ML20062B741
| ML20062B741 | |
| Person / Time | |
|---|---|
| Site: | Haddam Neck File:Connecticut Yankee Atomic Power Co icon.png |
| Issue date: | 10/22/1990 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20062B733 | List: |
| References | |
| NUDOCS 9010260106 | |
| Download: ML20062B741 (15) | |
Text
_ _ _ _ _ _ _ _ - _ - - - - -..
.. f* $9
- a UNITED STATES
,[
' [g NUCLEAR REGULATORY COMMISSION r,, ?
j WASHING TON. D. C. 20555 m.l
....+
SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION RELATED TO AMENDMENT NO.132 TO FACILITY OPERATING LICENSE NO. OPR-61 CONNECTICUT YANKEE ATOMIC POWER COMPANY HADDAM NECK PLANT DOCKET NO. 50-213
1.0 INTRODUCTION
i Pursuant to 10 CFR 50.90 and 50.91, Connecticut Yankee Atomic Power Company (CYAPCO or licensee) proposed to amend Operating License No.
DPR-61 for the Haddam Neck Plant.
By;1etter dated August 25, 1990,.and supplemented by letter dated August 30, 1990, CYAPC0 proposed to change the Technicel Specifications (TS) by providing'a clarification.on the definition of Operability of the automatic auxiliary feedwater (AFW).
initiation system for Cycle 16 operation only. CYAPC0 requested that' this license amendment be processed on an emergency basis tin:accordance with 10 CFR 50.91 since an increase in the power level of the'Haddam Ne'k Plant above ten percent up to the' full licensed power level is prohibid.ed until this amendment is issued. -In addition, CYAPC0 requested-that the NRC issue a Temporary Waiver of Compliance from the subject TS until 2he i
j proposed license amendment-is issued. By' letter dated August 30, 1990,.
i the NRC granted the Temporary Waiver of Compliance.
1 The NRC staff's safety evaluation of the auxiliary feedwater automatic initiation system Technical Specification (TS) change request _is presented in i
the following three sections:
Section 1.1 - Plant _ Systems' Branch Section 1.2 - Reactor _ Systems Branch j
Section 1.3 - Human Factors Branch 1.1 -Plant Systems Branch Evaluation 1.1.1 Introduction The auxiliary feedwater system (AFW) for the Ha'ddam Neck Plant consists of two turbine-driven pumps and one motor driven pump with associated piping:
and controls. Each turbine driven pump normally provides feedwater to a 4
9010260106 901022 gDR ADOCK0500((3
j
.g.
common header which is~ cross-connected so'that one pump is able to provide flow to any or all of the four steam generators should such need arise.
The j
motor driven (MD) pump-may be utilized to supply feedwater.to:the steam generators any time a sufficient steam supply,is not available for operation of the turbines for the turbine driven (TD)-pumps._-However, only the TD pumps are safety-related and are depended on for mitigation of accidents and transients.
The HD pump is not safety-related nor automatically initiated,.
thus cannot be relied upon to mitigate transients and accidents, i
The licensee, Connecticut Yankee Atomic Power Company'(CYAPCO), uncovered two problems with automatic initiation of_ the _ turbine driven pumps for which the plant requires relief in order to operate during Cycle 16 with power 1.n excess of 10%.
1.1.2 Evaluation j
The licensee conducted tests which show that sudden _ full-opening of the steam.
3 inlet valves.to the turbines of.the TD pumps causes turbine overspeed and pump j
trip.
In order to prevent overspeed, the licensee has adjusted-the inlet valve 1
to open partially.
This partial opening, however, does not permit the turbine q
to develop the speed to provide the flow required to mitigate the worst j
transient for which it has been designed - loss of feedwater (LOFW).
In order to provide sufficient flow, operato'r action is required within a 4 minute time.
frame to further open the steam inlet.
l This safety evaluation does not address the acceptability of the time frame 'in which to allow operator action nor with the affect of such action on the f requency of loss-of-coolant or other accidents nor on the. frequency. of resultant core damage; these issues are addressed _in Sections 1.2 and 1.3.
i As discussed above, the steam admittance valve to the AFW turbines should 'not i
be allowed to open fully on automatic initiation because the turbines will trip on overspeed.
In order to prevent this, the licensee has adjusted the; valves to allow incomplete opening on'AFW initiation.
This is' accomplished by using control air to limit the valve opening.
Any event requiring'AFW-I operation but which results in loss of control air could result in turbine overspeed and pump trip.
The Haddam Neck control air _ system has not been designed in accordance with safety-related criteria and, thus, is susceptible to failure in a-seismic event or because of damage by tornados.
