IR 05000285/1998012
| ML20237C327 | |
| Person / Time | |
|---|---|
| Site: | Fort Calhoun  |
| Issue date: | 08/17/1998 |
| From: | NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION IV) |
| To: | |
| Shared Package | |
| ML20237C324 | List: |
| References | |
| 50-285-98-12, NUDOCS 9808210094 | |
| Download: ML20237C327 (14) | |
Text
. _ _ - . - _ _ _ _ - _ _ _ _ _ - _ ___ _ __ _ _ _ _ _
_--
.
.
ENCLOSURE U.S. NUCLEAR REGULATORY COMMISSION
REGION IV
Docket No.: 50 285 License No.: DPR-40 Report No.: 50-285/98-12 Licensee: Omaha Public Power District Facility: Fort Calhoun Station Location: Fort Calhoun Station FC-2-4 Adm., P.O. Box 399, Hwy. 75 - Nor"1 of Fort Calhoun Fort Calhoun, Nebraska Dates: June 1-5,1998 (on site) through August 11,1998 (in office)
inspector: J. J. Russell, Resident inspector Approved By: W. D. Johnson, Chief, Branch B Division of Reactor Projects ATTACHMENT: Supplemental Information l
l l
I i
!
9808210094 980817 PDR ADOCK 05000295 G PDR
- 1bob 21W M
. - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _
.
.
EXECUTIVE SUMMARY Fort Calhoun Station
, NRC Inspection Report 50-285/98-12 poerations
-
The prejob briefing conducted for dynamic tuning of Auxiliary Feedwater (AFW)
Pump FW-10 was not in accordance with licensee management expectations. The briefing was incomplete because turbine overspeed backout criteria were not discusse The prejob briefing was for testing of a rebuilt and modified speed control circuit with the f turbine operating at various speeds. During the Pump FW-10 turbine start, the turbine
- oversped, and about 30 seconds elapsed before the turbine was secured. Operator i
actions to trip the turbine were satisfactory (Section 08.1).
Maintenance
The maintenance organization demonstrated poor work control practices. Two examples of a violation of 10 CFR Part 50, Appendix B, Criteria V, occurred. The first example involved machinists adjusting a Turbine-Driven AFW Pump FW-10 speed i- control linkage, and the second involved an instrument and control technician disconnecting a Pump FW-10 speed control air line. Both activities were performed without adequate procedural guidance. On a subsequent Pump FW-10 start, this i violation resulted in a Pump FW-10 overspeed occurrence. This nonrepetitive, licensee-identified and corrected violation is being treated as a noncited violation, consistent with Section Vll.B.1 of the NRC Enforcement Policy. A machinist did question the Pump i
FW-10 speed controller settings prior to the Pump FW-10 start, demonstrating an l
'
individual strength. The maintenance planning organization missed an opportunity to l identify the Pump FW-10 speed control deficiencies prior to Pump FW-10 start, by not performing testing to verify the overspeed governor setpoint (Section M8.1).
Enaineerina
Engineering organization actions to resolve an AFW single failure design issue were ineffective because of erroneous information supplied by a vendor, in 1990, the engineering organization identified that a particular failure of the Turbine-Driven AFW Pump FW-10 overspeed governor could overpressurize common AFW piping. Actions to correct the deficiency with a nardware modification were not taken until 1998 because i of vendor information which stated that the turbine would fail prior to over pressurizing
!
downstream piping. On a separate matter, system engineers provided incorrect information concerning adjustments to the Pump FW-10 speed controller to the maintenance craft, which resulted in an actual Pump FW-10 overspeed occurrenc This overspeed occurrence revealed the fact that the turbine would not fail prior to overpressurizing the downstream piping. Downstream piping was overpressurized but not damage * Based on a sample of five NRC information notices (ins) reviewed, licensee evaluation
! of their applicability to the turbine-driven AFW pump was complete and satisfactory, with L_______ _ _ _ _ _ _
.
.
one minor exception. The licensee had not evaluated minimizing the amount of discharge piping subjected to pump discharge pressure during pump starts, as described in a 1990 NRC IN (Section E8.2).
