ML20056B809
| ML20056B809 | |
| Person / Time | |
|---|---|
| Site: | Monticello  |
| Issue date: | 04/30/1971 |
| From: | Feierabend C, Thornburg H US ATOMIC ENERGY COMMISSION (AEC) |
| To: | |
| Shared Package | |
| ML20056B807 | List: |
| References | |
| 50-263-71-06, 50-263-71-6, NUDOCS 9102110379 | |
| Download: ML20056B809 (2) | |
Text
{{#Wiki_filter:' s- ?[ f.';.. .;. t t ' i o* ( U. S. ATOMIC ENERGY COMMISSION REGION III DIVISION OF COMPLIANCE Report of Inspection C0 Report No. 263/71-6 Licensee: Northern States Power Company Monticello Nuclear Generating Plant -r License No. DPR-22 Category B Dates of Inspection: March 25 and 26,1971 i Dates of Previous Inspection: March 15-17, 1971 Inspected By: C. D. Feierabend Responsible Reactor 4-24-71 Inspector J t tibub) Reviewed By: . D. Thornburg pr.ReactorInspector 4-30-71 Proprietary Information: None SCOPE Type: Boiling Water Reactor Power Level: 1670 Mwt Location: Monticello, Minnesota Type of Inspection: Announced The purpose of the inspection was to discuss the status of the feedwater system and to review the results of startup testing that has been completed. The inspector was accompanied by Mr. E. Brunner, who assisted in the inspection and contributed to portions of this report. ADOCK 05000263 '{- 9102110379 710430 f CF CF t --_____-_w.--a
j' h./ - ,e :. s - -
SUMMARY
Safety Items - None. Noncompliance Items - There were two instances of noncompliance with Technical Specification 3.8.C.3 which requires that two independent samples of each tank shall be taken and analyzed for gross beta-gamma activity, during discharge of laundry drain tanks. A Form AEC-592 was issued on April 21, 1971. (Section Q) Unusual Occurrences - During surveillance testing on March 12, 1971, the licensee discovered that the low condenser vacuum scram switch setpoints had drif ted significantly. (Section F.2) During surveillance testing on March 8,1971, the licensee discovered that one of the two reactor building to suppression chamber vacuum relief valves was inoperative due to a valving error. (Section K) During surveillance testing on March 26, 1971, the licensee discovered that one of the four low-low reactor water level switch setpoints was not within specification. (Section L) Status of Previously Reported Problems - Insufficient time has elapsed 4 for a reply to the Form AEC-592 that was forwarded to the licensee as a result of the previous inspection. Other Significant Items - The most recent feedwater pump failure has provided additional information concerning vibration and failure modes. The licensee and his contractors have determined that an improved design of the feedwater pump internals may be needed to provide a permanent solution to the feedwater system problems. (Section H) Management Interview - The inspector discussed the results of the inspection with Mr. Larson and at the conclusion of the inspection an exit interview was conducted. In addition to the inspectors, the following personnel attended: Northern States Power Company C. Larson, Plant Superintendent M. Clarity, Assistant Plant Superintendent G. Jacobson, Plant Results Engineer L. Eliason, Radiation Protection Engineer ( I-
~' ?[k,f - T f:. tr*.' e - General Electric Company J. Ed11er, Operations Manager G. Matte, Operations Superintendent The inspector stated that the information received concerning the feedweter system was of interest and that Compliance will follow the progress of resolution of the feedwater problens. The inspector asked questions concerning resolution of the recircu-lation system problems, stating that there are three areas of interest. 4 These include the present status of the recirculation pumps, seals, and seal cooling system; the status of the recirculation pump isolation valves; and the status of the recirculation pump start system. Mr. Larson stated that he has discussed the problems with GE and Bechtel, and that he believes that the actions in progress will resolve the problems. He also stated that the Operations Committee would review the status of the system before the plant would be restarted. The inspector expressed C0 concern regarding the several failures of limitorque type valves experienced recently at Monticello and elsewhere. Mr. Jacobson stated that Northern States Power is initiating a program to verify operability of all valves that have limitorque type operators. He stated that the valves will be tested in both hot and cold con-ditions. He stated that most of the valves that have safety significance already are on a surveillance test program and that surveillance testing is up-to-date. The inspector stated that his comments during the previous inspection exit interviewl/ concerning personnel following procedures were still i an item of concern, and that this will be again reviewed during future inspections. Mr. Larson stated that he had recently been informed that a decision has been made to proceed with upgrading the feedwater pump minimum flow system during this shutdown. This will include installing the self drag valve in the "A" loop and increasing the capacity of the piping. The 6" piping vill be replaced with 8" pipe. This will extend the shutdown for at least ten days. 1/ CO Report No. 263/71-5.