The licensee stated that.any failure resulting in slow loss of control air such asia small-
'1 break or compressor loss would not result in a sufficiently rapid loss of air
-so as to cause a complete, rapid opening of the steam admission valves to the AFW pump turbines and turbine overspeed and trip.
. In the event of AFW pump turbine overspeed and' trip it would be necessary to' restart the turbine manually.
As a_ mitigating action,' the licensee has agreed to have an operator _ dispatched to the local control' station for AFW pump:
turbine initiation'whenever the AFW system is initiated automatically.
The licensee committed to implement'a design change to resolve this issue during
-i i
. 3:.
i the Cy:le 16 refueling outage.
Therefore, permission to operate while relying on control air for automatic initiation of the AFW system would extend only through Cycle 16.
The licens% r.as added the following note to item 3. Auxiliary Feedwater, under tho heading " Functional Limit" in TS Tables 3.3-2, 3.3-3 and 4.3-2.
For Cycle 16 operation only, OPERABILITY of automatic initiation of auxiliary feedwater (AFW) is defined as including (1) credit for operator action to adjust AFW to full required flow following auto-matic initiation and (2) reliance on the contro1 air system.to ensure-
~
successful automatic AFW initiation.
Modifications will be implemented by the end of the Cycle 16 refueling outage, prior to startup for.
Cycle-17, to remove reliance on operator action and the control air system for successful automatic initiation of AFW.
The staff finds this acceptable, based on the discussion above.
1.1. 3 Conclusions The Plant Systems Branch staff finds.the licensee's proposal to require
. i operator action and the use of the control air system to operate the auxiliary feedwater system throughout Cycle 16 acceptable.- The licensee has added a note to TS Tables 3.3-2, 3.3-3, and 4.3-2 to indicate that' operability of automatic initiation of the AFW system requires operator action'and the control air system.
The staff finds the TS to be acceptable.
As stated in the TS notes, these modifications shall apply only throughout Cycle 16. opera-i tion prior to startup for Cycle 17.
i 1.2-Reactor Systems Branch Evaluation
- 1. 2.1 Introduction The auxiliary feedwater (AFW) system design of Haddam Neck Plant incorporates two turbine driven AFW pumps.
AFW flow is controlled by the steam admission valves to the turbines-and'by the feedwater bypass control valves on the pump discharge.
An automatic initiation system for AFW flow'was installed per the staff requirements after the Three Mile Island accident.
Rapidly, fully opening' the steam admission valve will result in a turbine overspeed trip.
Thus, the automatic initiation system was desir;.ied to only partially open the steam' admission valves. ~ The AFW flow rKe associated with. the partially open steam
,a admission valves was estimated to be sufficient for a design basis floss t of feedwater transient.
l.
DuringtheCycle~15refuelingoutage,ConnecticutYankeeAtomicPower[CompanyJ
(
(CYAPCO), the licensee for the Haddam Neck Plant, identified two issues..
regarding the automatic initiation of AFW flow..These two issues led the licensee to conclude that it is.now not in conformance with.its previous commitments on automatic initiation of.AFW flow. 'These issues are as follows:
l l
- 1) the recently identified nonconservatisms in the calculation of delivered l
i AFW flow have resulted in a reduction in the projected flow.
This has resulted in the determination that the calculated. flow rate achieved by automatic initiation of AFW alone is not sufficient to assure that the acceptance criteria of the design basis loss of.feedwater transient are met and 2) the recent testing confirmed that if there were a rapid depressurization of the control air system, the Terry turbines on the AFW t
pumps would trip on overspeed following their startup.
This leads to a conclusion that the safety grade automatic AFW initiation' system would-be relying on a non-safety grade control air system to perform its safety related l
function.
1 By letter dated August 25, 1990, supplemented by a letter dated August 29, 1990 and information telecopied to the staff on September 7, 1990, the licensee submitted a request for emergency changes to,the Technical-Specifications (TS) related to the operability of the automatic AFW initiation system.
The proposed TS changes would redefine the operatiility of the automatic AFW initiation by permitting operator actions to adjust AFW to full required flow following automatic initiation of.the AFW system and reliance on the control air-system to ensure successful automatic AFW initiation.
The licensee has requested that the proposed changes of TS would be effective l
during Cycle 16 operation only.
Modifications will be implemented by the end j
of the Cycle 16 refueling outage, prior to startup for Cycle 17, to remove reliance on operator actions and the _ control air system for successful l
automatic initiation of AFW flow.
1.2.2 Evaluation Following a loss of feedwater event, both of the turbine driven AFW pumps will receive a signal for automatic initiation.
Total AFW flow from two pumps will satisfy the design basis flow requirement.
However, if a. single failure is -
assumed on one pump, the remaining pump will deliver AFW flow slightly under the required flow rate.
