ML23109A082
ML23109A082 | |
Person / Time | |
---|---|
Issue date: | 04/13/2023 |
From: | NRC/OCIO |
To: | |
Shared Package | |
ML23109A075 | List: |
References | |
NRC-2022-000160 | |
Download: ML23109A082 (35) | |
Text
TM:i!l ~le eePJttt!11s SR:h'C:etf.
Duke Power C.ompany .
OCONEE NUCLEAR *STATION Unit3 Probabilistic Risk Assessment Volume2
3.3 FLOOD ANALYSIS 3.3. l INTRODUCTION Floods are external hazards that contribute to overall plant risk by producing multiple component failures when critical areas fill with water and submerge vital systems . These failures can lead to core melt. This flooding analysis encompasses both. internal and external flooding events .
Two potential events were found that could lead to externa.l flooding of the Oconee site . The first is a general flooding of the rivers and reservoirs in the area due to a rainfall in excess of the probable maximum precipitati on (PMP) . Since the Oconee site is well inland, the effects of hurricanes were not considered . The second source of external flooding is a possible random failure of the Jocassee Dam. Random dam failures include all causes other than a rain-induced failure (which will be discussed below) or an earthquake-induced failure (see Section 3.2) .
A break or breech of any of the many internal plant systems containing water could result in internal plant flooding. I (b)(7)(F), (b)(3).16 U .S C. § 8~4o-l'.d}
I
- 3. 3. 2 METHODOLOGY 3.3.2.l External Flooding from Precipitation
- l(b)(7)(F). (b)(3)16 U S.C § &2-!i,-l(d)
The effects of this PMP on reservoirs and spillways were evaluated in a study performed by the Duke Power Company (1966} . The results of this study demon-strated that the Keowee and Jocassee reservoirs are designed to contain and control the floods that could result from a PMP . Thus , i n order to flood the pl ant site , ra i nfall exceeding the PMP must occur. The frequency of exceeding the PMP was obtained from the analysi s presented in Reference 1 for the Oconee si te and was used as a bounding estimate of the frequency of core melt due to rain- induced external flooding.
3.3-1 Rev. l
The analysis yields the following frequencies of a PMP associated with the lower, median, and upper bounds of the probability distribution:
Cumulative PMP frequency probabi1 ity (per yr) 0.05 4.9E-8 0.50 2.9E-7 0.95 8.9E-7 The cumulative probability for a particular frequency interval is to be tnter-preted as the degr ee of certainty that the PMP frequen cy observed over a long period of time will be less than or equal to the upper value of that frequency interval.
Appendix A.13 describes the Oconee Standby Shutdown Facility {SSF) . \
(b)(7)(F). (b/(3).16 U S (' ~ 81-lo-l(dJ
\
Applying the SSF as a response to exceeding the PMP results in I (bJO\(F), (b)(}) 115 USC § 8'.!~o"l(dt
\ For both these reasons, it was concluded L------------------'
that precipitation-induced external flooding is a negligible contributor to core-melt frequency and public risk.
3.3.2.2 External Flooding From Dam Failure The Oconee site has a yard grade elevation a few feet below the full-pond level of Lake Keowee, which serves as the source of its condenser circul~ting water.
Lake Jocassee has a full-pond elevation about 300 feet above Lake Keowee. If a sudden failure of the Jocassee Dam were to occur, and a rapid enough release of the impounded water from Lake Jocassee into Lake Keowee resulted, the flood wave generated in Lake Keowee would overtop the Keowee Dam and the Oconee 3.3-2 Rev. l
intake dike, flooding the plant. This section presents the analysis performed to estimate the frequency of such a flood.
Frequency of Dam Failure The Jocassee Dam is an earth-rockfi11 structure approximately 400 feet high.
The dam was completed in 1972. and the reservoir was filled by April 1974. The spillway lies along one of. the abutments, about one-quarter of a mile from the dam, and is a concrete structure founded on granite.
An analysis was performed to determine an annual frequency of failure for earth, earth-rockfill, and rockfill dams due to events other than overtopping and earthquake ground shaking, which were considered in separate analyses.
