conservative.
Therefore, one may conclude that the development of the scaling data to and the resulting Additive Constant uncertainties used in the Safety Limit analysis are reasonable and appropriate.
References
- 1. "Application of Reactor Analysis Methods for BWR Design and Analysis", PL-NF-90-001-A, July 1992, plus Supplements 1-A (August 1995) and 2-A (July 1996).
- 2. NE-092-001A, Revision 1, "Licensing Topical Report for Power Uprate With Increased Core Flow," Pennsylvania Power 8 Light Company, December 1992, plus NRC SER on PP&L Power Uprate LTR (November 30, 1993).
- 3. EMF-97-010, Rev.
1, "Application of ANFB to ATRIUM'-10for Susquehanna Reloads",
March 1997.
a PaOP467any NCLOSURE TO PLAA605 58760 Page1 of6 I/O/y7 REQUEST FOR ADDITIONAL INFORMATIONON PP&L'S PROPOSED AMENDMENTNO. 166 TO LICENSE NO. NPF-22: UNIT2 TECHNICALSPECIFICATION CHANGES FOR ATRIUM~-10FUEL Question No. 1 In the standard SLMCPR analysis, the SLMCPR is calculated for a large set of statepoints throughout the burnup cycle and the maximum value is selected for determining the operating limit MCPR.
The proposed Susquehanna Unit 2 Cycle 9 operating limit is based on an evaluation of a relatively small set (of approximately 5) of operating statepoints, rather than an evaluation of the complete set of cycle-statepoints.
Provide a quantitative estimate of the effect of this simplification.
Response to Question 1
The original MCPR safety limit analysis for Susquehanna Unit 2 Cycle 9 evaluated each of the statepoints throughout the burnup cycle. This analysis identified a relatively small set of exposure specific conditions which had the potential to be the limiting conditions for the MCPR safety limit analysis.
Therefore, the flow dependent MCPR safety limit calculations were performed for each of the potentially limiting exposure speciTic conditions. The flow dependent MCPR safety limits would not change if all of the statepoint conditions throughout the cycle were evaluated for each core flow.
Question No. 2 The PP8L methodology for determining the flow-correction to the A00 dCPR employs a "more conservative off-line approach."
Describe this method in detail and how it insures a
conservative flow-corrected d,CPR.
Response to Question 2 The PPBL conservative off-line methodology for determining the flow correction to the AOO dCPR is described in detail below.
This methodology contains several conservatisms compared to the direct approach of incorporating the flow adjustment into PP8L's computer codes and re-performing the analyses.
These conservatisms are also described.
The PPBL off-line methodology employs calculations based on existing (unadjusted) dCPR analyses.
The detailed process consists of four steps.
First, for each event, the hot bundle flow and heat flux transients are examined to determine a conservative "minimum flow" for the event.
Second, the flow dependent ECPRs corresponding to the initial flow and the "minimum flow" are determined.
Third, the previously calculated unadjusted CPRs are modified using the concept of RCPR (fracttonat change in CPR) such that the minimum ~ad'usted CPR consistent with PP8L's NRC approved CPR methodology.
Finally, the modified CPRs are divided by the corresponding ECPRs (from Table 6.2 of EMF-97-010, Rev. 1).
The final resulting "adjusted dCPRe is then determined by subtracting the adjusted CPR at the minimum flow from the
NCLOSURE TO PLAP605 Page 2 of 6 REQUEST FOR ADDITIONAL INFORIIATIONON PP&L'S PROPOSED AMENDMENTNO. 166 TO LICENSE NO. NPF-22: UNIT 2 TECHNICALSPECIFICATION CHANGES FOR ATRIUM'-10FUEL adjusted time zero CPR.
These steps are described in detail below, using a sample calculation for illustrative purposes.
Step 1
Determine "Minimum Flow" for Event For a
typical limiting transient, such as a
Generator Load Rejection Without Bypass (GLRWOB), the time of minimum CPR occurs a short time after reactor scram.
Once the scram rod insertion occurs, the hot bundle heat flux decreases causing the calculated CPR to increase rapidly.
This rapid increase in CPR after scram produces a time of minimum CPR.
Since the flow adjustment to the calculated CPR would cause the CPR to decrease more rapidly as flow decreases, the time of minimum CPR would occur slightly after the original time of minimum CPR for a transient in which core flow decreases.
However, from the bundle flow, heat flux, and CHFR behavior, a conservative minimum flow can be determined.
Figure 1 presents a sample GLRWOB analysis using PP8L's methodology.
Scram rod insertion begins at approximately 0.27 seconds, producing a peak heat flux and minimum CHFR at approximately 1.0 second.
The time at which the adjusted minimum CPR would occur is determined to be sometime after the original time of minimum based on the trend in bundle flow.
After 1.0 second, the bundle flow decreases due to the effect of the End of Cycle-Recirculation Pump Trip. Following this rapid decrease in bundle flow, the hot bundle flow rate stays relatively constant.
