U-603092, Forwards Response to NRC RAI for License Amend Request Re Div 3 DG Droop Issue.Supporting Info,Encl
| ML20196A082 | |
| Person / Time | |
|---|---|
| Site: | Clinton  |
| Issue date: | 11/20/1998 |
| From: | Walter MacFarland ILLINOIS POWER CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| Shared Package | |
| ML20196A088 | List: |
| References | |
| LS-98-010, LS-98-10, U-603092, NUDOCS 9811270048 | |
| Download: ML20196A082 (12) | |
Text
.... _. _.. -
m.m.
Ilknois Power Cornpany Cisnton Power Stastion P.O Box 678 21 935 23 I
Fax 217 935-4632 Walter G. MacFarland IV I
Senior Vice President i
and Chief Nuclear offcer
.ILLINSIS P9WER An Illinova Company U-603092 Sh.100a i
November 20,1998 1
Docket No. 50-461.
10CFR50.90 10CFR50.59 Document Control Desk Nuclear Regulatory Commission Washington, D.C. 20555
Subject:
Clinton Power Station Response to Request for Additional Information and Supplement for License Amendment Request Renardinn Division 3 Diesel Generator " Droop" Issue (LS-98-010)
Dear Madam or Sir:
By letter U-603067 dated August 24,1998, Illinois Power (IP) requested -
amendment of Facility Operating License NPF-62 for the Clinton Power Station (CPS).
Per that amendment application, which is currently under review by the NRC staff, IP j
proposes to revise the acceptance criteria fer meeting the " ready-to-load" requirement of Technical Speciftcation (TS) Surveillance Requirement (SR) 3.8.1.17 for the Division 3 Diesel Generator (DG) via proposed changes to the TS Bases [and to the CPS Updated Safety Analysis Report (USAR)].
1 As part of the NRC staff's continued review of the subject amendment request,
. telephone conversations between IP and NRC staff personnel were conducted on September 28,1998, and on September 30,1998, to address NRC questions concerning '
the information provided in IP's August 24,1998 submittal. IP was asked to document these questions and provide a written response via a follow-up letter. As such, f to this letter contains the requested additional information, pursuant to the
'l noted telephone conversations.
)
In addition, further consideration of the changes proposed by IP in the August 24,1998, amendment application has resulted in the identification of some additional god l
changes for both the CPS Updated Safety Analysis Report (USAR) and the TS Bases.
I l
These changes are provided in Attaciments 3 and 4 of this letter, respectively. It should be noted that the additional changes are consistent with those proposed in IP's August 24 application, and that the Basis for No Significant Hazards Consideration provided in the August 24 appycptgn remains unchanged and applicable.
c~
- i 9911270048 981120 I
PDR ADOCK 05000461 P
U-603092 Page 2 4
' Finally, in support of the additional changes addressed in this letter and its attachments, an affidavit supporting the facts set forth therein is provided as Attachment 1.
Sincerely yours, 0
if Mf, h'? Walter G. MacFarland, IV Senior Vice President and ChiefNuclear Officer TBE/EET/krk l
Attachment j
cc:
NRC Clinton Licensing Project Manager NRC Resident Office, V-690 Regional Administrator, Region III, USNRC l
Illinois Department of Nuclear Safety 4
i e
i
1.
Attachment I to U-603092 Page1of1 i
l George A. Hunger, Jr., being first duly sworn, deposes and says: That he is Manager-l l
Clinton Power Station; that this letter has been prepared under his supervision and direction; that he knows the contents thereof; and that to the best of his knowledge and belief said letter and the facts contained therein are true and correct.
l Date: This (O[ day ofNovember 1998, i
Signed:
W h, MA,(o
' George A. Hunger, Jrh
' I i
c-- - -
j STATE OF ILLINOIS l
SS.
