ML20235N621
| ML20235N621 | |
| Person / Time | |
|---|---|
| Issue date: | 02/24/1989 |
| From: | Phillips L Office of Nuclear Reactor Regulation |
| To: | Hodges M Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 8903010308 | |
| Download: ML20235N621 (6) | |
Text
__-_ _ _ _ _
'. ,S e
- o,, UNITED STATES
[
g NUCLEAR REGULATORY COMMISSION
$ :y WASHINGTON, D. C. 20555
,,...! FEB 2 41989 MEMORANDUM FOR: M. W. Hodges, Chief Reactor Systems Branch Division of Engineering & Systems Technology FROM: L. E. Phillips, Section Chief Section C, Reactor Systems Branch t Division of Engineering & Systems Technology
SUBJECT:
FOREIGN TRAVEL TRIP REPORT - BWR STABILITY 1
From October 15 to November 1, Larry Phillips traveled to Europe to exchange information relating to experience with neutron operation and testing of Boiling Water Reactor (power BWR) oscillations plants. duringof The objective the trip was to obtain technical information concerning the conditions and behavior of instabilities in European BWRs, and to gain a better understanding
, of the European perspective on stability safety concerns and the need for automatic protection and control systems and procedural operating precautions to avoid or promptly detect and suppress operating instabilities.
Meetings were held with several foreign reactor owners, operators, fuel suppliers, and regulatory officials as part of the follow-up review of the generic implications of the March 9, 1982 power oscillations event at LaSalle Unit 2. Both jet pump BWRs of General Electric Company (GE) design and BWRs of other designs were discussed. Even though only a small portion of these discussions involved proprietary or sensitive information, the contents of this report are generalized and summarized to avoid unintentional disclosure of sensitive information. Details of the specific meetings and reactor experience may be discussed with Mr. Phillips. The summary follows.
European BWR Operating Experience With Neutron Flux Oscillations For the BWRs of GE design, core-wide neutron flux oscillation events were experienced during operation on several occasions commencing in 1982. One of the events occurred on a BWR3 during a Cycle 1 startup excursion into the extended operating region (above the 100 percent control rod - flow control line) for the first time. A BWR4 experienced two core-wide oscillation events durins its first operating cycle. During second cycle stability tests of the same raactor, cross-core out-of-phase oscillations also occurred. Another reactor of GE design experienced out-of-phase regional core power oscillations at 75% power /40% flow during plant startup testing. After a similar limit cycle event occurred inadvertently during a subsequent startup at 50%
power /32% flow (corresponding to the 100% control rod - flow control line),
controlled stability search tests were performed to map the regions of potential instability. Out-of-phase neutron flux oscillations were induced at several points at or beyond the 100 percent control rod line in the maximum extended operating domain (MEOD) region. These tests confirmed the margins available in the GE SIL-380 (see Reference 1) procedures, which were introduced at this reactor. Operating procedures were revised for all of these plants to incorporate GE SIL-380-type operating procedures, and extended
)] ' l(],
I 8903010308 PDR ORG
% pNU 24 k s
L .-.xm _ _ 4 L . _ . _ . . _ ~ _ . ~ -
.c u i M. W.-Hod.ges - 2- FEB 2 41989
' load line operation was suspended for one cycle in one of them. 'One of.the _ ,
plants that conforms to GE SIL-380 procedures relies arimarily on avoidance of I intentional entry into power / flow operating regions w1ere stability l surveillance is required. The plant can do this.because entry-into these regions during startups is not necessary since the reactor is fueled with pellet clad interaction (PCI) resistant barrier fuel. )
i Because scram protection of fuel thermal limits is not available during i out-of-phase oscillations, these plants have chosen from among the following. j
, actions to augment SIL-380 procedures: )
l
- -inserting preselected high-worth control rods to suppress oscillations if 1 they occur ,
- initiating a reactor scram or a manual select rod insertion (or partial scram)onthelocalpowerrangemonitor(LPRM)upscalealarm
- inserting the preselected rods upon trip of one or two recirculation pumps and upon-loss of feedwater heaters.
After incorporation of the SIL-380 procedures, core power oscillations have been avoided successfully for several operating cycles at several of-these plants (some experienced oscillations before the revision of procedures).