1 [ ~ ' 4'. , r.. c l _4_ The inspector discussed the items of noncompliance separately with Mr. Larson and again, later, by telephone. Mr. Larson stated that the occurrence would be reported to DRL within 30 days. He stated tha t the unusual occurrences that were discovered during surveillance i testing would also be reported to DRL. The inspector stated that a Form AEC-592 would be issued. DETAILS 1 A. Persons Contacted i i Northern States Power Company (NSP) R. Duncanson, Superintendent, Steam Plant Operation R. Jensen, Assistant Manager, Engineering C. Larson, Plant Superintendent M. Clariry, Assistant Plant Superintendent G. Jacobson, Plant Results Engineer M. Dinville, Test Enginee r L. Nolan, Engineer ) gg General Electric Company (GE) f ) ( 4 J. Violette, Project Manager jk J. Fdller, Operations Manager ]g G. Matte, Operations Superintendent i Bechtel Corporation (Be chtel) B. Dornaus, Pump Specialist W. Balodis, Chief Startup Engineer DeLaval Turbine, Inc. (DeLaval) F. Weaver, Manager, Engineering A. Gertz, Pumps Engineer C. Operations The reactor was again made critical on March 19, 1971, to continue startup testing. The plant was heated up and held at 150 psig while vibration and pulsation tests were being performed on the feedwater system. During this period, some additional testing was performed on the HPCI system. It was determined that the feedwater pump was not
.s_ .c",-- I( % $? -
- satisf actory for continued testing, so the reactor was shut down on March 22, 19 71, to complete repairs to the system.
(Discussion of feedwater system is included in Section H.) F. Reactivity Control and Core Physics 1. Control Rod Drive (CRD) System a. Startup Testing, STP-2, Control Rod Drives (Phase III) Phase III of the CRD startup tests included scram tests of the four slowest CRD's identified during open vessel testing at reactor pressures of 600 and 800 psig and of all CRD's at 1,000 psig. In addition, the four slowest CRD's were scram tested at 1,000 psig with zero accumulator pressure, were friction tested at 1,000 psig, and were timed at 1,000 psig. The results met the test criteria as follows: (1) Insert / Withdraw Timing Each CRD must have a normal insert or withdraw speed [ of 3.0 + 0.6 inches per second, indicated by. a full 12-foot stroke in 40 or 60 seconds. One CRD was retimed, as initial withdrawal time was 37 seconds. Af ter retiming, its withdrawal time was 51 seconds and its insert time was 52 seconds. All other CRD's tested were within the test criteria. (2) Friction Testing lf the differential pressure variation exceeds 15 psid for a continuous insert, a settling test must be performed, in which case the differential settling pressure should not be less than 30 psid nor should it vary by more than 10 psid over a full stroke. None of the CRD's' exceeded 15 psid. (3) Scram Testing Upon scramming, the mean of the insertion times of. all operable control' rods, exclusive of circuit response times, must be no greater than: t
,'[.{_,: }:. 5:,. ( Percent Inserted Seconds 10 0.70 50 2.05 90 5.00 None of the CRD's exceeded the criteria-for the mean insertion time of all CRD's. Maximum scram time observed for 100% insertion was 3.47 seconds. b. Performance CRD performance wag / The inspector reviewed records discussed briefly in the previous inspection report.m concerning CRD surveillance and testing. He also dis-cussed the test results and plans for subsequent testing with the licensee's cognizant engineer for the CRD system. The cognizant engineer has monitored CRD performance on a daily basis to identify any operating problems. Af ter the scram on March 10, 1971, all CRD's were exercised. It was noted that a total of 21 CRD's required increased drive water pressure to withdraw from position "00." The CRD's were then exercised daily, and all drives operated with normal water pressure on March 15 and 16, 1971. According to the licensee, the problem of withdrawing a CRD from position "00" appears to be associated with air in the drive lines, and is corrected by venting. A total of 25 drives have been noted as requiring venting and/or increased drive water pressure to withdraw on at least one occasion. Four of these were identified as having problems on several occasions. These CRD's have been identified for further surveillance. A review of the scram times for these four CRD's indicated that two of these four CRD's had scram tines shorter than the mean for all CRD's and that the slowest of the four had a scram time of 3.33 seconds for 100% insertion vs an allowable of five seconds for 90% insertion. 2/ 00 Report No. 263/71-5. ( ___________.-._.._A----_--