'O7ICIAL MAL
- p themes 8.smeed stomry Pubsc,semes eiminels DE WrrT COUNTY J
toycommission Empires 11/4e001 ;
'._:__-____mw, Subscribed and sworn to before me this 2 O th day ofNovember 1998.
l 0
(Notary Public) l
\\
)
to U-603092 l
LS-98-010
)
j Page 1 of 6 Responses to NRC Questions Regarding IP Request for License j
Amendment for Division 3 Diesel Generator " Droop" Issue (LS-98-010)
)
Backaround:
i As documented in Licensee Event Report (LER)98-023 for Clinton Power Station (CPS), a review of the closing documentation for previous LER 96-012 raised questions regarding whether
]
Technical Specification (TS) Surveillance Requirement (SR) 3.8.1.17 was being met for the Division 3 emergency diesel generator (DG). SR 3.8.1.17 requires verification that, with the emergency DG operating in test mode and connected to its bus, an actual or simulated emergency 4
core cooling system (ECCS) signal overrides the test mode by returning the DG to " ready-to-4 load" operation and automatically energizes the emergency loads from offsite power. " Ready-to-4 load" operation is defined in the TS Bases for SR 3.8.1.17 as the DG running at rated speed and 4
voltage with the DG output breaker open.
1 The requirement is currently interpreted to mean that the DG should be returned to a " ready-to-load" condition automatically, i.e., with no operator action required, upon receipt of a LOCA l
actuation signal while the DG is in the test mode. The Division 3 DG at CPS, however, is l
equipped with a mechanical governor and is operated in a " droop" mode during surveillance testing when the DG is operated in parallel with the offsite power source. By adjusting the droop control setting for the DG, the frequency is set down (by approximately three percent) for a
~
portion of the time that the DG is under test (i.e., while it is paralleled with the offsite power j
source for the bus). This facilitates control of frequency during testing. Upon completion of testing, the governor is reset, which includes returning the droop control setting to zero.
While the droop is set at three percent, it is recognized that the DG may be in a condition such that, if a LOCA actuation signal were received, the DG might not attain rated speed upon starting l
and loading in response to the LOCA signal, if called upon to supply power to the Division 3 bus
]
in the event of a LOCA concurrent with a loss of offsite power (LOOP). Thus, although receipt of a LOCA signal during testing of the Division 3 DG does override the test mode to the extent that the control logic for the DG output breaker resets and the DG output breaker is ready to close, the Division 3 DG mechanical governor may require resetting (operator action to return droop control setting to zero or adjust engine speed from the MCR) to ensure that the DG will operate at rated speed when supplying the Division 3 safety bus.
1 For this reason, the DG is considered to be inoperable while the droop setting is in effect during surveillance testing. At the same time, it is recognized that the need for operator action to reset the governor is in conflict with the interpretation that SR 3.8.1.17 requires the DG to automatically return to a " ready-to-load" status upon receipt of a LOCA signal while in the test mode. The proposed change to the TS Bases for SR 3.8.1.17 (as described in IP's August 24, 1998, application for amendment of the CPS Operating License)is intended to resolve this confHct so that the SR can be met, although as described in the response to Question 6, the diesel generator will continue to be regarded as inoperable during testing with the droop setting in effect.
2
____._______-.____.m to U-603092 LS-98-010 Page 2 of 6 In addition to the above concern, TS SR 3.0.1 states that failure to meet an SR constitutes a failure to meet the associated Limiting Condition for Operation (LCO). Thus, if an SR cannot be met, the affected equipment must be declared inoperable. Based on the current interpretation of the requirements of SR 3.8.1.17, it must be concluded that the SR cannot be met. That is, with respect to the Division 3 DG and the requirements of SR 3.8.1.17, the DG is inoperable until either the DG is re-designed to be able to meet SR 3.8.1.17, or until the requirements of SR 3.8.1.17 are revised or clarified to permit manual / operator action to fully return the DG to a ready-to-load status for completion of the SR (by resetting the governor). Until then, the DG is inoperable even when it is not in the test mode, due to the requirements of SR 3.0.1. The proposed change to the Bases for SR 3.8.1.17 will resolve this conflict and will allow the DG to be considered OPERABLE (except when the droop setting is in effect).
l IP has identified this issue as an unreviewed safety question since (1) the TS " ready-to-load" acceptance criteria denoted in the Bases, as it is now and conservatively interpreted, is l
inconsistent with the Division 3 DG design, and (2) the acceptance criteria of manually resetting the governor is not specifically documented in the licensing basis or in the NRC's Safety Evaluation Report (SER) from previous reviews of the CPS design. Notwithstanding the unreviewed safety question, IP believes that the proposed change is acceptable from a plant safety / risk point of view, as further explained in the responses to the following questions.
