ML20033F729
| ML20033F729 | |
| Person / Time | |
|---|---|
| Site: | Palo Verde  |
| Issue date: | 03/20/1990 |
| From: | Cintula T NRC OFFICE FOR ANALYSIS & EVALUATION OF OPERATIONAL DATA (AEOD) |
| To: | |
| Shared Package | |
| ML20033F728 | List: |
| References | |
| TASK-AE, TASK-T90-04, TASK-T90-4 AEOD-T90-04, AEOD-T90-4, NUDOCS 9003280176 | |
| Download: ML20033F729 (4) | |
Text
y Ty
-e
+_ -
AE00 TECHNICAL REVIEW' REPORT
- UNIT:
N/A' TR REPORT NO.: AEOD/T 90-04 DOCKET NO.:- N/A-DATE: March 20, 1990
~ LICENSEE:
N/A EVALUATOR /CONTACTOR:
T. Cintula NSSS/AE:-
N/A
SUBJECT:
REVERSE (BACKWARD)ACTINGVALVEMANUALHANDWHEELS
=
SUMMARY
A transient.at Palo Verde was further complicated by unsuspected reverse acting valve manual handwheels on~one of the two atmospheric dump valves on each steam generator. A NUDOCS search found only two other occurrences of confusion with the direction-of manual handwheel rotation at the nuclear power plants.
Neither.
.of these events had potential to significantly complicate plant operations.'
-DISCUSSION
- 1.
Background for Investigation - Palo Verde Unit 3 on March 3,1989 With the unit operating at 98 percent power, a fault on the grid caused the unit 3 main generator breakers to open and the unit separated from the grid.
All three units at Palo Verde were designed for 100 percent load rejection capability without a reactor trip or turbine trip, however, a failure of the steam bypass control system initiated a sequence of events which resulted in a:
a.
Reactor trip b.
Loss of all reactor coolant pumps c.
Safetyinjectionactuationsignal(SIAS)
Mainsteamisolationsignal(MSIS)gnal(CIAS)
Containment isolation actuation si d.
e.-
the transient, all four of the steam generator (S/G) atmospheric Also, during(ADVs) failed to respond when given an open signal from the control dump valves room and the one ADV tried from the remote shutdown panel also failed to respond. Therefore, the auxiliary operators (A0s) were dispatched to the main steam structure to open one ADV in the manual handwheel mode per S/G.
- Af ter reaching the main steam structure, the A0s were further hampered in their attempts to place the ADVs in the manual handwheel mode by several ongoing deficiencies. Some of these were:
- a..The S/G main steam safety valves (MSSVs) are in close proximity of the
'ADV platform. S/G safety valve 579 on No.1 S/G lifted several times while the A0s were attempting to manually open the.ADVs.
b.
There was no lighting of any significance which resulted in almost total darkness in the area of S/G number 2 ADV manual operators.
- This document supports ongoing AE00 and NRC activities and does not represent the position or reouirements of the responsible NRC program office.
-9003280176 900320 PDR ADOCK 05000530 3
PDR-
y
. va-m.
l'
~2-
. c.. The A0s had received almost no training on ADV~ manual operation in_ the t
'past'2 years,
- d. LProcedures and ADV component labeling were not clear.
~
In spite of these obstacles, the A0s gained control of ADV-178 on S/G No.1
-and the. valve was manually opened to approximately 7 percent open.
- However, 3
. several problems were encountered while trying to take manual control-of the two valves on S/G No. 2, ADV-179 and ADV-185.
First, the A0s. attempted to Pr9 ally open ADV-185, but were unable to get.it more than 3 percent open whe: che handwheel came off due to a loose valve
-handwheel set screw. Rather than wasting more time by attempting to fix the handwheel for ADV-185, the A0s decided to open ADV-179 by turning the manual handwheel clockwise, which was the direction required to open ADVs-178 and 185.
However, because of valve / handwheel orientation, the~open direction for ADV-179 was counter-clockwise. When they were unsuccessful in opening ADV-179 by hand,-
and believ_ing.the repeated lif ting of the MSSVs warranted extreme actions, the A0s secured a 24" pipe wrench for mechanical advantage.
Using the pipe wrench as a." cheater bar", they succeeded in rotating the. handwheel one full turn in.
the-wrong direction. At that point they heard a loud bang, and ADV-179 and the platform vibrated. They took out their flashlights and saw that significant damage had occurred to the top of ADV-179's valve operator causing the valve to be inoperable.
Subsequent attempts to restore the handwheel to ADV-185 were successful and operators were able to gain manual control of this valve and open it by rotating the manual handwheel in the counter-clockwise direction.
1 L
Part of the problem, is that at Palo Verde not all valve handwheels operate in L
the coventional rotation, i.e., counter-clockwise to open and clockwise to close.