Nevertheless, a " Select Rod Insertion" system was designed for one;of the plants and is intended to assist the operator to achieve a prompt power reduction via simultaneous insertion of a group of preselected rods by manual ;
scram action. In addition, an on-line stability monitoring system by noise analysis was developed for installation and testing on one of these reactors.
The reactors of non-GE design, including several with internal recirculation pumps, have had stability tests performed during startup and periodically or before operation in the extended power / flow range. Controlled stability tests performed at coolant flows lower than those permitted for normal operation are considered good practice to determine available core stability margin and to train the operators in the characteristics of core response behavior to rod insertion. Large (100 percent peak-to-peak LPRM amplitudes) out-of-phase limit cycle oscillations have been recorded for two of these reactors during natural circulation stability tests, the earliest in '1978. Another reactor experienced out-of-phase oscillations during low flow testing with some of the internal pumps not operating. In-phase limit cycle 3 sci 0 11ations have been observed for several others. The staff only recently became aware of these events. Typicalworstconditionsforinstability14reactorswithinternal recirculation pumps are 4 or 5 of 10 pumps running o'r 4 to 6 of 8 pumps running. These plants are less conscious of procedures to avoid instability than are the GE plants but are generally better equipped for detection and suppression from an instrumentation and control perspective. However, at least one. reactor has been denied permission to operate in the extended power / flow range until stability questions are resolved.
l I
M. W. Hodges I StabilityMeasurementandInstabilityAvoidanceTechniques
)
Detection of instability by observation of APRM and LPRM noise levels similar i to GE SIL-380 procedures is a well known technique. European reactors use l such procedures with somewhat more reliance on LPRMs for detection and action.
Several reactors have all LPRMs displayed in the control room. Some require operatoraction(includingscram)uponreceiptofLPRMupscale/downscale alarms. 3 Several personal computer based online stability monitoring systems have been developed for use as an operator aid in detection of instabilities and measurement of the existing stability margins. These systems are similar in ,
design and operate on process signals (APRM, LPRM, flow) to produce noise based decay ratio evaluations. The major differences in these systems relate to the design of the displays. The most prominent use of these systems is in core mapping of stability margins during test operations.
1 Six BWRs have flow instrumented fuel channels which provide valuable data for i detection and evaluation of thermal hydraulic instabilities.
Measures employed by various plants to detect and/or avoid oscillations include:
- partial scram based on power / flow setpoint to avoid mapped unstable regions,
- automatic insertion of groups of control rods based on Core Plate delta-p corresponding to a predetermined number of pumps not running, l
- manual insertion of groups of control rods on loss of feedwater heater (s),
- computer aided systems for runback of control rods at reduced flow, j
- unfiltered neutron flux signals to a power / flow scram circuit for protection against out-of-phase neutron flux oscillations,
- all LPRM signals displayed in the control room, l - history tapes which are continuously available in computer memory to generate high speed recordings of reactor data following an instability event.
- noise analysis decay rate measurement systems available for core mapping.
l Stability Characteristics and Sensitivities Some of the conclusions regarding stability behavior derived from the European experience follow:
---_.- -------- __ - -_------------ _ --- - _ _ _ _j
3m:-- = a . - . - . - . - -
EB 2 4 1989 M. W. Hodges (1) Operational parameters contributing to instability vulnerability of operating BWRs include:
(a) high power-to-flow ratio operating regime - most reactors are stable over most of the operating lifetime when in the normal operating regime below the 100% control rod line; many reactors are vulnerable to limit cycle oscillations when operating above the 100% control rod line and at low flow ( 40%) in the MEOD region; one reactor frequently tested for stability decay ratio at periodic burnup intervals reached a maximum stability decay ratio (most unstable) during the seventh operating cycle; (b) increase in subcooling - e.g., loss of feedwater heaters; (c) power distribution - very sensitive to bottom peaking of axial power distribution; also sensitive to higher radial peaking; and (d) the number of pumps running is directly associated with the power / flow vulnerability to unstable operating conditions; however, tests have not demonstrated a sensitivity to the distribution of operating pumps for internal pump reactors.