NRC Ouestion:
(1)
How often and for how much time is the Division 3 Diesel Generator (DG) in the " droop" condition?
Response
The normal periodicity of running the Division 3 DG and inserting the " droop" into the circuitry is monthly, primarily due to Surveillance Requirement (SR) 3.8.1.3 which necessitates paralleling the DG with the offsite power source while the DG is connected to the Division 3 bus. For each time that the surveillance is performed, the time between when the " droop" setting is placed into the circuitry and when the " droop" setting is removed is one and one-half to two hours. Not all l
of this time is of concern, however, since the only time that the lowered droop setting can result in reduced bus frequency (upon receipt of a LOCA signal / override concurrent with a loss of the offsite power source to the Division 3 bus) is when the DG is unloaded or lightly loaded. This condition thus exists for only a short time (on the order of several minutes) out of the entire period that the droop setting is set down during each test, since for most of the test, the DG is fully loaded to the required level.
The above does not include the 24-hour surveillance run performed pursuant to SR 3.8.1.14.
This surveillance, which is performed approximately every 18 months, adds approximately 25 i
i hours to the amount of time that the " droop" setting is in effect for the DG control system.
)
to U 603092 LS-98-010
-I Page 3 of 6 However, in light of the actual number of minutes that the droop setting puts the DG "at risk" for each monthly test, as well as for the 24-hour surveillance, a maximum time of 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> per year is a conservative assumption for the total time that the Division 3 DG and bus mty be consdered
" vulnerable" due to the droop setting. This is the total time assumed for the risk analysis performed for the Division 3 DG, as further described in the response to Question 5. Specifically, on a per-year basis, the fraction of time that the DG droop setting is "in effect" is as follows:
f = (24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> / year)/(8760 hours0.101 days <br />2.433 hours <br />0.0145 weeks <br />0.00333 months <br /> / year) = 2.74E-3 NRC Ouestion:
(2)
What is the shutdown risk associated with the 24-hour surveillance as a result of having the " droop" setting in effect for the DG control system?
ResD0nse As described above, the only time during testing that the lowered droop setting can result in reduced bus frequency upon receipt ofLOCA override signal (concurrent with a loss of the offsite source for the bus) is when the DG is unloaded or lightly loaded. Throughout the 24-hour run, the DG is loaded at, near, or in excess ofits nominal rating, and is unloaded or lightly loaded only for a short period of time (primarily during the startup and initial loading phase). Again, this time would be only on the order of several minutes. When this risk is comb ~med with the reduced ri:k of a LOCA that exists during shutdown conditions, the shutdown risk associated with the droop j
condition is very small.
Because of the very small risk involved, no specific, quantitative analysis of the impact of the potentially reduced Division 3 bus frequency has been performed. The droop setting normally used during testing of the DG could result in a reduction of bus frequency of up to 3% (to 58.2 Hz) if the offsite source was separated at a particular moment during testing (as described above).
Pump speeds would thus be reduced comparably. The most obvious impact would be reduced flow from the HPCS pump. However, required HPCS flow is based on the LOCA analyms which are based on conservative assumptions and worst-case or bounding conditions, including the assumption that the plant is at full power. Since the 24-hour run required by SR 3.8.1.14 is only performed during shutdown conditions, conditions existing during such testing would be much less limiting then those assumed for the LOCA analyses, and the impact of any slightly reduced flow would not be significant.
FRC Ouestion:
(3)
What procedure guidance is there for control of removing DGs from service?
to U-603092 LS-98 010 Page 4 of 6
Response
During plant operation, when all three DGs are required to be operable, the process of scheduling maintenance and surveillance testing at CPS includes controls to prevent simultaneous testing of the Division 3 DG when another DG is out of service (OOS). Specifically, the on-line work management process specifies that applicable Technical Specifications are complied with, impact matrices are reviewed, and limiting conditions of operation are met. These controls do not allow for surveillance testing on one DG while another is OOS.
