ML20083D197
| ML20083D197 | |
| Person / Time | |
|---|---|
| Site: | Waterford  |
| Issue date: | 09/25/1991 |
| From: | Barkhurst R ENTERGY OPERATIONS, INC. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| Shared Package | |
| ML20083D199 | List: |
| References | |
| W3F1-91-0481, W3F1-91-481, NUDOCS 9109300080 | |
| Download: ML20083D197 (9) | |
Text
,
Entergy
'"i m -u'~. ' -
o
-.=
Operations
. _ _ _. _ _ - _ ~. - _. -,, _. _. _ _ _ _ _.
R, P. Bath hurst W3F1-91-OlS1 A4.05 9^
September 25, !991 U.S. Nuclear llegulatory Commission ATTN: Document Control Desk Washington, D.C.
20555 Subject :
Waterford 3 SES Docket No. 50-382 License No. NPF-38 Technical Speelftention Cimnge Itequest NPF-38-117 Gentlemen:
The attached description and safety analysis supports a modification of Table 4.8-1 of the Waterford 3 Technical Specifications. This table, entitled " Diesel Generator Test Schedule," describes criteria used to determine test frequency of the emergency diesel generators (EDGs). The criteria are based on the frequency of EDG survelilance test failures. As a result of an August 20, 1991 failure, the Waterford 3 Train "A" EDG is presently on an act elerated test schedule. Due to the nature of this specific failure and because it is believed that, under these circumstances, accelerated testing would be inappropriate for augmenting EDG reliability, Entergy is requesting this chango to exclude this failure from the count that determines the adjustment of test frequency for the
" A" Train EDG.
This is a plant specific TS change that has potential safety significance. In order to reduce the potential for increased wear associated with the present accelerated test schedule, Entergy is req"esting this as an exigent change request. Please direct any questions or comments to D. A. Rothrock on (501) 739-GG93.
Very truly yours,
_ ))
\\, _
s' N
! 4 ThjT)'\\
M1O FB / DAR/asf Attachments:
Affidavit NPF-38-ll7 cc:
Il.D. Martin, NRC Region IV D.L. Wigginton, NRC-NRR R. B. McGehet N.S. Reynolds NRC Resident inspectors Office Administrator Radiation Protection Division (State of Louisiana)
American Nuclear lusurers y00 fI \\
9t09300080 910925 PDR ADOCK 05000382
-P PDR 4,( U U U i
~..
=
.i '.
UNITED STATES OF AMERICA NUCLFAR REGULATORY COMMISSION
.f
?
In the matter of
)
)
Entergy Operations, lucorporated
)
Docket No. 50-382
- Waterford 3 Steam Electric Station.
)
AFFIDAVIT R.P.' Barkhurst, being duly sworn, hereby deposes and says that ho is Vice President Operations - Waterford 3 of Entergy Operations, incorporated; that he is duly authorized to sign and file with tite Nuclear Regittatory Commission the attached Technical Specification Change Roquest NPF-38-117; that ho is familiar with the content thereof; and that the matters set forth therein are true and correct to the best of his knowledge, information and belief.
AIL u
R.P. Barkhurst Vice President Operations - Waterford 3 STATE OF LOUISI ANA
')
) ss PARISil OF ST, CIIARLES
)
Subscribed and sworn to before me, a Notary Publje in and for the Parish and Stato above named this A day of
.6 7tL lo q
,1991.
r; All. ' u e - !;)4 kr l'
- j
//
Notary Public My Commission expires M[ 4/. (L /1 1:
no l -
/
i 1
DESGRIPTION AND SAFETY ANAhYSIS OF PROPOSED CllANGl: N PF-38-l l 7 This elmnge adds a footnote, "a" to Table 4.h-1 of the Waterford 3 Technical Specifications (TSs). Table 1.S-1, " Diesel Generator Test Schedule," describes the criteria used in determining the test frequency of the Waterford 3 cmergency diesel generators (EDGs). The footnote allows the August 20, 1991 failure to be excluded from the count that determines the test frequency for the "A" Train EDG, and that after fourteen successful tests, this failure is trans-v-dued to zero and this provision, Footnote "a," no longer applies.