Also, based on dam design information, structural failure of the spillway during discharge and failure associated with seepage along an outlet works have been eliminated as a possible failure mechanism. The following principal modes of failure were considered:
- 1. Piping.
- 2. Seepage.
- 3. Embankment slides.
- 4. Structural failure of the foundation or abutments.
These failure mechanisms are referred to collectively as random failures. Only failures resulting in the complete collapse of the structure and the uncon-trolled release of the reservoir's contents were considered to have the poten*
tial for flooding the Oconee plant.
Previous investigations into the frequency of dam failure indicate that it decreases with later years of constructio~ (Baecher et al., 1980). This is generally attributed to improvements in the methods of design and construction.
Therefore, another criterion* considered in developing a data base and failure-freQuency estimate was the period of construction.
The age of a dam is another factor that has been identified as having an effect on the rate of dam failure. Approximately half the dam failures occur during 3.3-3 Rev. 1
the first 5 years of operation (Reference 3). Therefore, age was a1so considered in developing a data base.
Size, type of construction, realistic failure modes, period of construction, and age were the major considerations used to define a data base for use in estimating the failure frequency of the Jocassee Dam.
The data base characteristics that were attributed to the Jocassee Dam are:
Characteristic Jocassee Dam Location (country) United States Year completed 1972 Age (years in operation) 13 Height (feet) 400 Type Earth-rockfill Data Various catalogs were used to develop the data groups that were studied. Each group consisted of large earth, earth-rockfill. or rockfill dams (more than 45 feet high, Reference 15) in the United States that were in operation 6 or more years when they failed. Of the references used in this study, not one is a complete catalog, and therefore they were used collectively. At present. these references represent the best readily available infonnation. From the various listings of failures, cross-checks were made when possible.
A data base uniquely suited in every major respect to the Jocassee Dam was unattainable because of a scarcity of the number of the earth-rockfill type.
The data base ultimately developed reflects discussions with Duke Power eng1-neers familiar w;th the characteristics of the Jocassee Dam. It was decided that the data base should include only the failure modes that could occur at Jocassee. The two major failure types excluded from the data set were fai1ures resulting from piping at a conduit passing through the dam and structural failures of the spillway during the flood discharge. Neither of these failures 3.3-4 Rev. 1
can occur because the necessary physical conditions do not exist at Jocassee.
(Note, however, that dams in the data base do include those that can fail in either or bot~ of these modes. This is proper because the experience from these dams represents realizations of nonfailure for other failure modes, such as embankment piping, foundation failure, and slope failure.}
Because of limitations in the historical rec0rd 1 it is possibly only to develop a data set that takes into account a limited number of the specific properties of the Jocassee Dam. Since Jocassee is a structure designed and constructed in recent times, it can be assumed that state-of-the-art technology was used in its design. In addition, because of its role as a critical facility of large size and importance, other aspects of the dam, such as seepage monitoring and inspection programs, are important factors that decrease. the likelihood that the dam will fail. The following specific characteristics of Jocassee are identif;ed as relevant factors that will affect the frequency of failure:
- l. Quality maintenance and inspection programs.
- 2. Monitoring of the dam (i.e., seepage, settlement1 .etc.).
- 3. Presence of personnel at the site.
- 4. Responsiveness of the owner to potential problems (i.e., implementation of emergency plans).
- 5. Detailed geologic investigations conducted before site selection.
- 6. Experience of earth-rockfi11 dams in Jacassee'.s class (w;th respect to random failures).
- 7. Design techniques.
These factors notwithstanding 1 the data were examined, and the best available data base for application to Jocassee Dam was extracted and used.
The data base covered the period of dam construction from 1940 to 19B7. In this period U.S. dams in operation 6 or more years at the time of failure and 45 feet or more in height were considered. The dams were of three types--earth, earth-rockfill, or rockfi11--and only catastrophic failures were included. Table 3.3-1 lists the failures consi~ered in the analysis.
3.3-S Rev. 1
The number cf dam-yea rs cf operati on was determined from data on the rate of construction provided in References 15 and 16. Table 3.3-2 1.ists chrono-logically for the period of construction the year a failure occurred, the interval between each failure in years, the cumu1ative number .of dam-years to the year of failure, and the number of dam-years between failures.
lb11i\(fl (b)(31 !61J H h!.4.:i-ltdj
________________________________ I _.
To the extent that this class is representati~e of the Jocassee Dam, these results can be interpreted as the predicted annual failure frequency of the Jccassee Dam .from causes other than eart_hquakes or overtoppi ng.