The root cause assessment, performed as a result of the May 27,1998, Pump FW-10 overspeed occurrence, was thorough and self-critical (Section E8.3).
I
!
!
!
______ .
.-
!-
- Report Details
!
1. Operations 08 Miscellaneous Operations issues l 08.1 ' Incomplete Preiob Briefina i- 1 l- , Inspection Scope (93702)
{
I On May 27,1998, the licensee held a prejob briefing in preparation for conducting j dynamic tuning of AFW Pump FW-10 speed control system using Maintenance Work 1 Order (MWO) 980691, Revision 0," Functional Analysis of Pump FW-10 Speed Control Loop." The inspector interviewed operations, maintenance, and system engineering personnel. The inspector also reviewed licensee expectations for prejob briefings as stated in generic prejob briefing critique sheets used by maintenance personnel, MPEM-3.03, " Maintenance Policies and Expectations Manual," dated August 14,1997, and incident free performance hand cards used by operations personnel. The inspector also reviewed Pump FW-10 pump discharge pressure data recorded during the Pump FW-10 start for the dynamic tuning.
!
L . Observations and Findinas l ' The prejob briefing was attended by maintenance, operations, and system engineering personnel. The briefing was conducted in preparation for a dynamic tuning of the speed control system for Pump FW-10. The normal speed controller had been essentially rebuilt during the outage. The turbine-driven pump was to be started from the control room and various parameters were to be checked and calibrated as turbine speed was !'
changed. The initial turbine start was to be performed with the normal spaed controller
-
in service, and then later in the procedure the air supply to the normal speed controller was to be decreased. The overspeed governor was to then assume speed control of the turbine. The inspector found, based on interviews, that overspeed backout criteria l
[ were not discussed during the prejob briefing. This was the first time the licensee had 'j conducted dynamic testing of the Pump FW-10 speed controls. Since the testing l involved turbine speed control systems that had been modified and rebuilt during the j outage, and since the testing was to be conducted while varying turbine speed, the .
inspector found that failure to include overspeed backout criteria was a poor practice, and contrary to licensee management expectation .When Pump FW-10 was started, the turbine oversped until stopped from the control l' room. Normal discharge pressure for the dynamic tuning conditions was approximately 1100 psig. Actual discharge pressure had reached approximately 2309 psig when control room operators tripped the turbine. Prior to the turbine start,- the caged area e enclosing Pump FW-10 was evacuated of personnel, which was normal practice for turbine starts. Remote discharge pressure recording equipment had been located i
. outside the caged area. An equipment operator, along with the system engineer, :
noticed the pump discharge pressure rising past the anticipated maximum pressure and the turbine sounds continuing to indicate increasing speed. The equipment operator ,
,
h I
(i l
.- - - _ _ - - - _ - . - _ - - - - - - _ _ _ - - _ - -
.
-2- l contacted the control room to have the turbine stopped. It took approximately 30 seconds from the time discharge pressure reached 1100 psig, until the turbine was trippe Conclusions The prejob briefing conducted for dynamic tuning of AFW Pump FW-10 was not in accordance with licensee management expectations, because turbine overspeed backout criteria were not discussed. The prejob briefing was for testing of a rebuilt and modified speed control circuit with the turbine operating at variable speeds. During the Pump FW-10 turbine start, the turbine oversped, and about 30 seconds elapsed before the turbine was secured. Operator actions to trip the turbine were satisfactor I 11. Maintenance M8 Miscellaneous Maintenance issues (93702)
M8.1 Poor Work Control Practices Inspection Scooe (93702)
i The inspector reviewed the maintenance organization's involvement with an overspeed occurrence of Pump FW-10 on May 27,1998. The inspector interviewed maintenance supervision, machinists, and an instrument and control technician. The inspector visually inspected the Pump FW-10 normal speed control and overspeed governor components. The inspector reviewed portions of MWOs 980691 and 981951 and Procedure IC-CP-01-1039, Revision 1, " Calibration of Auxiliary Feedwater Pun'p FW-10 Speed Control Loop 1039." The inspector also reviewed portions of Procedure MM-PM-AFW-0001, Revision 1, " Pump FW-10 Speed Lirniting Governor Check," and Vendor Technical Manual 1438.0010, Revision 9, " Technical Manual for Coffin Turbo Pump l Auxiliary Feed Pump." Observations and Findinas The speed control for Pump FW-10 consists of a normal speed controller and an overspeed governor. The normal speed controller is air operated and senses differential ,
pressure between the Pump FW-10 pump discharge and the steam generators. The j I
normal speed controller attempts to maintain a nominal 145 psid by positioning one end of a control beam (a lever arm). As demanded differential pressure is less than actual differential pressure, the normal speed controller positions the lever arm downwar The lever arm is attached to a control oil pilot valve in the center and the overspeed governo: positions the other end of the lever arm. The overspeed governor is controlled by flyweights on the turbine shaft, which cause the overspeed governor to position the lever arm upward as turbine speed increases. The overspeed governor, nominally set to limit turbine speed to a value less than 8600 rpm, was set at 7580 rpm as of the end of this inspection period (Procedure MM-PM-AFW-0001, completed on June 2,1998).
