ML20045G468
| ML20045G468 | |
| Person / Time | |
|---|---|
| Issue date: | 07/13/1993 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20045G456 | List: |
| References | |
| NUDOCS 9307140014 | |
| Download: ML20045G468 (11) | |
Text
..
,i+
ENCLOSURE SAFETY EVALUATION REPORT ON "BWR OWNERS' GROUP LONG-TERM STABILITY SOLUTIONS LICENSING METHODOLOGY" NEDO-31960 AND SUPPLEMENT 1 1
INTRODUCTIC E The Boiling Water Reactor Owners' Group (BWROG) has submitted to the U.S. Nuclear Regulatory Commission (NRC) the Topical Report NEDO-31960, "Long-Term Stability Solutions Licensing Methodology," (Ref. 1) and Supplement 1 (Ref. 2) for staff review.
The long-term solutions described in these reports consist of conceptual designs for automatic protection systems developed by the BWROG with its contractor, General Electric Company (CE), to either prevent stability-related neutron flux-oscillations or to detect and suppress any oscillations that may occur.
The reports also describe the methodologies that have been developed to establish setpoints and demonstrate the adequacy of the protection systems to prevent violation of the critical power ratio (CPR) safety limits in compliance with General Design Criteria (GDC) 10 and 12 in Appendix A to Part 50 of Title 10 of the Code of Federal Regulations (10 CFR Part 50).
The BWROG has requested that the NRC staff review these reports and approve the following:
The overall regional exclusion and detection and suppression methodology, including the overall treatment of uncertainties.
The solution concepts and associated licensing approach for each option.
The application of the methodology to the solution concepts.
9307140014 930713 p
PDR ORG NRRB k
PDR y
\\
--j
2 The BWROG/GE position concerning safety classification of new and existing hardware; that is, the safety classification of all. existing and interfacing _ equipment should not change when new or modified stability long-term solution hardware is installed.
The BWROG plans to proceed with the selection of options and specific hardware design based on NRC's approval of the proposed concepts and associated methodologies.
2 EVALUATION The solution options proposed by the BWROG in NEDO-31960 are:
I f2LcQuplon Region A region in the high-power / low-flow area of the power / flow map outside of which instabilities are very unlikely is calculated for each representative BWR type using well-defined procedures.
If the reactor is operated within this exclusion region, an automatic protective action is initiated to exit the region.
This action is based exclusively on power and flow measurements; the presence of oscillations is not required for its initiation.
Four solutions for the Type I option have been proposed by the BWROG, although not all have been completely developed:
I-A 1mmediate protective action is taken when entering _the exclusion region.
This action can be either a scram or a select rod insert (SRI).
I-B Same as I-A, but.the protective action when entering the exclusion region can be bypassed.if a stability monitor is operational and detecting sufficiently stable conditions (e.g.,
a decay ratio less than 0.6).
. =-
y.
3 I-C Protective action is taken if the following two conditions both exist: (1) the reactor is operating inside the exclusion region and (2) an average. power 1
range monitor (APRM)_ oscillation.(of small magnitude) is detected.
i e
I-D A few small-core plants with tight inlet orifices have a reduced likelihood of out-of-phase instabilities.
For these plants, the existing unfiltered, flow-biased APRM scram provides' sufficient protection.
In addition, administrative controls are proposed to maintain the reactor outside the exclusion region.
II Quadrant-Based APRM Scram In a BWR/2, the quadrant-based APRM is capable of detecting both in-phase-and out-of-phase oscillations with sufficient sensitivity to initiate automatic protective action to suppress the oscillations before safety margins are compromised.
III LPRM-Based Detect and Suppress.
Local power range monitor (LPRM) signals or combinations of a small number of LPRMs are analyzed on line by using three diverse algorithms.
If any of the algorithms detect an instability, automatic protectiv action is taken to suppress the oscillations before ca aty margins are compromised.
Two options have been proposed by the BWROG:
Option III and Option III-A.
The main difference between the two is in the hardware implementation.
Option III requires a new Class 1E computerized system.
For Option III-A, newly designed digital replacements of the existing APRM amplifier cards will be used and a smaller number of LPRM detectors in a revised configuration will be required.
Ccoceptually, the algorithms are similar in both solutions.
+.
2 4
The NRC contractor, Oak Ridge National Laboratory (ORNL),
assisted the staff in reviewing the topical reports.
ORNL has provided a technical evaluation report (TER) that is included as.
The TER describes the results of the staff's review of the functional performance criteria for the proposed 4
protection systems and of the assumptions, principles, and models inherent in the methodologies used to define protection system stability boundaries and setpoints.