In order to evaluate the contributi on of random failure of Jocassee Dam to the frequency of core melt at Oconee Unit 3, three factors must be quantified. The first factor was the frequency of random dam failures. The remaining two factors are the following:
- l. The conditional probability of flooding of the Oconee site given a failure of the Jocassee Dam.
- 2. The conditional probability of core melt given floe.ding at the Oconee site.
3.3-6 Rev . l
The data base includes only catastrophic failures by modes believed app l icable to Jocassee. Most of the catastrophic
. failures reported in .the . literature for earth or earth-rockfi 11 dams were total (i.e., the entire reservoi r emptied) ,
and most t imes to fa i lure were i n the range of l to 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />. However, the f ail ures i n the data base (Table 3. 3-1 ) occurred onl y in earthen dams; thus, there is some Questi on as to how app l icable their rates of erosion and depths of failure are to the Jocassee Dam, an eartn-rock.fi 11 structure.
The Hell Hole Dam, an earth-rockfill structure, failed during construction in 1964 when extreme precipitation caused overtopping. Since construction had not been completed and since the dam was overtopped, it did not satisfy the data-base characteristics and was not included in the failure set . . However, its material of construction and size were very similar to those of Jocassee, and its failure behavior can therefore be used as one point of reference in judg-ments about the possible failure behavior of Jocassee. The overtopping of this partly constructed dam lasted for more than 40 hours4.62963e-4 days <br />0.0111 hours <br />6.613757e-5 weeks <br />1.522e-5 months <br /> before a catastrophic failure occurred. This suggests that a long warning time may be available, at least in some cases, before the failure of an earth-rockfill dam. Once the dam was breached, however, the breach propagated the full depth of the dam, down to the foundation, in about 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />. Thus, though more warning time may have been available, once failure began, the time and depth to complete breach were quite similar to those observed for earthen dams.
In addition to the uncertainty in the important parameters described above, a spectrum of calculated flood levels is not available. Thus, it is not possible to determine where the sensitive range for the key parameters lies. Given this lack of information, a bounding value of 1 was used for the conditional proba-bility of flooding of the Oconee site given a catastrophic failure of the Jocassee Dam.
3.3-7 Rev. l
The probabi1 ity of core melt given flooding at the Oconee site depends on the warning t;me , actions taken by operators, the depth of the flooding and other factors. The discussion of the possible role of the SSF in Section 3.3.2. 1 would also apply to this set of sequences, given *.the discussion of the time to complete failure of the Hell Hole Dam. Three SSF recoveries are appropriate for external flooding sequences: I I
3.3.2 . 3 Auxiliary Building Flooding Review of Historical Data LER, SER and SOER data were reviewed through December, 1987 to determine how many incidents of Auxiliary Building flooding had occurred in U.S. pressurized water reactor plants. Table C-1 shows the information on U.S. PWR and BWR operating experience, while Table 3.3-3 contains descriptions of applicable Auxiliary Building flooding events.
Classification of Initiators Poss1ble Auxiliary Building floods were classified as follows :
- 1. ~ - Flooding on the order of hundreds of gallons (e.g . , valve-pit flooding, flooding of an instrument, or flooding within a compartment) .
- 2. Moderate. Flooding on the order of several thousands of gallons (e.g., a few feet of water on the floor of a typical pump room).
- 3. Large. Flooding on the order of tens of thousands of gallons of water (e.g., a few feet of water in large rooms, very deep water (more than 10 feet) in a typical pump room).
3.3-8 Rev . l
- 4. Verx large. Flooding on the order of hundreds of thousands of gallons (e.g., floods involving the ruptures of major CCW or service-water piping).
It should be noted that these severity levels are based on the volume of spilled water . A different severity measure is used for the CCW-induced floods in the turbine bui 1ding . Severity can be def; ned in many ways . For example, it could be defined in terms of spill rate and duration of spill. It should be noted that the extent of damage or impact on the plant safety margin is a plant-specific issue and cannot be attributed to a particular severity level before the plant-equipment 1ayout is analyzed.