Existing Specifications See Attachment A Proposed Specifications See Attachment B Description This submittal requests the following footnote to be added to the first column in Waterford 3 TS Table 4.8-1, " Diesel Generator Test Schedule," as Footnote "a"*
"The valid failure to start the "A" Train Diesel Generator on 08/20/91 due to the loss of the power dropping resistor in the governor circuitry is not to be counted toward the adjustment of test frequency for the " A" Train I
EDG. If the first 14 valid tests on the "A" Train EDG following the l
08/20/91 failure do not result in a failure caused by the new power dropping resistor, the valid 08/20/91 failure is trans-valued to zero and this provision no longer applies."
On August 20, 1991, Waterford 3 Emergency Diese! Generator ( EDG) " A" was manually started in order to perform EDG operability verification. Following a i
successful start, diesel generator speed started to drop slowly. After the l
frequency decreased to fifty-five hertz, the voltage began to decrease The I
operator reported that the generator speed started to fluctuate during the test, dropping to as low as 280 rpm. Shortly af ter the voltage decrease, the generator l
surged returning its speed, voltage and frequency to normal. Af ter a few seconds, the generator repeated the smne sequence of fluctuations. This resulted from a failure in one cf the two voltage dropping resistors in the power supply to the generator's electele governor speed control unit. Af ter one more cyche of fluctuations, the generator was secured.
j This qualifies as a valid failure as defined tu Regulatory Guide 1.105, the second in the last twenty tests on the " A" Train EDG. In accordance with Technical Specification (TS) 3/4.8.1, A.C. Sources - Operating, and Table 4.h-1, Diesel Generator Test Schedule, this requires the test frequency to be increased from once every thirty-one days to once every seven days. TSs require this frequency to be maintained until seven consecutive failure-free demands have been performed and the number of failures in the last twrotv valid demands has been reduced to one. Under existing circumstanres, this mbans mm test every seven days until another fourteen tests are completed.
1
The intent of accelerated testing is to provide a means of identifying systematic deficiencies in equipment or practices that might result in equipment failures.
The aim is to identify persistent problem areas and assure their correction, if employed correctly, it can significantly augment equipment reliability, llowever, if the test frequency becomes excessive, if accelerated testing is applied in an inapprepriate situation where no benefit is gained, or if the accelerated test period ~1s longer than necessary to verify reliability, the value of accelerated testing is either reduced or lost. Justead,-it represents excessive, unnecessary
-vear that can compromise, rather than enhance, the reliability of the component.
In recent years, the nuclear industry has regarded the starting and loading of EDGs as the harshest events associated with their normal operation. Recognizing this, Entergy has been proactively purstting the reduction of unnecessary starts of the EDGs at Waterford 3.
Last year, Entergy requested and was granted a TS change enabling Waterford 3 to combine EDG operability testing with other tests such that surveillances could be run concurrently producing only one EDG start rather than one for each test. More recently, Entergy requested a change to the Waterford 3 TSc that relaxes the criteria that determines when accelerated testing of the EDGs is required. This change was consistent with recent developments by NUMARC and the NRC in the revision t_o Regulatory Guide (RG) 1.9.
However, due to delays in the revision of this RG, this change has not yet been granted. As such, Entergy has evaluated this recent falhtre of the Train "A" EDG to determine if accelerated testing will represent reliability enhancement and not just additionet unnecessary wear that could lead to degradation.
l
.Waterford 3 has two EDGs manufactured by the Cooper Bessemer Reciprocating Company, Two methods exist to govern the speed of the :e generators, one electrical and one mechanical. In the event that the electrical governor falls such that the speed starts to increase, the back-up mechanical governor assumos the function of speed control. In the power supply circuit for the electrical governor there is a resistor assembly that functions to reduce the 125 V, system voltage to twenty-four V This resistor assembly has two parallel, three hundred ohm, x.
seventy watt resistors. As discussed in NRC Information Notice 90-51, " Failures of Voltage-Dropping Resistors in the power Supply Circuitry of Electric Governor
- Systems," there have been several occurrences at other plants identical to the Waterford 3 failure where one of these parallel resistors has failed. In fact, a similar failure of one of these resistors has previously occurred on the same EDG in May 1986. These failures resulted in generator behavior consistent with the August 20,.1991 failu re.