In this case, operator actions inside the control room are needed to adjust AFW flow to the required AFW flow rate.
The licensee asserted that these operator actions could be accomplished within four minutes.
Since a loss of feedwater event.will-cause a reactor. trip very quickly, the operator would enter the emergency operating procedures (EOPs) E-0, " Reactor Trip or Safety Injection," and on the fourth. step of E-0, transfer to ES-0.1, " Reactor Trip Response," or transfer to FR-H.1,~ " Response'to Loss of Secondary Heat Sink." For 'both ES-0.1 and FR-H;1, one.of. the first steps is to verify adequate AFW flow and to take steps to manually achieve the required flow.
Therefore, taking manual control of the AFW system is covered -
in training and practiced routinely on the simulator.
The' licensee has performed a walkdown under this scenario following E0P steps and the required _
operator actions were accomplished in less than 3.5 minutes.
In_the design basis loss of feedwater analysis, there is a four minute time delay assumed for AFW flow delivered to the steam generators following the initiation of_ the loss of feedwater event.
This conservative assumption in th'e existing FSAR is i
i
y L
5-now supporting the operator action time required for AFW flow adjustment.
Also, the licensee stated that, by a best estimate. analysis, approximately 15 minutes are available for the required operator actions and the acceptance criteria for the transient are still met.
During the NRC staff review, a concern was raised that failure of the operator to increase AFW flow within the required time may cause the pressurizer to fill and result in the PORV opening with liquid relief.
This is a contributor to the frequency of a small break LOCA.
In response to the staff concern, the licensee has estimated that the small break LOCA probability will l
be increased by approximately nine percent.
The staf f judges that this <
increase in the calculated small break LOCA probability at Haddam Neck Plant is acceptable for one cycle.
In addition, as a compensatory measure, the-licensee has agreed to dispatch an operator to the AFW pump room whenever AFW is automatically initiated to ensure that local control.of the AFW system is available if required.
i The licensee stated that while the control air is a non-safety grade system, i
operating experience indicates that the system is highly reliable.
However, a l
rapid depressurization of control air could result from a break of a major pipe in the system.
The air line tubing is comprised primarily of copper tubing or stainless steel tubing, both of which are relatively ductile materials.
The staff agrees with the. licensee's assessment that a-catastrophic failure of the tubing is not likely. to occur during one fuel cycle.
Also, a walkdown performed by the: licensee has demonstrated that the operator actions needed to establish the AFW flow following tripping of-turbine driven AFW pumps due to control air system failure could be accomplished within 15 minutes.
Thus, in the best estimate scenario,.the pressurizer will not be filled and result in the PORV opening with liquid j
relief.
1.2.3 Conclusions l
l Based on the NRC Reactor System Branch staff evaluation in Section 1.2.2 i
above, the staff concludes that the licensee proposed Technical Specification i
Tables 3.3-2, 3.3-3, and 4.3-2 are-acceptable only during: Cycle 16 operation.
The staff requires that plant modifications'be implemented'by the end of the Cycle 16 refueling outage to remove-reliance on operator actions and control air system for successful automatic. initiation of'.AFW flow..
1
'I 1.3 Human. Factors Assessment Branch Evaluation 1.3.1 Introduction i
The revised AFW design basis analysis assumes that required AFW' flow would not be achieved until four-(4) minutes following;a. loss of feedwater event.
The-analysis assumes that within four minutes, required AFW flow would be provided by' automatic AFW initiation followed by operator action;to take manual control of the AFW system at the main. control board to increa'se flow, as necessary.
]
To support their position that operators could adjust AFW' flow manually within the analyzed time, the-licensee' stated that' operators are currentlyL It s
i
6-instructed in applicable Emergency Operating Procedures (EOP) to verify and,.
If necessary, to achieve the required AFW flow.
The licensee further stated.
that taking manual control of AFW and accomplishing the actions to increase flow are-covered in operator training and practiced routinely on the simulator.
To verify that operator actions and response _ times assumed in the design basis analysis are achievable, the licensee performed 'a walk-through of the new AFW initiation process.
In addition, the licensee performed a:
walk-through of the back-up operator response to an event requiring operator' actions to take control of the Terry' Turbine locally because of an overspeed y
trip.
The licensee also performed an AFW initiation event analysis with best estimate assumptions using guidance provided in ANSI /ANS 58.8, " Time _ Response
-i Design Criteria for Nuclear Safety Related Operator' Actions."-
i The docketed results of the licensee's walk-through times of required operator 1
actions, and the analysis of an AFW initiation event, were' submitted for staff.
review on September 19, IJ90.
The licensee performed walk-throughs and analyses for two cases.