Identification of Critical Areas A critical area was defined as an area where a flood could cause an initiating event, fai 1 the re 1ated mitigating systems. or cause both with a high frequency relative to nonflood contributions. This implies that there is a high poten*
tia1 for damage and a credible source of flooding.
The initiating events can be divided into two groups : LOCAs and transients (e . g . , a reactor trip) . The LOCAs can occur through four ~ai1ure modes: RCS pipe rupture, inadvertent opening of isolation valves, a relief valve openi ng and failing to reclose, and reactor coolant pump (RCP) seal failure . The first two are very .unlikely to result from a flood. No mechanism could be identified to relate flooding to pipe rupture. Furthennore, a LOCA leading to containment flooding does not necessarily lead to the failure of the LOCA-mitigating functions. \
I
. The fa i1 ure of a pressurizer relief valve to reclose is unlikely to be affected by a flood. This event has been included whenever a transient initiating event is followed by failures resulting in an increase in RCS pressure to the relief 3.3-9 Rev. 1
valve setpoint. \
(b)(7)(F). (b)(3)*16 U .S C. § 8140-l(d; A transient initiating event may result when any equipment needed to ~ustain power operation is affected. It is judged that *th~ likelihood of a manually initiated reactor trip becomes very high in some cases where *the failure does not cause a direct trip.
Based on a review of plant systems, interdependencies, and the physical layout of the plant, the following areas were identified as potentially critical flooding areas:
(bX1)(F). (b)(3)c16 U.S.C. § 82~o-l(dJ 1.
2.
3.
~.,c~n 4.
5.
6.
7.
- b)/7)(F), lbX3)16 U.S.C § 82~o-l(d)
SEHSffPlE SECUR:ffY R£1:A'Pim IHfORMA'PION Cilt'PtCAL EHEROWEtEC'PRtCA-t lHfRA-S'fftUC'fU~f: tNf'OftMA'ftOtq 3.3-10 Rev. 1
(b)(7J(F), (b)(3)1 6 US.C. § 8~4o-l(d)
- 3. 3-11 Rev . l 5!!1~~1TIV~ ~~CUl".IT'Y-~~Ll<T~f511q!-5~19l~Tl5rq C~l+IC!\~ li~Jlig~¥.libi!~lRICo'\L l~lrRl~e+RUG+URe IN~Qj;lMA:r1mi
r X7)(F) (bJ(J):16 USC ~ 81~0-l1d)
Using Auxiliary Building layout Industry data for Auxiliary Building floods were searched for events of this magnitude. Table C-1 shows a calculation of years of operating experience for light water reactors. From Table 3.3-3 1 four large Auxiliary Building flooding event$ are applicable to this analysis. I (b)(7)(F), \bJ(3>.16 USC H2 lo l(dl I
The first scenario was given a conditional probability of l.OE-01, since the HPI iump room is about one of ten areas in the Auxil iuy Building where there ire sufficient flood sources of this size. The second scenario was also assigned a conditional probability of .l.OE-01. Any flood elsewhere in the build1ng would have to spill 1n excess of 150,000 gallons, but the h1storical data does not have any Auxiliary Building floods of near this size. Neverthe-less, engineering judgment leads us to assign this value to account for possi-ble future floods .
(b)(71(F) \1>)(3116 lJ SL i ~24o-11d)
The event tree in Figure 3. 3-la was constructed to estimate the frequency of core melt due to HPI pump room fl coding. r )(7)(F),(b)(3)I 5 U.SC §S24a-i(dt 3.3-12 Rev. l
Flood of the Equipment Room There are no cases of moderate or large Auxiliary Building floods due to pipe failures. As stated previously, there are no welds, valves, bends or flanges i n the single pipe which passes through this room. To estimate the probability of this pipe failing and flooding or spraying the equipment room, the empirical relationships reported by Thomas (Reference 13) and given in Table 3.3-7 were used. r X7)(F), (b)(3)'16 USC § S~Jo- l(d) I The equipment-room accident sequences are difficult to delineate since so many failure combinations are oossible. I (l>l(7)CF1. (b)(3)*16 USC § 8240-l 1d1 I
(b)(7)(F), (l>X3):16 U S .C § &24o-i(d1 3.3*13 Rev. 1
(b1(i)(F), (b)(3J 16 USC ~ S2-lo-l(dJ
.A flooding incident may also involve water sprays. _ l r_
... )(7_)(F)
_ .(b_)(3 16
_>_* _ u 824 1
_sc_ §__ 0-_ _<d_) _________ I lj was judged that there is i small likelihood that a redundant division would be affected by the sprays .