The staff's evaluation of hardware safety classification follows.
At a meeting on March 26, 1992, with the NRC staff, the BWROG proposed the" for the long-term stability solution options relying on "APRM flow biased scram" recirculation drive flow signals, the use of existing hardware be allowed in the new protection system.
The recirculation flow drive system, although highly reliable, is not designed to Class 1E standards.
The staff, therefore, requested additional information on. plant-specific arrangements of the existing recirculation drive flow instrument channels, channel integrity and independence, the failure rate data for each component in the flow channels, and the failure indication alarms.
The response, which the BWROC transmitted with a July 17, 1992, letter (Ref. 3) provides the results of a survey among 9 licensees for 12 operating plants.
In general, redundant flow channels exist in these arrangements.
The failure history of the channel components (from eight BWR units covering 84 reactor-years) shows the failures to be random and the failure rate to be insignificant.
For failure indication, the output signal from a flow channel is compared to the output signal from another flow unit. 'The comparator activates an alarm when two flow signals differ more than the i
specified tolerance.
Alarms also are activated when the comparator fails high or low.
Isolators are provided between i
flow units and between the comparator and the APRM circuitry and the elarm circuitry.
However, the survey results indicate that many operating plants do not meet the configuration in BWROG
~-c.
5 Viewgraph 3/26-7, " Drive Flow Signal Path," which was shown at the meeting on March 26, 1992.
4 The staff will review the hardware design details on a plant-specific basis.
In general, it finds the proposed concept to be acceptable, but may require modifications for some plants.
3 CONCLUSIONS a
i The staff has reviewed the licensing basis for long-term solutions to BWR stability proposed by the BWROG and adopts the recommendations described in the attached TER.
The regulatory positions with respect to the specific approvals requested by the BWROG are summarized below:
(1)
Methodoloav The exclusion region calculation methodology described in NEDO-31960 and its Supplement 1 is acceptable for defining the Option I-A exclusion region and the Options III and III-A exclusion boundaries outside of which the detect and suppress action may.be deactivated.
The overall treatment of uncertainties is acceptable for the selection of initial conditions and for the selection of oscillation contours and the treatment of failed LPRM sensors for Options III and III-A.
The methodology is acceptable for evaluating the protection provided by the Option II quadrant-based APRM Specific procedures for application of the scram.
methodology consistent with documentation and calculations submitted for this review should be developed and documented by BWROG.
(2)
So_1_ution Concents (a)
Options I-B and I-C have not been developed in
~.. _..
m
-.e--
m.
~-. -. - - -
i.
6 detail by the BWROG and, therefore, will not be considered acceptable as long-term solutions until fully developed by the BWROG and reviewed and approved by the staff.
option I-D is still under review and its acceptability as a long-term solution i
depends, to a large degree, on the details of calculations that are not yet available.
Attachment 1 identifies some concerns about the option I-D reliance on predictive calculations to conclude that the out-of-phase mode of oscillation will be avoided.
To address these concerns and to provide reasonable assurance that out-of-phase oscillations will be avoided by I-D plants, it may be necessary to incorporate strict operational controls on axial and radial power distribution and to enhance the capability to recognize operating conditions that are approaching instability by other means such as on-line stability monitoring.
Core stability sensitivities are illustrated by experience with the instability event on August 15, 1992, at Washington Nuclear Power Unit 2, in which oscillations developed outside of the stability exclusion regions because of~
a combination of fuel, core design and control rod patterns which resulted in conditions unfavorable to core stability, and conditions that an NRC inspection team concluded to be vulnerable to out-of-phase instability (Ref 5).
The staff will evaluate the acceptability of option I-D when the calculations for-the lead plant are submitted.
If the lead plant analyses are acceptable, the staff will evaluate detailed calculations for all plants that may propose option I-D (e.g. Duane Arnold, Vermont Yankee, Monticello, and FitzPatrick).
If individual. plant analyses are inconclusive because of large uncertainties involving assumed operating
+,
, 04 7
conditions, the quality of administrative controls and available core monitoring to reduce instability vulnerabilities will be considcted in evaluating the Option 1-D acceptability for a specific plant.
(b)
The implementation of Option I-A is an acceptable long-term solution for any type of BWR, subject to the following conditions:
(i)
Specific reload confirmation procedures should be developed so that for every reload, the licensee can either confirm the applicability of old exclusion region settings or set a new exclusion region boundary.
(ii)
The exclusion boundary setpoints for this option shou)d be sufficiently bounding to avoid changes on a cycle-by-cycle basis.
Major setpoint changes should be expected only if the fuel design changes significantly.
