ML17329A067

DC Cook Unit 2 Low Temp Overpressure Protection Sys Setpoint Evaluation.

ML17329A067
Person / Time
Site:	Cook
Issue date:	06/30/1989
From:	WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP.
To:
Shared Package
ML17329A065	List: ML17329A064 ML17329A066 ML17329A067
References
NUDOCS 9106270146
Download: ML17329A067 (687)

Text

AMERICA'N ELECTRIC POWER SERVICE CORPORATION 9206270146 9 F'DR ADOCK 05000315 P

PDR g APPROVED tN GENERAL C1 APPROVED EXCEPT AS NOTED Cl NOT APPROVED CI FOR REFERENCE ONLY OA~ >

D. C. COOK UNIT 2 LOW TEMPERATURE OVERPRESSURE PROTECTION SYSTEM (LTOPS) SETPOINT E'IALUATION JUNE 1989 MESTENGHOUSE ELECTRIC CORPORATION 8921 o: 1 d/OSQBB9

'k

$ pic I

P P

M4 ,v 'I 40&M 4j I r '7+4

• st kak wt4 elf,

>} a E<

4i, 1 4

t '

INDEX Section ~Pn e No.

I ntroductlone ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ e ~ ~ ~ ~ ~ ~ ~ ~ e ~ ~ ~ ~ ~ e ~ ~ ~ ~ ~ ~ ee ~ ~ ~ ~ e ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ eS ~ 1 S ummary of Results......................................................S.2 1.0 Description of the Low Temperature Transients.......................l.l 1 .1 General Oescription..................................,...............1.2

~ e ~

1 .2 Operation....................................... ~ e ~ e ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 1 ~2 1.3.Potential Overpressure Transients............... ~ ~ ~ ~ ~ ~ e ~ ~ ~ ~ ~ ~ ~ 1 ~ 5 ~ ~ ~ ~ ~ ~

1.3.1 Summary of Hass Input Transients.......... ~ ~ ~ ~ e ~ e ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ e le 5 1.3.1.1 Inadvertent Safety Injection...... ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ele5 1.3.1.2 Charging/Letdown Flow Hismatch.... ~ ~ ~ ~ ~ ~ ~ ~ ~ e ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ le 7

, 1.3.2 Summary of Heat Input Transients,......... ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ el 7~

1.3.2.1 Actuation of Pressurizer Heaters.. ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ele7 1.3.2.2 Loss of RHRS Cooling.......................... ~ ~ ~ ~ ~ ~ ~ ~ 1 e8 1.3.2.3 RCP Startup With Temperature Asymmetry........ ~ ~ en ~ ~ ~ ~ 1~8 1.3.2.4 Relative Severity of the Heat Input Transients ~ ~ ~ ~ e ~ ~ le 10 1.4 Summary of Transient Evaluation.............................. ~ ~ ~ ~ ~ ~ e le 10 2.0 Description of the LTOPS Setpoint Algorithm. ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ e ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ e2e 1 2.1 Pressure Limits Selection.....,............. e ~ ~ e ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 2~1 2.2 Hass Input Considerations...,............... ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 2e 7 2.3 Heat Input Considerations................... ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ e2e 7 2.4 Final Setpoint Selection.................... ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ e ~ ~ 2~8 3.0 LTOPS Setpoint Analysis for D.C. Cook Unit Z. ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ e3 ~ 1 3.1 Operational Limits..............,............. ~ ~ ~ ~ ~ ~ ~ 0 ~ ~ ~ ~ ~ ~ ~ ~ ~ e ~ ~ ~ I 1302 3.2 PORV Stroke Time............................. ~ e e ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ e ~ ~ ~ ~ ~ e3 ~ 5 3.3 PORV Operation........................,...... ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ e ~ ~ ~ ~ ~ e3e5 3.4 Hass Input Considerations.................... ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 3~6 3.5 Heat Input Considerations.................... ~ ~ ~ ~ e ~ ~ ~ ~ ~ ~ ~ e ~ ~ ~ ~ e ~ ~ ~ e3 ~ 11 3.6 Specifications for Hass Input Transients..... ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 3 ~ 18 3.7 Specifications for Heat Input Transients..... ~ ~ ~ e ~ ~ ~ e ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 3e 19 3.8.Setpoint Evaluation.......................... ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ e 3 ~ 20 89216:1d/080889

I' 1

~ ~

INDEX (Cont'd)

I'ection ~Pa e No.

4.0 Correlation to HOG LTOPS Setpoint Hethodology... ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 4~1 4.1 HOG Hethodology LTOPS Setpoints................. ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 4~1 4.2 LOFTRAN/MOG Correlation......................... ~ o ~ ~ ~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 4 ~ 10 4.3 Impact of Steam Generator Tube Plugging......... ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ o ~ ~ ~ 4 ~ 14 8S21 e:1d/060S89

Wgg I

~g

~,

SI Ae ~ gr, I!., =p. ~4, --y > ~ ~ i p )fjg+g

<~ "%. lj L V t+;

'YflQE,<+4)l '~"pl 'W'ltk~k ~ ~ ~ ++a <<~'I A(

D. C. COOK UNIT 2 LOW TEMPERATURE OVERPRESSURE'PROTECTION SYSTEM SETPOINT ANALYSIS INTRODUCTION USNRC Regulatory Guide 1.99 Revision 2, "Radiation Embrittlement of Reactor Vessel Materials," dated Hay, 1988 became official with it's publication in the Federal Register on June 8, 1988. The guide revises the general procedures acceptable to the NRC staff for calculating the effects of neutron radiation embrittlement of the low alloy steels currently used for light water cooled reactor vessels.

Appendix G of 10 CFR Part 50 provides the fracture toughness requirements for

'"" reactor>pressure vessels under certain conditions. To ensure that the Appendix G limits are not exceeded during any anticipated operational occurrence, technical specification pressure-temperature limits are provided 0 during low temperature operations. The embrittlement algorithm specified by revision 2 of Regulatory Guide 1.99 is more conservative than revision 1, and requires that these limits be re-calculated.

The Low Temperature Overpressure Protection System (LTOPS) provides protection against exceeding the vessel ductility limits, as expressed by the Appendix G pressure-temperature limits, during cold shutdown, heatup, and cooldown

"""'operatio'ns." 'The 'limits resulting from implementation of the new revision to Regulatory Guide 1.99, requires that the LTOPS setpoints be re-evaluated. The purpose of this report is, in part, to document the re-evaluation.

This report includes a study of the sensitivity of the LTOPS setpoint on pressurizer PORV opening time, and a correlation that benchmarks the results of the analysis to that of the algorithm described in the report "Pressure Mitigating Systems Transient Analysis Results" (July 1977). This report was prepared for the Westinghouse Owners Group (acronymed "WOG") on Reactor Coolant System Overpressurization by Westinghouse Electric Corporation. The purpose of the correlation is to provide American Electric Power Corporation a 8921 e:1d/060789 S.1

k V

sf'tff i

4 '4 of determining LTOPS setpoints equivalent'o those obtained from the

~ ~

means

~ ~

relatively sophisticated LOFTRAN based analysis described here, by using a

~ ~

...simple methodology; i.e., the HOG report. The correlation remains valid as

~ ~ ~ ~

~

~ ~

long as certain plant parameters are unchanged. These are: instrumentation time delays, pressurizer PORV flow characteristics, and pressurizer PORV full flow C(v).

The organization of this report, apart from the introductory comments and the summary statement, is in four sections: the first and second sections describe, respectively, the justification for the design basis transients and the algorithm used for the setpoint analysis; the third section documents the analysis specific to 0. C. Cook unit 2, and the fourth section provides the correlation to the HOG methodology referenced in the introduction to this repor t.

Summar of Results The result of the analysis is summarized by F1gures S.1 and S.2, illustrating, respectively, the dependency of the maximum allowed LTOPS setpoint on PORV opening time for reactor vessel'xposures of 12 EFPY and 32 EFPY. A minimum setpoint limit, for RCP number 1 seal protection, does not exist. This is due to the fact that, given the 4 second PORV closure time, there is not enough ,

separation (white space) between the steady-state'pressure-temperature limit and the minimum RCS pressure requirements for an RCP start, to umbrella the pressure swing resulting from either a heat injection or mass injection event. As the closure time decreases, the pressure undershoot would become less severe. At the current 435 psig setpoint, with single PORV operation, the peak pressure would remain below the Appendix G limit provided the PORV stroke open times remained below 6.5 seconds, with vessel exposures to 12 EFPY; or 3.5 seconds, with vessel exposures to 32 EFPY.

Figure S.3 provides the correlation that benchmarks the results of the "LOFTRAN" based analysis to the results obtained from application of the algorithm described in the "MOG" report. These curves are used to compensate for some of the conservatisms or nonconservatisms, depending on the selected PORV stroke open time, inherent in the HOG methodology. These correlations 8921 e: 1 d/060989 S.2

'i

\

'M ll I

1

~

.are independent of, Appendix G limits,

~ ~

and will remain valid

~ ~

as long as

~

certain

~

plant parameters (pressurizer PORV flow characteristics and C(v.), and

~ ~ ~

instrumentation signal delays) are unchanged. ~

Utilization of the curves requires that the LTOPS setpoint first be determined using the "MOG" methodology. At the PORV opening time corresponding to that selected for the "MOG" calculation, determine the "LOFTRAN" analysis setpoint from the ordinate by linearly interpolating between the two curves bounding the predetermined "MOG" setpoint.