With regard to testing of the Division 3 DG, and as noted in the response to Question 6, the Division 3 DG is considered to be inoperable during testing with the droop setting in effect.
NRC Ouestion:
(4)
What operator actions can be taken to recover from the " droop" setting if the DG were needed to supply the Division 3 safety bus?
Response
Considering the scenario where a LOOP or LOCA occurs during the time the Division 3 DG
" droop" is placed in the circuitry, it is a simple operator action to reset the DG " droop" locally in the DG room. This action can be accomplished in less than a couple of minutes. In addition, an operator in the main control room can overcome the " droop" setting by increasing the DG speed to bring bus frequency up to 60Hz. It should also be noted that if an offsite source were to become available after a LOOP, power to the bus could be recovered during the event, and bus frequency would be restored to 60 Hz from the offsite power source.
Given the conditions under which the above operator actions would have to occur, however, credit was not taken for them in the analysis discussed under Question 5. In addition, since the CPS emergency operating procedures (EOPs) are symptom-based, during accident conditions plant operators would be concentrating on reactor power, reactor vessel level, and reactor vessel pressure. Therefore, the fact that the HPCS pump is not producing full flow may not be immediately apparent to the operator. The redaced flow at a slightly reduced bus frequency may not even be noticeable. However, if the reduced flow would cause the reactor water level to continue to drop or not to recover, operators could start other systems as directed by the EOPs rather than troubleshoot the reason for reduced HPCS flow.
NRC Ouestion:
(5)
Provide a risk assessment of the scenario where the Division 3 DG is in the " droop" mode for testing and the would be required to provide AC power for the Division 3 4.16kV safety bus in the event of a LOCA.
l
. - -. - ~. - -. - - -
to U-603092 LS-98-010 i
Page 5 of 6 Ets.22n.s.g A risk assessment was conducted to determine the likelihood of the HPCS being unable to perform its function of reactor coolant inventory makeup due to the Division 3 diesel generator being in the " droop" condition. As long as the Division 3 bus is being supplied by the offsite power source, HPCS operation cannot be affected by the " droop" condition. Therefore, the HPCS system can only be adversely affected if a LOOP occurs resulting in the Division 3 bus being supplied by the diesel.
Without the inventory loss associated with a LOCA, only a small fraction of the HPCS rated flow would be required for inventory makeup to the reactor vessel. Therefore if the HPCS were run at the reduced frequency associated with diesel generator " droop" conditions, it would be able to supply adequate inventory makeup to avert core damage under non-LOCA accident initiator conditions. Even under realistic LOCA analysis, the slight reduction in HPCS flow would not necessarily result in any increased consequences. However, for purposes of this risk assessment,
)
it is assumed that HPCS would be unable to meet LOCA inventory makeup requirements ifit g
were being supplied by the diesel generator under " droop" conditions. Therefore, the only time HPCS is assumed to fail for purposes of this analysis is if the Division 3 diesel generator is in the
" droop" condition concurrent with a LOOP and a LOCA.
Normally, in the context of PRA analyses, a LOCA and a LOOP are considered to be independent events. However, the conditional probability of a LOOP given a LOCA can be determined as an input to a risk analysis. Given a LOCA initiating frequency, therefore, a frequency of occurrence can be calculated for the event in which a LOOP and LOCA occur concurrently with the Division 3 DG in a " droop" condition. This is conservative in light of the fact that, as previously noted, the DG is really only susceptible to operation at a reduced frequency (in the event of LOCA and associated LOOP) for a small fraction of the time that the DG is in a " droop" condition (as defined by the DG droop control being set to the 50% setting).
CPS LOCA Freauency of Occurrence:
The LOCA initiator is divided into three magnitudes for analysis considerations, along with an inadvertent open relief valve (IORV). If a small LOCA or IORV initiator is postulated in the CPS PRA, the High Pressure Core Spray (HPCS) system is capable of providing adequate makeup, even if the " droop" setting were in effect. On this basis, the medium and large LOCA and the Interfacing System LOCA (ISLOCA) initiators are considered to be the components of a LOCA for the risk analysis performed for the " droop" issue. The combined initiating frequency for these events is 9.05E-5/ year.