Since the heat flux is decreasing rapidly during this time, as evidenced by the increasing CHFR, a conservative "minimum flow" for the event was selected to be 0.63 (0.0754 Mlb/hr). One of the conservatisms in the PP8L off-line approach is that no credit is taken for the increase in the CPR due to the heat flux decrease after the original time of minimum (approximately 1.0 second in this example) an examination of the increase in calculated CHFR clearly indicates the conservatism of this assumption.
This technique is utilized for all AOOs to determine a minimum flow for the event.
This conservative minimum flow is used to adjust the calculated CPR as described below.
Step 2 Determine ECPR Adjustment Factors Given the time zero flow and the minimum flow from Step 1, the ECPR for each flow can be determined.
The ECPRs corresponding to these two flows are taken from Table 6.2 of EMF-97-010, Rev. 1.
For bundle flows above 0.1 Mlb/hr, the ECPR is less than 1.0, indicating the correlation is conservative (Table 6.2 of EMF-97-010, Rev. 1).
In the PP8L off-line approach, no credit is taken for this conservatism in ANFB and the values are used directly from Table 6.2, which conservatively overestimates the bCPR for events with hot bundle flows above 0.1 Mlb/hr.
For the sample GLRWOB illustrated in Figure 1, the initial and minimum hot bundle flows are 0.1190 and.0754 Mlb/hrwhich correspond to ECPRs of
, respectively.
NCLOSURE TO PLA I605 Page 3 of6 REQUEST FOR ADDITIONAL INFORMATIONON PP&L'S PROPOSED AMENDMENTNO. 166 TO LICENSE NO. NPF-22: UNIT 2 TECHNICALSPECIFICATION CHANGES FOR ATRIUM'10FUEL Step 3 Establish Unadjusted Initial and Minimum CPRs As stated in PP8L's NRC approved MCPR methodology (Appendix B to PL-NF-89-005-A and PL-NF-90-001, Supplement 2-A), calculated dCPR is determined by iterating on hot bundle power until the transient minimum CPR equals 1.0.
Therefore, in conformance with PP8L's NRC approved methodology, the initial value of CPR is increased such that the minimum
~ad'usted CPR equals1.0.
This is done using a parameter which defines the fractional change in CPR during the event, referred to as RCPR.
RCPR is defined as:
RCPR = d CPR / Initial CPR, RCPR = (Initial CPR - Final CPR) / Initial CPR RCPR is used to modify the previously calculated CPRs so that when the final CPR is adjusted for the flowdependence (dividing by the ECPR from Step 2), the adjusted minimum CPR will be 1.0.
As discussed in Step 2 for the sample GLRWOB illustrated in Figure 1, the ECPR corresponding to the minimum flow is'n order for the adjusted CPR to equal 1.0, the unadjusted minimum CPR must be equal to the calculated ECPR at the minimum flow (from Step 1). For the sample GLRWOB:
RCPR = ECPRUM(gUgied / (1 + dCPRttttgdittsigd ) = 0.383 / (1 + 0.383)
=.2769 Since RCPR is constant, Step 4 RCPR
= 0.2769
(Initial CPRmodirtedbypcpp Initial CPR
Adjust CPRs for Flow Dependence
) / Initial CPRmodiftedby RcpR Using the ECPRs from Step 2, the Initial CPR (adjusted by RCPR as described in Step 3) is divided by the ECPR corresponding to the initial hot bundle flow rate (see Step 2).
For the sample GLRWOB:
Initial CPRad~usied
= Initial CPRmodirtedby ricpp / ECPR i p p
NCLOSURE TO PLA-4605 Page 4 of 6 REQUEST FOR ADDITIONAL INFORMATIONON PP&L'S PROPOSED AMENDMENTNO. 166 TO LICENSE NO. NPF-22: UNIT 2 TECHNICALSPECIFICATION CHANGES FOR ATRIUM'-10FUEL Since the adjusted minimum CPR is equal to 1.0, the adjusted dCPR is:
BCPR,djusied
= Initial CPRadjusied
- 1.0 Conservatisms in PPBL Off-LineA roach As discussed
- above, there are two major conservatisms to the PP8L off-line approach compared to the direct approach (i.e., modeling the flow dependence of ECPR directly into PP8L's computer codes and re-analyzing each event).
First, the increase in CPR following the original time of minimum CPR is not credited (see Step 1).
Second,'in order to calculate conservative ZCPRs for high initial hot bundle flows, the conservatism in ANFB for flows above 0.1 Mlb/hr is not credited.
These conservatisms assure that the PP&L off-line approach remains conservative compared to the direct approach.
4 Question No. 3 What uncertainty allowance was included for the additive constants at low flow in the Susquehanna Unit 2
Cycle 9
Technical Specification MCPR safety limit calculations (Figures 2.1.2-1 and 3.4.1.1.2-1)'P Was an uncertainty allowance Response to Question 3 The additive constant uncertainties which were used to establish the flow dependent MCPR safety limits in the Susquehanna Unit 2 Cycle 9 Technical Specification are presented in Table 1.
The additive constant uncertainties which were used in the MCPR safety limit calculations for assembly flows were more conservative than the uncertainties documented in Table 6.2 of EMF-97-010 Revision 1.
Question No. 4 The increased additive constant uncertainty at low flows is to account for the lack of data below 0.05 Mlb/hr.