Case 1 involved operator action in_the control room to adjust AFW flow to greater than 320 gpm'(as required by procedure)'to demonstrate that operator actions can achieve the required design basis AFW flow in the time required.
Case 2 involved-the back-up operator response to an AFW Terry Turbine overspeed trip requiring local manual actions ~to start an i
AFW turbine-driven pump to supply-the required AFW flow.
Following are summaries of walk-throughs:
I Case 1 - Simulating a_ loss.of feedwater, a normal shif t complement of
- 3 operators responded to the event using plant emergency operating procedures, j
Operators were required to read all steps of the procedures, simulate all.
required actions, and were instructed not to rush through the procedures, q
In addition, operators were not allowed to take action to restore the required AFW flow until the procedural step was reached that addresses _
1 establishing AFW flow.
Normally, operators are: allowed to proceed t'o a subsequent instruction in-a-procedure before a required task.is' fully.
completed, provided that there is assurance that the task is progressing satisfactorily (reference CYAPC0 procedure ACP 1.2-6.15;" Emergency Operating
]
Procedures User's Guide").
Timing was stopped when operators reached the point where the simulated AFW flow reached 320 gpm.'
The total' time to.
-i complete this simulation was three (3) minutes-and 24 seconds.
The.
design basis analysis time to reach the required AFW flow is-four (4) y minutes.
i Case 2 - Simulating a loss of feedwater event, a normalLshift complemen't
[
of operators complied with the committed action of paging an, Auxiliary Operator to proceed tn the Terry Turbine Room after recognizing that tne j
event had occurred.
Timing of this event included a 30 second delayi i
between the event initiation-and operator recognition that automatic AFW
-l initiation had occurred.
For worst case purposes, the Auxiliary Operator
~~
was assumed to be in the Screen House.
Upon arrival at'the: Terry TurbineL Room, the Auxiliary Operator was advised that both tu.rbines had tripped.on 4
L J
7 overspeed.
Timing was stopped after the Auxiliary Operator had simulated.
taking local control of a Terry Turbine.using the appropriate procedures j
in place and coordinating with control room operators to establish the -
required 320 gpm.
The total time required to complete this simulation was four (4) minutes and 38 seconds.
The design basis analysis time to reach-the required AFW flow does not apply to.this case.
The licensee's ANSI /ANS 58.8 analysis assumed that for the two cases considered, plant systems parameters were at realistic operating conditions rather than at the design basis analysis conditions.
Essentially this-difference changed the time allowed to establish greater than 320 gpm AFW flow-
~
f rom four minutes per the design basis analysis to :fif teen' minutes using.best estimate conditions.
A summary of the licensee's ANSI /ANS= 58.8 analysis;.is as follows:
Case 1:
Loss of feedwater event'with control room operators adjusting AFW.
I flow to the required amount,' included the follcwing elements:
Event initiation - Both main feed pumps trip.
Event alarm - Reactor trip due to low steam generator level coincident' with steam flow / feed flow mismatch (seven seconds af ter ' event initiation).
Operator actions alarm (indications to operator that AFW is needed);- Heat l
sink criticai safety function _ in a ' red path condition (critical safety.
l function in jeopardy-- immediate operator action required)..
1 L
Time margin -- time interval between event alarm and when ' operator ist ready to take action (five minutes).
' ~
Time margin complete -- Earliest time af ter the' event initiation that the operator can be credited with taking action (five minutes and -seveni seconds for this event).
Complete operator action / safety function - Time operator ' action and-safety function must be completed to ensure design criteria are' not exceeded (AFW flow greater than 320 gpm) (fif teen minutes)?
Latest time to initiate operator action =- Time 1 required;to complete safety-function minus time required to complete manipulations to!
establish required AFW flow..
The results of this analysis indicated that-there would be.a maximum.of. fifteen minutes available to complete the safety. function (i.e., establ..ish" required AFW flow).
The analysis also established that four minutes would be required for operators to complete the actions necessary to establish required AFW flow, which means that the latest operators could. initiate taction without exceeding design limits would be'11 minutes into the event.-, Based;on the above elements, the earliest operator actions could be taken din this event is adequate time for operators to establish ' required AFW flowland. remain within best case estimate limits.of fifteen minutes, i
i
,\\.
3 *
~
4
.g.
Case 2:
A seismic event causing loss of offsite power, which causes auto AFW initiation and loss of control air requiring local manual actions at' the AFW Terry Turbine to establish required flow, contained the same elements as in case 1.
The results of the analysis indicated that there would be.15 minutes-available to complete the safety function and six (6) minutes required for completion of-operator actions.
This means that the latest time operator action could be -
initiated is nine (9) minutes (15 minus 6) after the beginning of the event.