8921 e:1d/060789 S.3

~ L F l~ go qg @14 1

Notes t 1 ) l4 Pressut ~ lnstrueent Error

2) Stngle PORV Operet ton

3) PORV Closure Ttne ~ 4.I Sec.

4) Reector Vessel Exposure ~ 12 EFPY 6

s I58.8 Oeq F RCS Te~p. i 85.8 Oeg F Figure S.l PORY LTOP Setpoint vs. Ya1ve Opening Time at 12 EFPY 8921 ~:1d/060289 S.4

0 4A. lilt t ~ I .pleiad 5 .

Agf +

tl 1

I'

Roice- 1) Ho Pressure lnstrunent Err of

2) Single PORV Oper ation

3) PORV Closur e Tine ~ 4.8 Sec.

4) Reactor Vessel Exposure ~ 32 EFPY I

C C

)Sb.b Oeg F RCS Temp. =

, E., 85.b Oeg F Figure S.2. PORV LTOP Setpoint vs. VaIve Opening Time at 32 EFPY d921 ~:1d/0602S9 S.S

F

( OI V

gr.y 5 p e eP i,,

V I W i ~

I t pC ~

t

M06 LTOPS Setpnt (paiq ) ~

Figure S.3 LOFTRAN/MOG LTOPS Setpoint Correlation vs.

Pressurizer PORV Opening Time M21e:1dj060289 S.G

a 0

~r.

1.0 OESCRIPTIQN OF THE LOW TEMPERATURE PRESSURE TRANSIENTS Overpressure protection for the reactor coolant system (RCS) is achieved by means of self-actuated steam safety valves located high in the steam space of the pressurizer. These safety valves have a set pressure based on the RCS design pressure and are intended to protect the system against transients initiated in the plant when the RCS is operating near its normal temperature.

To avoid brittle fracture at low reactor vessel metal temperatures, the allowable system pressure is progressively reduced from the nominal system design pressure as temperature is decreased. Therefore, supplemental overpressure mitigation provisions for the reactor vessel must be available when the RCS and hence the reactor vessel, is at reduced temperatures. This supplemental protection, utilizing the power operated relief valves (PORYs),

.-is known. as the Low Temperature Overpressure Protection System .(LTOPS).

The PORVs are designed to. limit the RCS pressure during normal operational, transients when the reactor is at power by discharging steam to the pressurizer relief tank (PRT), thus avoiding the need for the code safety valves to.function. The, flow capacity and stroke time of the PORVs is selected to avoid a reactor trip during a large step load decrease. In addition, the valves are uti lized for pressure relief (water, gas, or a mixture). as a part of .the LTOPS,,and when performing this function will also.

discharge to the PRT. The setpoint for the LTOPS function is selected so that

.if. one, valve fails,.to actuate when required, the, second valve will be able to mitigate the transient.

Normally, when the RCS is at a temperature below 350 F, the RCS is open to the Residual Heat Removal System (RHRS) for the purposes of removing residual heat from the core, providing a path for letdown to the purification subsystem, and controlling the RCS pressure when the plant is operating in a. water solid mode. The RHRS is provided with selfactuating water relief valves to prevent overpressurizing this relatively low pressure system due to events originating either from within the system itself or from transients transmitted from the RCS. The RHRS relief valves will mitigate pressure transients originating in the RCS,to maximum pressure values determined by the relief valves set 89216:1d/0602S9

0

,h >~

$A V>

f-

$ eH

4 pressure plus a pressure accumulation above the set pressure dependent on the liquid volume magnitude of the transient.

The low pressure RHRS is normally isolated from the high design pressure RCS during operation at temperatures above 350'F by two isolation valves in series. Because of the NRC required automatic closure feature designed to prevent inadvertent overpressurization of the RHRS, spurious closure of the RHRS isolation valves is a creditable event. The LTOPS is intended to provide overpressure mitigation for the RCS, considering those transients which might occur when the RHRS isolation valves are inadvertently closed, thus isolating the RHRS water relief valves from the RCS.

1.1 GENERAL OESCRIPTION

'~'The "LTOPS'"is'designed'o provide the capability, during relatively low temperature reactor coolant system operation, to prevent the RCS pressure from exceeding 10CFR50 Appendix G limits. The LTOPS is provided in addition to the administrative controls to prevent, overpressure transients and as a supplement to the RCS overpressure mitigating function of the residual heat removal system water relief valves. The system is designed with redundant components to assure its performance in the event of the failure of any single active component.

The power operated relief.valves located near the top of the pressurizer, together 'with'additional actuation logic. from the wide range pressurizer channels, are utilized to mitigate potential RCS overpressure transients which might occur- if the RHR water relief valves are isolated from the RCS. LTOPS provides the additional'relief capacity for the specific transients which would not be mitigated by the RHRS relief valves and thereby maintain the system pressure below the limits specified by Appendix G requirements. ,

1. 2 OPERATION During normal plant heatup, the RCS is open to the RHRS and is operated in a water solid mode unti I the steam bubble is formed in the pressurizer. During 8821 a:1d/060289

I lt 4

1~

tt '

I IC

~

• r~ IJ%r')>

~ .~) ~ P I

I-- ig l7 these low-temperature Iow-pressure operating conditions, the LTOPS is armed and in a ready status to mitigate pressure transients which might occur if the RHRS is inadvertently isolated. After the steam bubble is formed and the pressurizer water level is at its'ormal value for no"load operation, the RHRS is manually isolated from the RCS and the plant heatup continues..

Pressure surge control is provided by the steam bubble. tthen the reactor coolant temperature has increased above a preset value (152'F, in the case of D. C. Cook Unit 2) the LTOPS is disarmed.

During a normal plant cooJdown,r the LTOPS is armed as the reactor coolant temperature is decreased below the preset value. At this time there is a steam bubble in the pressurizer and the water level is at the normal level for no-load operation., The RHRS has been placed in service by opening the suction

~ isolation valves; thus making the RHRS water relief valves available to

'itigate" ressure transients. Rhen the coolant temperature has been reduced to abou 160'y the operation of the RHRS, the steam bubble may be quenched and the reactor coolant pumps stopped. From this point on in the cooldown, the plant is water solid and if the RHRS becomes inadvertently isolated, the LTOPS will be in an active status ready to mitigate those pressure transients that might occur.

Mhen the RCS, is operated in the water solid mode, the pressure is automati-..

cally controlled by the low pressure 'letdown valve .in 'the= Chemical- and Volume Control System (CYCS). This valve senses the pressure in the letdown line (ref Figure 1.1) and maintains the pressure at the selected control value by throttling the letdown flow from the RCS. At this time, the charging flow into the RCS is set at a constant value. and is controlled by the charging flow control valve. It should be noted that. the pressure being controlled is that in the letdown line which then indirectly controls the pressure in the RCS.

However, if the pressure drop through the RHRS and the bypass line into the CVCS is changed by throttling of valves or changing the flow rate through the RHRS, the RCS pressure will also change since the location of the controlled pr essure is in .the letdown line.

8921 a:1d/060289 1.3

~ ~

'I g;

4I 1 r

~ j\

Charging F1' Conlrol l~

Pressure Purl f lca tie P 0~

RIIR IIX IIX eCV Rel lef Yalve IIC f g I.ethel 4 Pressure tontrol RIIR IIX RIIR Pump I'igure 1.1 Typical Low Pressure Coolant/Purification Flog Path 892Ie:1d/06078$

'E el II

'I g I

1;fi '~~-w g 0; t

7 1.3~ POTENTIAL OVERPRESSURE TRANSIENTS

",During~

low temperature operations, reactor coolant system overpressurization transients can be caused by either of two types of events: mass input or heat input. Both types result in more rapid pressure changes when the RCS is water solid. However, at reactor temperatures below 350'F, the RCS must be aligned to the Residual Heat Removal System to remove core decay heat and the water relief valves will be available to mitigate pressure transients which might occur. Also, there will generally be a steam bubble in the pressurizer when the reactor coolant temperature is above about 150'F during plant cooldown so that water solid conditions are limited to relatively low temperature conditions. Therefore, the descriptions of the two types of transients imply that the RCS is water solid at a low temperature.

1.3.1 Summar of Mass Input Transients 1.3.1.1 Inadvertent Safety Injection Inadvertent actuation of safety injection events include full system (both trains), single train, or single component within a train actuation. Each of the three types of events are discussed separately. Full system actuation would include the opening of the isolation valves on all SI accumulators, startup of all low-head and high-head safety injection pumps and isolation of

,=;the normal. letdown path <to -,the Chemical.and .Volume .Control System. Such an event would result in unacceptably large volumes of water being forced into the RCS. Therefore, such events must be prevented by strict administrative controls. These controls require the blocking of the automatic SI actuation circuits, immobilizing the SI accumulator motor operated isolation valves, and locking out power to the high-head SI pumps. In a typical westinghouse design, the low-head safety injection pumps (the RHR pumps) are normally in operation and aligned to take their suction from the RCS, and not from the refueling water storage tank, during low pressure and low temperature plant operations. Therefore, even if a spurious start signal was received, the low-head SI pumps would not function to inject RMST water into the RCS.