t The control oil pilot valve acts to port control oil to Pump FW-10 Steam Throttle Valve MS-361. On a normal turbine start, the turbine speed would initially be controlled l
_
. . - . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ - _ - _ -
.
-3-by the overspeed governor (although not at overspeed setpoint) because the normal speed controller would be demanding full open on Valve MS-361, and then, as demanoed differential pressure approached actual differential pressure, the normal speed controller would assume contro During the 1998 refueling outage, in accordance with an upgrade plan that system engineering had developed to improve the performance of the normal speed controller, Normal Speed Controller Air Positioner YC-1039-2 and Air Actuator YC-1039-3 had been rebuilt, Differential Pressure Transformer DPT-1039 had been replaced, and various air pressure test connections had been installed, as documented in MWO 980691. On May 26,1998, when Pump FW-10 was started for testing, speed readings of approximately 7300 rpm and differential pressure readings of approximately 195 psid were recorded. The pump was shut down because of steam leakage from newly installed turbine casing drains. On May 26,1998, machinists adjusted Air Actuator YC-1039-3 linkage (the linkage between the air actuator and the lever arm)
twice. The linkage was set first so as to result in an approximate 0.3-inch measurement (between a nut on the air actuator and a support) under MWO 980691. A second time the linkage was set so as to result in an approximate 0.4-inch measurement under MWO 981951. On May 27,1998, Pump FW-10 was again started for approximately 3 minutes to cycle Steam Stop Valve YCV-1045, in an attempt to stop steam leakby from that valv For both adjustments of the air actuator linkage, the system engineer directed the machinists to perform the adjustment with Steam Chest Throttle Valve MS-361 ope The system engineer, unsure of the correct position of Valve MS-36i, had telephoned a representative from Coffin Turbo Pump Inc., the pump vendor who had given this guidance. The correct position of Valve MS-361 for the adjustment was close l MWO 980691 contained no specific work steps to perform this adjustmen l MWO 981951 stated to perform the adjustment in accordance with the vendor technical j manual and the directions of the system engineer. The vendor technical manual directed, on page D-48, to shut Valvo MS-361 and directed the linkage adjustment on page D-49. None of the machinists performing this adjustment had attended vendor training, although this same adjustment had been performed satisfactorily during the last j refueling outage. The machinists performed the adjustment with Valve MS-361 open, resulting in the overspeed governor being set above the 8600 rpm setpoint.10 CFR Part 50, Appendix B, Criterion V, states, in part, that activities affecting quality shall be prescribed by procedures or instructions appropriate to the circumstances. Contrary to this, the machinists set the Pump FW-10 linkage, using MWO 980691, which contained no guidance for the setting, and MWO 981951, which contained contradictory guidance for the setting. These procedures were not appropriate to the circumstances. This is the first example of a violation (NCV 50-285/98012-01). i l
After the second Pump FW-10 start described above, and before the overspeed, the i
'
system engineer directed an instrument and control technician to measure the valve travel of Valve YC-1039-3. The steam supply to the turbine had been tagged out for work on Valve YCV-1045. The technician disconnected an air line between Valve YC-1039-2 and Valve YC-1039-3 in order to use an external air supply to stroke Valve YC-1039-3. This was performed using Step 8.3 of MWO 980691. This step referenced Procedure IC-CP-01-1039, which contained no guidance for disconnecting the air lines i
t _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _
_ _ _ _ _ _ _ - _ _ _ -_-_- _ _-_ ______-__ __ _
.