Because ANSI /ANS 58.8 does not allow credit for operator action outside the control room to be taken_for at least 30 minutes into the' event, the earliest time operator action could be initiated, in accordance with the Standard, is 30 minutes and two (2) seconds into the event (time' margin-complete element).
With the best case estimate time of fif teen minutes, there would-not be adequate time available, according to ANSI /ANS 58.8-guidance, for operators to locally establish AFW flow greater than 320 GPM, 1.3.2 Evaluation The licensee has demonstrated that operator actions can be taken within sufficient time to obtain the required AFW flow following automatic system initiation during a design basis event.
The licensee verified, via a-walk-through, that under design basis conditions for a loss' of feedwater event, a typical operating crew using existing emergency operating procedures could.
take manual control of the AFW system in the control room and establish the
.,f required AFW flow.
An additional walk-through was performed in which an.
automatic AFW initiation occurred, but control room functions were assumed not available because of an AFW Terry Turbine overspeed. trip. - This required an auxiliary operator to be dispatched to the Terry Turbine to reset the overspeed trip mechanism, and manually start and control ~the AFW turbine:while'in communication with control room operators.
Operators _ demonstrated that they' could perform the actions required by this event within realisticLoperating conditions (i.e., best case estimate time limit' of fif teen minutes)'.. This l
l event is outside the licensee's design basis analysis and therefore, the operator action times required by their design basisJanalysis di not apply.
The licensee also analyzed the two plant event scena-ios that'would require AFW-initiation assuming realistic rather than design basis conditions.
The analyses -
4 were done using the guidance provided in' ANSI /ANS 58.8,c Time Response Design Criteria'for Nuclear Safety Related Operator Actions.'" _ Thellicensee applied the s
guidance provided in ANSI /ANS 58.8 appropriately, and. demonstrated that control; room operators could accomplish the actions _ required to manually adjust AFW flow to the required levels to successfully mitigate the consequences of a loss of feedwater event, and prevent the condition from degrading into"an accident-condition.
In the scenario requiring operator action outside the control room to manually start and control the AFW turbine, because ANSI /ANS:58.8' states, that operator actions outside the-control. room cannot be taken for'30 minutes after the indication of an event,"the analysis showed that the required AFW flow would not be act.ieved within the' time assumed pacessary to prevent the
- event from degrading into an accident condition.'
' 11 ;
I j
y
4 i
' i
?
1.3.3 Conclusions Based on reviewing the results of the licensee's verification walk-throughs and analyses, the staff concludes that the licensee has provided reasonable assurance that the Haddam Neck Plant AFW automatic initiation system, including credit taken for operator actions, can meet the design basis analysis required flow rates within the response times necessary for a loss'of feedwater event.
The staff is satisfied with the licensee's statement that an event of this type is run routinely on the plant simulator for all operators, and is' thoroughly covered in operator training.
The staff ietermined that operating 1
the AFW system in this manner provides an acceptable level of safety, does not I
present undue risk to public health and safety, and is' an acceptable interim approach to be in effect-for cycle 16 only.
- 2. 0 EMERGENCY CIRCUMSTANCES U
Pursuant to 10 CFR 50.91(a)(6), CYAPC0 by-letter dated August 25, 1990,.
requested the NRC to approve this proposed amendment under emergency
(
circumstances.
By letter dated August 30, 1990, the NRC issued a Waiver of Compliance for specification 3.3.2 Table 3.3-2 Item 3 allowing power ascension above 10% with the auxiliary feedwater automatic initiation ~ system inoperable.-
In that letter the NRC informed CYAPC0 the NRC would-process.the proposed amendment under emergency circumstances as the waiver would allow the plant to progress with power ascension until the TS could be processed.
The Haddam Neck Plant is in the~ process of returning to operation following the Cycle 15 refueling outage.
As part of.the AFW system testing performed' during the preparations for. and during start-up, CYAPCO has identified two--
issues for which CYAPC0 concludes that.the plant is'not in conformance with-the TS requirements regarding operability of the:AFW automatic initiation system.
CYAPCO has determined that the calculated flow rate achieved by.
automatic initiation of AFW alone is not sufficient to assure that the criteria of the design basis loss of feedwater analysiscare met'and'that the safety grade automatic initiation system relies on' alnonsafety grade control-l air system.
TS Table 3.3-2 Item 3 restricts power level to'below 10% when the i
AFW automatic initiation system is inoperable.-
1 Emergency approval is necessary because an emergency situation' exists in that failure to act in-a timely way would prevent the increase >in power output up to the plant's licensed power level.
The ' emergency situation could not have.
been avoided.
The testing of the AFW pumps which demonstrated the reliance of the AFW system on control air system was conducted on August 12, 1990.