'During such a failure when this resistor fails open, the higher voltuge drop on the remaining resistor causes a lower reference voltage at the amplifier. As the generator approaches its operating speed, the lower reference voltage produces a high amplifier output voltage which causes the EDG to lose speed. Normally, if the electronic governor fails and starts to overspeed, the mechanical governor will assume control to regulate speed, flowever, in this scenario, speed is decreasing rather than increasing; the mechanical governor does not u v to intervene. This is viewed as a design deficiency in that the failure of on of these resistors, although not desired, is designed to compromise the oport 'lon of -
the EDG.
l e
2
.e
-,e
-,..---..--,m
.-,.,._ve
.,,.g.
---.--,--__y m-~-
y wg y
Af ter the 196G failure, other occurrences in the industry indicated tiuit this was not an isolated event. Based on discussions with the vendor and industry history, eighteen months were chosen as a conservative value for a minimum lifetime for the components. As such, Entergy chose to replace these components, and to do so every eighteen months thereafter. The most recent replacement was in June of 1990 -- only fif teen months before it failed. As indicated, Entergy was aware of this failure scenario and therefore was actively pursuing correction of the deficiency.
Cooper Bessemer redesigned the resistor assembly such that a single resistor replaces the parallel resistors. Failure of this improved resistor assembly causes the electronic governor to speed up rather than slow down. In this situation, the
(
mechanical governor will be able to assume control and regulate the speed. Such L
a design is inherently safe since it does not compromise the operation of the EDG.
Entergy submitted a purchase order for these components before the manufacturing had even begun. As such, the components were not available.
Only recently, two had become available to Waterford 3.
One was used to replace the damaged component in the " A" EDG following the failure on August 20, 1991.
1 The second one, received later, was used to replace the old component in the "B" Train EDG.
Special circumstances surround the August 20, 1991 failure of the " A" Train EDG. Reviewing these circumstances, Entergy concludes that an accelerated test schedule is urswarranted since it does not represent an enhancement of reliability.
Several factors support this.
The effect of the resistor failure in the electronic governor circuitry on the operation of the EDG is clearly understood by Waterford 3 pe rsonnel. Our knowledge is based on documentation by both NitC and vendor, as well as previous experience with the 1986 failure.
There is no interdependence between the failed resistors (or the newly designed replacements) with other components outside of the governor circuitry. The only effect of such a failure is n reduction in the speed of the EDG. It is believed that no other failure would result based on defective operation of these resistors. T herefo re,
increasing test frequency based on the August 20, 1991 failure only d
serves to focus on the operation of these resistors. liowever, the mere presence of this component in the governor circuitry, in and of itself, does not necesFitute necelerated testing. As such, there is no recognized benefit in the detection of related or resulting failures by increasing the test frequency The only course of action to correct the design flaw such that a similar failure would not compromise the operation of the EDG in the replacement of the deficient component with the new design. Since these have not been available, Entergy has had :10 recourso to correct the deficiency and avoid a possible failure, despite our knowledge of and experience with this issue Had they been available, the deficient components would have been replaced, the failure would not have occu'.' red and an accelerated test schedu!e would not be necessary for the " A" Train EDG.
i 3
j
- Based on an eighteen month lifetime, these components would have been replaced with the newly designed components before reaching the end of their minimum life. Entergy has been responsive in replacing and monitoring the life of these components. in accordance with Waterford 3 TSs, accelerated testing is required for approximately the first three months of the new component's life.
Since this component successfully completed a load test following ita installation, it is improbable that any aging problem would be detected in such a short period during the.beginning of the l
component's life.
2 The newly designed component now in the "A" EDG is fail-safe such that it will not compromise the operation of the EDG. The failure mechanism associated with the previous failure is no longer present.
Increased test frequency of this EDG would not detect a Gimilar failure mechanism. Therefore, there would be no recognized advantage.
Entergy feels these concerns support a position that accolorated testing based specifically on the recent failure of the Waterford 3 "A" Train EDG is inappropriate. Recognizing the effect on reliability that excessive testing may have under these circumstances, it becomes apparent that accelerated testing may diminish rather than enhance EDG reliability, As such, Entergy is requesting this change to exclude the August 20, 1991 failure from the count that determines test frequency. -Presently; Waterford 3 is on an accelerated EDG test schedule.
?In the interest of avoiding as much potential wear as possible, Entergy is requesting this to be reviewed as an exigent change request.