Conditional Probability of LOOP Given a LOCA:
The current CPS PRA does not include evaluations of Generic Safety Issue (GSI)-171 regarding a LOCA with delayed LOOP or LOOP with delayed LOCA. In the current PRA, and as noted previously, these initiators are treated as independent events and are not triggered by the occurrence of the other. Howtver, NUREG/CR-6438, " Evaluation of LOCA With Delayed
to U 603092 LS 98410 j
Page 6 of 6 LOOP and LOOP With Delayed LOCA Accident Scenarios," contains an estimated probability of a LOOP given a LOCA (LOCA/ LOOP), which is 6.0E-2 for BWRs.
Result:
i The probability of occurrence (Pw) of the " droop" condition existing concurrent vAh a LOOP and LOCAis calculated as follows:
Fraction ofTime LOCA (Initiating Conditional P=
Droop Condition x
Ewnt) Frequency x
Probability of 4
Is in Effect LOOP with LOCA P=
(2.74E-3)(9.05E-5/ year)(6.0E-2) l P=
1.5E-8/ year This frequency is sufficiently low so that further analysis of the censequences of reduced HPCS flow in this situation was not warranted. Even ifreduced HPCS flow is assumed, this case does j
not represent a significant risk.
Risk Assessment:
The probabilistic approach shows that the occurrence of the " droop" condition existing concurrent with a LOOP and LOCA is acceptably low and does not represent a significant risk.
Further, as noted previously, there are different operator actions that can be performed to recover from, or overcome, the Division 3 DG " droop" effect, though such actions were not credited in the above analysis.
NRC Ouestion:
(6)
Per IP's requested change, would the division 3 diesel generator be considered OPERABLE with the droop setting in effect?
Response
I
- As briefly noted in the Background section of this attachment, even after issuance of the license amendment proposed by IP, the Division 3 DG would continue to be considered inoperable while the droop setting is in effect. That is, although the proposed amendment would enable the
~;-
requirements of SR 3.8.1.17 to be met by crediting operator action to reset the DG governor and thus complete the surveillance, the DG must be considered inoperable with the droop setting in effect. This is based on the possibility of the DG not being able to automatically attain rated
.spee od r frequency if a LOCA and LOOP condition occurred during testing of the DG with the droop setting in effect.
4 N
W
. -.-~. _ -... - - - ---
to U-603092 LS-98-010 Page 1 of16 Proposed Changes to Clinton Power Station (CPS)
Updated Safety Analysis Report (USAR)
Backaround l
With respect to the design, testing, and operation of emergency standby diesel generators (DGs),
l guidance is provided by applicable industry standards published by the Institute of Electrical and l
Electronics Engineers, Inc. (EEE). These applicable EEE standards are endorsed, with certain exceptions or clarifications, in applicable Regulatory Guides (RGs) published by the NRC, including RG 1.9, " Selection, Design and Qualification of Diesel-Generator Units Used as Standby (Onsite) Electric Power Systems at Nuclear Power Plants," RG 1.32, " Criteria for Safety-Related Electric Power Systems for Nuclear Power Plants," and RG 1.108, " Periodic Testing ofDiesel-Generator Units Used as Onsite Electric Power Systems at Nuclear Power Plants."
Section 5.6.1.4 ofIEEE 387 ("EEE Standard Criteria for Diesel-Generator Units Applied as Standby Power Supplies for Nuclear Power Generating Stations") addresses governor design for standby diesel generators, including both the isochronous and droop modes of operation. It states, "If the diesel engine is equipped to operate in either the isochronous or the droop mode, provisions shall be included to automatically place the engine governor in an acceptable mode of operation when the diesel generator unit is required to operate automatically." In addition, Section 5.6.2.2, " Automatic Control," states, "Upon receipt of a start-diesel signal the automatic control system shall provide automatic startup and automatic adjustment of speed and voltage to a ready-to-load condition." Part (1) of this section further states, "A start-diesel signal shall override all other operating modes and return control of the diesel-generator unit to the automatic control system." These requirements are consistent with EEE 308 ("IEEE Standard Criteria for Class 1E Power Systems for Nuclear Power Generating Stations"), Section 6.2.6(2), which states,
" Provisions shall be made for automatic switchover from system test mode to operate mode in case of an accident signal." (This particular IEEE 308 requirement is the one referred to in the TS Bases for SR 3.8.1.17.)