The plant went critical for the first time since the. outage. started on August 15,-
1990 and went subcritical again on August 17; 1990 for some minor repairs.
On August 18, 1990 the plant went critical and progressed to 9% power.
CYAPC0 held-power at 9% while they were determining their course:of. action because of AFW automatic initiation system problems.
On August' 22,.1990, CYAPC0 submitted a 50.72 report noting the deficiencies ofL the AFW automatic initiation system and declaring the system' inoperable.
As noted above, with t
4
.l
y
' l l
l I
the AFW automatic initiation system inoperable the plant cannot proceed above.
10% power.
By letter dat.ed August 22, 1990 CYAPC0 proposed to modify the TMI order to extend the TMI order for one cycle for a fully automatic'and safety.
grade AFW automatic initiation system.
The staff reviewed this request and stated this was not an appropriate avenue for redefining the operability I
requirement in the TS. The staff recommended that the TS be amended to redefine the operability requirements of the AFW automatic initiation system for one cycle until the system could be modified.
Therefore,.on August 25,-
1990, pursuant to 10 CFR 50.91(a)(5) CYAPC0 requested NRC emergency i
authorization and approval of the proposed amendment to define the operability l
of the AFW automatic initiation system as relying on operator action and the nonsafety control air system.
Thus, the staff does not believe that the I
licensee has abused the emergency provisions by failing to make a timely application.
Accordingly, the Commission has determined that emergency circumstances exist warranting prompt approval by the Commission,-in that failure to act will limit plant output below the design output, the situation could not'have been avoided and the amendment, as discussed in Section 5.0, does not involve a significant hazards consideration.
3.0 FINAL NO SIGNIFICANT HAZARDS CONSIDERATION
DETERMINATION l
The Commission's regulations.in 10 CFR 50.92 state that the Commission may.
l make a final determination that license amendment involves;no significant I
hazards considerations, if operation of. the facility, in accordance with the amendment would not:
l (1) Involve a significant increase in the probability of conse-l quences of any accident previously evaluated; or (2) Create the possibility of a new'or different kind of accident from any accident previously evaluated; or (3) Involve a significant reduction in a margin or safety.
This amendment has been evaluated against the standards in 10 CFR 50.92.. The does not involve a significant hazards consideration because the changes would l
not:
1.
Involve a significant increase-in the probability or consequences of an accident previously evaluated.
Although virtually all design basis accidents except large break loss of coolant accidents (LOCAs) either. explicitly or implicitly rely on auxiliary feedwater (AFW) for decay heat' removal, only the. loss of normal feedwater (FW) accident needs to be reviewed in detail;for impact since this accident is the most limiting from.the standpoint of AFW system' performance both in terms of timing and' minimum. flow req 0lrements.
I
+
. II.,
The proposed changes will not have any impact on the consequences of a i
loss of normal FW' accident.
Assuming that control air is available during system startup and that the operator manually increases flow, system performance will be within the assumptions of the loss of FW analysis.
No system parameters.till be affected.and system response will be as previously assumed.
Therefore, the proposed changes will not have any impact on consequences.
{
3 The proposed changes will have no significant impact on the probability of occurrence of any design basis accident..The proposed changes can af fect only 'the response of. AFW to a system challenge.
The changes cannot have any impact on the probability of a loss.of normal FW or any-other accidents.
However, the failure modes associated with'the proposed changes will slightly increase the probability of failure of AFW.
That is, wither a rapid loss of control air'during system initiation or a failure of the operator to' increase flow within four minutes.(assuming a successful start) would result in failure of the AFW system to meet its design basis requirements.
l
}
The reliance on control air to prevent an AFW pump turbine trip on_
l overspeed lasts for only a' very brief. period of. time.
A loss of control i
air af ter the pump has started will not result in an overspeed> trip.
I Similarly, a slow loss ~of control air, as would be expected following l
loss of a running compressor, will likely not result in an overspeed trip. Only a rapid loss of air would result in a trip of.the pump i
turbine.
Even if an overspeed trip were to occur, the pump.can be manually restarted by the operator locally in a very short time frame.
CYAPCO has committed to dispatch an' operator to_the. Terry turbine building in the-event of an automatic AFW system initiation.
This commitment provides:
further assurance that should an overspeed trip occur, the. pump can be.
manually restarted quickly.'
f i
It should be stressed that manual start of the AFWL umps does_not result I
p in AFW turbine trip.. Actual experience supports the conclusion that AFW pump turbines _ do not overspeed and trip during manual actuation.
Most failures-of the control air system would result sin a slow lossiof-l air pressure and would have no impact on1AFW' pump;startup..There are
{
very few failure mechanisms which could. result in airapid-loss,of air pressure.