8921 e:1d/060289 1.5

P,. j g~>Ãi ~

'I I 1

l ':p44Dg+ A f- A1 II t MC

The probability of a'single train actuation is about the same as a full system actuation, since the signals which call for safety injection, both manual and

'automatic, are normally processed through the engineered safety feature logic circuits such that a signal, whether spurious or not, will impact both trains. Therefore, since the SI system is essentially immobilized at low temperature, single train inadvertent actuation is considered no more likely than full system actuation.

For those plant designs in which the charging pumps serve the second function of high head SI pumps, e.g. the Cook units, a spurious safety injection signal would realign the valving to transfer from the charging function to the safety injection function. Therefore, the one operating charging pump which is not locked out will deliver through the safety injection flow path to the RCS.

~'-'~ ~'~',~-However,,'~the~RHRS.would. remain open to the RCS at this time and the RHRS relief valves would mitigate the resulting RCS pressure transient.

Inadvertent actuation of a single component would require that an operator selectively unlock the electrical power to the component and then cause the component to be energized. The most probable way for this event to occur would be during periodic surveillance testing required by the technical specifications or during post maintenance checkout of the component.

Deliberate opening'f an SI accumulator, isolation valve while ..the accumulator..

is pressurized with gas is not considered probable because there is no technical-'specification",to.test the isolation valves at shutdown, and prudent maintenance procedures for the valves would likely require that the compressed gas in the accumulator be removed.

Post maintenance checkout or periodic surveillance tests however, might be attempted on a high head safety injection pump providing an opportunity for operator error to cause an inadvertent single pump injection event.

Therefore, a single safety injection pump startup event during surveillance testing or following maintenance .is considered a potential mass input transient. Since the RHRS would be open to the RCS at this time, the RHRS relief valves would mitigate the resulting RCS pressure transient.

89216:1d/060289 1.6

%or e Al 0

~ vltos M ~ g ~ II% ~ S m~ r in

&4 i ghltl > 4 I C a ~ Sled ~ p i ~ '* rw 's ww w <<j "ii e ea I ~

1.

~

1.3.1.2

~ ~ Charging/Letdown Flow Mismatch Charging/letdown flow mismatch events can be postulated to occur in a number of ways. One way would involve the complete termination of letdown: closure of the letdown control valve, isolation of the RHRS/CVCS crossover path, or closure of the RHRS inlet isolation valves caused by malfunctions of the control systems. A second way would involve an increase in the charging flow by either operator or instrument error such that the charging flow exceeds the prevailing letdown flow.

The most severe mass input transient would occur if the letdown flow controls failed to the. zero flow condition while the charging flow controls failed to the full flow condition. This failure mode would result in the maximum

'-"-"'"" '"'" charging/'letdown'flow-mismatch event but does not result in the isolation of the RHRS relief valves from the RCS. Therefore, the RHRS relief valves would mitigate this transient and prevent an overpressure condition in either the RHRS or the RCS.

The most likely way that a charging/letdown flow mismatch would occur is for the RHRS (and relief valves) to be inadvertently isolated from the RCS by spurious closure of the RHRS inlet isolation valves. Such a spurious closure is credible'due'o the- presence'f.::the automatic closing..signal, required by the NRC. Since this spurious valve closure event causes an RCS pressure transient-by" stopping" the 'letdown flow and concurrently isolates the RHRS relief valves from the RCS, it is considered a "design basis" transient for the LTOPS.

1.3.2 Summar of Heat Input Transients 1.3.2.1 Actuation of Pressurizer Heaters The inadvertent actuation of the pressurizer heaters when the pressurizer is filled solid will cause a slow rise in the water temperature with a consequent increase in pressure of the constant volume RCS, if the installed automatic

'ressure control equipment is not in service. Since the pressure transient is 8 921 e:1d/060269 1.7

~ CP <<So n

'llli l

f f

v~a a bC

very slow, .the operator should recognize and terminatethe transient before an unacceptable pressure is reached. The RHRS is open to the RCS whenever the pressurizer is filled solid, in accordance with administrative controls.

Therefore, the RHRS relief valves will be actuated, negating the need for the LTOPS, if the operator does not intervene and stop the transient. This case

's not considered significant to the design of either the RHRS relief valves or the LTOPS.

1.3.2.2 Loss of RHR Cooling A loss of residual heat removal cooling while the pressurizer is filled solid could be caused by a loss of flow malfunction in the component cooling water or the service water systems, or the closure of the RHRS inlet isolation valves."..The continual .release of core residual heat into the reactor coolant, with no heat rejection into the environs, would cause a slow rise in the coolant temperature and pressure. Since the transient is slow;- the operator should respond and either restore the RHRS valves to their open position, restore cooling, or limit the RCS pressure by venting the pressurizer. This transient. is not considered significant to the design .of, the LTOPS since it is a relatively slow transient compared to the heat input transient resulting from the startup of a reactor coolant pump in combination with an RCS/SG temperature asymmetry.

.. '., 1.3.2.3.,RCP.Startup Ni.th Temperature Asymmetry During plant heatup and cooldown operations, typical administrative controls require that at least one reactor coolant pump be maintained in operation whenever the reactor coolant temperature is greater than 160'F. Therefore, the'arge volumetric =flow throughout the RCS will maintain an isothermal condition in the RCS. The steam generator secondary side water immediately surrounding the tubes will also remain at a temperature near that of the circulating reactor coolant on the primary side.

During normal cooldown operations, when the reactor coolant temperature has decreased 'below 160'F the reactor coolant pumps may be stopped. Subsequently, 8921 e:1d/060789 1.8

E v P+,va t, @~~+A ~ ! I ~, ~ ~ Ali~~

~< k1 4 ey lip" ~ I t gl 4 QQ i ~

W F

~1

isothermal-conditions may no longer exist. The-.reactor, coolant temperature will be decreased below 160'F by heat rejection through the RHRS. The steam

'generator contained water (both primary and secondary) may remain at a relatively constant temperature, greater than the RCS temperature, due to little or no circulation through ihe tubes. Therefore, a significant temperature asymmetry may develop between the primary water contained in the steam generator and the rest of the RCS. If a reactor coolant pump were to be started, the sudden heat input into the reactor coolant from the steam generator would cause a rapid increase in reactor coolant temperature. If the event were to occur while the pressurizer is filled solid, a rapidly increasing pressure transient would occur.

In accordance with typical administrative controls, the plant will be under

-water..sol;id;.conditions..only while the RHRS .is. in service, and at least one reactor coolant pump will be in operation at reactor coolant temperatures above 160'F. Therefore, this type of heat input transient will be limited to initial coolant temperatures below 160'F. Since it is not practical to determine a representative temperature for the large stagnant volume of

-<< secondary water in the steam generator,,the, operator will not be aware that the temperature may be substantially different from the remainder of the reactor coolant. From the initial isothermal temperature of 160'F when the RCP is. stopped, the bulk. reactor .coolant temperature is unlikely to decrease ...

below 110'F without some extraordinary cooling means, while the steam

-<'.=-':"~.. gene'rator..water,.may, remain, near..160'F,...Therefore,.the differential temperature is not expected to be greater than 50'F for this type of heat input transient.

If the RHRS is inadvertently isolated from the RCS by closure of the isolation valves, as bj a spurious operation of the required auto/close "interlock, while, the plant is water solid and in mode 4, typical technical specifications require that a reactor coolant pump be restarted within one hour if an RHR loop cannot be returned to ser vice. During the potential one hour delay period, a temperature asymmetry in the reactor coolant lo'ops, due to the continued input of cold seal injection water, could develop and not be S921e:1d/0602BB 1.9

lP'g ~,

(s r

~ yp"y ~wp~ y I g" ~ 44.= ~ > a I%~~

.$ 4.

~

CCQ + ~ su., ~m gyjpy k pr,, a A

~ 4 apparent to the operator. Then when the reactor coolant, pump, is. restarted, an increasing pressure transient will occur as the cold water contained in the

~ ~

'steam generator/RCP cross"over pipe mixes with the rest of the RCS water..

1.3.2.4 Relative Severity of the Heat Input Transients Figure 1.2 compares the relative severity of the pressure transients resulting from the heat input cases discussed in the previous paragraphs, as analyzed for the Hestinghouse Owners Group. From an inspection of the figure, it is evident that the heat input cases from pressurizer heaters and decay heat are less significant than those for the cases with a loop asymmetry. Therefore, these less significant cases are not considered for the setpoint analysis.

Similarly, the loop seal (cross-over pipe) asymmetry case is seen to result in

<-a "relative'ly,.small pressure transient-:compared-,to the potenti'al excursion possible from the steam generator/RCS temperature asymmetry cases. The "design basis" case for the setpoint determination is therefore the temperature asymmetry between the steam generator and the RCS.

1.4

SUMMARY

OF TRANSIENT EVALUATION Based on the previous discussion, most of the identified mass input and heat input transients which might occur while the plant is water solid, will be mitigated by the water relief valves in the RHRS. However,-for those remote

"'-"'~ " cases~which".occur or~.are. caused by..the.RHRS having,.become -isolated from the RCS, the LTOPS may be called upon to mitigate certain increasing pressure transients. Specifically, the LTOPS design basis transients are: 1) the mass input transient caused by a normal charging/letdown flow 'mismatch after termination of letdown flow, and 2). the heat input transient caused by the restart of a RCP when the RHRS is not open to.the RCS.

8921 e:1d/060289 1.10

~1 4

4 II I '

5J

~ "

4'1 4tw

",S" ups ~ ig + 4 l ~E 44+ ><y

• IL r>

t,:,

II p

P 0'

C J

" 4)l Yp mksa.p L sl q

.1

~ ~

Single RCP atart vitae no relief valve acutataen 4'tA + "f Ig '%' why RCS/S6 ~ l56 F/2BB F Loop Saai/RCS '86 F/218 F Cor ~ Oacay Ht h ddt t ion ht ti Figure 1.2 Relative RCS Pressure Transients Resulting From Heat Input Events

$ 82l e:1d/060289

Al I J,'7

2 .0 DESCRIPTION OF THE LTOPS SETPOINT ALGORITHM 4

The determination of the low temperature overpressure protection setpoint is based on a local version of the LOFTRAN code. The LOFTRAN code predicts plant transient thermal/hydraulic behavior by modeling the reactor coolant system, including the steam generators, pressurizer (including PORVs), and reactor coolant pumps, as well as the control and protection systems, selected valving, and, some balance of plant systems. Two versions of the LOFTRAN code were utilized: the first version, used for the mass input calculations, collapses the several RCS loops into a single 'loop model; the second version, used for the heat input calculation, models each loop explicitly.

The selection of the proper LTOPS setpoint requires the consideration of a number of system parameters. Among these are the following:

1: 'Volume'of the reactor coolant involved in the transient.

, 2., RCS pressure signal transmission delay.

3. Volumetric capacity of the relief valves vs. opening position.

4. Stroke time of the relief valves (opening and closing). If the pressure undershoot is important, the closure time is required.

5.'ass input rate into the RCS.

1

, G. Heat transfer characteristics of the steam generators.

7. Initial temperature .asymmetry between the RCS and,steam generator secondary water.

8. Mass of steam generator secondary. water.

9. RCP startup dynamics.

10, RCP No. 1 seal delta P requirements. Important if a lower setpoint limit is to be specified for RCP seal protection.

11. Appendix G pressure/temperature limits for the reactor vessel.

2-1 PRESSURE LIMITS SELECTION The function of the LTOPS is to prevent the RCS pressure from increasing above the PORV piping/structural analysis limits and the limits prescribed by the allowable pressure-temperature characteristics for the specific reactor vessel 8921 e:1d/072089 2.1

C" Qt

~1 f2%; <<1 C

F Et~

C E a

material in.,accordance, with the;rules .given in .Appendix..G .to 10CFR50. For the Cook units, a constant pressure setpoint, independent of temperature, is employed so that the limits considered for,this particular case must meet the most restrictive segment of the Appendix G curves when compared to the overpressure transient as a function mf ~perature. The PORV piping/

structural limits are well above those pressures of concern here,'nd are generally considerations only for those plants whose PORY setpoints are functions of temperature.

A characteristic pressure-temperature relationship is shown in Figure 2.1, i llustrating the allowable system pressure increases with increasing temperature. This type of curve sets the nominal upper limit on the pressure, which should not be exceeded during JKS increasing pressure transients. When

<Qlg~ 1 j't jl -.a'relief. val.ve".is..actuated to mitigate an mcreasing pressure transient, the release of a volume of coolant through the valve will cause the pressure increase to be slowed and reversed as illustrated by Figure 2.2. The system pressure then decreases, as the relief valve discharges coolant, until a reset pressure is reached where the valve is signalled to close. Note that the pressure continues to decrease below the reset assure as the valve closes.

The nominal lower limit on the pressure during the transient is selected based

. on a requirement of the reactor coolant pump No. 1 .seal to maintain a nominal 200 psi differential pressure across the seal faces.

'<< '..~.:~>>..The nominal;;upper.."l.hami,t,(basedonthe..minimum .of.AppendixG, requirements or the PORV piping limitations) and the nominal RCP No. 1 seal lower I limit pressure values create an acceptable pressure range into which the PORV setpoints must be fit. An illustration of these limits al'ong with the setpoint selection range is shown in Figures 2.3 and 2.4. In the event that the setpoint selection range is insufficient to accommodate both the Appendix G,.and the RCP No. 1 seal limit, the Appendix G limit will take precedence.

S921e:1d/060789 2.2

s.gy

"*it II

)

'I

~

r

'id 4,~

'ii 'I tkt4>$ 48'iick'laity %i" i car~

~i Pi<

vg 2SOO I I I I I I I ~ I I 2250 I I I I I I I I I s I I I I I I

~ I I ~ I I I I I I I I I I I I 2000 I I i I I I I I I I I I I I I I I I I I I I I I I I I I

...1'750, I I I

I I I I I I I I I I I I I I I I ~ I I 1500 I I I I I I I I I I I I I I I I I line cceptabl e 12SO Operation Acceptabl e I I I Operation I I I I I I I I I I 1000 1 I I O

4l I I

750 I I I I II~, '- C001 dONl Rates 500 'F/Hr 0

I '

20 I I I 250 40 ~

60 I I I ~ I 100 I I I I ~ I 0

50 100 '50 200 250 300 1NOlCATED TEK1'ERATURE (DEC.F')

FigUre 2.1 Typical Appendix G P/T Characteristics 8921 e:1d/060289 2.3

J

!v

'4 r lr

<<~i t

~ .

-'PA

P Over

'SETPO?RT

- RESET Under TlNE Figure 2.2 "Typical Pressure Transient d921e:1d/060269 2.4

APPENDIX G MAXIMUM liMIT L YilN P

MIN L

RCP fl SEAL MINI YlN LIYilT SETP" >" RAN"~

PDRV SKTPOlNT, l'SIG Figure 2.3 Setpoint Determination (Mass Input) d$ 21e:1d/060289 2.5

0 4 4/WPM - 1 lf D

<<y

Appendix G Maximum Limit '

NAX BIN RCP fl Seal Ninimvm Limit SETPOINT RANGE PS PORV SETPOINT ~ PS IG Figure 2.4 Setpoint Determination (Heat Input)

SS21e:1d/0602S9 2.6

0 1

l1 t, '"+ 'A 'b" t f1~

Itt QJ" P l -V I A I

'I <

~g

t 2.2 MASS INPUT CONSIDERATIONS For a particular mass input transient to the RCS, the relief valve will be signalled to open at a specific pressure setpoint. However, as shown in Figure 2.2, there will be a pressure overshoot during the delay time before the valve starts to move and during the time the valve is moving to the full open position. This overshoot is dependent on the dynamics of the system and the input parameters (e.g. mass flow rate), and results in a.maximum system

. pressure somewhat higher than the set pressure. Similarly, there will be an undershoot while the valve is relieving, both due to the reset pressure being c below the setpoint and to the delay in stroking the valve closed. The maximum and minimum pressures reached in the transient are a function of the selected setpoint and must fall within the acceptable pressure range shown in Figure 2.3. A number of mass input cases are run at various LTOPS setpoints

."'"<'. *'-"se'1'ecte'd'Co"bound*the expected 'setpoint range and over a range of mass injection'ates. From these runs, a locus of the maximum and minimum pressure values is generated over the expected setpoint range, as shown in Figure 2.3.

The shaded area represents the acceptable range from which to select the setpoint. The massinjection cases are conservatively analyzed at low

~ ~

~

temperature where "the bulk modulus of the fluid is greatest. The resulting overshoot is therefor worst case, and evaluating mass injection events at higher, temperatures is not required.

2.3 HEAT INPUT. CONSIDERATIONS The heat input case is analyzed in about the same way as the mass input case except that the locus of transient pressure values vs. selected setpoints are determined for several values of the initial RCS temperature. This heat input, evaluation provides a range of acceptable setpoints dependent on the reactor coolant temperature, whereas the mass input case is limited to the most restrictive low temperature conditions only. The shaded area on Figure 2.4 highlights the acceptable range for a heat input transient for a particular initial reactor coolant temperature.

8921I:1d/060889 2.7

tJ I

',2.4 FINAL SETPOINT SELECTION

.By,.superimposing the results of the several mass input and heat input cases evaluated (from a series of figures such as 2.3 and:2.4), the range of allow-able setpoints can be determined that will satisfy both mass input and heat

'nput considerations. As previously stated, the seTection of the pressure setpoints for the PORVs is based on the use of nominal upper and lower limits.

Use of nominal values is justified based on the recognized high degree of conservatism inherent in the 10 CFR 50 Appendix G pressure-temperature limits.

8921 I:1d/060789 2.8

0 p~, I'.~II't> ~ Pic.ALC E,

\

I

/

4

'3.0 LTOPS SETPOINT ANALYSES FOR THE D. C. COOK UNIT 2 The.LTOPS setpoint analysis presented in this section was developed for the American Electric Power Corporation s D. C. Cook unit 2 using the algorithm described in the previous section. The analysis evaluates the impact of PORV opening times up to 10 seconds on the value of the setpoint required to maintain overpressures below the 10CFR50 Appendix G limits. PORV closure time is assumed to be 4 seconds for all cases. In addition, this section documents the development of the correlation benchmarking the results of the analysis to that of the algorithm described in Mestinghouse Owners Group (MOG) report referenced in the introduction to this report.

The LTOPS currently installed at Cook Unit 2 features a constant value setpoint program (i.e. a program independent of temperature) with an

. '"enabl'e/disabl'e"reactor coolant system temperature of 152'F.- The program setpoint is currently 435 psig for both PORV's.

The analysis assumes overpressurization transients resulting from either mass injection or heat input events under 4-loop water solid conditions, with the replacement model 51F steam generators.'he mass injection transient occurs as a result of the operation of a single centrifugal charging pump concurrent with a spurious loss of letdown. The allowable charging configuration is limited by technical specification to a single .centrifugal=charging pump in operational modes 5 and 6, ;The safety injection pumps are required to be racked out in these'odes.'he heat injection event results from the start of a reactor coolant pump assuming that the primary water in the steam generator is 50'F warmer than the water contained in the rest of the reactor coolant system. The temperature asymmetry can develop during a cooldown maneuver following the shutoff of the reactor coolant pumps and continued cooling with the RHR system. No credit is taken for the RHR relief valves for either of the overpressure scenarios. The setpoint selection is based on the most restrictive of either the mass injection or heat input cases.

A constraint 'required for the analysis is the assumption of the failure of one of the PORV's. The design basis for the overpressure events require that 8921 e:1d/06028S 3;1

0 iz

~

'f I' '

either

~

PORV provide adequate relieving capability in the event of

~

a single valve failure.

~

~

Code Description The evaluation of the cold overpressure mitigation system setpoints is based on'ocal versions of the LOFTRAN code. Two versions of the code were utilized: LOFT12, used for the mass input calculations, collapses the several RCS loops into a single loop model; and'OFT4, which models each loop explicitly, for the heat input calculation.

3.1 OPERATIONAL LIMITS The pressure-temperature limit curves (Appendix G curves) based on revision 2

'""of 'USNRC Regulatory*Guide 1.99, have been generated for reactor vessel exposures of 12 and 32 effective full power years (EFPY). The setpoint evaluation is based on the steady-state cooldown limit, where the technical difference between a steady-state cooldown limit and a steady"state heatup limit is the assumed location of the flaw; i.e., inside or outside of the

~ ~

vessel. The steady-state limit provides the greatest operational margin and has been accepted by the NRC with the justification that "most pressure transients have occurred during isothermal metal conditions." The steady-state limits"are shown in Figure 3.1 and, in digitized form, in Table .

3.1. The curves also include the 800 psig PORV piping load limit.

The analysis assumes nominal values; i.e., pressure instrumentation uncertainties are not included. This is consistent with standard Restinghouse practice. Use of the Appendix G limits without instrumentation uncertainty is justified on the basis of the large amount of conservatism (recognized by the NRC) inherent in the development of the limits.

The 800 psig pipipg limit results from an analysis of water hammer effects on relief valve pipihg for certain classes of rapidly opening relief valves (e.g., Garrett valves) under water solid conditions. The Garrett valves are solenoid operated with a rapidly increasing flow vs. stem position curve.

8921e:1d/060789 3.2

lt l'

I o,t lr.

~ ~ '~ 1f

'pl P

P0RV P 12 E FPY Figure 3.1 I I I IIII Reactor Coolant System Steady-State Pressure Temperature Limits at 12 and 32 EFPY 89216:1d/072089 3.3

0

'W'tÃ1tC&a'we ' 4 4i I l

~ Wj 4;, I1 t g k b 1t I l>% l << ~

y p

<<t

TABLE 3.1 STEADY"STATE COOLDOMN PRESSURE/TEMPERATURE LIMITS RCS Pressure si RCS Pressure si RCS RCS Tem de F 12 EFPY 32 EFPY Tem de F 12 EFPY 32 EFPY 85.0 509.2 490.2 165.0 629.1 568.3 90.0 513.3 492.8 170.0 642.1 576.9 95.0 517.8 495.7 175.0 656.3 586.1 100.0 522.6 498.8 180.0 671.5 595.9 105.0 527.7 502.1 185.0 687.7 606.6 110;0

'" 533 505.7 190.0 705.2 618.0 115.0 2'39.0 509.6 a95.o 724.1 630.4 120.0 545.4 513.8 200.0 744.2 643.5 125.0 552.2 518.2 205.0 766.0 657.7 130,0 559.6 523.0 210.0 789.3 673.0

'35.0 567.5 528.2 215.0 814.5 689.3 140.0 576.1 533.8 220.0 841.4 707..0 145.0 585.1 539.6 225.0 870.4 726. 0 150.0 594'.9 546.0 230.0 9oa.6 .746.3 155.0 605.5 552.9 235.0 935.0 768.2 160.0 616.9 560:4 240.0 970.8 791.7 245.0 1009.6 817.0 8921 e:1 d/060788 3.4

I

,", L) 0

~ IL LI I e

1 W

I lJ 'i 'W+

( Lf

. I 'I VL

When a characteristic curve of this type is combined, with Garrett's typically rapid stroke'time (less than two seconds) .the "valve becomes effectively full open or full closed within a few tenths of a second, thus setting the conditions for a water hammer.

'he flow through an air operated relief valve, when compared to solenoid valves, is much less sensitive to stem position. Hhen combined with the relatively slow opening and closing times characteristic of these types of

'valves, the water hammer effects will be much reduced, if not effectively eliminated. Evaluation of water hammer forces on the piping of air operated valves has not been performed by Hestinghouse. The practice has been to assume the conservative position of taking the worst case results (the Garrett analysis) and applying them to all LTOPS setpoint evaluations, regardless of the relief valve employed.

3.2 - PORV STROKE TIME As instructed by American Electric Power Corporation, the LTOPS analysis also

~

includes a parameter study on valve opening time in order to determine the

~

relationship between opening time .and LTOPS setpoint. The parametric study was performed assuming 6 different valve opening times, ranging from 1.0 sec.

to 10 seconds. This was considered a broad enough range to cover all reasonable"opening time measurements;.

"" " " '" "'""'The" PORV clo'sing time selected for -the analysis was set at 4.0 seconds, independent of the opening time. The closure time selection has no impact on the overpressure transient (an Appendix G consideration), but does impact the underpressure transient (important for the protection of the reactor coolant pump No. 1 seal). The characteristic of the transient is such that as closure time increases, the underpressure becomes more severe.

3.3 PORV OPERATION The design basis for an overpressure event requires that either PORV provide adequate relieving capability in the event of a single valve failure.

8921 e:1 d/060189 3.5

L

'I'

~ o 4~

ro

'herefore, the setpoint analysis is based on the assumption of single valve operation.

The setpoint analysis includes the effect of time delays associated with the transmission of the wide range r'eactor coolant system pressure signal. A conservative value of 0.95 seconds was utilized for the analysis. The breakdown of the time delay is as follows:

Pressure sensing line transport delay . , . . 0.15 sec.

Pressure transmitter delay . . 0.25 sec.

Electronics delay . . . . . . . . . . .' . . 0.10 sec.

Solenoid actuation delay . 0.10 sec.

Valve Chamber (Pneumatic)',delay . . . . , 0.35 sec.

Total . . . . 0.95 sec With the exception of the pressure transmitter delay, the factors comprising the total'elay time are Westinghouse generic estimates for air operated pressurizer relief valves.'. C. Cook Unit 2 features two Foxboro wide range (Model N-EIIGH"HIM2-A) pressure transmitters. Westinghouse instrumentation group had no information on Foxboro transmitters, so the delay time of 0.25 seconds was obtained directly from Foxboro's Product Information Group.

The flow characteristics (valve C(v).'vs. valve'stroke)- for the-pressurizer.

power operated relief valves were obtained from American Electric Power Co.

and are shown by Figure 3.2. The percent at full flow C(v) corresponding to selected percent of valve stroke values is listed in Table 3.2.

3.4 MASS INPUT CONSIDERATIONS The mass injection transient is assumed to occur as a result of the operation of a single centrifugal charging pump in combination with a sudden loss. of letdown. The allowable charging configuration is-. limited by technical specifications in operational modes 5 and 6, where the Cook units LTOPS is enabled. The charging flow, assuming that the normal and alternate flow paths

,. are. both open,, as a function of reactor coolant system pressure is shown in Figure 3.3. The date in tabular form is given in Table 3.3.

BSZl e:1 d/0607BS 3.6

('4

~ i 0

I p

f~

~ 'I 4

+'

4

+1~

C t

fW I'i III l'lp~

II C~

Iloaonoilon Linear hir Oporotod Relief Volvo Full Flov C(v) ~ 46 Figure 3.2 Pressurizer PORY Flo~ Characteristic Curve 8921 a:1d/060789 3.7

t 1

~*

TABLE 3.2 PRESSURIZER PORV CHARACTERISTICS~

PORV enin

'A Valve 'A Valve 'A Valve 5 Valve 0.0 0.0 '0.0 56.0 10.0. 9.0 70.0 66.0 30.0 26.0 80.0 77.0 40.0 35.0 90.0 88.0 50.0 45.0 100.0 100.0 PORV Closin 5 Valve 'A Valve  %%d Valve .5 Valve 0.0 100.0 50.0 45.0 10.0 88.0 60.0 35.0 20.0 77.0 70.0 26.0 30.0 66.0 90.0 9.0 40.0 56.0 100.0 100.0 "Based on the characteristic curve shown in Figure 3.2.

8921e:1d/060789 3.8

,{;%V Q) '~W~

r4

'4 f

k

'1 P

• t

Chargln g Oua to ~ Sln g l ~ Cantrlfu p al Charging Punp Figure 3.3 Mass Injection Flow vs. RCS Pressure 892)a: 1 d/060789 3.9

84 1

~ I %O e E II

TABLE 3.3 MAXIMUM CHARGING FLOW AT RCS PRESSURE~

RCS Press. Charging RCS Press. Charging

+>~si ~ Flow m si Fl ow 'm 400 439.2 1100 358.3

,500 429.7 1200 344.1 600 420.2 1300 329.8 700 410.7 1400 315.2 800 399.9 ~

1500 300.5 900 386.2 1600 285.6

'='1000 372.3 1700 268.9 1800 251.5 "Based on single centrifugal charging pump operation.

SS21 e:1d/0607&9 3.10

~ tt

~,>>

I

~ t~

>>~

I ~

'he b mass input case was conservatively analyzed at low temperature, 85'F,

'here the'pressure transient resulting from a mass input shows the greatest overshoots and undershoots. No other, calculations were performed at different temperatures, for the reason that this is the worst case situation and the D. C. Cook Unit 2 LTOPS setpoint is independent of temperature.

Mass injection rates ranging from 100 to 600 gpm were selected for the mass input parameter study portion of the analysis. This range was based on analyses performed for units similar to D. C. Cook Unit 2, and provides a good definition of the relationship between mass input and the resulting pressure transient. The results of the LOFTRAN runs are presented in Figures 3.4 through 3.9 for the both overpressure and underpressure transients.

3.5 . HEAT INPUT. CONSIDERATIONS The heat input mechanism, from the discussion in Section 1.3.2, is based on a.

single reactor coolant pump startup with a temperature asymmetry existing

~

between the primary water held up in the steam generator tubes and the water,

~

in the remainder of the reactor coolant system. The magnitude of the

\

asymmetry depends on the previous plant operation which allowed the asymmetry to develop. For this study, it was considered conservative to assume a maximum asymmetry of 50'F as the design basis, since much higher differences would be difficult to develop.

The heat input cases were analyzed at RCS temperatures of 85'F and 150'F (steam generator temperatures of 135'F'and 200'F respectively). The characteristic behavior of the overpressure transient resulting from a heat input event is to become more severe with increasing RCS temperature. This required .the additional LOFTRAN runs at increased temperatures in order to provide assurance that the incr'ease in pressure overshoot with temperature did not exceed that of the Appendix G limit, as the heat injection events began to dominate.;.

S921e:1d/060789 3.11

9 1

Wq<c, pe ~ver, ~ 0>> 4-v~ C 1

4 p

'I t V

+I \e\f fl h II gl++lg) E k 4 ~ 4 J we )

OI I O

OA ;C lo INe F 00 '

D 0

a I/1 Figure 3.4 Overshoot and Undershoot dP Va1ues vs. Mass Injection Rate at PORV Stroke open time of 1.0 Sec.

$ 921e:1d/OSOSS9 3.12

0 tg V

s ~

I'4 I

"4 ')tA I ll'I 1 g%*g) ~ III I fp 1

gC

O los tie l. ~ eec.

0 O

0 a

illa'igure 3.5 Overshoot and Undershoot hP Yalues vs. Mass Inject)on Rate at PORY Stroke Open Time of 2.0 Sec.

8921 e:1d/080889 3.13

0

~ ~ ':mViw'4'-

eC e

a

~ - i,4~'gjj e S

ll P 4

g,i I

Q C'P

I s

!)

PO aiuro ~C Q

a C

0 a

Ill UIUIIWIIUUIWWllilWUlilllUIIIIIUllll Figure 3.6 Overshoot.and Undershoot hP Values vs. Mass Injection Rate at PORV Stroke Open Time of 4.0 Sec.

8921e:Ud/OBOS89 3.14

III 7P t@p4X I

t 1'" 9 'h k Il C

Q<t .sp,e,w. "~ '

4+'f

'!i I

4

'1

lo C ~

0 D

0 a

Q el ra C

0 a

Figure 3.7 Overshoot and Undershoot hP Values vs. Hass Injection Rate at PORY Stroke Open Time of 6.0 Sec.

8921e:id/080889

e 4i I h 1

I .

OR V C10 SU LhO ~0 ~

C

~ t 0

a a

D C

0a.

Figure 3.8 Overshoot and Undershoot hP, Values vs. Mass Injection Rate at PORV Stroke Open Time of 8.0 Sec.

8921 e:1d/060889 3.16

Cj i

l1 h

w) s ~ c 1

I pl/ b\ w t

$ -~" p

'r I

- POII re ~ ee.

D 0

C 0

a O

e D

C 0

a Figure 3.9 Overshoot and Undershoot hP Yalues vs. Hass Injection Rate at PORV Stroke Open Time'of 10.0 Sec.

892) e:1d/080889 3. 11

plyI s 0

P

~ ~

l

%4'li g tl .+ r ps inc'v "

H

'.6 SPECIFICATION FOR MASS INPUT TRANSIENTS

~Yt Reactor Coolant System temperature = 85'F Reactor Coolant S stem Volume:

RCS volume = 12,509 cu. ft. for the D. C. Cook units Initial Reactor Coolant S stem Pressure:

The initial RCS pressure was assumed to be 2QD psi less than the setpoint

.... pressure. This is conservative and assures that the transient is well defined by the time the PORV setpoint is reached. Given this definition, the overpressure is essentially insensitive to the initial pressure selection.

Reactor Coolant Relief Ca acit  :

The transient is analyzed assuming the failure of one PORV.

PORY Characteristics:

LTOPS Setpoints = 400, 500, 600, 700 psig PORV flow characteristics are shown in Figure 3.2 Opening Time = 1.0, 2.0, 4.0, 6.0, 8.0, 10.0 sec.

Closing time = 4.0 sec.

C(v) = 46 8921 e:1d/060789

lfl V

~ lU 3 (

4 I

,I Mass In'ection Flow Ca abilit:

Mass input Rates = 100, 250, 400, 600 gpm Maximum charging flow corresponding to LTOPS setpoint pressure.

(Ref. Table 3.3)

Pressure Si nal Transmission Characteristics:

Time delay to PORY stem motion = 0.95 sec.

3.7 SPECIFICATION FOR HEAT INPUT TRANSIENTS

-" '" ','".;,lS stem;Tem eratures:

SG/RCS temperature difference = 50'F Steam generator (heat source) temperature = 135'F, 200'F Reactor Coolant S stem Volume:

The same as that used for the mass injection cases.

'"Initial.Reactor'Coolant S stem Pressure:

The initial RCS pressure = 200 psi less than the LTOPS setpoint pressure Reactor-Coolant S stem Relief Capabilit:

The same as that for the mass injection cases PORV Relief Characteristics:

The same as "that for the mass injection cases.

89210:1d/060789 3.19

"1

~At k

C

1 Steam Generator Desi n Characteristics:

"S/G tube heat transfer .surface. area =,54,500 sq. ft.

S/G type = 51F (D. C. Cook Unit 2 replacement steam generator)

Reactor Coolant Pu RCP type = 93A Pressure Si nal Transmission Characteristics:

The same as that for the mass injection cases.

3.8 SETPOINT EVALUATION The impact of variable mass injection rates and PORV opening times on overshoot and undershoot values is provided by Table 3.4, The information provided by this table is based on the pressure overshoot/undershoot relationship with mass injection rate shown in composite Figures 3.4 through 3.9. The overshoot value was determined from the overshoot curves presented in these figures for the maximum mass injection rates consistent with single centrifugal charging pump injection flow (from Figure 3.3, or .the tabulation

<<>> "".~shown in, Table 3.3),for. the. indicated.,PORV. setpoint., The undershoot value was conservatively estimated from the maximum delta P value below the indicated setpoint. In some cases (PORV opening times of 1 sec. and 2 sec.) the minimum delta P was estimated by extrapolating the undershoot to zero mass injection.

This provides a basis for estimating the setpoint requirement for reactor coolant number 1 seal protection.

The overshoot and undershoot pressures resulting from the heat injection events are .shown in Tables 3.5 and 3.6 as functions of setpoint pressure and PORV opening time. The tables summarize the results of LOFTRAN runs performed for RCS temper'atures of 85 and 150'F, respectively. The evaluation was 8921 e:1d/060789 3.20

0 C,

D D

t4 a

ass

'4 D

tO TABLE 3.4, I Max Overshoot/Undershoot Delta P Values vs PORV 0 enin Time Due to Mass In ection D. C. Cook Unit 2 Overshoot/Undershoot Values si vs. PORV enin Time sec

• Setpt Press. Mass Inj. 1.0 2.0 4.0 6.0 8.0 10.0 st ~Rate e Over . Under Over Under Over Under Over Under Over Under Over Under 400.0 439.2 439.0 231.0 446.0 232.0 460.0 282.0 473.0 296.0 487.0 310.0 499.0 318.0 500.0 429.7 538.0 302.0 542.0 303.0 555.0 372.0 566.0 388,0 578.0 403.0 588.0 407.0 600.0 420.2 635.0 381.0 641.0. 388.0 651.0 460.0 560.0 482.0 671.0 492.0 681.0 493.0 700.0 410.7 733.0 465.0 738.0 479.0 747.0 547.0 755.0 576.0 765.0 577.0 773.0 591.0

• PORV Closure time = 4.0 sec.