-4-in order to check the valve movement. The technician did not reconnect the air lines after he measured the valve movement, resulting in a loss of air supply to Valve YCV- '
1039-3, and lhe resultant loss of the normal speed control function.10 CFR Part 50, Appendix B, Criterion V, states, in part, that activities affecting quality shall be prescribed by procedures or instructions appropriate to the circumstances. Checking the valve travel of Valve YC-1039-3 was an activity affecting quality because these components were classified as quality controlled equipment in the licensee's computer data base. Contrary to this, the instrument and control technician used MWO 980691 to perform the evolution but this work order contained no guidance to perform the evolution. This is a cecond example of a violation. This nonrepetitive, licensee-identified and corrected violation is being treated as a noncited violation, consistent with Section Vll.B.1 of the NRC Enforcement Policy (NCV 50-285/98012-01). The licensee's ,
corrective actions for this violation included: (1) developing a specific procedure for '
adjusting the air actuator linkage for Pump FW-10; (2) developing and administering a machinist training class covering Pump FW-10 air actuator linkage adjustment; (3) reinforcing management expectations regarding procedure usags; and (4) developing a case study regarding this event to be used in training. Corrective actions were to be tracked by Condition Report 19980124 The inspector found that the two examples of the violation described above were caused, in part, by erroneous information given to the craft by the system engineer, who in turn was given erroneous verbalinformation from the vendor. The system engineer gave instructions directly to the craft, bypassing the work control and planning organization The inspector also found that Procedure MM-PM-AFW-0001 provided direction to verify the actual setting of the Pump FW-10 overspeed governor in a controlled manner. This procedure was for a refueling interval surveillance. Although the speed control system was modified and rebuilt, as described above, this procedure was not performed prior to the two pump starts and then the pump start resulting in the overspeed. The inspector found that failing to verify the proper setting of the overspeed governor was a poor practic Prior to the modification and rebuilding of the normal speed controller, a machinist took various as-found measurements. These measurements included an actuator linkage distance, which he recorded as about 0.64 inches. This data was not required to be ken by any procedures. Subsequently, this machinist questioned the as-left measurements of 0.3 inches and then 0.4 inches. Although this questioning did not prevent the occurrence, the inspector found that taking the measurements and then questioning the as-left measurements was a strength displayed by the machinis Conclusions The maintenance organization demonstrated poor work control practices. Two examples of a violation of 10 CFR Part 50, Appendix B, Criteria V, occurred. The first example involved machinists adjusting a Turbine-Driven AFW Pump FW-10 speed control linkage, and the second involved an instrument and control technician disconnecting a speed control air line. Both activities were performed without adequate y procedural guidance. On a subsequent Pump FW-10 start, this violation resulted in a t
_ _ - _ _ _ - - - - - _ l
- - _ _ _ _ _ _ _ _ _ _ _ _ _ __ ____-___ _ _ _ _ _ _ _ _
.
l
l l -5-Pump FW-10 overspeed. This nonrepetitive, licensee-identified and corrected violation is being treated as a noncited violation, consistent with Section Vll.B.1 of the NRC j
! Enforcement Policy. System engineering contributed to this violation by providing j 1 erroneous information to the craft. The maintenance planning organization missed an j
- opportunity to identify the Pump FW-10 speed control deficiencies prior to Pump FW-10 ]
l
'
start by not verifying the overspeed governor setpoint by using a postmaintenance tes {
A machinist did question Pump FW-10 speed controller settings prior to Pump FW-10 i start, demonstrating an individual strengt Ill. Enaineerina E8 Miscellaneous Engineering lasues (93702) ,
!.
l E Weak Enaineerina Actions to Resolve a Desian Issue i Inspection Scope (93702) "
The inspector assessed the engineering aspects of an overspeed occurrence of l Pump FW-10 on May 27,1998. The inspector reviewed licensee operability evaluations I for the Pump FW-10 turbine, pump, downstream piping, and downstream components, I contained in Condition Report 199801246 and discussed verbally in conference calls with the NRC Office of Nuclear Reactor Regulation and Region IV on May 28 and 29, j 1998. The inspector also reviewed portions of the Updated Final Safety Analysis )
Report (UFSAR), Section 9.4, " Auxiliary Feedwater System," and a change to I l Section 9.4 of the UFSAR approved by the Plant Review Committee dated November 6, l 1998. The inspector also reviewed Engineering Analysis EA-FC-90-028 and vibrational analysis provided by Bentley Nevada, Inc., contained in a letter from Mr. Mark Dimond, Principal Engineer, Bentley Nevada to Mr. Jim Allen, System Engineer, Omaha Public !
Power District, dated June 2,1998. The inspector also reviewed LER 90-16, " Potential for Overpressurization of Auxiliary Feedwater Piping." The inspector visually inspected the outboard row of Pump FW-10 turbine blades, and walked down portions of the AFW piping from the Pump FW-10 pump discharge to the containment isolation valves, HCV-1107B and HCV-1108B. The inspector reviewed portions of Modification Package MR-FC-98-008, Revision 0, and Construction Work Order 980069, " Perform testing to ensure that Discharge Overpressure Feature Functions as Designed," and observed portions of testing conducted in accordance with this construction work order on June 2, 199 Observations and Findinas Following the errors described in the Maintenance section of this report, on May 27, i
< 1998, Pump FW-10 was started for postmaintenance testing and it overspe l
!
Downstream piping was aligned up to containment isolation Valves HCV-1107B and HCV-11088, and miniflow was aligned to the emergency feedwater storage tari '
Discharge pressure reached about 2309 psig, which was equivalent to about j 10,672 rpm. The duration of the overspeed was about 23 to 30 seconds. Discharge i piping is 4-inch Class 1600 (ASTM A106 Schedule 80 Grade B carbon steel) which branches off into two 3-inch Class 1600 lines to the two steam generators. The design
!
t I
_ - - ._- ____ _ _ _ _ _ _ _ - _ _ - - __ - _-_
-
i
.
-6-pressure of this piping is 1660 psig. However, Engineering Analysis EA-FC-90028 demonstrated, using an assumed minimum pipe wall, that this piping thould withstand 2520 psig (4-inch piping) and 2870 psig (3-inch piping) without damage. Both of these sizes of piping were exposed to the overpressure. The downstream flanges and valves are rated at 900 psig or higher. In Engineering Analysis EA-FC-90028, the licensee stated that these components were actually required to be hydrostatically tested to 3250 psig. The Pump FW-10 pump casing had been hydrostatically tested to 2400 psig, as discussed in a licensee conference call on May 28,1998. Since these hydrostatic and calculated pressures were all above 2309 psig, and since the inspector walkdowns j revealed no visible signs of damage, the inspector found nc basis to question the
[ licensee's analysi The Pump FW-10 turbine is vendor type " DEB" with two disks. The disks are keyed to a
, tapered shaft, with one blade segment per disc being held in place with a lock pin. The nameplate rated speed is 10,000 rpm. In Engineering Analysis EA-FC-90028, the licensee had previously documented maximum turbine speed as between 10,200 and 10,500 rpm prior to failure. As a result of this occurrence, the vendor supplied additional information that demonstrated that actual turbine failure would not occur until between 14,000 and 15,000 rpm. This was the calculated speed at which the locking pins would fail. Design aspects of the erroneous 10,200 to 10,500 rpm assumed failure speed are discussed below. To demonstrate turbine operability, the licensee performed a visual inspection of the turbine blades and discs, measured bearing movement, performed a '
J 24-hour endurance run for Pump FW-10, conducted enhanced vibrational analysis utilizing a' contractor (Bentley Nevada, Inc.), and performed a monthly and quarterly inservice test. Based on review of data from these tests, the inspector found that the turbine probably had not been damaged by the overspeed occurrence. However, the
,
licensee was unable to provide requested analyses for blade design speeds and l?
associated margins.
l The licensee had previously submitted LER 90-016, in which it was described that a l single failure of Pump FW-10's speed limiting governor could overpressurize L downstream piping. This piping is common to the only other safety-related AFW pump (Pump FW-6). As a result of this issue, the licensee developed Engineering Analysis EA-FC-90-028. This analysis referenced a letter from Coffin Turbo Pump, Inc., Mr. John Amicucci, Sales and Service Manager, to Omaha Public Power District, dated
,
. November 16,1989. This letter stated, in part, that between 10,200 and 10,500 rpm,
!
' turbine key buckets would shear, causing a high vibration such that the unit would drop off line or suffer a shaft failure. Engineering Analysis EA-FC-90-028 referenced a
'-
record of a telephone conversation between Mr. Jon Ressler, Omaha Public Power District, and Mr. Jon Amicucci, Coffin Turbo Pump, Inc., dated March 20,1990. M Amicucci had stated that the key buckets would fail, thus failing the turbine between 10,200 and 10,500 rpm, and that the failure had been confirmed via testing around the time the turbine / pump was designed (late 1950s). The licensee used this information to assume a maximum discharge pressure of 2235 psig and, using this pressure, performed a safety analysis for. operability (90-012) approved on May 13,1990, to justify acceptability of plant operation. The licensee did not provide complete and accurate information to the Commission in LER 90-016, because the LER stated that "The turbine key buckets on Pump FW-10 are designed to shear at a speed between 10,200 rpm and 10,4500 rpm . . ." This information is not accurate, as described above, and as
_ _ - _ _ - _ _ - _ - - _ _ - _ _ _
!
.
l .
, -7-l documented in an analysis done by Coffin Turbo Pump, Inc., dated May 29,1998, which concluded that the turbine key buckets would shear at between 14,000 rpm and 15,000 rpm. Because the licensee was provided inaccurate information by Coffin Turbo Pump Inc., in the 1989 letter and the 1990 telephone call, a condition adverse to quality l was not promptly corrected. A design deficiency that could have enabled a single failure of the Pump FW-10 overspeed governor to overpressurize piping common to all AFW pumps, identified in 1990, was not corrected until 1998, by installing a modificatio During this inspection, the inspector reviewed a draft white paper titled, " White Paper on Peak Pressure Associated with an Overspeed of FW-10," dated July 8,1998,
! concerning tw potential for a single failure to disable the AFW system. The white paper concluded *.ci,if the Pump FW-10 oversped such that the Pump FW-10 steam chest throttle valve were full open, the maximum discharge pressure would be "less than 2700 psig." The 2700 psig number assumed that the AFW containment isolation valves were somewhat open before the Pump FW-10 reached a maximum speed. The white paper also concluded that Pump FW-10 overspeed, with the pump recirculating and the AFW containment isolation valves full closed, would result in a maximum discharge pressure of 2815 psig. The white paper then concluded that 2815 psig is below code l allowable for downstream piping but above the pressure at which the Pump FW-10 casing had been hydrostatically tested. The pump casing would remain intact, although i a casing gasket might leak.
l l
The inspector found that the 2815 psig discharge pressure was not below " code l
'
allowable" for downstream piping. Licensee engineering Calculation EA-FC-90-028, performed in the 1990 time frame, used the original construction code
- (USAS B31.7-1968) to yield a maximum permissible pressure of 2520 psig for the 4-inch l downstream piping common to Pump FW-10 and electric-driven AFW Pump FW-6.
l The 4-inch piping is Class 1600, with a design pressure of 1660 psig. In addition,
'
2815 psig is above the hydrostatic pressure for the pump casing, which gives little l con 6dence that the pump casing itself would not fail.
l l
l The inspector will review a final version of the white paper described above, in order to
'
further assess the validity of a sing!a failure of the Pump FW-10 disabling the AFW l system. Closure will also involve issuance of any enforcement actions (if appropriate)
l- associated with the erroneous statements in the 1990 LER and the consequent lack of corrective actions, also described above. This is an inspection followup item (IFI)
(IFl 285/98012-02).