(iii)
When establishing reactor trip setpoints for the power / flow exclusion. region scram, operational restrictions on other parameters important to stability (e.g.,
radial and axial power distribution during low flow power maneuvering) that are consistent with the assumptions of the exclusion boundary analyses should be addressed, including the need for technical specifications, and factored into the'setpoint evaluation.
(iv)
Select rod insert (SRI) may be used in conjunction with option I-A, but a full scram should occur if the reactor does not exit the
~.. -
, r.
8 region within a reasonable period of time (about a few seconds).
(c)
Option II is an acceptable long-term solution for implementation in BWR/2s, which have quadrant-based APRM scrams.
For implementing Option II, plant-specific analyses should show that the quadrant-based APRM scram provides sufficient protection against out-of-phase instability modes to avoid the violation of CPR safety limits.
(d)
Options III and III-A are acceptable long-term solutions for implementation in any type of BWR, subject to the following conditions:
(i)
All three algorithms described in HEDO-31960 and Supplement 1 should br: used in Option III or III-A.
These three algorithms are high LPRM oscillation amplitude, high-low detection algorithm, and period-based algorithm.
(ii)
The validity of the scram setpoints selected should be demonstrated by analyses.
These e
analyses may be performed for a generic representative plant when applicable, but should include an uncertainty treatment that accounts for the number of failed sensors permitted by the technical specifications of the plant's applicant.
(iii)
Implementation of Option III or III-A will require that the selected bypass region outside of which the detect and suppress action is deactivated be defined in the technical specifications.
~..... -
-~
. - ~.
- 1 *.
9 (iv)
If the algorithms detect oscillations, an automatic protective action should be initiated.
This action may be a full scram or an SRI.
If an SRI is implemented with Option III or III-A, a backup full scram must take effect if the oscillations do not disappear in a reasonable period of time or if they reappear before control ~ rod positions and operating conditions have been adjusted in accordance with appropriate procedural requirements to permit reset of the SRI protective: action.
(v)
The LPRM groupings defined in NEDO-31960 to provide input to the Option III or III-A algorithms are acceptable for the intended oscillation-detection function.
These LPRM groupings are the oscillation power range monitor for Option III or the octant-based' arrangements for Option III-A.
The requirements for a minimum operable number of LPRM detectors set forth in NEDO-31960 are acceptable.
(e)
Options I and II do~not protect the fuel against single-channel instability, and the protection provided by Options III and III-A for single-channel instability is not highly reliable.1 When implementing F
the long term solution, a procedure to review the thermal hydraulic stability.of lead use assemblies i
t (LUA) in a. core reload should be established.
The review should ensure that inclusion of the LUA as proposed in the core reload is very unlikely-to result in single-channel instability.
1
,,..,,_,,a--,_,---.--
m.__
10 (3)
Safety Classification As a minimum, the recirculation drive flow channel should comply with the requirements of the Electrical and Electronics Engineers, Standard 279 (Ref. 4),
which include the single-failure criterion, component quality, channel independence, and the capability for test and calibration.
Isolation devices are required to be qualified for their application.
No credible failure at the output of an isolation device should prevent the associated protection system channel from meeting the minimum performance requirements specified in the design bases.
The plant-specific submittal should include the specification documentation for the isolation device.
In addition, because Solution I-A involves an automatic reactor scram function, any modification to the reactor protection system trip function requires a submittal to the NRC proposing a change in the technical specifications.
The plant-specific technical specification change should include limiting conditions for operation, action statements, allowable out-of-service times, surveillance tests, and test frequency commensurate with the importance to safety of the system.
The detailed technical specification requirements should be addressed generically during review of the detailed hardware design.
4 BEEERENCES 1.
NEDO-31960, "BWR Owners' Group,Long-Term Stability solutions Licensing Methodology," May 1991, l
-. ~ _. - -...
i 11 2.
NEDO-31960, Supplement 1, "BWR Owners' Group Long-Term Stability Solutions Licensing Methodology," March 1992.
3.
Letter from C.
L.
Tully (BWROG) to A.
C. Thadani (NRC),
" Response to RAI on Stability Report NEDO-31960, dated June 5, 1992," July 17, 1992.
4.
Institute of Electrical and Electronics Engineers, Standard
- 279, Criteria for Protection Systems for Nuclear Power Generating Stations."
5.
Letter from J.
B.
Martin, NRC, to A.
L.
Oxsen, Washington Public Power Supply System, "NRC Augmented Inspection of Washington Nuclear Project, Unit 2",
September 29, 1992.
,-=w-=---r-v.-
---= -- -- = r w v e r t w ervtv-,---
~,-
r-w e, - ~
=e.--
-, - - -.. - - - - - - - -