'-~ ~ ' ^ - - -~-' r (([,.>... N. i ( Information received by telephone from the facility sub-sequent to the inspection indicated that the licensee will remove and inspect several of the CRD's during the current shutdown. The inspector will follow CRD operation and maintenance during future inspections. 2. Low Condenser Vacuum Scram Switch Setpoint Drift On March 12, 1971, while performing a regularly scheduled surveillance test, it was discovered that the trip settings of the four low condenser vacuum scram pressure switches were significantly off'. the ' desired ' setting of 23" Hg vacuum. The ~ switches were recalibrated to the correct settings and an-investigation of the problem was initiated. At the time of the discovery of the occurrence, the reactor was shut down and depressurized and all control rods were fully inserted. The "as found" trip settings of the four switches were found to be: PS 5-11A - 20.2" Hg Vacuum PS 5-11B - 19.2" Hg Vacuum ( PS 5-11C - 20.2" Hg Vacuum PS 5-11D - 20.1" Hg Vacuum Prior to the calibration performed on March 12, the four. vacuum switches were previously calibrated on February 5, 1971. At that time, three of the four switches were found in calibration; the fourth switch required only a small adj us tment. The inspector discussed the occurrence with licensee instrumenta-tion and supervisory personnel. The personnel stated that the February 5 and March 12 calibrations were performed by the same instrument man applying.a standard procedure. The vacuum test gage which was. used has been verified to be in calibration. 1 hey stated that it is very improbable that the-setpoint change was caused by incorrect setting during the February.5 calibration. 9
.[. L a ' - i et!* i ~ p-. ll- - They also stated that, although the cause of the drif t problem has not been positively identified, checks performed on the switches following the occurrence indicate that the switches were most probably affected by forces applied to the switch bodies through the instrument sensing lines. The four switches are rigidly mounted to a concrete shield wall and the sensing lines were supported at the switch by means of the switch body. Each sensing line was connected directly to the metal j diaphragm capsule. Checks performed subsequent to the occur-rence demonstrated that the switch settings could be changed by several inches Hg by merely pulling down or pushing up on the sensing line. Although the vacuum switches were calibrated several times prior to the occurrence, setpoint drif t only occurred subsequent to operation of the condenser under vacuum conditions. It is suspected that the application of condenser vacuum caused the sensing lines to exert forces on the switches, causing the switch settings to be affected. I Actions taken to prevent recurrence include installation of a flexible sensing line at the pressure switch and increased surveillance testing to determine whether the setpoint drif t has been corrected. Minimum surveillance testing requirements i t, for calibration of the switches is once per quarter. The licensee intends to verify calibration when condenser vacuum is initially estchlished during the next startup. H. Power Conversion System l 1. Feedwater Pumps ( Subsequent to the failure of the feedwater pump P-2B, described l in the previous inspection report,3/ the licensee again proceeded to rebuild the pump and prepare the system for additional te s tin g. This included search and removal of pieces in the feedwater system, inspection of the control valves, and minimum flow valves, etc. The pump was reassembled and the plant was restarted to continue testing. Pulsation testing performed on March 20, 1971, indicated that j feedwater pump P-2B had a resonant frequency in the 2000-3000 l gpm range. The pump was shut down, disassembled, and inspected. No evidence of physical damage was noted. The licensee then 3!/ Ibid. ( i