(?) Core design parameters contributing to instability vulnerability of BWRs include:
(a) ratio of single phase core pressure drop to two phase core pressure drop - a lower ratio as obtained by reduced fuel entrance losses across the entrance orifice and greater hydraulic resistance across the fuel bundle decreases stability; (b) smaller fuel pin diameters with reduced time constants for heat transfer decrease the stability; and (c) a smaller water-to-fuel ratio resulting in a more negative void coefficient decreases the stability.
Stability Predictive Methods Frequency domain codes of the GE FABLE type are considered not reliable for prediction of core stability margins because of the large uncertainty in choice of the appropriate operating parameters. Two instances of approximately 40% error in the predictive margins are known (i.e., the calculated decay ratios were about 0.60 as in LaSalle 2 and instabilities were experienced). In one case, most of the error was attributed to bottom power peaking far in excess of the values assumed in predictive calculations. In the other case, the oscillations were out-of-phase which was not predictable by the code used.
Some success has been achieved in the application of three dimensional transient codes such as TOSDYN-2 and RAMONA-3B to reproduce the out-of-phase
-. =i.z a :-----:; z = . = 2 -.. 2-
. ~
M. W. Hodges G t 4 tggo power oscillations observed at a few reactors. Several other codes, some with one dimensional neutronics and one dimensional thermal hydraulics capability have been used for stability analyses with varying degrees of success. One study using one dimensional code capability predicted that neutron flux power could reach levels as high as 2500% if scram protection is not available. A study of out-of-phase oscillations versus APRM signals indicated that the 120%
indicated power level would not be reached and thus would not provide scram protection during out-of-phase neutron flux oscillations.
Fuel Performance There were general discussions regarding current fuel performance experience in European BWRs. The incidence of fuel failures reported is very low and most were either clear cases of PCI or suspected cases of PCI. As in the USA, even though most reload fuel is designed for resistance to PCI, much of the fuel remaining is vulnerable. This is important for stability considera-tions since fuel management startup procedures to avoid PCI require reactors to pass through low flow /high power regions of low stability margin.
There was also discussion of a recent event where several fuel rods failed due to prolonged operation at dry out conditions. Detailed information regarding this event is available.
In general, BWR operators who were visited gave high priority to avoidance of fuel failure or continued operation with failed fuel and stressed the importance of maintaining a clean primary coolant system.
Code Verification There were discussions about obtaining some of the stability test data from reactor startup tests for purpose of benchmarking the NRC analytical codes, LAPUR and RAf10NA38. The data are especially valuable because the characteristics of a new core are well defined and the tests were well documented. As follow-up, to these discussions, initial data on a startup test involving large (>100 percent peak-to-peak LPRM amplitudes) out-of-phase l limit cycle oscillations have been transmitted to NRC. Discussions are I
continuing about cooperative programs for analyses of these data.
/ s , h p L. E. Phillips, S 4 tion Chief Section C, Reactor Systems Branch Division of Engineering & Systems Technology l.
cc: See next page 1
- v"E ,,
- .. .
- . - - ~ . . .; .-..z__.;;.;..,--------------------------
L...
- s. .
DISTRIBUTION a YCentral F11eN6 SRXB R/F NRC PDR V.Stello(17G-13) 1 J.Scarborough(16H-3)-
T. Murley (12G-18)
J.Sniezek(12G-18)
F. Miraglia (12G-18)
D.McPherson(12G-18)
E. Rossi (11E-4)
L. Rubenstein (13H-24)
E. Jordan (MNNB-3701)
C. Heltemes (MNNB-3701)
S. Fabic. (MNNB-3701)
J. Rosenthal (MNNB-2104)
- 8. Sheron (NL-007)i L.Shotkin(NLN-353)
H. Scott (NLN-353).
D.Bessette(NLN-353)
K. Burke (3H-5)
H.Schechter(3H-5)
P. Boehnert (P-315)
J. March-Leuba'(ORNL)
J.Carew(BNL)
R. Jones T. Collins
'H.-Abelson T. Huang L. Kopp H. Richings G. Schwenk S. L. Wu L. Phillips L. Phillips R/F i
i
_ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _