ML18051A544

From kanterella
Jump to navigation Jump to search
Evaluation of Palisades MSIV Single-Failure Backfit.
ML18051A544
Person / Time
Site: Palisades Entergy icon.png
Issue date: 08/15/1983
From:
CONSUMERS ENERGY CO. (FORMERLY CONSUMERS POWER CO.)
To:
Shared Package
ML18051A543 List:
References
TASK-15-02, TASK-15-2, TASK-RR NUDOCS 8308230379
Download: ML18051A544 (51)
v • d • e


Text

ENCLOSURE 1 CONSUMERS POWER COMPANY PALISADES PLANT DOCKET 50-255 EVALUATION OF PALISADES MSIV SINGLE-FAILURE BACKFIT August 15, 1983 50 Pages

/ -- 8308230379 830815

' PDR ADOCK 05000255 P PDR NU0783-0181AB-TP20

1

  • ENCLOSURE 1 EVALUATION OF PALISADES MSIV Single-Failure Backfit INTRODUCTION As a part of SEP Topic XV-2 "Spectrum of Steam System Piping Failures Inside and Outside Containment," a previously unanalyzed transient was identified for the Palisades Plant which assumed a rupture of one of the two main steam lines inside containment with a concurrent single-failure of the main steam isola-tion valve (MSIV) in the other main steam line. This series of events results in the blowdown of both steam generators into the containment. Figure 1 shows the plant configuration for this event .
  • In the Integrated Assessment for the Palisades Plant, the staff presented the results of an analysis indicating that the peak containment pressure expected for this event was 1.53 times accident design. Because the Palisades contain-ment was designed with a 50% margin over expected accident pressure conditions and because the increase in leakage or failure probability was negligible at these pressures, the staff concluded that backfitting Palisades for this transient for containment pressurization reasons was not required.

The staff did predict, however, a severe detrimental impact on the ability to remove decay heat during the event. A variety of reasons was presented as a basis for their concerns:

NU0783-0181A-TP20

2

  • A potential for operator interference with the feedwater system exists in attempting to control the cooldown rate.

Difficulties may be encountered in controlling heat removal given the existence of open steam system piping.

  • The ability to feed High Pressure Safety Injection (HPSI) water to the primary coolant system under these conditions has not been demonstrated.

As a resolution to the deficiencies identified in SEP Topic XV-2, Consumers Power Company (CPCo) committed to backfit the Palisades Plant eliminating the MSIV single-failure. A series of alternative modifications have been evaluated; the most feasible alternative appearing to be the replacement of the existing air-operated check valves in the main steam line with fast acting gate valves having stored energy actuators. Closure of either MSIV will then result in the isolation of the intact steam generator following a main steam line break. The valves will be capable of full closure within several seconds of actuation thereby restricting the final containment pressure to a value below the original accident design pressure. The size of the new valves requires.reevaluation of the stresses in the main steam line under a variety of conditions including seismic loading. Total cost of this backfit is estimated to be on the order of $2 million.

Due to the cost of this backfit, CPCo is reexamining the need for this modification. The initial results of this examination are presented as a part of this report. These examinations take the form of a reevaluation of the assumptions made in concluding the need for this backfit, an evaluation of plant response to this transient initiator in the form of an event tree and a preliminary cost-benefit analysis of the modification using Probabilistic Risk Assessment (PRA).

NU0783-0181A-TP20

3

  • EVALUATION OF PLANT RESPONSE For clarity in examining plant response to the steam generator pipe rupture transient with MSIV failure, an event tree has been developed. The event tree and a description of its branch points are presented as Attachment 1. In developing this tree, a reevaluation of the assumptions made in the SEP Review was performed.

The assumption that HPSI could not be demonstrated to be effective in cooling the primary system without the assistance of continuous heat removal through the steam generators remains valid. The event tree does not reflect this mode of core cooling.

At the time of the SEP evaluation, the assumption that the loss of both steam generators resulted in only a single motor-driven auxiliary feedwater pump re-maining was valid. An additional motor-driven pump, however, is being added to the auxiliary feedwater system improving the system's reliability over that assumed in the original analysis. The configuration of "the auxiliary feedwater system as it will exist is reflected in Figure 1.

With respect to the time frame in which to establish heat removal following steam generator dryout and the operator's role in establishing the cooling, several assumptions have been reconsidered:

1.) Auxiliary feedwater is automatically actuated within minutes of attaining low steam generator level. It is recognized, however, that depending on the size of the rupture, the steam generators may have completely blown down by the time automatic actuation occurs.

2.) Given that low steam generator levels may exist concurrent with an extreme cooldown rate thereby providing conflicting information to the operator, it is recognized that the operator may interfere with the operation of the auxiliary feedwater system in an attempt to control cooldown.

NU0783-0181A-TP20-NL01

4

  • 3.) As backup to the feedwater system, the condensate system can be aligned manually to the steam generators to establish cooling provided the steam generators are sufficiently depressurized. This method of cooling may be effective only in the long-term following steam generator blowdown.

4.) The time frame in which the operator must establish cooling through the steam generators in order to prevent core damage is dictated by the amount of time it takes to boil off primary coolant inventory to the top of the core through decay heat generation and safety relief valve actuation. This time frame is in excess of three hours.

5.) Finally, it is recognized that introducing cold water to the steam generators after a significant period of time has elapsed since dryout can thermally shock steam generator components and may compromise the integrity of the primary steam through steam generator tube failures. It should be noted, however, that in order to result in loss of primary system inventory through the steam generators, it is necessary that the operator attempt to reestablish decay heat removal by way of the steam generators by reenabling the auxiliary feedwater or condensate pumps.

The event tree in Attachment 1 reflects these considerations. The sections of the tree which are applicable to this topic begin with Branch Point 19. The auxiliary f eedwater and condensate pumps are shown as systems available for removal of decay heat from the primary system. Given that these systems are

  • placed in service following steam generator dryout, the potential for steam generator tube rupture is reflected. Depending on the size of this rupture, the necessary combinations of ECCS are developed in terms of HPSI, Low Pressure Safety Injection (LPSI) and Safety Injection (SI) Tanks.

(SI).

Development of the logic behind the plant response to this event beyond that considered in the original evaluation of this SEP Topic has been accomplished and is generally presented in the event tree of Attachment 1. A more detailed description of this logic is presented as a part of the attachment. While the steam line break with coincident MSIV failure is an admittedly severe NU0783-0181A-TP20

5 transient in terms of its ultimate effect on plant hardware, it can be seen that there are installed systems within the plant which respond automatically to this event and are designed to provide adequate core cooling throughout the course of the transient.

QUANTIFICATION AND COST BENEFIT In evaluating the effectiveness of the proposed modification toward reducing the risks associated with MSIV single-failure, an attempt has been made to quantify the applicable branches of the event tree in Attachment 1. This quantification is presented in Tables 1 through 4 of this report. Several important points require explanation in reviewing the information on these tables.

Table 1 contains a list of all piping which at this time is assumed to lead to a blowdown of a steam generator should a rupture occur. These lines are attached to the secondary side of the steam generators and include all piping four inches in diameter (or greater) between the shell and the first isolation valve. It has been assumed for this analysis that auxiliary feedwater is capable of maintaining steam generator inventory for smaller pipe ruptures.

Although the SEP Topic relates primarily to steam line ruptures inside containment, it should be recognized that any of the pipes attached to the steam generator can lead to steam generator blowdown if ruptured and such ruptures are possible outside as well as inside containment. The table there-fore, contains piping from main steam lines, feedwater lines, auxiliary feed-water lines, steam feed lines to the steam-driven auxiliary feedwater pump and steam lines to the atmospheric dump valves. The number of pipe segments in each system was estimated from in-service inspection records.

Table 2 contains the quantification of each event heading which follows Branch Point 19 in the event tree. The initiating frequency for breaks inside and outside containment is developed.from the data presented in Table 1. The probability of the single-failure of the MSIV is derived based on Palisades historical data. Two contributors to auxiliary feedwater failures are derived in Table 2. The first is the failure of the system to operate given that an NU0783-0181A*TP20

6 actuation signal exists. The second failure probability reflects operator failure to manually initiate the system given that it has been disabled and given that several hours exist in which to reestablish the steam generators as a heat sink before core damage occurs. Quantification of the auxiliary feedwater heading requires that these two failure modes be considered as independent contributors to auxiliary feedwater failure as it will be assumed that the operator will always attempt to control cooldown early in this transient by taking manual control of the auxiliary feedwater system (perhaps disabling it). The condensate system is assumed to be useful only after mechanical or electrical failure of the auxiliary feedwater system. As it is a manually actuated system, it shares a common mode failure with the auxiliary feedwater system (operator failure to actuate) which was quantified as a part of that system. The quantification of this tree will assume that if either auxiliary feedwater or condensate water are injected to the depressurized steam generators, it is extremely likely that tube rupture will occur due to thermal shock; thus the high probability of steam generator tube failure .

  • Given LOCA conditions through the steam generator tubes, a demand on HPSI or LPSI will be expected depending on the size of the rupture.

motor-driven pumps and isolation valves.

Like auxiliary feedwater, each of these systems consists of two independent trains of An unavailability factor for each of these systems similar to the auxiliary feedwater failure probability was thus assigned. The notes following Table 2 provide additional explanation of the quantification of these event tree headings.

Table 3 contains the final estimate of the likelihood of core damage and con-tainment failure for the steam generator pipe rupture with coincident MSIV failure sequences. The quantification is presented in the form of an event tree. As HPSI and LPSI ECCS have been assigned similar failure probabilities, they are represented in this quantification as a single system. The operator actions disabling auxiliary f eedwater early in the transient and reenabling the steam generators as a heat sink prior to core damage have been represented by their own headings to clarify the quantification of the tree.

The total contribution of these sequences to core damage is estimated to be 8

approximately 10- /yr with the potential for a significant radionuclide 8

release also being 10- /yr.

NU0783-0181A-TP20

7 A sensitivity study on the qualification of several of the event tree headings is presented in Table 4. Order of magnitude variations in the availability of ECCS or steam generator tube integrity have only a marginal affect on the probability of core damage. Variations in the probability of operator actions in disabling and reestablishing the steam gene,rators as a heat sink have the largest affect on the dominant sequence quantification. This might be expected as operator action is the single dominant contributor to auxiliary feedwater unavailability.

In order to determine the cost-effectiveness of the MSIV backfit, it is necessary to evaluate the consequences of the release from this accident. An estimate of the consequences is presented as Attachment 2 to this report. The cost-benefit analysis may then be performed as follows:

x= c (CRF) (M-R) (Y)

  • Where:

X = Cost-Benefit of the Backfit (~/manrem)

C = Cost of Modification ($2 x 10 )

CRF = Frequency of a Significant Release Resulting From These Sequences (9.8 x 10- 9/Yr) From Table 3 M-R = Consequences of a Significant Release 6

Given This Transient (1.46 x 10 manrem) Attachment 2 Y = Time Remaining in the Operating License for the Palisades Plant (23 years) x= 6

($2 x 10 )

(9.8 x 10- 9 /yr) (1.46 x 10 6 manrem) (23 yr)

NU0783-0181A-TP20

8 6

= 6.1 x 10 $/manrem, which is significantly in excess of the recommended $1000/manrem suggested as a part of the Commission's proposed safety goals.

SUMMARY

AND CONCLUSIONS Elimination of the Palisades MSIV single-failure, identified as a part SEP Topic XV-2, has proven to be a substantially greater backfit than had originally been anticipated. Due to the cost of the backfit and the apparent infrequency of the transient, CPCo questions the effectiveness of the modifi-cation proposed as a resolution to the Palisades MSIV single-failure issue.

PRA cost-benefit evaluation of the sequences in question appears to confirm the necessity to reassess the need for the MSIV backfit.

Several reasons ~xist for the apparent lack of cost-effectiveness associated

  • with this modification. Primarily, emergency systems already exist at the Palisades Plant which actuate automatically in response to this event provid-ing a means for adequate core cooling throughout the transient. These systems include auxiliary feedwater and, if necessary, emergency core cooling systems in the form of HPSI or LPSI. Several hours exist from the beginning of the transient until sufficient primary coolant inventory loss results in core uncovery providing ample time for operator action in establishing adequate core cooling. These systems are designed to be at least as reliable as an additional MSIV in precluding core damage as a result of this transient. As a result, the steam line break scenario with concurrent single-failure of an MSIV appears to be more of a plant availability problem in attempting to recover from the transient than a safety issue.

CPCo proposes to defer the installation of the new MSIV for an additional cycle beyond its current commitment which presently requires completion of this modification early in 1985. In the interim, CPC intends to confirm the reliability of the systems which are now believed to prevent inadequate core cooling from occurring during this event. CPC also intends to perform more detailed and transient specific containment response and consequence analyses NU0783-0181A-TP20

9 than those which have been provided in this report. These analyses are expected to be complete by Mid 1984. At that time, CPCo expects to submit final justification for cancellation of this modification or propose alternate more cost-effective resolutions to the MSIV single-failure issue should any exist .

  • NU0783-0181A-TP20

i F!"l'Jm air

  • " supply r

-~

~ ...

r;..:> * -*

  • ... From Cond'!n~r J.

I From CST F,igurel Palisades main system and feec~a~cr system

    • . .... *- - Palisades SC:P 4-47
~.::-: **. ' .

---~::

10 TABLE 1 Major Piping Systems Attached to the Secondary Side of the Palisades Steam Generators Number of Welds Number of Welds System Steam Generator A Stream Generator B Inside Cont Outside Cont Inside Cont Outside. Cont Main Steam Line (36 11 ) 10 29 10 29 Main Feedwater Line (18") 14 2 15 2 Auxiliary Feedwater Line (14") 38 5 34 2 Auxiliary Feedwater Pump Steam Feed 27 I lQ Atmospheric Dump Valves (8 11 ) 29 25 Total 62 92 59 68 Total Inside = 121 Total Outside = 160

  • NU0783-0181A-TP20

TABLE 2 Palisades MSIV Backfit Event Heading Quantification Event Quantification Notes

-10 Steam Generator Blowdown Inside (121 Segments) (10 /Segment-hr) (8760 hr/yr) 1, 2

-4 Containment = 1.1 x 10 /yr

-10 Steam Generator Blowdown Outside (160 Segments) (10 /Segment-hr) (8760 hr/yr) 1, 2

-4 Containment = 1.4 x 10 /yr MSIV Failure to Isolate (3 Failures)/(196 Demands) 3, 4, 5

-2

= 1.5 x 10 /Demand

-4 Auxiliary Feedwater Unavailability 10 /Demand (Auto) 6, 7 2.4 x 10- 3 /Demand (Operator Error)

Condensate System Unavailability 1.0 8 Steam Generator Tube Rupture Given Cold 1.0 9 Water Addition to a Dry Steam Generator NU0783-0181A-TP20

Event Quantification Notes

-4 HPSI Unavailability 10 /Demand 10

-4 LPSI, SI Tank Unavailability 10 /Demand 10 Containment Failure Probability Given 1.0 11 SG Tube Failure Containment Failure Probability Given 1.0 12 SG Blowdown Inside Containment Containment Failure Probability 13 Given SG Blowdown Outside Containment 1.0 NU0783-0181A-TP20

1 NOTES ON TABLE 2 I. Table 1 contains the number of pipe segments between the steam generators and the first automatic isolation valve which can lead to steam generator blowdown if ruptured.

2. Failure rate per section of pipe > 3" diameter is from WASH-1,400, Table III, 4-1.
3. Since 1971, there have been three failures of MSIVs to operate:

September 21, 1972 due to Solenoid Valve Failure May 19, 1973 due to hardened packing August 12, 1973 due to hardened packing

4. Each MSIV is normally stroked twice during each cold shutdown; once to isolate the steam generators from the main steam auxiliaries and once to verify valve timing prior to return to power. Each MSIV is stroked an additional time during refuelings as a result of surveillance testing.
5. Since 1971, there have been 40 cold shutdowns and six refuelings.
6. Auxiliary feedwater for this transient is limited to the two motor-driven trains (the steam-driven auxiliary feedwater pump is assumed to fail).

Preliminary fault tree analysis indicates that the failure of each train is dominated by motor-driven pump failure to start plus associated valve

-2 -4 failures (~ 10 /Demand/Train or 10 /Demand/System).

7. Manual actuation failure of auxiliary feedwater is taken from the Handbook of Human Reliability" (NUREG/CR-1278) assuming two control room operators and a senior reactor operator are present under a highly stressful situation with several hours available to reestablish the steam
  • generators as a heat sink.

NU0783*0181A-TP20

2

-2 (Op Fails to Act Correctly After Several Hours)

C02 HEP = .32 (Medium to High Dependence)

SRO HEP= .75 (High to Complete Dependence)

Reference:

Table 20-25, 20-24 and 20-1 NUREG/CR-1278.

8. The cond_ensate system is manually operated. Its iinavailability is dominated by human error which was developed in Note 7.
9. Steam generator tube rupture is assumed to occur due to thermal shock whenever the steam generator is allowed to dryout and cold water is sub-sequently initiated.
10. HPSI and LPSI are initiated automatically early during the transient due to the rapid cooldown of the primary coolant system. Like auxiliary feedwater, each system consists of two independent trains of pumps and

-2 -4 automatic valves (~ 10 /Demand/Train or 10 /Demand/System assumed).

11. Loss of containment integrity is assumed as a result of steam generator tube rupture after injection of cold auxiliary feedwater or condensate water. It should be recognized that this release path is tortuous as compared to containment failure paths resulting from containment over-pressure.
12. Blowdown of both steam generators inside containment yields a containment pressure 1.53 times design. This does not by itself fail the contain-ment. However, containment failure is assumed if additional pressures associated with core melt and hydrogen ignition are added to the initial blowdown.
13. Those sequences in which the pipe rupture occurs outside containment and the primary system integrity is not challenged by introduction of cold water to the steam generators, are assigned a large c~ntainment failure probability. It should be recognized that this containment failure mode
  • occurs significantly later in time than does the initial over-pressure failure.

NU0783-0181A-TP20

Table 3. Palisades MSIV Backfit Quantification of Dominant Sequences Man S.G.

MSLB Man Ree st Integrit S.G.

with Aux FW Aux FW No Integrit MSIV Auto (No Op or Cond Large No Failure Aux FW Interf) I Pump LOCA Sm LOCA HPSI LPSI 0

0 1-? I Ii. o I rulO -4 1.0  ?

I io-4 I

2.4x10-3

( 2. '5x10-4 )( . 01 '5) 0 1-? I I 1.0 I -4 I .-vlO

  • "'10-4 2.4x10-3

?

.-v lo-4 4

Core Damage= (2.5xlo- 4yr- 1 ) (.015) (2.4xlo- 3 + lo- )

Probability ?:!I _8/yr 10

  • * (

Table 4. MSIV Backfit Sensitivity Analysis Manual Auto Manual Reset S.G. Tube ECCS Core Damage Factor Case MSLB MSIV AFW AFW AFW/Cond Integrity HPSI/LPSI Probabiliti Change Base -4 -2 10- 4 -3 10-4 10- 8

1. 2.5x10 l.5xl0 1.0 2.4x10 1.0
2. S.G. In!Ig- Base Base Base Base Base 10-1 Base -9 9.6x10 1 rity 10 1.04
3. ECCS _ Base Base Base Base Base Base 10-3 1. 4x10 -8 1.4 3

a) 10 2 10- 2 -8 b) 10- Base Base Base Base Base Base 5.2xl0 5.2

4. Operator Base Base Base 10-1 Base Base Base 10- 9 1 Failur= 1of 10 AFW 10
5. Operator Base -4 -9 Base Base Base 2.4x10 Base Base 1.4x10- 1/7 Restor AF!i 4

a) 2.4xl0

-2 -2 -8 b) 2.4x10 Base Base Base Base 2. 4x10 Base Base 9.6xl0 9.6 NU090983A-NL01

1 ENCLOSURE 1 ATTACID1ENT 1 MAIN STEAM LINE BREAK EVENT TREE PALISADES PLANT NU0783-0181A-TP20

2

Initially, a steam line break transient will result in a rapid cooldown of the primary coolant system (PCS) due to the rapid energy extraction produced by the steam blowdown. Once the initial inventory in the steam generators is depleted, additional water must be supplied to provide decay heat removal. If additional water is not supplied following dryout of the steam generators, the PCS will heatup until system pressure exceeds the set point of the pressurizer relief valves which will result in steam and water being released from the PCS to containment. The release of primary inventory by the pressurizer relief valves will stabilize PCS temperature and pressure; however, the high-pressure charging system will not be able to replenish all the inventory lost through the pressurizer relief valves. Ultimately, the fuel will be uncovered and core melt will occur.

Should sufficient cooling water be supplied to the steam generators subsequent to a steam line break, core decay heat will be removed and the plant sta-bilized. The normal mode of cooling following a break would call for isolat-ing both the feed and steam lines to the faulted steam generator, and cooling the plant by supplying water to the intact steam generator and dumping steam through the atmospheric dump valves (ADVs). Should the atmosperic dump valves not be available or fail to operate, the code safety valves (SRVs) would provide a path for energy removal.

Failure to isolate the faulted steam generator or failure of the ADVs or SRVs to reseat on the intact steam generator w'ill not necessarily produce a core

  • melt sequence. Decay heat can still be removed from the core as long as cooling water is supplied to the steam generators.

NU0783-0181A-TP20 Energy will be removed by

3

  • steam release through either the broken steam line or the failed relief valve(s).

The normal source of cooling water following a steam line break is the auxil-iary feedwater system (AFWS). The system consists of three high-pressure, full-capacity pumps with cross-tie headers to allow any one pump to supply both steam generators. These pumps are capable of supplying sufficient water for decay heat removal at all steam generator pressures up to the steam generator safety valve set point pressure.

Should the AFWS fail to provide flow to the steam gener_ators, it may be possi-ble to align the condensate system to provide flow through the main feed lines. This system consists of two full-capacity pumps; each of which is capable of supplying either steam generator provided the steam generator is not pressurized above the shutoff head of the pumps.

KEY ASSUMPTIONS MSIV: MSIV failure refers to failure of the main steam isolation valve in the steam line to the intact steam generator. The Palisades MSIVs are of a stop-check design and are unable to prevent flow in the reverse direction.

Therefore, failure of the MSIV in the intact steam loop will allow steam from the intact steam generator to flow through the steam line cross-tie header, down the broken steam line in a reverse direction, through the other MSIV and out the ruptured line. The result is that both steam generators will blow down given a steam break upstream of one MSIV and failure of the MSIV in the opposite steam line. It should be noted that for the MSIV application study, questions regarding containment integrity will not be addressed. The containment response, however, will be included in the overall risk assessment.

SG Isolation: Steam.generator isolation refers to automatic actuation of the AFWS isolation valves or action taken by the operator to terminate the flow of

NU0783-0181A-TP20 Concurrent signals indicating

4 excessive differential pressure between steam generators and low-level (approximately 447 inches above the bottom of the steam generator support skirt) in the depressurized steam generator will initiate isolation of the depressurized steam generator by closing corresponding motor-operated isolation valves in the AFW supply lines to the depressurized generator.

Successful isolation will result in a requirement that either the atmospheric

~

dump valves or code safety valves on the intact steam generator open to relieve steam and remove core decay heat. Without isolation, decay heat can be removed through the faulted steam generator provided sufficient cooling water is supplied.

It is assumed that the feedwater regulating and f eedwater regulating bypass valves to both steam generators close on steam generator low-pressure or containment high-pressure. Therefore, to establish flow to either steam generator with the condensate pumps requires the operator to manually open the feedwater regulating bypass valve.

AFWS: Successful operation of the auxiliary feedwater system implies that a minimum flow of 300 gpm be established. Decay heat removal can be accomplish-ed by supplying this flow to either one or both steam generators.

ADVs: Failure of the ADVs to open includes the failure to automatically open as well as failure of the operator to open them. Failure of the SRVs to open is not considered in this event heading since it is probabilistically insigni-ficant (ie, this would require all 12 of the SRVs failing to open).

Failure of the ADVs or SRVs to close does not necessarily affect the ability to remove decay heat. If an ADV fails to close (either automatically or manually), or one of the SRVs fails to reclose, the intact steam generator will depressurize and preclude the use of the steam-driven auxiliary feedwater pump.

Condensate Sys.tern: The two condensate pumps are full-capacity pumps capable of supplying the steam generators through the main feed line. The condensate system can be aligned to supply either steam generator by opening the NU0783-0181A-TP20

5 feedwater regulating bypass valve to the specific steam generator. The con-densate pumps take suction from the condenser hotwell (inventory of approxi-mately 50,000 gallons). In addition, the condensate storage tank can be aligned to supply the pumps through the condenser hotwell.

Long-Term Cooling: Long*term cooling in this event tree represents many events. The heading includes the possibilities of continuing to stabilize the plant using only the AFWS or a condensate pump. It also includes the possi-bility of switching at some point to the low-pressure shutdown cooling system.

The options considered or available depend on the particular sequences. For example, should cooling be provided by an isolated intact steam generator using only the code safety valves, use of the shutdown cooling system would be prohibited because steam pressure would remain at or above the steam generator safety valve set point pressure. Consideration must also be given to the long-term water sources available to each of the cooling systems when evaluat-ing this heading.

An important aspect of the long-term cooling ~eading is that it also includes consideration of the ability to makeup inventory lost from the primary coolant system during the course of the transient. The usual pathways for losing inventory are through the control rod drive mechanisms and the reactor coolant pump seals. The leakage rates are generally low but may increase due to seal degradation over time due to effects such an inadequate seal cooling.

Another consideration of this event is the need for borated water injection due to xenon decay. This will be required probably within a day and a half of the initiating event. Depending on the sequence, the Chemical Volume and Control System, HPSI or LPSI will have to be used.

HPSI: The HPSI pumps are available to provide low-volume, high-pressure makeup to the primary coolant system in the event of a small loss of coolant accident. Success of this system implies actuation of one of two HPSI pumps injecting coolant from the Safety Injection Refueling Water (SIRW) Tank through any of the four primary coolant loops.

NU0783-0181A-TP20

6 LPSI: The LPSI pumps are available to provide high-volume, low-pressure makeup to the primary coolant system in the event of a large loss-of-coolant accident. Success of this system implies flooding from three of the four SI Tanks early during the_ PCS depressurization period and operation of one of the two LPSI pumps.

NU0783-0181A-TP20

7

  • EVENT TREE DESCRIPTION The initiating event for this event tree is assumed to be either a rupture of a main steam line upstream of the MSIV or a rupture of either a main or auxiliary feedwater line downstream of the last check valve in that feed line.

Moreover, the break is assumed to be sufficiently large such that the blowdown empties the steam generator with the broken line prior to the time that auxiliary feedwater flow can be initiated. If steam line isolation fails, this assumption is applied to both steam generators.

The following provides a brief description of the event tree illustrated in Figure 1. The write-up will first concentrate on the success and failure paths of Branch Points one through seven. A discussion of the remaining branch points will follow.

BRANCH POINT 1

  • At this point a pipe break has occurred and a reactor trip condition has been reached sending a signal to the reactor protection system (RPS) to initiate automatic insertion of the control rods into the core. This action, if successful, will place the core in a subcritical state.

Depending upon the core moderator coefficient and the ability to inject borated water into the core using the charging and safety injection pumps, the core may be returned to a critical state by the cooling effects of steam blowdown if a stuck control rod occurs. In the event of core recriticality due to the cooldown from the pipe rupture, the power level reached in the core will be controlled by the rate of blowdown and subsequently by the rate at which coolant from the AFWS or condensate system is supplied to the steam generators.

Failure of the control rods to insert into the core either automatically or manually due to operator intervention will lead to core damage due to fuel

  • undercooling.

NU0783~0181A-TP20

8 BRANCH POINT 2 At this point the reactor has successfully tripped. The continuing reduction in steam pressure due to the broken line will quickly produce a steam line isolation signal. The MSIVs are fast-acting, air-operated, stop-check valves.

The valves are actuated by de powered solenoid valves that receive signals from the RPS. Steam line isolation signals are generated based on the receipt of different inputs. For example, high containment pressure or low steam generator pressure will initiate MSIV closure.

There is a single isolation valve located just outside of containment in each of the two main steam lines. Successful closure of both of these valves will result in isolation of the steam generators from any rupture occurring down-stream of the valves. For breaks between one of the MSIVs and a steam genera-tor, successful closure of the isolation valve in the intact steam line will isolate that line and its associated steam generator from the break. As a result, only the steam generator feeding the broken line will continue to blow down. However, if the MSIV in the intact steam line fails to close, both steam generators will blow down through the ruptured pipe. This will occur because the stop-check design of the MSIVs cannot prevent flow in a reverse direction, and steam from the intact steam generator will move through the steam header crossover line, through the MSIV in the broken line in the reverse direction and ultimately out of the rupture. It is this situation that characterizes the failure path from Branch Point 2.

Continuing the development of the failure path on Branch Point 2, it can be seen that the operation of the SG isolation is not important since both steam generators blow down as a result of MSIV failure (Branch Point 19). Inter-locks on the automatic isolation equipment of the AFWS prevent more than one steam generator being isolated .

  • NU0783-0181A-TP20

9 BRANCH POINT 3 Following successful closure of the MSIV in the intact steam line, continued loss of fluid through the broken line will result in an auxiliary feedwater isolation signal to the faulted steam generator. This signal is based on the concurrent occurrence of excessive differential pressure and low level (approximately 447 inches above the bottom of the steam generator support skirt). Failure to isolate the faulted steam generator would result in the decay heat removal path through the broken line, (Branch Point 15).

Although the AFWS would likely receive a signal to start prior to steam generator isolation, it is assumed that by the time feed flow is established the faulted steam generator will have isolated.

At this branch point the feedwater regulating and feedwater regulating bypass valves on the main feed line to both steam generators are assumed to close upon receipt of a signal indicating low steam generator pressure or contain-ment high-pressure.

BRANCH POINT 4 Following successful closure of the MSIV in the intact steam line, continued loss of fluid through the broken line would result in automatic AFW system actuation based upon detection of low water level in the faulted steam genera-tor. The motor-driven pumps, P-8A and P-8C, receive automatic initiation signals. The steam control valve CV-0522B, which supplies steam from steam generator E-SOA to the turbine-driven pump P-8B, receives an 80-second timed delay actuation signal on low-water level in a steam generator. Operator action is required to supply steam to the turbine-driven pump from steam generator E-SOB. Following an auxiliary feedwater actuation signal (AFAS),

generated by low-water level in either steam generator, start signals to the two motor-driven pumps are generated in a timed sequence. Motor-driven pump

  • P-8A starts first after a 5-secondy delay, and the pump running signal permits the opening of flow control valves CV-0727 and CV-0479.

NU0783-0181A-TP20 Any one pump is

10 capable of supplying sufficient flow to allow removal of core decay heat subsequent to reactor trip. As a result, successful operation of the AFWS at this branch means that at least one of the three pumps has started and sufficient flow is being delivered to either the faulted or the intact steam generator to allow for removal of decay heat.

The AFW system pumps normally take flow from the condensate storage tank.

Should the condensate storage tank be unable to supply water to the pumps, the Fire Protection System (FPS) can be aligned to supply the suction header of pumps P-8A and P-8B, and the Service Water System (SWS) can be aligned to supply the suction header of pump P-8C.

If secondary cooling is not established, steam relief will continue for 15 minutes to an hour at which time the intact steam generator will boil dry.

Primary system temperatures would then rise until the pressurizer relief or safety valves open. Relief through these valves would continue until either

  • cooling is reestablished in the secondary system or until the core is uncovered due to the loss of primary fluid and core melt occurs.

BRANCH POINT 5 Following successful isolation of the faulted steam generator and successful operation of the AFWS, the normal path for decay heat removal will be via dumping of steam through the atmosperic dump valves (ADVs) on the intact steam generator. Failure of the valves to open would result in a steam pressure increase to the code safety valve set point pressure.

BRANCH POINT 6 As energy is removed from the secondary system by the ADVs or the SRVs on the intact steam generator, pressure in the generator will oscillate above and below the set point pressure of the ADVs or SRVs, causing them to cycle open and closed. This branch point represents the possible failure of the ADVs or

  • SRVs to reseat during one of these cycles.

NU0783-0181A-TP20

11 Continued successful opening and closing of the ADVs or SRVs will allow sus-tained removal of plant decay heat and ultimately a transition to some form of long-term cooling as discussed under Branch Point 7.

Failure of the ADVs or SRVs to properly reseat will result in depressurization of the intact steam generator. However, because coolant flow and stable boiling have already been established at this point, decay heat removal is not prevented. There can st~ll be a transition to long-term cooling and plant stabilization. The primary difference in this sequence relates to the impact of the depressurization of both steam generators on the ability of the turbine-driven auxiliary feedwater pump to operate. Consequently, failure of the ADVs or SRVs to reseat would result in loss of the turbine-driven pump.

BRANCH POINT 7 This point, defining sequence 1, represents the eventual need to switch to

to the shutdown cooling system.

Long-term cooling may be accomplished by continued use of the systems already operating or by switching As was mentioned previously, part of the long-term cooling success requires that necessary makeup be supplied to the PCS to replenish water lost through seal leakage. Long-term cooling success also requires that backup sources of water be made available to the cooling system in use.

Also, the HPSI or LPSI system will be required to provide borated water injec-tion within a day and a half of the initiating event. Failure to inject borated water will result in a reactivity insertion due to xenon decay.

Failure at Branch Point 7 indicates an inability to continue plant cooling or an inability to provide makeup. Either event will eventually lead to uncovering the reactor fuel and a core melt.

NU0783-0181A-TP20

12 BRANCH POINT 8 Branch Point 8 is similar to Branch Point 7, except the ADVs or SRVs have failed to reseat, thus both steam generators are depressurized.

BRANCH POINT 9 If the ADVs on the intact steam generator fail to open, secondary pressure will increase until the code safety valve set point pressure is reached. At this point, the SRVs will cycle to maintain steam generator pressure. As in Branch Point 7, some form of long-term cooling must be established.

BRANCH POINT 10 At this point the AFWS has failed. In order to supply the intact steam generator with the condensate pumps, the ADVs must be 1 opened to decrease steam generator pressure. Failure at this point precludes use of the condensate pumps to feed the intact steam generator.

BRANCH POINT 11 At this branch point, the AFWS has failed and the intact steam generator pres-sure is being controlled by the ADVs. In order to align the condensate system to supply the intact steam generator, the operator must use the ADVs to lower the steam generator pressure. If the pressure of the intact steam generator cannot be lowered, the operator must attempt to align the condensate pumps to feed the faulted steam generator. Failure of the condensate pumps at this branch point leads to eventual core damage once the pressurizer relief valves lift and begin to release primary inventory in containment. Success results in use of the condensate pumps until long-term cooling can be established.

BRANCH POINT 12

  • This point represents the ability to establish long-term cooling by continued use of the condensate pumps or the shutdown cooling system.

NU0783-0181A-TP20

13

  • As in Branch Point 7, either LPSI or HPSI must be utilized to provide borated water injection to prevent a reactivity insertion due to xenon decay.

BRANCH POINT 13 If the AFWS fails and the intact steam generator cannot be depressurized by use of the ADVs, the operator must align the condensate system to supply flow to the faulted steam generator. Success at this branch point results in the condensate pumps supplying the faulted steam generator until long-term cooling can be established.

BRANCH POINT 14 This point is the same as Branch Point 12, except that the condensate pumps are supplying the faulted steam generator .

  • BRANCH POINT 15 MSIV closure has occurred, however, steam generator isolation has not taken place. Success at this point means that the AFWS is supplying flow to both the intact and faulted steam generator. Failure at this branch will require use of the condensate pumps.

BRANCH POINT 16 This point is the same as Branch Point 7, except that the AFWS is supplying both the intact and faulted steam generator.

BRANCH POINT 17 Failure of the AFWS to supply flow to the steam generators will result in the operator attempting to supply flow to the intact steam generator. Success at this branch point requires the operator to align the condensate system to feed

  • either the intact or the faulted steam generator. If the intact generator is to be supplied by the condensate pumps, the ADVs must be opened to lower the NU0783-0181A-TP20

14

BRANCH POINT 18 This point is the same as Branch Point 12.

BRANCH POINT 19 Upon failure to close of the MSIV associated with the intact steam generator, the AFWS is aligned to supply flow to the intact steam generator. The flow to the faulted generator will be blocked, since it will be the first to experi-ence low-level and pressure. Steam will be blown down from the intact steam generator through the steam line break. Success at this branch point is defined as successful injection of auxiliary feedwater to either of the steam generators. If operator action disables this system in an attempt to control cooldown rate, success implies reestablishing auxiliary feedwater injection prior to boil off of primary coolant inventory above the core.

BRANCH POINT 20 Blowdown of both steam generators is occurring (or has occurred) and auxiliary feedwater injection is successful to one of the two steam generators. At this branch point, the integrity of the primary coolant system is being challenged by way of thermal shock of the steam generator tubes. Success of this branch point assumes no major tube rupture occurs (ie, no large LOCA through the steam generator tubes).

BRANCH POINT 21 .

This point is the same as Branch Point 20 except that success now implies no tube rupture occurs at all.

NU0783-0181A-TP20

15

  • BRANCH POINT 22 This point is the same as Branch Point 7 except perhaps for the shutdown cooling system which may not be available due to environmental reasons if the pipe rupture occurred inside containment.

BRANCH POINT 23 Thermal shock of the steam generator tubes has caused some of them to rupture resulting in a small loss of coolant accident through the steam generators.

Success at this branch point implies successful HPSI actuation to the primary coolant system.

BRANCH POINT 24 The primary system has been cooled by way of the AFWS with makeup to the

  • primary coolant system being accomplished by HPSI to accommodate inventory loss through the steam generators. Long-term cooling at this stage requires continued use of HPSI for makeup or LPSI plus cooling by way of auxiliary feedwater to maintain sufficiently low reactor pressures for low-pressure injection success. As backup to auxiliary feedwater, the condensate system can be placed in service given that the steam generators are depressurized.

BRANCH POINT 25 Thermal shock of the steam generator tubes has caused many of them to rupture resulting in a large loss of coolant accident through the steam generators.

Success at this branch point implies successful safety injection tank opera-tion, flooding the reactor vessel as the primary system cools to LPSI pres-sures. Failure is assumed to lead to temporary core uncovery until LPSI has actuated with a potential for limited core damage .

  • NU0783-0181A-TP20

16 BRANCH POINT 26 The primary coolant system is depressurizing as a result of auxiliary feed-water and the tube ruptures in the steam generators. Flooding of the primary system by the safety injection tanks has been accomplished. LPSI is now required to maintain primary coolant inventory.

BRANCH POINT 27 This point is the same as Branch Point 24.

BRANCH POINT 28 Both steam generators have depressurized as a result of a main steam line break and failure of the MSIV in the intact steam generator to close. Auxil-iary f eedwater has failed to actuate or has been disabled in an attempt to

  • control primary system cooldown rate and for some reason will not restart.

Success at this branch point is manual alignment of the condensate system to the steam generators to provide a means of cooling. Failure implies a heatup of the primary system to the safety relief valve set point, loss of primary coolant through the safety relief valve and eventual core uncovery.

BRANCH POINTS 29 THROUGH 36 These points are the same as Branch Points 20 through 27 except that references to the AFWS are now applicable to the condensate system .

  • NU0783-0181A-TP20

\

FIGURE 1 PALISADES PLANT MSLB EVENT TREE MSLB RPS MSIV

~.G.

~=~~ AFWS ADV ADV Cond Open Close Pump S.G. JS.G. J Integ Integ SI

<larg <smal Tank j HPSI LPSI LTC 6

8 5

I I 9 4 I I 12 11 10 I 14 3

I 13

16 15 I 18 17 I

2 II 22 21 23

~

20 I 27 26 25 I 31 19 I 30 1 I 33 32 29

  • 36 35 28 34

ENCLOSURE 1 ATTACHMENT 2 MAIN STEAM LINE BREAK CONSEQUENCE ANALYSIS Palisades Plant

  • NU0783-0178A-TP20

1 Determination of the Radiation Dose to the Public Summary: The dose consequences for this accident (Main Steam Line Break with MSIV, ECCS and Auxilariy Feedwater Failures) were analyzed using Palisades average meteorology, population distribution for the sector affected, WASH-1400 PWR-2 core release and accident specific containment release fractions and dose conversion factors from the GASPAR Code. l.46E+06 manrem resulted from this accident.

I. THE SEQUENCE OF EVENTS The sequence of events for this postulated accident are:

A. Main steam line break inside containment with failure of the MSIV on the undamaged steam line B. Reactor trip on low steam generator pressure and/or hi containment pressure C. Blowdown of both steam generators results in a peak containment pressure above design D. Steam generators dry out and the tubes uncover resulting in poor primary to secondary heat transfer E. Primary coolant system pressure rises to the setting of the safety valves F. Decay heat dissipation results in the boil-off of primary coolant G. Reactor vessel water level drops below top of core and core heatup begins - melt begins as the water level drops H. Molten core drops into the reactor vessel bottom and begins melt through I. Molten core melts through the vessel and drops into sump J. Containment over-pressure occurs after melt due to steam explosion (from reaction of molten core with the remaining water in the reactor vessel, or the water in the sump), or overpressurization (nonexplosive) due to the gas generation from the core debris -

concrete reaction or the core debris - water reaction NU0783-0178A-TP20

2 The timing of the events were taken from WASH-1400, Appendix V. The event above is similar to TMLB - ~ or TMLB - o. Page V-97 of Appendix V shows the case TMLB -o. The melt starts at 2.6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> and ends at 3.6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br />. Overpressurization is at 3.7 hours8.101852e-5 days <br />0.00194 hours <br />1.157407e-5 weeks <br />2.6635e-6 months <br />. Similar times are given on Page V-88 for TMLB - ~

II. RADIONUCLIDE SELECTION It was thought that this release would be similar to PWR-2 shown on Page V-4 of WASH-1400. The release fractions were:

PWR - 2 Fraction of Core Isotope Inventory Released Xe, Kr 0.9 I 0.7 Cs, Rb 0.5 Te, Sb 0.3 Ba, Sr 0.06 Ru 0.02 La 0.004 To determine which isotope had the greatest effect on each organ of interest, the dose conversion factors (Rem/Ci) for each organ as a result of radiation emitted from each isotope in Regulatory Guide 1.109 were obtained. The Dose Conversion Factors were multiplied by the total curies in the core at 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> post shutdown (from the ORIGEN Code) and then multiplied by the release fraction. This gave the relative radio-logical importance of each radionuclide on the organ of interest. The following radionuclides were chosen for this analysis based on their affect on the body and abundance outside the core after melt.

Kr - 83m I - 131 Sr - 91 Kr - 85m I - 132 Sr - 92 Kr - 85 I - 133 Cs - 134 NU0783-0178A-TP20

3

  • Kr - 87 Kr 88 Xe - 133m I -

I -

134 135 Te - 132 Cs - 137 Cs - 138 Ba - 140 Xe - 133 Rb - 88 Xe - 135m Sr - 89 Xe - 135 Sr - 90 III. CODE Neither the CORRAL Code nor the CRAC Code used in WASH-1400 were avail-able for this analysis. However, the GASPAR Code was available. GASPAR calculates dos.es to the entire population living within SO miles of a nuclear power plant due to radionuclide releases at that plant. The meteorology models in each.code were reviewed and found to be similar.

Both consider decay and deposition during transport. One difference is that GASPAR uses annual average and sector averaged X/Q where CRAC does not. (The section entitled Meteorology shows how this was overcome.)

The dose models were reviewed and found to be similar. Both codes determine the concentration of radionuclides in the air, on the ground and on the vegetation at the area of interest considering decay and depletion during transport. GASPAR assumes that all of the particulates are deposited within 0-50 miles so that the deposition is independent of meteorology. Regulatory Guide 1.111 shows that for a ground level release, 35% of the nuclides remain in the plume at 50 miles. CRAC calculates ground deposition as a function of meteorological conditions.

Use of GASPAR deposition values is therefore conservative. Both con~

sider transport of radionuclides into the body similarly (although CRAC appears to contain a more complicated milk and food pathway model), and both calculate dose from concentration using dose conversion factors 3

(typical units mr - cm ) .

µCi - hr IV. METEOROLOGY NU0783-0178A-TP.20

4

  • CRAC uses the actual meteorological conditions present at the "start" of the accident. CRAC calculates the results from many of these start times (91 in the Big Rock Point case) and then finds the mean and the maximum dose to the population. GASPAR uses joint frequency distributions to calculate the dose to the entire population within a 50 mile radius. To force GASPARs meteorological conditions and dose calculation to be like that in CRAC:
1. The X/Q used in the analyses was based on average meteorological conditions for the year 1982. The value used corresponded to a

.6.7 m/sec wind blowing from the WNW, Pasquill D conditions and a 1,000 meter mixing height.

2. GASPAR was run with annual and sector averaged X/Q but with population removed except for that in the ESE quadrant.
3. It was mathematically determined that the difference between the annual and sector averaged X/Q and the X/Q from Item 1 above is that the X/Q from Item 1 is a factor of 22 greater. This factor is used in Section VII.

V. ATMOSPHERIC RELEASE FRACTIONS To determine the dose to the population, the fraction of the core inventory released to the environment must be determined.

The release fraction for noble gases was taken to be 1.0.

The release fraction for particulates was determined using the equations in WASH-1400, Appendix VII. The removal mechanisms for particulates are decay, sprays and gravitational settling. Decay was not considered since the time between release and immersion is small as compared to the half-life. Inhalation or ingestion is small compared to the half lives of most nuclides. The containment sprays were assumed to have failed so NU0783-0178A-TP20

5

  • that the only mechanism is settling.

is:

A = UA The removal constant for settling v

Where U Settling Velocity A Cross Sectional Area of Room of Settling V = Volume of Room in Which Settling Occurs Settling of particles occurs in both the reactor vessel and the contain-ment. For conservatism, the containment area and free volume were used in the analyses. The particle diameter came from WASH-1400, Page VII -

190 which states that soon after fission product release, the particle diameter was 15 microns and a few hours later the size was 5 microns in diameter. Therefore a particle size of 15 microns in diameter was chosen from Page VII - 194 of WASH-1400:

U d2 (pp - P m) g 18 u Where d = Partical Diameter pp,m = Density of Particle and Medium Respectively g = Acceleration of Gravity u = Viscosity

-1 In reality, U was found to be 0.10 ft/sec and A is then l.2E-03 sec

  • A would be much larger because the particulates will agglomerate to a much larger size than assumed here.

The release fraction was determined by assuming that the radionuclides are released uniformly over the 60 minute melt time and that over this time, removal due to settling occurs. So:

  • NU0783-0178A-TP20

6

  • 60 Where N The Number of Atoms of the Radionuclide of Interest Not The Uniform Release Rate 60 (so that at 60 min N = N 0

if no settling had occurred)

N 0

= Core Release Fraction Times the Number of Atoms in the Core t = Time Numerically solving this equation yields (at t = 60 minutes) N 0.25 N

  • Then for 0.1 hours1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> (according to WASH-1400), further settling 0

occurs with a minimum of further airborne production

  • As a result, the particulate curies available for release are 10% of the curies released from the core.

It is assumed that since the melt is occurring in a confined area (the intact reactor vessel) and steam is present due to the boil-off of the primary coolant system, the iodine will react with the Cs to form CsI and be removed and released like particulates.

In Summary:

Containment Fraction of Core Fraction of Core Release Inventory Released Isotope Inventory Out of Fuel Fraction to Environment Xe, Kr 0.9 1.0 0.9 Rb, Cs 0.5 0.10 0.05 Ba, Sr 0.06 0.10 0.006

  • I Te NU0783-0178A-TP20 0.70 0.30 0.10 0.10 0.07 0.03

7 Note that the accident at SL-1, which was a core melt and steam explo-sion, resulted in the release of 0.005 of the core inventory of iodine.

(Reference NSAC-14 "Workshop on Iodine Releases in Reactor Accidents").

The release fractions would most likely be less for the case in which the steam generator tubes failed resulting in a release outside containment. This is due to the larger surface-to-volume ratio and greater holdup time which is due, in turn, to a much smaller differential pressure in the auxiliary and turbine building.

VI. CONTAINMENT RELEASE The containment was assumed to remain intact until after fuel melt (see WASH-1400, Appendix V). The failure mechanism is either steam explosion due to the reaction of the core melt debris with water in the bottom of the vessel or the water in the sump, or hydrogen burn. The explosion was assumed to result in a rapid depressurization of the containment.

Therefore, all of the radionuclides which have not settled are assumed to be released.

VII. TOTAL DOSE 0-50 MILES Dose conversion factors for each nuclide listed in Section II of this report were obtained by use of GASPAR (1 curie of each nuclide). This gave population manrem to the whole body from immersion in the plume, shine from the ground, inhalation and ingestion for each nuclide. The equations for each dose pathway were investigated to determine how changing X/Q changed the manrem. It was found that the dose due to immersion and inhalation were linearly affected by a change in X/Q (ie, doubling the X/Q doubled the dose) and that dose due to shine from the contaminated ground and ingestion did not vary with X/Q. (See Regula-tory Guide 1.109 and Section III of this report). Therefore, since the X/Q for this analysis (Pasquill D, 6.7 m/sec wind speed, 1000 m mixing height) differed from the X/Q used in the GASPAR runs by a factor of 22, the manrem/Ci for immersion and inhalation were ~de a factor of NU0783-0178A-TP20

8 22 higher. The manrem/Ci for immersion, inhalation, shine from the ground and ingestion were added to obtain the total manrem/Ci. The dose was obtained by multiplying for each nuclide: (Core Inventory Curies at t =3 Hours from ORIGEN) times (Fraction of Core Inventory Released to the Environment) times (Total Manrem/Ci).

The isotopes of each element were summed:

0-50 Mile Element Total ManRem Kr 10,376 Xe 7,696 I 91, 762 Rb 200 Cs 627,000 Sr 612,000 Ba 2,000 Te 400 1,351,434 Note that if X/Q were from Pasquill F (l.Om/sec rather than annual 6

average X/Q) the total would be 8 x 10 manrem *

  • NU0783-0178A-TP20

9

  • VIII.TOTAL DOSE 50-200 MILES It is assumed that all of the iodines and particulates are deposited over the first 50 miles (GASPAR assumption). Using X/Q from Pasquill D, 6.7 m/sec and l,OOOm mixing height, the cummulative dose due to immersion at 50 miles is 0.4R; at 100 miles is 130 mR; and at 200 miles is 50 mR. The analysis was stopped at 200 miles since the cummulative dose here is 1/2 of the average annual United States background. To get the total population dose, the plume was drawn out to 200 miles.

For conservatism, the south sector was used as more people live in this sector. The 0-50 mile region in the south sector stops south of South Bend, Indiana. To get the people in the 50-200 mile area, the population of the major areas are needed as well as the average population of the "rural" areas. The "rural" area population density is given by:

(Total State Population) - (Sum of All of the States Major Area Populations)

Total State Area The dose for 50-200 miles is given by the sum of: 1.) the sum of the population in the major population areas within the plume (2cr y in width) times the dose in those areas plus 2.) the plume area times the "rural" area population times the dose over the "rural" area. For this case, 111,000 manrem was delivered. lf the X/Q was from Pasquill F (l.Om/sec) the dose would be about 4.5 x 10 manrem.

IX.TOTAL MANREM FROM THIS CASE The total manrem from this accident is:

0-50 Miles 1,351,000 manrem 50-200 Miles 111,000 manrem Total 1,462,000 manrem 6

For Pasquill F (1.0m/sec) it could be 12.5 x 10 manrem.

X. COMPARISON WITH BIG ROCK POINT RESULTS

1. Big Rock Point PRA used CRAC to determine the manrem to the popula-tion. CRAC analyzed the dose consequences from 91 different meteorological conditions. The mean total manrem and the maximum
  • were obtained. It is difficult to compare the atmospheric dispersion factor used in the Big Rock Point PRA with that used here.

NU0783-0178A-TP20

10

  • 2. The Big Rock Point case (PEF C -2), with the average consequences s

(56 latent fatalities for the mean case), released the same quantity of noble gases, 3 times more iodine and 25 times more Cs, Ba and Sr. Using the Big Rock Point release fractions instead of those listed in Section V of this report would result in a 0-200 mile total manrem of 3.0E+07.

XI. INDEPENDENT ESTIMATION OF HEALTH CONSEQUENCES FOR CORE MELT SEQUENCES WITH EARLY CONTAINMENT FAILURE AT THE PALISADES NUCLEAR PLANT To allow verification of the simplified consequence analysis presented in this Attachment, an independent estimate of the public health con-sequences associated with the core melt sequences of interest was per-formed. To facilitate this analysis, use was made of the results of the Sandia siting study presented in NUREG/CR-2239 (Technical Guidance for Siting Criteria Development). This report developed Complimentary Cumulative Distribution Functions (CCDFs) for a variety of health effects for all current and planned nuclear power plant sites in the United States. Although these results were developed for a generic 1120MWe PWR using generalized source terms and evacuation characteris-tics, they were of value in estimating the mean manrem exposure for postulated accidents at Palisades.

The approach taken in this verification calculation was to select two source terms utilized in the siting study and to estimate the manrem exposure to the population surrounding Palisades by making the follow-ing assumptions:

a. Mean*latent fatalities from the siting study were converted to 4

manrem exposure by multiplying by 10 manrem per fatality

b. The power at Palisades was taken into account by a normalization factor of 740MWe/1120MWe
  • NU0783-0178A-TP20

11

  • The resulting exposures calculated for the SST-1 and SST-2 source terms 7 5 were 1.1 x 10 and 5.9 x 10 manrem, respectively.

The SST-1 and 2 source terms are shown in Table XI-1 along with source terms from other studies. As shown for most isotope groups, SST-1 is consistent with other severe accident source terms associated with early containment failure in the absence of radionuclide depletion mechanisms (eg, PWR-1 and 2 from the Reactor Safety Study and Z-1 from the Zion study). If, on the other hand, a significant radionuclide de-pletion mechanism is available (eg, containment spray or aerosol settl-ing with sufficient time), the source term SST-2 is more appropriate and consistent with the Z-3 and Z-5 source terms from the Zion study.

Table XI-2 presents a comparison of the data used in the MSIV study with the NUREG siting studies and Table XI-3 presents the comparison of the results .

  • Given the potential for both containment spray and aerosol settling in the subject accident sequences at Palisades, it appears that the SST-2 source term would be more appropriate. This source term produces an integrated population exposure which is consistent with the values cal-6 culated earlier in Attachment 2 (ie, approximately 1.46 x 10 manrems).

Thus, the appropriate basis for cost-benefit analysis is a population 6

exposure of approximately 1.46 x 10 manrems *

  • NU0783-0178A-TP20

TABLE XI-1 SOURCE TERM COMPARISON Core Melt Source Terms from Various Studies (Fraction of Core Inventory)

Reactor Safet~ Stud~ Sandia Siting Stud~ Zion PRA Radionuclide Group PWR-1 PWR-2 PWR-3 SST-1 SST-2 SST-3 Z-1 Z-5 Z-3 Xe-Kr Group 0.9 0.9 0.8 1.0 0.9 0.006 0.9 0.8 0.8 I Group 0.7 0.7 0.2 0.45 0.003 0.0002 0.7 0.2 0.2 Cs-Rb Group 0.4 0.5 0.2 0.67 0.009 0.00001 0.1 0.25 0.2 Te-Sb Group 0.4 0.3 0.3 0.64 0.03 0.00002 0.35 0.25 0.025 Ba-Sr Group 0.05 0.06 0.02 0.07 0.001 0.000001 0.05 0.025 0.02 Ru Group 0.4 0.02 0.03 0.05 0.002 0.000002 0.21 0.015 0.015 La Group 0.003 0.003 0.003 0.009 0.0003 0.000001 o*.003 0.004 0.003 NU0783-0178A-TP20

13

  • TABLE XI-2 Comparison of the Data Used in This Study With NUREG/CR-2239 A. SOURCE TERM Percent of Core Inventory Released to Environment Nuclides SST-1* SST-2* This Study Xe-Kr 100 90 90 I 45 0.3 7 Cs-Rb 67 0.9 5 Te-Sb 64 3 3 Ba-Sr 7 0.1 0.6 Ru 5 0.2 La 0.9 0.03 NOTES: This study and SST Core Melt, Containment Failure is a large direct breach in which no safety systems operate. SST Core Melt, Containment Failure is loss of isolation in which containment sprays operate.

B. POPULATION DENSITY 50-200 MILES NUREG/CR-2239 423 People/Sq Mile (50-100 Miles)

(Table D.1-1) 148 People/Sq Mile (100-200 Miles)

This Study 100 People/Sq Mile + a Population Center of 750,000 People at 170 Miles NU0783-0178A-TP20

14

  • C. METEOROLOGICAL DATA NUREG/CR-2234 This Study Mixing Height 1250 Meters (Fig A.3-1) 1000 m Stability D (Table A. 3-2) D Wind Speed 75 m/s (Table A.3-2) 6. 7 m/s
  • NU0783-0178A-TP20

15

  • TABLE XI-3 Results of the Analyses Man rem NUREG/CR-2239 Case SST-1 for Palisades l.1E+07 NUREG/CR-2239 Case SST-2 for Palisades 6.0E+OS This Study l.46E+06 For Comparison NUREG/CR-2239 Case SST-1 for Big Rock Point 3.8E+05 Big Rock Point PRA - Average Case 6.0E+OS Notes:
1. NUREG/CR-2239 data is from Table C-1 with P 1 and P 2 set equal to 1.0 and the result multiplied by 10,000 manrem/fatality.
2. On Table XI-2 it is interesting to note that the major contributors to population dose (Cs-Rb and Ba-Sr) are analyzed to be on the order of a factor of six higher in containment release fraction in this study than in the SST-2 (or PWR-5) source terms. These two radionuclide groups represent approximately 92 percent of the total population exposure. This observation, together with the above results, leads _to the observation that the GASPAR analysis used in this study produced a manrem exposure of approximately one half that predicted by CRAC-2 in the Sandia Study.
3. The release fractions (except for noble gases) in SST-1 are about 11 times higher than that used in this study. Since the contributors to the dose are Cs-Rb and Ba-Sr, the total manrem for this study would be l.61E+07 if the SST-1 release fractions were used. This compares well with the l.1E+07 manrem obtained in SST-1.

NU0783-0178A-TP20

Retrieved from "https://kanterella.com/w/index.php?title=ML18051A544&oldid=2507223"