.f.i;(.r' - .g : \\ *[ ' k decided to shut down the plant and perform additional upgrading of the feedwater system. This will include replacing the "A" loop cinimum flow control valve with a "self-drag" valve as was described for the "B" loop in a previous inspection report,I and replacing the 6" feedwater pump minimum flow lines with 8" lines, to increase minimum flow capacity. NSP representatives met with GE, Bechtel, and DeLaval repre-sentatives in San Francisco, California, on March 16, 1971, to discuss permanent solution to the feedwater system problems. NSP management persennel were briefed on March 24 and the Safety Audit Committee was briefed on March 25. 2. Safety Audit Committee Briefing The licensee requested that its contractors brief the Safety Audit Committee (SAC) concerning the history of the feedwater pump failures and the actions planned to resolve the problems. The inspectors attended a portion of the SAC meeting on March 25. Discussion of the feedwater pump problems was presented by ( DeLaval Turbine, Inc., representatives followed by questions and discussion by SAC members. Mr. F. Weaver, Manager, Engineering, DeLaval, discussed the history of purp failures through the most recent (March 7,1971) failure. Mr. A. Gertz, Pump Engineer Specialist, discussed the modes of failure of the pump impellers and the contributing causes and the actions planned to eliminate the problems. Mr. Weaver briefly reviewed the history of the Monticello pump problems. He stated that in reviewing the problems and comparing them with previous DeLaval experience in feedwater pump design and operation, it appears that the most significant difference between Monticello and the conventional plant feedwater systems is that in the fossil fueled installation, the system operates at low flows for a very short time, during plant startup, i.e., when the pump starts it quickly goes on up to rated flow. Mr. Gertz presented a technical discussion of the modes of f ailures due to vibration and the modes of vibration amplifica-tion experienced in rotating components, and traced the investigation and results to date. 4/ C0 Report No. 263/71-3. I
. [ - Q.'! ' ~ .-(f a:. {..- The most significant portion of the presentation was that the most recent failure of the pump identified a failure different from those experienced earlier. Mr. Gertz stated that this type of failure had not been previously experienced by DeLaval in other pumps or compressor impellers. This failure included major damage to the second stage impeller. All previous failures were in the first stage. The first stage impeller was intact, but a crack near the impeller hub was in an area that had not previously indicated any problem. Subsequent vibration analysis of a similar impeller in DeLaval's research ldboratory indicated that the impeller had a natural frequency response of 1230 Hg comparing with the excitation frequencies expected due to pump operation. DeLaval found that this frequency was nearly identical to the third harmonic of the frequency expected for a seven-vane impeller at normal operating speed. i According to Mr. Gertz, there are two ways to eliminate this type of design problem. One is to change the natural frequency of the impeller, the other is to increase the strength of the components so that failure does not occur. ) In discussing the projected course of action, Mr. Gertz stated that this has been divided into an intermediate and a permanent type modification. The purpose of the intermediate l modification is to allow startup testing to continue because the time cycle to supply new components for the permanent modification would cause undue delays. a. Intermediate Modifications (1) New impellers will be installed. (2) Thickness of impeller hub plates will be increased from k" to -v3/8". 4 (3) Outer diameter of impeller will be reduced by one half,# to increase the impeller to diffuser gap.