These would include seismic or tornado induced failures.
Although the air tubing is.not seismically. designed, it is comprised primarily.-of copper tubing and-in some areas stainles's steel tubing, both of which are relatively-ductile materials.. Original-plant design
~
required the tubing to be supported-for deadweight and. thermal' loadings,-
1
- which provides for a very. flexible support system capable of.
accommodating large deformations; Therefore,=although the tubing is not j
d]
E d
~
12 -
seismically qualified, based on-engineering judgmen_tLit is concluded that a catastrophic tube failure, and rapid loss 0.f control' air pressure, is' unlikely to occur during'a seismic event.
The potential for failure during a seismic event or tornado cannot be.
ruled out entirely.
However, in letters dated September-15 and October 14,1980, CYAPC0 provided the NRC staff with justifications for' continued '
operation until required seismic upgrades have been completed.
i The Haddam Neck Probabilistic Safety Study (PSS) evaluated the
~
probability and consequences of-loss of main FW and loss of control air initiating events.
The PSS also considered the dependency 'of control air.
on support systems such as AC power and service water.
Although not specifically quantified, only a fraction of complete loss. of. air events could result in a sudden depressurization of;the. air system, based.on the design.
l q
The potential for AFW pump turbine trip on;overspeed due to a' sudden. loss' of control air was not included in 'the PSS'since this tiss'ue' was raised =
i after the study was completed.
However, this iss~ue isi expected.to have a i
modest impact on the core melt frequency.
The net effect of this~
increase, in combination with the -increase in frequency due to external
.j events, is a potential 3E-5/ year increase in. total core melt frequency for the nest operating cycle.- Based on this,'it;is concluded lthat.
I although this aspect of the change (reliance on controltair).will result in an increase in the probability of-failure 1oftthe'AFW, the increase in j
the failure probability.is ' acceptably low and is not'signi ficant.-
.{
i Even if there is no failure'in-the control' air system:and_the AFW pump starts successfully, operator action is still requires'within four' minutes from the time main FW-is lost in. order to successfully mitigate
.l the ' accident without resorting to m' r'e/ extreme means4of reactor coolant o
system heat removal (i.e., feed and bleed).
For accidents'other~than a loss.of normal FW, operator action would still be' required to increase flow although the time frame for this. action.would'be4slightly, increased, on the order of ten minutes.
If the operator fails toiperform'this action in the required time, AFW will not satisfy itsiintended' safety
?
l function.
Since the proposed ch'a'ge is effectively changing an'(assumed)'
{
n i
fully automatic system response to one which requires operator. action in :
a relatively short period of time, the probability thatythe" system will
~
failtits safety function is increased, j
i Although the probability of. system failure 'is increasedgdue to reliance.
1 on-operator action, the increase.in failure probability'is not i
significant.
One of the first actions in Emergency 0perating Procedure.
(E0P) ES-0.-1, " Reactor Trip, Response,"~ is to increase ATW flow to 320 gpm.
This E0p would be entered ?very quickly from E-0J" Reactor Trip or -
Safety Injection,"~during any accident.for which AFWrflow rate'would'be.
1 1
.)
I i!I k
- s l
13 critical because a loss of Fw event causes a reactor trip very quickly.
Taking manual control of AFW and the actions'which can be taken to increase flow are covered in training and practiced: routinely on the plant-specific simulator.
Experience has shown that these actions are taken very quickly.
Therefore, although the probability'of system failure is increased somewhat by reliance on short7 term operator action, the increase is judged to be small.
To summarize, the need for full automatic AFW flow in four minutes following a loss of normal FW event is a worst case analysis. assumption.
It includes conservative initial conditions and the' failure of one of the AFW pumps.
If one assumes more representative operating conditions and-that both AFw pumps.are available, the operator would have.significantly more time to ensure full AFW flow.
As -demonstrated on' the plant-specific simulator, the typical operator manually' initiates AFW -in approximately 30 seconds following the reactor trip.
This. is an evolution that is practiced of ten at the simulator during operator requalification training.
Therefore, if the operators have demonstrated that.they obtain_
.j full AFW flow in approximately 30 seconds and= analysis demonstrates that it is not needed for four minutes in the conservative case, then expecting operator action to manually adjust AFW flow after automatic initiation and within four minutes of the event is reasonable, o
The AFW operating station on the main control board was evaluated in'the Haddam Neck Plant Detailed Control Room Design: Review-(DCRDR).
Although no major problems were identified,_'four Human Engineering; Deficiencies-l (HEDs) were written.
Two HEDs pointed out that the'AFW controls and 1
displays were not immediately adjacent to-each other, although;the displays could be read from the control station.