Notes: 1. Hass injection rate obtained from RCS pressure vs. single pump charging flow.

2. Overshoot obtained from max delta P overshoot Vs. mass injection flow curve,

3. Undershoot obtained from max delta P undershoot vs. mass injection flow,

'l~

1 0

' 'I

~

V E'," ~u ~

J'I, ~

r>

tO a

TABLE, 3.5 I

Max Overshoot/Undershoot Delta P Values vs PORV enin Time Dde to Heat In ection D. C, Cook Unit 2 Overshoot/Undershoot Values si vs.. PORV enin Time sec

• Setpoint 1.0 2.0 4.0 .6;0 8.0 10.0 v

Over, Under Over , Under Over Under Over Under Over Under Over Under dIOO.O 418.0 265.0 420.0 266.0 423.0 289.0 427.0 311.0 431.0 323.0 435.0 330.0 500.0 517.0 329.0 518.0 335.0 521.0 372.0 525.0 400.0 528.0 416.0 531.0 426.0 600.0 617.0 422.0 . 618.0 426.0 621.0 463.0 624.0 496.0 627.0 513.0 629.0 523.0 700,0 716.0 502.0 717.0 510.0 720.0 557.0 723.0 593.0 726.0 610.0 728.0 621.0

"- Valve Closure time = 4.0 sec.

RCS temperature = 85.0'F S/G temperature = 135.0'F

4 0

I I

A/4 j'

TABLE 3.6 Max Overshoot/Undershoot Oelta P Values vs PORV enin Time Due to }feat In ection

0. C. Cook Unit 2 Overshoot/Undershoot Values si vs. PORV enin time sec

• Setpoint 1,0 2.0 4.0 6.0 8.0 10.0 Over Under Over Under Over Under Over Under Over Under Over Under 400.0 438.0 278.0 445.0 279.0 459.0 295.0 475.0 311..0 490.0 311,0 5D5.0 324.0 50D.O 540.0 338.0 547.0 342.0 562.0 381.0 576.0 395.0 589.0 407.0 602.0 408.0 600.0 639,0 419,0 646.0 433.0 659.0 468.0 672.0 487.0 685.0 493,0 697.0 495.0 700;0 "' 739.'0 512.0 745.0 516.0 757.0 563.0 770.0 579.0 782.0 581.0 793.0 585.0

• Valve Closure time ~ 4.0 sec.

RCS temperature = 150.0'F S/G temperature 200,0'F

lc n

f'I a wn ~>>

i'd%

S hg) *

'eVq ~ -t R ~ i;e

stopped at this point (150'F) since

~

it became

~

evident that the most limiting

~

temperature with respect to Appendix

~

G criteria was at 85'F and that the mass

~ ~

~ ~ ~

...,..injection events are dominant.~

~

The data from these tables (both mass injection and heat injection) is pre-sented in Figures 3.10 through:3.21, showing the maxima and minima system pressure as a function of setpoint pressure for the valve opening times selected for the study. Figures 3.10 through 3.15 show the pressure extrema at an RCS temperature of 85'F, and Figures 3.16 through 3'.21 show the extrema-at 150'F. Also. included on these figures is the Appendix G pressure. limit at 12 and 32 EFPY, and the minimum system pressure for an RCP start. The Appendix G limits (ref. Figure 3.1) correspond to the RCS temperature assumed for the calculation shown by the figure. The RCP seal limit was selected to

'" ...correspond,.to.the minimum. system pressure specified by D. C. Cook for reactor coolant fill and vent operations (325 psig). At other times, "the system pressure is governed by the requirement to maintain 200 psid across the number ,

1 RCP seal." It is assumed that the differential pressure requirement across the seal plus the volume control tank pressure and the static head in the number 1 seal leakoff line will be no more than 325 psig; the value specified for fill and vent.

The intersection of these limits (Appendix G and RCP seal) with the most limiting of the maxima and minima -curves due to mass and/or heat injection

,~ =; ., '..events forms the..basis. for the construction of..Figures 3.22 and 3.23. These figures show, respectively, the dependency of the maximum allowed LTOPS setpoint on valve opening time for reactor vessel exposures of 12 EFPY and 32 EFPY. The minimum setpoint limit (RCP protection) has not been shown, since the "white space" is generally not large enough to meet both the RCP number 1, seal'equirements and. the Appendix G limit.

8921e:1d/060669 3.24

ling "C

g

~ 'V 5 I V r.

V L=

1 A ~

C C

4

-Z 4 't

0

$$ $$$ III3$$ t$ $ II E$ $ $ $ $ $

Heat In)ection Extreaa Pressurixcr PORV Opening Tine 1.1 Sec.

Pressurixor PORV Closing Tine ~ 4.6 Sec.

RCS Temperature ~ BS.I Oeg F Steam 6cnerator Tesperatur e ~ 136. ~ Oeg F Append 6 Lie at 12 EFPY Append 6 Lie at 32 EFPY c

RCP Seal Limit pr r Figure 3.10$ $ RCS Pressure Extrema vs. Setpoint Pressure at PORV Stroke Open Time of 1.0 Sec. and RCS Temperature of 85'F 8921 c:1d/060789 3$ 25

0 k;

F

>>+ g" I' I

It 0>> I te 4

~

I

tlass Intestate Eetrena Hest tntaetten Eeireaa Preseur liat PORV Opening Tice 2.0 Sec.

Pt eaaur izer PORV Claalng Tine ~ 4.8 Sec.

RCS Tenperature > 85.0 Deg F Steam 6enerator Tenperatur ~ ~ 135.I Deg F hppend 6 Lln ~ t 12 EFPY hppend 6 Lln at 32 EFPY RCP Seal Limit Figure 3e 11 RCS Pressure Extrema vs. Setpoint Pressure at PORV Stroke OPen Time of 2a0 Sec. and RCS TemPerature of 85'F

$ 921 e:1d/060789 3.26

l WI ~

0

)g+C,g' Wg'I

'g I

i, "1

li. s <<lk v g' II

'I

ttass Zntsctton Extras>

toast Iotas iion Extrsxa Pressur ixer PORV Opening Tine aa 4.1 Sec.

Pressurizer PORV Closing Tine 4.b Sec.

RCS Tenper atura ~ 85. ~ Deg F Stean 6enarator Tenperatur ~ sa 135. ~ Deg F hppand 6 Lin at 12 EFPY hppand 6 Lie"at 32 EFPY RCP Seal Limit x Figure 3a~ 12 RCS Pressure Extrema .vs. Setpoint Pressure at PORV Stroke Open Time of 4.0 Sec. and RCS Temperature of 85'F BS21e:1d/0607BS 3.27

1 0

gt t ~ 'g tiJ ' II 0

lt 4 4 I 4'e+ t.'

Ilats Intacttcn Entrant Heat In>ection Extrena Pt eaeuriaer PORV Open}ng Tine 6.8 Sec.

Pr eaauri?er PORV Closing Tine ~ 4.6 Sec.

RCS Tenperature ~ 85.S Oep F Stean 6ener atOr TeEEper atur e Es 135.1 Deg F hppend 6'Lin at 12 EFPY hppend 6 LiEE at.32 EFPY I E ~

RCP Seal Limit Figure,.3..13., RCS.Pressure Extrema,vs. Setpoint Pressure at PORV Stroke Open Time of 6.0 Sec. and RCS Temperature of 85'F 8921 e:1 d/060789 3a28

I 4 g t t

CJ

+\

g $ ,1 N t

Vg 'A D .

tlaaa Intaattnn Eatraaa Hast Znjaattnn Eatraaa Pressurizer PORV Opening Tine ~ B.b Seen Pressurizer POAV Closing Tine ~ 4.8 Sec.

RCS Tenperatut ~ ~ 86.b Deg F Stean 6enerator Tenper ~ ture ~ l3S. ~ Deg F i.tataa 'l t>> an" 't a hppend 6 Lie at 12 EFPY hppend 6 Lin at 32 EFPY RCP Seal Limit

""""Figure-3.14" "RCS "Pressure .Extrema vs. Setpoint Pressure at PORV Stroke Open Time of Sa0 Sec. and RCS Temperature of 85'F BB21a:1d/060789 3.29

4t r

1't J,

r k

1, ji

Raae Inleatlan Entrant Heat Infection Extreme Pressurizer PORV Opening Tine ~ IB.B Sac Presaur ixer PORV Cloaing Tine ~ 4.B Sec.

RCS Tenperature ~ 85eB Oeg F Stean Generator TeRRperatur e ~ 135.B Oeg F hppend 6 Lin at 12 EFPY hppend 6 Lin at 32 EFPY, RCP Seal Ltntt Figure 3.15 RCS Pressure Extrema vs. Setpoint Pressure at PORV Stroke Open Time of 10.0 Sec. and RCS Temperature of 85'F SS21e:1d/060789 3.30

I III I

~ ~

>>>>'1 y g>>p ~ ping'(>>>>'">>I>>.

ness inteotton Entrees Hest Intent ton Eatr eaa Pressurizer PORV Opening Tine an 1.6 Sec.

Preaaur izer PORll Closing Tine ~ 4.8 Sec.