(b)(7)(Fl. (b)(}):16 USC § S24o-l(dl The event tree in Figure 3.3-lb was constructed to estimate the frequency of
. core melt due to equipment. room flooding. \
\
3.3.2.4 Turbine Building Flooding The three Oconee units share I single Turbine Buildfog. The basement of this building is a large open area that contains \
(btt7i(FJ. (b)(3)*16 U.S C § S~~o-lidJ
\ Table 3.3-5
J shows this equipment, and divides it among two critical levels: equipment that wi 11 fa i 1 when (bX7XH (b)(3) 16 u S .C. s8::'4o-J(d)
(b)l7l(f1. (b)(3;-!6 U .S.C. S S::'k I
Figure 3.3-2 is a simplified floor plan of the basement.
During a 1976 Unit 3 refueling outage, a de power supply failed and a mechani-cal jack failed, allowing one of the six Unit 3 condenser discharge valves to open. Water from Lake Keowee flowed from the CCW discharge piping through this valve, and out open waterbox manways. The Turbine Building basement flooded to 3.3-14 Rev. 1
a depth of 21 inches in the middle (about 16 inches at the Turbine Building/
Auxiliary Building wall) before power to the valves was restored, and the flood was isolated. As a result of this flood, a six-foot diameter drai n was in-stalled in the south wall of the basement. Following further analysis in NSAC-60, several more modifications were carried out, including J I
CCW System Flood Sources The CCW System consists of four pumps for each Oconee unit. The pumps. are located on the intake structure in the intake canal. Under normal conditions.
three of the pumps are running. Figure 3.3-3 gives a schematic diagram of this svstem. /
Figures 3.3.-4 through 3.3-7 present other views (schematic., plan, and e1eva-ti on) of the CCW System that are useful in the discussion of CCW flood sources.
Figure 3.3-4 illustrates the major flow paths and interconnections associated with the CCW piping for all three Oconee units. Figure 3.3-5 is a plan view of the arrangement of the CCW buried pipe for the Unit 3 condenser. Figure 3.3-6 illustrates the cross-section of the Turbine and Auxiliary Buildings showing the elevations of the CCW pipes relative to the buildings and equipment .
Figure 3.3-7 shows an additional representation cf the CCW piping and compo-nents that is useful in understanding the nature of the initiating events.
3.3-15 Rev. 1
A survey of the CCW System was conducted-to identify important factors regard-ing possible flood sources. Of principal interest were the following:
- 1. Type of affected component (pipe, valve, etc . ).
- 2. Elevation in the building (hydrostatic head).
- 3. Estimated flow rate.
- 4. Type of isolation available.
- 5. Flow after procedural isolation steps are taken .
. Table 3.3-6 su11111arizes the review of the CCW system. I
.__ _______________________ (b)(7)(F), (b)(3),16 USC§ 82-lo-l(d)
~
J The results of this survey were used to develop the flood initiating event categories dis.-
cussed below.
Initiating Event Analysis Based' on the information in the preceding sections, the following initiating events were developed for the turbine building flood analysis.
- \.