jo,
- f,3_,. __
4 lL3,. l 11 - i k. b. Permanent Modifications DeLaval is redesigning the p: mp internals taking into account all experience to date. The newly designed impeller will have thicker hubs, thicker side plates, and modified The above changes will alter the natural frequency vanes. responses. This will be verified by testing to assure - that the new natural frequency responses do not coincide vf. h excitation frequencies. Menbers of the SAC asked several questions concerning the previous failures, previots DeLaval experience, and the proposed course of action. General discussion ensued with participation by all present. The general concensus appeared to be that the approach being taken appears sound, but that previous experience indicates that further problems may be expected with the pumps before a satisfactory solution to the problems has been achieved. The committee recognized that whatever action is taken must be supported by performance demonstration. The committee was most - l interested in obtaining assurance that the testing cA1-be safely accomplished and that no unevaluated safety-l considerations will exist. ( i The inspectors did not attend SAC discussions af ter the DeLaval representatives departed; however, the inspector was subsequently informed by Mr. Larson that the licensee had retained pump specialists from The Franklin Institute to previde an independent evaluation of the problem and the solutions to the problem. K. Containment Valving Error, Reactor Bui3 ding to Suppression Chamber Vacuum Relief System On March 10, 1971, during a check of the reactor building to suppression chamber vacuum relief system, it was discovered that the manual instrument isolation valve for one of the two differential pressure switches was closed. The switch should open a vacuum breaker valve if the differential pressure reaches ten inches of water. The instrument had been previously checked on February 9,1971, as part of an established sur-veillance test; the instrument isolation valve was apparently secured in the improper position at 'that time. .w. --ww,-. e e-i
[ .m ig-t , ( i At the time of the discovery of the occurrence, the reactor was shut down and all control rods were fully int.erted. j Immediately following the discovery of the closed instrument isolation valve, the surveillance test on the nystem was completed to insure that the vacuum relief sys tem was functioning properly. The manual isolation valve for DPI! 2573 was seal wired open. Two pressure suppression chamber - reactor,uilding vacuum breakers are required to be operable at all times when the l primary containment integrity is required, except that, from the first date that one of the two vacuum breakers is found to be inoperabic, reactor operation is permissible only during the succeeding seven days unless it is sooner made operable. Ihe licensee stated thac during the period of February 9,1971, to March 10, 1971, containment integrity was required on approximately 19 days. Although vacuum breaker A0 2380 would J have been prevented from performing its function if a vacuum had developed in the primary containment, vacuum brea'ker A0 2379, which is also designed for 100 percent vacuum relief capability, was operational and would have provided vacuum ) relief if required. The valve was made operable immediately after it was discovered to be inoperable. Action taken by the licensee to prevent recurrence or similar 1 occurrences included review of the occurrence and discussions with all members of the instrumentation crew to further emphasize the significance and importance of properly perforaing calibra-tion and surveillance testing. In addition, a program was initiated to revise surveillance i procedures to include instrument system checks to assure that each system has been returned to its normal operating condition following the completion of the surveillance test. The licensee intends to report the occurrence to DRL within 30 days. ~ LJ
s !' .;[. a ^ gh. { i L. Emergency Core Cooling System (ECCS) Low-Low Reactor Water Level Switch Setpoint Deviation j On March 2C 1971, during a regularly scheduled surveillance test, one of the four reactor low-low water level switches in i the ECCS initiation logic did not trip at the correct setpoint. l l The four level switches used in the ECCS initiation logic are Yarway instruments which use mercury wetted, magnetic switches. 1 Each of the four level switches has two separate switches i in the ECCS initiation logic circuitry. The desired switch I setting is -48" with an allowable deviation of -3". One switch was found to trip at -55 1/4". A visual inspection by instrumentation personnel revealed that i the switch was loose in its counting clip. Previous calibration experience had demonstrated that a relatively small displacement l 01 similar switches could change the trip value by several I inches. Investigation of other similar switches revealed that normally one lead of the switch was routed through the mounting clip to keep the switch secure. The switch was removed and i reinstalled tightly with one lead routed through the mounting l clip. The switch was then recalibrated and functionally tested. j l The inspector verified that this was not the same switch onat had been replaced as described in a previous inspection report.5/ The two occurrences are apparently unrelated. Although the switch exceeded the allowable deviation for low-low reactor water level initiation of core spray and RCIC, the remaining low-low reactor water lerel switches would have initiated core spray and RCIC at the desired trip setting of -48". The low-low reactor level initiation of LPCI and HPCI was unaffected by this occurrence since separate switches are used in the initiation circuitry for LPCI and HPCI. The licensee stated that the occurrence would be reported to DRL within 30 days. 5/ Ibid. t
}' {jr. ~ q:y,
- P.