However, the PP8L response'(Page 3 of 16) appears to indicate that this increased uncertainty was only applied What is the effect on the SLMCPR of applying the increased additive constant uncertainty
?
ENCLOSURE TO PLA-4605 Page 5 of 6 REQUEST FOR ADDITIONAL INFORMATIONON PP&L'S PROPOSED AMENDMENTNO. 166 TO LICENSE NO. NPF-22: UNIT2 TECHNICALSPECIFICATION CHANGES FOR ATRIUM'-10FUEL Response to Question 4 The additive constant uncertainty which is applied to each assembly in the MCPR safety limit calculation is determined by Therefore, the additive constant uncertainty increases as the assembly flow decreases.
Using the more conservative additive constant uncertainties summarized in Table 1
above (see
Response
to Question
- 3) results in an increase of in the flow dependent MCPR safety limits compared to using the additive constant uncertainties in Table 6.2 of EMF-97-010 Revision 1.
NCLOSURE TO PLA-4605 Page 6 of 6
$99 - F'~0 P R Ib7 ~+/
ENCLOSURE A TO PLA-461't J10t68
(/tb (9 7
ATTACHMENTTO PLA-4611 Page1of3 Question No. 1 Demonstrate the validity of using the previously identifie limiting subset of exposure specific conditions for generating flow dependent MCPR Safety Limits at lower core flows.
Response
To demonstrate the adequacy of using the previously identified limiting subset of exposure specific conditions for generating flow dependent MCPR Safety Limits, two-loop MCPR Safety Limitcalculations were throughout the burnup cycle.
Based on the verification calculations described
- above, as well as an understanding of the statepoint conditions which are important in the MCPR Safety Limit calculations, it is concluded that the flow dependent MCPR Safety Limits would not change.
Thus, the subset of exposure statepoints chosen to perform the core flow dependent MCPR Safety Limits analyses is appropriate to produce a conservative MCPR Safety Limitat each core flow analyzed.
Question No. 2 Confirm that the hot bundle flow behavior shown in Figure 1 of Reference 1 is typical of the AOOs for which a conservative minimum flow is determined in the PP&L offline bCPR adjustment process.
Response
The calculated hot bundle flow behavior has been examined for the limiting Anticipated Operational Occurrences (AOOs) for Unit 2 Cycle 9 (Generator Load Rejection Without Bypass and Feedwater Controller Failure events).
The bundle flow behavior shown in Figure 1 of Reference 1 is typical of these events.
After an initial decrease, the bundle flow rate remains relatively constant or begins to increase.
This behavior allows the selection of a conservative minimum flowfor use in the PP8L offline dCPR adjustment process.
Question No. 3 Provide information on the effect of implementing the additive constant uncertainty below bundle flows of as a step increase rather than a linear interpolation.
The effect on MCPR Safety Limitand the Pump Seizure Accident should be addressed.
Response
ATTACHMENTTO PLA-4611 page2of3 In the calculations supporting the Unit 2 Cycle 9 Technical Specification flowdependent MCPR Safety Limits, the additive constant uncertainty applied to each assembly is determined Therefore, the additive constant as the assembly flow decreases.
The effect of implementing the uncertainty below bundle flows was investigated by performing the following two loop Safety Limit calculations.
A step increase in the additive constant was analyzed and the results compared to a calculation with the same inputs except for a linear increase in the additive constant uncertainty This two loop Safety Limit evaluation was performed, ', ", As expected for the step increase case, the calculatednumberof pins inboiling tiansition
- However, this calculation demonstrated
- that, the calculated Technical Specification MCPR Safety Limit (
) does not change.
Additional two loop Safety Limitanalyses were performed further investigate the impact of an increase in additive constant uncertainty The number of pins calculated to be in boiling transition in each case.
The cases at demonstrate that the submitted Technical Specification Safety Limit would not increase.
For the
, however, the number of pins calculated to be in boiling transition increased to (the analysis requirement is s 0.100 %).
Since the derivation of MCPR Safety Limits is done in increments of 0.01, use of the step change to would have increased the calculated Safety Limitby 0.01. A subsequent case which assumed a
increase in additive constant uncertainty to below bundle flows of demonstrated that the submitted Technical Specification Safety Limit (
) would not change.
The effect of implementing the uncertainty below bundle flows of on the Single Loop Pump Seizure Accident was also investigated.
Although for the step increase case, the calculated number of pins in boiling transition (
, the calculated offsite doses were still within the NRC acceptance criteria (small fraction of 10CFR100 limits).
The above results indicate that use to determine the low flow additive constant uncertainties,
, does not significantly affect the results.
Therefore, the approach used in the calculation of the Unit 2 Cycle 9 Technical Specification MCPR Safety Limits and the Pump Seizure analysis is justified.
Reference PLA-4605, "Susquehanna Steam Electric Station Response to NRC Request for Additional Information on PP8L's Proposed Amendment No. 166 to License No. NPF-22: Unit 2 Technical Specification Changes for ATRIUM'-10Fuel", April 9, 1997.
ENCLOSURE B TO PLA-46't'1