The change itself consists of two sentences that will be incorporated as a footnote in the current TS. - The first sentence essentially reduces the significance of the
. August 20, 1991 failure to zero in determining the frequency of testing for EDG Train "A". Since this will reduce the number of failures in the last twenty tests from two to one, accelerated testing would no longer be necessary. The second sentence states that if there are no failures of the power dropping resistor in the "A" Train governor control circuitry in the first fourteen tests following the August 20, 1991 failure, then the count for the August 20, 1991 failure can be
' permanently trans-valued to zero and -the provision would no longer apply. In the present specification, the triple asterisk footnote requires fourteen
. successful EDG tests to reduce the previous test failure count to zero following an overhaul to like-new conditions, provided that the overhaul, including
. appropriate post-maintenance operation and testing, is specifically approved by the manufacturer and that acceptable reliability has been demonstrated.
c
- Although this is not a complete diesel overhaul, it is equivalent to an overhaul of.
l 1the governor circuitry since the damaged component was reolaced by a ' newly-
. designed, functional component, and the failure was in a component that shares no interdependence with other components outside of the governor circuitry.
This rendered the governor electronic circuitry to "like-new" conditions. As previously stated, this work was : approved by Cooper Bessemer (the manufacturer) and was verified for post-mtJntenance operability. As such, trans-valuing _the August 20,_1990 failure to zero, as stated in the requested g
amendment, is supported assuming the next fourteen tests do not r suit in 'a failure of the electronic governor control circuitry. At that time, the provision of this. amendment _would no longer be applicable.
4
... ~.. - -.. - ~.-. -.. - _.. -
Safety Analysis:
The proposed changes described above shall be deemed to involve a significant hazards consideration if there is a positive finding in any of the following areas:
1.
Will the operation of the facility in accordance with these proposed changes -
involve a significant increase in the probability or consequence of any
_ accident previously evaluated?
Response
No Previously analyzed accidents that are potentially affected by this change are those that postulate a loss of offsite power to occur coincidentally with the accident (e.g., a loss of coolant accident with a losc of offsite power).
To significantly increase the probability or consequence.of such an accident, this change would have to negatively impact the reliability or performance of the EDGs. As indicated in the discussion above, accelerating the test schedule for tho'"A" Train EDG following the August 20, 1991 failure does not represent an enhancement of reliability.
llowever, excessive testing may harm the hardware adversely affecting reliability and performance. _ Since this change represents a reduction in the number of challenges to the system, it anytrung, there wili be un increase in rel' ability of the EDG and potential improvement in performance.
Based on the above information, this addition to Tablu 4.8-1 will not adversely affect the reliability or performance of the EDGs. Consequently,_
operation of Waterford 3 in accordance with the proposed changes does not involve a significant increase in the probability or consequence of any accident previously evaluated.
2.
Will the operation of the facility in accordance with these proposed changes 7reate the possibility of a new or different kind of accident from any a 'cident previously evaluated?
Response
No To crea*e a new or different kind of accident, these changes must introduce a new failure path. This change addresses an isolated incident.
Although a hardware change is being made to the EDGs, it reprerents the removal of a failure path rather than the introduction of one. Operation of
-the EDGs (and the rest of the plant) will remain unaltered. Consequently, a new failure path cannot exist as a result of the proposed amendment and the current plant safety. analyses remain complete and accurate in addressing licensing basis events and analyzing plant response.
Therefore, the proposed amendment cannot create the possibility of a new and different kind of accident than previously evaluated.
5
~. -
_. -.. _.. _ ~. _. - _ __
>4' 3.
Will the operation of the facility in acebrdance with these proposed changes 4
-involve a significant reduction in the margin of safety?
Response
No i
l This change does not alter the operation of any equipment installed at Waterford 3. Therefore, existing margins of safety are retained, and the operation of Waterford'3 in accordance with this_ proposed change will not' involve a significant reduction in a margin of safety.
Safety and Significant llazards Determination I
Based on the above Safety Analysis,'it is concluded that: (1) the proposed change does not constitute a significant hazards consideration as defined by 10 CFR 50.92; and (2) there is a reasonable assurance that the health and safety of the public will not be endangered by the proposed change; and (3) this action will not result in a condition that significantly alters the impact of the station on the environment as described in the NRC Final Environmental Statement.
h A
a I.
I f
lL l
V t
l-6
A
-4 p..;
l 1
-1 r
NPF-38-117 ATTACHMENT A 4
9 I'
l i-6 P
l ~.
.