The other-two HEDs-i noted that the operator was required to add four' individual flow meters
'j to obtain a total AFW flow number, although;a. total: AFW flow number is provided on the Safety Parameter Display System..CYAPC0 has concluded a
that these HEDs are relatively minor and=is evaluating' the need for:
correction within the Integrated Safety Assessment. Program - As an additional compensatory measure, operator training during the current operator requalification classroom sessions-will emphasize thi's concern ~
with the AFW system,
,i Proper adjustment of control air to limit the" travel' of (the steam.
admission valves is established by plant procedure; Plant procedure SUR 5.1-141, " Functional Test for Auto Initiation Scheme for AFW,"-
specifically lists as a prerequisite forfproper setpoint of the control knobs'to ensure that the steam admission valves open to.the correct j
position.
This is established based'on. Terry turbine steam inlet i
pressure.. The adjustment knobs are on the main 1 control board with-indicators that are used to verify that: the pneumatic signal-is set at -
i the appropriate value.
Also, procedure SUR 5.1-141, Step 6.2.12-,
instructs the operator to-record and verify the correctness of the i
turbine steam inlet pressure during actual. testing and.take action, if-necessary, to bring it back into the proper rangee 3
- 14 4
The proposed changes themselves have no impact on-the performance of the i
AFW system.
The changes only. consist of' reliance on control-air to prevent overspeed trip and on the operator-to manually increase' flow.
Actual system responses are not affected,- Therefore, the proposed changes are concluded to not result in a significant increase in the probability or consequences of an accident previously evaluated.
2.
Create the possibility of a new or dif ferent kindrof accident from any previously evaluated.
The proposed changes will have no impact on plant response to.any transient or accident.
Therefore, the proposed changes cannot create a new accident.
The recently identified dependence of the automaticfinitiation of AFW on
~
both operator action to adjust flow and the control air system.has; identified a new equipment failure sequence leading to a total 11oss of FW.
However, the probability of.such an equipment failure sequence (i.e., a catastrophic failure of theLcontrol air system or failure of'the operators to take action) has been shown to.be acceptably small.
Therefore, it is concluded that the failure modes do:not create a new or different type of accident from any previously evaluated.
3, Involve a significant reduction in a margin of safety.
Since there is no impact on the consequences of any accident,-there:can.
be no impact on any of the protective boundaries.
Crediting operator action to increase AFW flow within four minutes. of' a loss of FW : event Jand~
relying on the integrity of the control air system ensure that'the consequences of the design basis accident analysis'are unchang'ed.~
This-assures that none of the acceptance limits are exceeded and; represents no significant reduction :n the margin of safety.
3 Accordingly, the Commission has determined that this amendment ~ involves no significant hazards considerations.
4
4.0 STATE CONSULTATION
In accordance with the Commission's regulations, efforts were madefto contack the Connecticut State representatives. The state. representative was contacted on August 31, 1990 and had no-comments.
~
q
5.0 ENVIRONMENTAL CONSIDERATION
This amendment changes a requirement with respect to the' installation or use-1 of a f acility component located.within,the restricted area as def.inedtin 10 i
CFR Part 20.
We have determineds that the amendment-involves no significant increase in the amounts, and no significant change in the. types, of any-effluents that may be released offsite,-and that there is no.significant
'y
(,.
t
-. }
0?
/ii
't I
e 15 -
increase in individual or cumulative occupational radiation exposure.
The
~
staff has made a final no significant hazards consideration finding with.
l respect to this amendment.
Accordingly, the-amendment meets the eligibility criteria for categorical exclusion set forth in 10 CFR $51.22(c)(9).
Pursuant-to 10 CFR SSI.22(b), no environmental impact statement or environmental assess-
~
ment need be prepared in connection with the issuance of the amendment.
6.0 CONCLUSION
We have concluded, based on the considerations discussed above, that (1);there is reasonable assurance that the health and safety of the public will not1be endangered by operation in the proposed manner, and (2).such activities will be conducted in compliance with the Commission's regulations, and (3) the issuance of the amendment wili not be inimical to the common defense and' security or to the health and safety of the public.
.7
7.0 REFERENCES
1.
Letter from W.D. Romberg of Connecticut Yankee Atomic' Power Company to
~
USNRC " Proposed Changes to Technical Specifications on AFW Actuation System," August 25, 1990.
2.
Letter from W.D. Romberg of Connecticut Yankee = Atomic Power Company'to USNRC " Request for Additional: Information on AFW Actuation System,"
August 29, 1990.
Dated: October 22, 1990 1
Principal Contributors:
N. Wagner C. Liang J. Bongarra R. Correia if
'}
j t
- )
I^
- n a
..