Radiation Protection The inspector reviewed the results of Radiation Measurements tests (startup test No. STP-5) that have been completed. The tests showed that the total gamma plus neutron dose rates at hot standby condition ( + 5% reactor powe-and at 15% reactor power for all points on the survey were < C.1 mrem /hr except for one point, located near-a 1 curie Pu-Be source. That point was measured to have a dose rate of ~ 0.2 mrem /hr. Survey instruments used for the measurements were: Eberline Neutron REM Detector PNR Neutron Eberline Fast / Slow Neutron Detector PNC Neutron Cutie Pie - Gamma Victoreen Radector III - Samma Q. Radioactive Waste Systgw_ On two occasions during recent operations, discharging the laundry drain tanks to the discharge canal involved inadvertent additional releases from a tank not selected for discharge at the time. Both occurrences are considered to be in noncompliance with paragraph 3.8.C.3 of the Technical Specifications. ( 1. On Saturday, March 13,1971, - shortly af ter. initiating-a release of liquid waste from Laundry Drain Tank (T-21A), the operator discovered that the liquid waste in Laundry Drain Tank (T-21B) was also being released. He immediately isolated Tank (T-21B). A check of the previous log entry for tank level indicated that approximately 50 gallons of liquid waste had been released to the discharge canal. Both laundry drain pumps discharge into a corn.m header. At the beginning of the release from (T-21A), the. liqu!.d from Tank (T-21B) was being recirculated to obtain a final sample prior to its release. The operator had failed to suspend the recirculation of flow from Tank (T-21B) prior to initiating the release from Tank (T-21A). Analysis of the waste in Tank (T-21B) showed a radioactive concentration of 5.86 x 10-8 uci/ml. t
,. g f-1 .[. y. ; ~ im, Ik t f, ( l d 2. On Monday, March 15, 1971, while releasing liquid radwaste from Laundry Drain Tank (T-21A) to the discharge canal, the operator noted that the level of the Floor Drain Sample Tank (T-32) had dropped. The operator immediately isolated Tank (T-32); however, approximately 345 gallons from'the floor drain sample tank had been released to the discharge canal before it was 1solated. The water remaining in the tank was analyzed and showed a radioactive count of less than the normal back-ground of the count room instrument. The inspector reviewed the occurrences and discussed them with plant management and radiation protection supervisory personnel. The procedure for discharging the laundry drain tanks did include a caution that the discharge from tank not selected for discharge be isolated. The intent appeared to be that when one tank was full, - the emptv tank would be put into service to collect l=undry waste while the full tank was being processed. It did not considr: e.ha t i both tanks may be full at one time. j ( The siphoning of liquid from the floor drain sample tank was an operator error, in that he did not adequately check the valve lineup to assure that the tank selected for discharge was the only source of water being discharged. ) I Actions taken to prevent recurrence included supplementing the operating procedures, discussions with shif t supervisors and operators, and implementation of valve lineup checklists. The valve lineup includes that manual isolation valves be closed for all system 3 discharges except for the tank selected. In addition, there is now a requirement for verification of the valve lineup by a second knowledgeable operator prior to release of each batch of liquid waste. The licensee stated that a report will be forwarded to DRL. A Form AEC-592 has been forwarded to the licensee. i a 8'e e> m.@mbietin,=he* =.}}