- FVLJ--Very large (300,00 gpm) flood that is isolable. Only floods through the largest lines (78 inches) in the system are placed in this category.
rXWJ, lbJOJ 16 Us c § 8"~1(d1 FVLN--Very large (300~000 gpm) flood that is not isolable . I (b)(7)(F), (b)(3) 16 USC § 82-lo- l(d)
I
. FLII--Large (75,000 gpm) isolable 'flood on the inlet side of the condens-er. I (bX7)(F), (bX3):16 U,S C § 82-lo- l(d) 3.3-16 Rev. l
FLIO--Large (75 , 000 gpm) isolable flood on the outlet side of the condens-er. I (b)(7){F), (b)(3):16 U.S.C § 8240-l(d)
I FLN--Larae (75 ,000 gpm) flood that is not .i solable . I (b)(7)(F). (bX3)*J6 USC ~ 8240-l(d)
I FMII--Medium (30 . 000 aDm) flood on the inlet side of the condenser that is i sol able. I CbX7){F), (b)(3)*16 U.S C § 8240- l(d)
FMIO--Medium (30,000 gpm) flood on-the outlet side of the condenser that is i so 1ab 1e. f (b)(?)(F). (b)(J) 16 U SC § &24o-l(dl I
flood that is not isolable.
(b)(7)(F), (bXJ) 16 u.s C § 8240-l(d.)
Note that the information on flow rates ~efore and after pump trip, and isola-tion steps required resulted in distinguishing between inlet and outlet floods .
The pressure drop through the condenser would result in different flow rates for inlet and outlet floods , depending on what isolation actions were taken.
The fault tree shown in Figure 3.3-8 was constructed to quantify the Turbine Buil ding flood initiators. Information from Table 3. 3-7 was used to quantify
- 3. 3* 17 Rev. l
each event . . Tables 3.3-8 and 3.3-9 contain CCW System information and Duke Power _data on cond~nser expansion joints that were used in developing some of the data i n Table 3.3-7. Results of quantification of the Figure 3.3-8 fault tree are given in Table 3.3-10.
- 3. 3.3 EVENT TREE The event tree in Figure 3.3-9 was ~onstructed for Turbine Building flood sequences . This event tree was closely patterned after the transient event
- tree discussed in Section 2.2.
Top logic for this event tree closely resembles the transient event tree top 1og i c dt scussed in Section 2.2 . The flood top logic was constructed by l r-------
CbX7)(F), (bl(3);16 U.S C. § 8~-lo-l(d)
I Domi **,,n.a nt minimal cut sets for the Turbi ne Building flood analysis are given in Appendi :,
x D.
~:*
Sequence FlOsX
\bX7)(F). CbX3) 16 U.S.C § 8240-l(d) 3.3-18 Rev. 1
Sequence FlQsU (b)(i)(F), (b)(.3):16 USC. § 8240-l(d)
Sequence FlQrX CbX7J(F), (b)(3)' 16 U.S.C. § 8~4o-l(d)
Sequence FlBX (bX7)(F), (b)(3):16 U.S C § 8240-lid\
3.3-19 Rev . 1
(bX7)(F), (bX3):16 USC § 8240-l (d)
- . Sequence Fl BU (b)(7)(F), (b)(3):16 US.C S8~~0-l(d)
Sequence FlBOsX (bX7)(F), (b)(3):16 lT SC § 8240-l (d) 3.3*20 Rev. 1
[ "' ' " 0)<3) !6 u s.c § " ' * ' ' "
Sequence FlBQsU (b)(?)(F), (bJ(3):16 U S.C § 82~o-l(d)
Sequence FlBQrX (b)(7)(F). (b)(3)*16 USC § 8:40-l(dl 3.3-21 Rev . 1
Sequence FlBQrU (b)(7)<_'F), (b)(3J:16 USC § 82~0-l(d)
SSF Recovery Human Errors (b)(7)(F), (b)(3):l6 L- SC s8240-l(d) 3.3.4 LIMITATIONS OF THE ANALYSIS Auxiliary Building Flooding As was discussed in the analysis, no Auxiliary Building flooding events have occurred, to date, that would have been large enoYgh to flood the HPI pump room. It was assumed that the largest of these historical floods, however, fit into this category for the pyrpose of calculating an initiating event fre-quency. I (b)(7J(F). (b)('.\)*16 ll SC § 82-lo-l(dJ I
3.3-22 Rev. l
Turbine Building Flooding A flood in the Turbine Bu11ding wou1d affect all three Oconee units. Since this analysis is concerned only with Unit 3. possible negative effects result*
ing from things such as manpower needs on Units land 2, could not be accurate-ly factored into the analysis. Since unit "staffing requirements do consider station-wide events, suc:h as fires, however, i t is felt that any such contribu-tions would not be significant.