[ The nameplate rated speed of the Pump FW-10 turbine is 10,000 rpm. The inspector
[ found that licensee engineering judgement was weak when the vendor stated that the l turbine would fail between 10,200 and 10,500 rpm, and the licensee accepteo this
'
statement. A failure speed of only approximately 2 percent above rated speed was an abnormally small margin to failure for a rotating machin On June 2,1998, the licensee completed a modification to the control air for Valve YCV-1045. A pressure switch set at 1450 psig was installed on the discharge of Pump FW-10. The pressure switch acts to provide air to the Valve YCV-1045 actuator, closing the valve and stopping steam to Pump FW-10. The inspector found that the 10 CFR 50.59 safety evaluation contained in Design Change MR-FC-008 was l
l
!
l L______ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . . _ . _ _ . . _ _ _ . _ _ _ _ . _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _
.
.
-8-satisfactory, that no unreviewed safety question resulted, and that the modification would correct the design deficiency described above. The inspector observed postmodification testing on June 2,1998, which was performed satisfactorily and demonstrated a less than one second response time for the closure of Valve YCV-104 On June 2,1998, the insp?ctor walked down portions of this modification, prior to the modification being declared operable. The inspector noted an approximate 5-foot length of electrical conduit, from a wall-rnounted junction box to a solenoid, installed as a part of the modification. The conduit was secured to a stanchion, midspan, with plastic tie wrap. Although it was not necessary to seismically secure the conduit, use of is wrap ,
for permanent support was contrary to licensee management expectation for securing i electrical wiring outside switchgear or cable trays. In response, the licensee removed the tie wrap and installed a permanent clam Conclusions Engineering organization actions to resolve an AFW single failure design issue were ineffective because of erroneous information supplied by a vendor. The engineering organization identified the design issue in 1990 that a single failure of the Turbine-Driven AFW Pump FW-10 overspeed governor could overpressurize common AFW piping, but failed to correct the deficiency until 1998, with a hardwarc modincation. The erroneous vendor information, that Pump FW-10 would fail prior to overpressurizing downstream pip;ng, also caused an LER submitted as a result of the design deficiency to contain erroneous information. System engineers provided erroneous information to j the maintenance craft concerning adjustments to the Pump FW-10 speed controller, j resulting in an actual Pump FW-10 overspeed occurrence, revealing the design issu Downstream piping was overpressurized but not damage E8.2 incorporation of NRC Information Notices inspection Rcope (93702)
In order to assess licensee incorporation of NRC ins regarding turbine-driven AFW pumps, the inspector reviewed the following NRC ins, the licensee's evaluations of their applicability, and any actions taken: IN 86-14,"PWR Auxiliary Feedwater Pump Turbine Control Problems"; IN 88-09," Reduced Reliability of Steam-Driven Auxiliary Feedwater Pumps Caused by Instability of Woodward PG-PL Type Governors"; IN 90-45,
"Overspeed of the Turbine-Driven Auxiliary Feedwater Pumps and Overpressurization of the Associated Piping Systems"; IN 90-76," Failure of Turbine Overspeed Trip Mechanism Because of inadequate Spring Tension"; and IN 94-66, Supplement 1, "
Overspeed of Turbine-Driven Pumps Caused by Binding in Stems of Governor Valves."
b Observations and Findinas Based on review, the inspector found that the licensee's evaluations of the above referenced NRC IN's were complete and satisfactory, with one minor exception. The i
inspector observed that IN 90-45 recommended that licensees minimize downstream l piping that could be subjected to overpressure when testing turbine-driven AFW pump This aspect of the IN had not been reviewed for incorporation into the Ucensee's program. The inspector reviewed the valve alignment used for the Pump FW-10 l
I
- - - _ _ - _ _ - _ _ _ _ - _ - _ . _ _ _ - _ _ _ _ _
.
.
. dynamic tuning that resulted in the overspeed occurrence and the valve alignment routinely used for monthly Pump FW-10 surveillance starts. Each of these provided for Pump PN-10 discharge pressure up to containment isolation Valves HCV-1107B and HCV-11088, which were shut for these evolutions. The inspector also noted that manual Pump FW-10 discharge isolation Valve FW-172 was left open during these evolutions. Valve FW-172 could be shut while still providing recirculation of Pump FW-10. Valve FW-172 is located between the Pump FW-10 discharge and the containment isolation valves. The licensee agreed to evaluate shutting Valve FW-172 during Pump FW-10 starts, as appropriate and in accordance with the UFSAR, in order to minimize the amount of downstream piping that could be subjected to Pump FW-10 discharge pressure. The inspector considered this to be a satisfactory response to the concer Conclusions Based on a sample of 5 NRC ins reviewed, licensee evaluation of their applicability to the turbine-driven AFW pump was complete and satisfactory, with one minor exceptio The licensee had not evaluated minimizing the amount of discharge piping subjected to pump discharge pressure during pump starts, as described in a 1990 NRC I E8.3 Licensee Root Cause Assessment inspection Scope (93702)
The inspector reviewed the licensee's assessment of the root cause for the Pump FW- 1 10 overspeed occurrence, contained in SRG 98-014," Pump FW-10 Overspeed Event," !
dated June 16,199 l Observations and Findinas Based on interviews and review of records as described in this report, the inspector found the event description (and time line) contained in Root Cause !
Assessment SRG-98-014 to be accurat The licensee listed the root causes as: inadequate maintenans planning, a lack of formal training, the engineering team not answering questions concerning why linkage measurements were not consistent, over-reliance on vendor information supplied over :
the phone, and some vendor manuai guidance being unclear (the vendor manual called for adjusting the air actuator linkage while holding a piston assembly up, which appeared to be facilitated by having Valve MS-361 open; which was the incorrect position). An enclosed causal factor chart showed starting Pump FW-10 with the speed control loop disconnected and the speed limiting governor set at too high a speed as causes for the overspeed occurrence. Based on interviews and review of records as described in this report, the inspector found that these root causes and causal factors i
were accurate. In addition, the root cause analysis was self-critical and thorough. The
! root cause assessment did not contain corrective actions.
!
w__ ._ _ _ _ . . _ _ _ _ _ _
_ - _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ -
.
.
-10-c. Conclusions The root cause assessment performed as a result of the May 27,1998, Pump FW-10 overspeed occurrence was thorough and self-critica IV. Manaaement Meetinos X1 Exit Meeting Summary l
The inspector presented the results of the initial onsite inspection to members of licensee management at an exit meeting on June 5,1998. The licensee acknowledged j the findings presente The inspector asked the licensee whether any materials examined during the inspection should be considered proprietary, One proprietary document was identified, an analysis done by Coffin Turbo Pump, Inc, dated May 29,1998, which concluded that the Pump FW-10 turbine key buckets would shear at bety een 14,000 rpm and 15,000 rp The conclusions of this document were presented to the inspector as nonproprietary and the specifics of the document are not mentioned in this report (referenced in i Section E8.1 of this report).
The licensee was informed of the inspection findings, as described in this report, in a telephonic exit meeting on August 11,199 _ -
_ _ _ _ _ _ _ . ___ ___ __ _ ______-_ _ _ _ _ _ - _ -
.
.
ATTACHMENT SUPPLEMENTAL INFORMATION PARTIAL LIST OF PERSONS CONTACTED Licensee J. Chase, Plant Manager R. Clemens. Manager, Maintenance S. Gambhir, Division Manager, Nuclear Assessments J. Gasper, Manager, Nuclear Projects G. Gary, Vice President, Nuclear A. Hackerott, Supervisor, Systems Analysis R. Hamilton, Manager, Radiation Protection B. Hansher, Supervisor, Station Licensing .
J. Herman, Manager, Planning and Scheduling R. Jaworski, Manager, Design Engineering Nuclear R. Ridenoure, Supervisor, Operations R. Wylie, Manager, Nuclear Construction Management INSPECTION PROCEDURES USED IP 93702: Onsite Response to Events ITEMS OPENED 50-285/9812-01 NCV inadequate Maintenance Procedures (Section M8.1)
50-285/9812-02 IFl Auxiliary Feedwater Design Deficiency (Section E8.1)
l
i J
l