In the CPS USAR, design differences between the Division 3 DG and the Division 1 and 2 DGs are noted due largely to the fact that the Division 3 DG is dedicated primarily to a single large load (the high pressure core spray system). Consequently, differences exist in the sections of the USAR that identify what regulatory guidance and design requirements are met by the Division 3 DG design versus the design fer the Division 1 and 2 DGs. For the Division 3 DG, an important design and licensing basis document referred to in the USAR is NEDO-10905, "High-Pressure Core Spray System [HPCS) Power Supply Unit," by General Electric. Exceptions to applicable design and regulatory basis documents are taken in the USAR for the Division 3 DG based on differences between the design described in NEDO-10905 and the requirements specified in general regulatory guidance documents for DG design, operation, and testing, such as RGs 1.32 and 1.9. Further, evaluation of the HPCS DG design against some of the industry standards, including IEEE 387, is also included in NEDO-10905.
to U-603092 LS-98-010 Page 2 of16 Although the design basis for the Division 3 DG is firmly established within the licensing basis for CPS, as described in the CPS USAR and as acknowledged in the NRC's Safety Evaluation Report (NUREG 0853) for CPS, further review ofNEDO-10905 and the CPS USAR has confinned a need to clarify the Division 3 DG design as described in the USAR, including conformance to applicable IEEE standards and/or regulatory guidance, with regard to the i
" droop" mode of operation associated with the Division 3 DG due to its mechanical governor.
Currently, this aspect of the Division 3 DG design is not sufficiently addressed in the USAR.
Description of Proposed Channes i
j-The attached changes are based on the droop mode of operation for the Division 3 DG and its
~
potential impact on the capability of the DG to attain a ready-to-load status after automatic switchover from the test mode upon receipt of a LOCA initiation signal, as described in. The attached changes are summarized as follows.
(1)
The description of compliance for Regulatory Guides 1.9 and 1.32, contained in USAR 4
]
Section 1.8, are being revised to reflect changes to the applicable USAR Subsections.
j.
Another change to USAR Section 1.8 is the clarification being added to the description of compliance for Regulatory Guide 1.108. The description of the CPS Project Position i
relative to Regulatory Guide 1.108 is being changed by adding a section pertaining to compliance with Section C.I.b(3). Specifically, operation of the Division 3 DG in the 1
droop mode during testing may require operator action to reset the governor to ensure bus frequency is within required limits if a LOCA occurs concurrent with a loss of offsite power.
(2)
USAR Section 8.1.6.2.13 is being revised to describe that, except as also clarified in j
USAR Subsection 8.3.1.2.2, the Division 3 diesel generator meets the applicable i
requirements ofIEEE 387.
(3; Page 23 of USAR Section 8.3.1.1.2 is being revised to describe that the test controls associated with periodic testing of the Division 3 diesel generator are overridden during a i
LOCA event, except as described in USAR Subsections 8.3.1.1.2.1 and 8.3.1.2.2.
(4)
USAR Section 8.3.1.1.2.1 is being revised by adding two paragraphs describing that operation of the Division 3 DG in the droop mode during testing may require operator action to reset the governor to ensure bus frequency is within required limits if a LOCA occurs concurrent with a loss of offsite power.
(5)
USAR Section 8.3.1.2.2 is being revised by adding a paragraph to clarify that relative to RG 1.9, Position C.7, operation of the Division 3 DG in the droop mode during testing may require operator action to reset the governor to ensure bus frequency is within required limits if a LOCA occurs concurrent with a loss of offsite power.
p
[
i to U-603092 i
LS-98 010 Page 3 of16 Natt The attached, proposed changes to the CPS USAR are consistent with, but supersede, the i
proposed changes provided in IP's application for amendment dated August 24,1998 (IP letter l
U-603%7).
i.
i t
,