3.3.5 INSIGHTS Several modifications were made to Unit 3 f o11 owin o the NSAC-60 analvsis. Several of these modifications f --------
\
(b)(')(Y), (b)(3U6U.S.C § 82-<o-l(d) 3.3-23 Rev. l
(b)(i)(F), (b)(3\.16 USC§ 8'.'A-0-l(d) 3.
3.6 REFERENCES
- 3. 3.6.l Documents
- 1. Oconee PRA: A Probabilistic Risk Assessment of Oconee Unit 3, NSAC-60, Electric Power Research Institute, Palo Alto, CA , June, 1984 .
- 2. Babb, A. 0., and T. W. Mermel , 1968. Catalog of Dam Disasters, Failures and Accidents, U.S. Department of Interior, Bureau of Reclamation, Wash-ington D.C.
- 3. Baecher, G. B., M. E. Pate, and R. De Neufville, 1980. "Risk of Dam Failure in Benefit-Cost Analysis,- Water Resources Research, Vol. 16, No.
3, pp. 449-456.
- 4. Benjamin, Jack R., and Associates, 1982. A Database for the Evaluation, of the Frequency of Random Dam Failure, Report 120-010-01, Palo Alto, Calif.
- 5. Biswas, A. K. , and S. Chatterjee, 1971. -Dam Disasters: An Assessment,"
Jngineering Journal, Vol. 55, No. 3.
- 6. Craig , P. S., 1977. Turbine Building Flood Study, Duke Power Company, Charlotte, N. C.
- 7. Duke Power Company, 1966. Flood Study 1 Jocassee and Keowee Reservoirs, Charlotte , N. C.
- a. Duke Power Company, 1976. Oconee Nuclear Station 1 Incident Investigation Report, 8-536, Charlotte , N.C.
3.3-24 Rev. l
- 9. Gruner, E. , 1964. "Dam Disasters," Institution of Civil Engineers, Vol.
- 1 24, paper 6648 .
- 10. Gruner, E., 1967 . "The Mechanism of Dam Failure, 11 paper presented at the
. ICOLD Conference, Istanbul, Turkey.
11 . Jansen, R., 1980. Dams and Public Safety, U. S. Department-of the Interi-
~r, Bureau of Reclamation , Washington , D.C .
- 12. Middlebrooks, T. A. , 1953.* "Eart-h-Dam Practice in the United States,"
Transactions of the American Society of Civil Engineers, centennial volume.
- 13. Thomas, H. M. , 1981. Pipe and Vessel Failure Probability," Reliability Jngineering, Vol . 2, pp .83-124 .
- 14. USCOLD Comittee on Failures and Incidents to Large Dams of the United States, 1975, "Lessons from Dam Incidents," American Society *of Civil Engineers .
15 . USCOLD, 1988, "Lesson f ro!!' ~.am Incidents USA-II," American Society of Civil Engineers.
- 16. Nash, J . A. , Memo to File, Uedate of the Random Failure Frequency of the
~oca~see Dam, File No. OS-203, October 3, 1989.
17 . USNRC (U.S . Nuclear Regulatory Comission), 1975. Reactor Safety Study--
An Assessment of Accident Risks in U.S. Co111nercial Nuclear Power Plants, WASH-1400 (NUREG-75/014), Washington, D.C.
3.3. 6.2 Procedures ONS AP/3/A/1700/10, Change 0: Uncontrollable Flooding of Turbine Building ONS AP/3/A/1700/19 , Change 0: Loss of Main Feedwater
- 3. 3-25 Rev. 1
ONS EOP/3/A/1800/01. Change 3: Emergency Operating Procedure 3.3.6.3 Drawings OF0-133A-1.1 thru OFD-133A-1.5: Flow Diagram of Condenser Circulating Water, Unit 1.
OF0-133A-2.l thru OFD-133A-2.3: Flow Diagram of Condenser Circulating Water. Unit 2.
OFD-133A-3.l thru OFD-133A-3.4: Flow Diagram of Condenser Circulating Water. Unit 3.
0-201 thru 0-238: Turbine Building Substructure.