--Of the 12 types of air operated valves ~ at Palo Verde. 5 have nonconventional handwheel rotation.
In the specific case:of the ADV's at each unit at Palo Verde,
'the change in handwheel rotation direction resulted from flip-floping.the hand-wheel from one side of the valve actuator to the other so that the handwheels for both ADVs on a S/G would face the platform. This change resulted in a configuration in which an operator standing on the ADV platform has a handwheel on his'left hand and a handwheel on his-right, i.e., facing-each other. Although this makes the handwheels more.assessible, it also-results in having identical
' valves that manually open and close in a reverse handwheel direction from each '
other.
L 2.
Purpose of Investigation The additional problems encountered with valve manual handwheels that operate in different directions of rotation further complicated the licensee's ability l'
to mitigate the effects of an ongoing transient.
It seemed appropriate to find l-
_other cases of complications caused by reverse acting manual handwheels to deter--
mine how widespread the practice is at the nuclear plants, and if the reverse acting handwheels are contributing to complications within the nuclear industry.
1 i
r:
L W,g.
A?
g:
I 3. ' Search Technique The Nuclear. Document Control. System (NUDOCS) with a variety of related search
'j terms was used to search for events describing a problem with a reverse acting-valve manual handwheel. Two other events were found, they were:
'4.
Other Events (a). Limerick-1onApril 23, 1989 The unit was in a refueling outage and a local leak rate test (LLRT) was being
' performed on the reactor water cleanup return containment isolation valve. After completion of valve positioning and tagging, draining of the LLRT test volume began. Shortly thereafter, persornel in the control room observed a decrease in
]
the reactor vessel water level, and the LLRT crew was notified to terminate draining and to reverify proper valve lineup. During reverification of valve lineup, the LLRT crew discovered the reactor vessel boundary valve handwheel' was mounted backwards resulting in the valve being open rather than closed as required by the test. The reactor vessel water level had decreased approximately 1
i 2" dur.ing the one hour that the valve was inadvertently in the open position;-
this would correspond to a loss of_ approximately 400_ gallons of-inventory.
The consequences of this event were thought to be minimal by the licensee. The 1
. licensee estimated the maximum rate of draining through the 3/4" open test line i
to be about 25 GPM. When the reactor inventory decrease was noted in the g
control room, the normal letdown flow was isolated. With a normal control rod
(
1 l
drive make-up flow of 5_0 GPM, a letdown flow of 0 GPM and a drain path flow of l
-25 GPM, the reactor vessel inventory would have been restored had the drain path I
not been discovered in a timely fashion, j
(b) Brunswick Units 1 and 2 on July 11, 1989.
1 A service water (SW) system modification to provide minimum flow protection for 1
the SW pumps required the conversion of valve SW-V103 (in-both units) from a-l manual-to a motor operated valve. The modification to V103 would provide &
L erated valve.
In redundant motor operated valve to each units SW-106 motor op(81%) on a LOCA signal.
addition, SW-103 and --106 were modified to partially shut Eight days later, the NRC Resident Inspector observed a caution tag on SW-V103 that the valve operates backwards when operated manually, i.e., clockwise to open and counterclockwise to shut. The caution tag was needed to warn operators that the valves' handwheels were not painted orange, the Brunswick convention for identifying reverse operating valves.
Knowledge of correct manual operation was particularly important to SW-V103, since after the modification the motor operator only partially shuts the valve and only manual operation can fully shut the valve.
J
?c ;g,9 s 6
- .=
~vm
,4, I
-The 'e't day, the Resident inspector was informed that SW-V103 did not operate n
backwards. Apparently, during the modification, the valve faceplates were-incorrectly reversed, -1.e.,.the'"0" and "C" were mispositioned. With the f ace -
_' plates reversed, the valve would appear to operate backwards (unlike gate or-globe valves, these butterfly. valves had no other local indication for. valve
- position).
However, in the interim, the inaccurate manual operation of these valve's had been incorporated into formal training. Also, other nperating shifts received t
-the same incorrect information as it had been incorporated into the operations c,
real time-training.
The safety significance of this specific problem was small.
Had an operator attempted to fully shut one of these valves to isolate a leak, thus unknowingly L
. opening it, the valve would have automatically returned to the 81 percent shut position with a LOCA signal present.
CONCLUSION The event search found only two other recent occurrences of confusion'with direction of manual handwheel rotation. -Each of these events was of limited safety significance and neither event had potential to significantly complicate the effects of an ongoing transient as the precursor event. The ADV problems at Palo Verde have undergone extensive investigation, and it would appear that the generic aspects associated with nonconventional rotation of valve handwheels-do
._ not require _ f urther investigation at this time, i
i l
l L
l r
I
,