0-331 thru 0-338: Condenser Cooling Water, Piping Layout.
O-4DOA thru O-4OOW: Turbine Building Basement Floor, Pi~1ng Layout, Unit 1.
0-14O0A thru 0-1400W: Turbine Building Basement Floor, Piping Layout, Unit 2 0-2400A thru 0-2400W: Turbine Building Basement Floor, Piping Layout, Unit 3.
3.3-26 Rev. l
Table 3;3-2 Chronological Order of Dam Failures Year of Years Cumulative Dam-years failure between failures dam-years between failures Period of Construction: 1940-1987 1963 23 32,207 32,207 1975 12 74,782 42,575 1987 No failure 126,435 No failure
Table 3. 3-1 Dam Failures Used in This Study Year Year Dam compl~ted failed Baldwin Hillsa 1951 1963 Walter Boudinb 1967 1975 aData from Babb and Merme1(1968) and USCOLD (1975) .
bData from Jansen (1980) .
8.1.l.4 Steam Generator Tube Ruptures The calculated annual core-melt frequency resulting from a steam generator tybe rupture ( SGTR) U 13~1~,i~ § This accounts f_or approximate 1
lyl~~J!of the internally-initiated core-melt freQuency and less than ~.) f the total core-melt freQuency.
8.1.1.5 Interfacing Systems LOCA An interfacing-systems LOCA occurs when an RCS pressure boundary failure, usually through a valve, results in the overpressurization of an interfacing s stem. Althou (b)(7)(r) (b)(i) 16 USC § &24o-l(d) discussed in Section 2 6 (b)( ), (b) sequence of
1 As * (3).16 USC failures leads to an interfacing-systems LOCA ~i~h a frequency greater than
/b)(7}(F). (b)(.l):)6 U SC § 8240-l (d) The sum of the probabilities of these sequences is ~x;~'tc~ (
8.1.1.6 Reactor Pressure Vessel Failure The calculated annual core-melt frequency resulting from a reactor pressure vessel failure is CJN}1J:tl 98 This accounts for approx.imately l~
(b1 *
,Ff t~e inter-nally-1nitiated ccre-melt frequency and iess -than FH f the tcta1 core-melt frequenc \b)(i)(F), (b1(31.l6 USC § 8'.!~o- l(d*- - - - - - - - - - - - - - - - '
Section 2.1 discusses the analysis for this event.
8.1.2 EXTERNAL EVENTS
)\7/
Approximately ~-~' f the calculated core-melt frequency for Oconee Unit 3 is attributable to external initiating events. The dominant component failures 8.1-6 Rev. l
and/or operator actions required to produce a core melt for each initiating event are described below. A detailed listing of the external event -results is provided in Appendix D.
- 8. 1.2.1 Seismic Events The calculated annual core-melt frequency resulting from a seismic event is (b)(7)(F). )
3):16 c.sc § s Th ,. s accounts f or approximate. 1 1>)(3).16 .. ex t erna l , y-,n,t
) 7 F), f tne .
- i ated 1,(7) core-me 1t frequency and appro>:16 U.S C. § 82.io-l(d) 8.1-15 Rev. l
Table 8. 1-1
- Rev . 1
- Surm,ary of Mean Annual Core-Melt Frequencies (Page 1 of 2)
Total Event FreQuency FreQuency Internal
- Events l(b.)(1 J(F), (b)(3) 16 U S C I§S2.lo-l(d)
I Plant Transients (b)(7J(F). (bX3)16 U.S.C § 8240-l(d/
Reactor/Turbine trip
- Loss of off-site power Loss of main feedwater Loss of condenser vacuum Excessive feedwater Steam line break Feedwater/Condensate line break Loss of instrument air Loss of service water Other transients Plant Transient Total Loss of Coolant Accidents Sma l 1 break LOCA Large break LOCA Medium break LOCA LOCA Total Anticipated Transients Without Scram Steam Generator Tube Ruptures Interfacing Systems LOCA
Table 8.1~1 Rev. 1 Surrmary of Mean Annua 1. Core-Melt Freguenc*; es {Page 2 of 2)
Total Event Frequency Frequency RPV Rupture External Events Seismic Turbine Building Flood External Flood Tornado Fire Total