RCS Tellperature ~ 150.8 Deg F Stean 6ener ator Temperature 288.0 Deg F Append 6 Lie at 12 EFPY Append 6 Lie at 32 EFPY RCP Seal Limit

'igure 3.16 RCS Pressure Extrema vs. Setpoint Pressure at PORV Stroke Open Time of la0 Sec. and RCS Temperature of 150 F 892le:1d/060789 3.31

aa r

~

ah

~a r

Kf a

I!axe Inteattnn Extrene Heat Iniaatinn Exir ane Pr eaaurixet PORV Opening Tine ~ 2.8 Sec.

Preasur izer PORV Closing Tine ~ 4.8 Seen RCS Temperature ~ 150.8.8 Oeg F Stean 6enerator Tenperatur e 288.8.8 Oeg F Append 6 Lla at 12 EFPY Append 6 Li~ at 32 EFPY nCp teal Ltnti

" Figure 3.17 RCS Pressure Extrema vs. Setpoint Pressure at PORV Stroke Open Time of 2.0 Sect and RCS Temperature of 150'F 8921 e:1d/060789 3.32

I I

0

"\

"\ j 4

l

liana Intaattan Entrana Hast Intanitan Entrant Pressurizer VORV Opening Tine ~ 4.8 Sec.

Pr assur izer PORV Closing Tine 4.8 Sec.

RCS Tenper ature ~ 158.8.8 Deg F Stean 6anarator Tenper ~ tur ~ ~ 28B ~ 8.8 Deg F hppend 6 Lill at 12 EFPY hppend 6 Lin at 32 EFPY RCP Seal Limit n Pr r (

Figure 3.18 RCS Pressure Extrema vs. Setpoint Pressure at PORV Stroke Open Time of 4.0 Sec. and RCS Temperature at 1SO'F 8921 c:1 d/060789 3.33

0 sl

~ (

V I

P

• P4 h ~

• t

naaa Intention Exir axe Hant Intention Extrana Preasur izer PORV Openinp Tine ~ 6.8 Sec.

Preeeur izer PORV Cloainp Tine ~ 4.S Sec.

RCS Temperature ~ 15I.I Oep F h

Stean 6enerator Temperature 2BB.B Oep F

~,

h a

Append 6 Lin at 12 EFPY Append 6 Lie at 32 EFPY RCP Seal Lintt Figure 3. 19 RCS Pressure Extrema vs. Setpoint Pressure at PORV Stroke Open Time of 6;0 Sec. and RCS Temperature of 150'F 8921e:1d/D60789 3.34

e

~ ~

n

'iana Pressurizer Heat PORV Intaatten Entreat Inteetien Eetl Opening Tine ~

aaa 8.8 Sec.

Pressur izer PORV Closing Time ~ 4.8 Sac.

RCS Tenparatur ~ i 150.8 Deg F Stean 6ener ster Temper atua ~ 288. ~ Deg F Append 6 Lie at 12 EFPY Append 6 Lie at 32 EFPY RCP Seal Limit

"- Figure" 3;20 RCS Pressure."Extrema vs. Setpoint Pressure at PORV Stroke Open Time of 8.0 Sec; and RCS Temperature of 150'F 8621 e:1d/060789 3.35

~ QQ.

0 I

4 ee 1

P

hans Iniaoiion Extras'ast Inianiion Exis ann Preaeur iver PORV Openinp Tine ~ 1B.B Sec.

Pressurizer PORV Cloainp Tine aa 4.B Secs RCS Temperature ~ ~ l5I.~ Dep F Stean 6ener ator Tenperatur ~ 2BB.B Oep F Append 6 Lie at 12 EFPY Append 6 Lie at 32 EFPY RCp Oaal Lsnst ~

Figure 3a21 RCS Pressure Extrema vs. Setpoint Pressure at PORV Stroke Open Time of 10.0 Sec. and RCS Temperature of 150'F 892le:ld/060789 3.36

J 0

I II r5' II 1

rw~

'Y

~

v 4Md6%

Notea. 1) No Preaau ~ Inatrunant Error

2) Single PORV Operation

5) PORV Cloaura Ttne ~ i.b Sec.

i) Reactor Veaael Expoaur ~ ~ 12 EFPY

)SB.B Oep F RCS Teno. ]

BS.B Oeg F Figure 3.22. PORY LTOP Setpoint at 12 EFPY vs. Valve Opening Time B921 e:1d/060/89 3.37

g~

C

Roice: 1) No Pressure Instrument Error

2) Single PORV Operation

3) PORV Cloaur ~ Tine ~ i.b Sec.

4) Reactor Vessel Exposure ~ 32 EFPY 1

a 158.$ Oeg F RCS Temp. ~

&S.O Deg F Figure '3.23" PORV LTOP Setpoint at 32 EFPY vs. Valve Opening Time 8921e:1d/060789 3. 38

~%!la 0

' ( I%I 4r>>

1 V

'.0 CORRELATION'O MOG LTOPS SETPOINT METHODOLOGY As part of the LTOPS'etpoint analysis"performed for D. C. Cook Unit 2, American Electric Power Corporation requested a correlation that benchmarks the results of the analysis to that of the algorithm described in the report .

prepared for the Mestinghouse Owners Group (acronymed MOG) on Reactor Coolant System Over-pressurization by Mestinghouse Electric Corporation. The reason for the request was to provide a means to determine the equivalent LOFTRAN derived setpoints from the execution of a relatively simple algorithm;.i.e.,'he MOG report.

The correlation assumes a mas's injection event, only. This is justified on the following bases: 1) the LTOPS>> at D. C. Cook unit 2 features a setpoint

'!'-~"::-~>'independent,"of".,temperature,')'he-most limiting'condition is at low temperature, and 3) the mass injection event dominates at low temperatures.,

Under these conditions, as will be shown, the correlation takes the form of a series of curves of "LOFTRAN" setpoints plotted against pressurizer PORV opening time, with;the "MOG" setpoints as a parametric.

4.1 MOG METHODOLOGY LTOPS SETPOINTS

, The,MOG methodology, for mass .injection events, determines the. resulting. over-,

pressure from a 4-factor formula; the derivation, of which, is provided in the

","'"MOG report.'referenced in"the introduction. to this report:

Delta P = Delta P(Ref)~F(v)"F(s)~F(z)

The factors comprising the formula are- determined from a series of linear rela-tionships with known (or assumed) plant parameters. For convenience, these relationships are reproduced here from MOG: report, and are shown as Figures 4.1 through 4.4. For the parameter study performed here, it is convenient to express these figures analytically:

~ " (Hass 82 Delta P(Ref) = Injection Rate, ibm/sec)

B921e:1d/060ZSB 4.1

~ \

'11 ~

a C '4 k,W . ',S~ ~

tt>>

~

'P "tt"q r 1~

• ." tt

.~

~ ~

• t i4>>

'I

+

~RES S III t FI6URE i.t. 1 I

hPREF

~ Refcftnee Overshoot mme=.A..

k+

j -"~i"8

."... x tkss Input Rate - lb/see "".F Figure 4.1 Oelta P (Ref) vs. Mass Injection Rate 8921 e:1d/060789 4.2

)IT 0

r.

tr,s I

4 1

~y ~ 4>> ~

I

<<gs

I

~SS In )It F16URE Volte Factor M ~ ~~ FV RCS i.2.2,'4:Al MM=--2 KK- ~

~ \

~ ~ ~

~ ~ ~

~ ~ ~ ~ ~

~ ~~

r

'I

~t

~ 'I

~ "I~

OOO

~ I

'S OOO

'il .: ".Bi:.!ii i:- Ii:.: .-'z-.:

V. Total RCS Vol~ - ft Figure 4.2 Volume Factor, F(v), vs. RCS Total Vo1ume 8921 e:1d/06D789 4.3

0 A 11 4

1 ~

~ C t

lht p

d

amass Input I'

FICHE a,2 F> - kcllef Hive

~ ~ ~

Opening TksIe Factor I

'" I r>>

~~ I f ~~ 0 )' 'I 'I ~ ~ ~ ~ ~ I

LI

~ ~ ~ ~ ~~

I

~ ~ ~

~~

~

ll L

- .L

~ ~

%11 >>a s>>I >> <<I<<y 4 I,>>>>1<>

I<<I

IMP ~I -<<0 <<n q<<1 g>> ~ M" .I ...~

L

>>~

1-i=

~ >>

I

= Z, kelfef Valve Opening TfaIe, seconds IiiiiI:'II I0 Figure 4.3 PORY Opening Time Factor, F(z), vs. Valve Opening Time 8921 e:1 d/060889 4.4

0

~w ~

C

~ at+ I "' 8 IraM'" Ye 1'f p~'4

'.e r Et) i r lg

~k~

~llnnn In nt

- Relief

=I:.:II::." .:

. FS Valve Setpo)nt fattnp'll."'I:-'-:.-, ~

-.I::".

~ ~

.:i--'i:-:I.;::i: ~

'. I W

n

.- J.- ~ ~

-w4

~ ~

I---->>'I I + fl + YII. ~ l>>nn Y~ >>Lfn + ng +}Q I f Y

>> .II

i. ~

~

~

I

~ ~

M=~:.

~ A

~ d)II Y

NN W ~

n Y n ~

\ ~

"-:IL', Rel 1af Valve Setpoknt - paly I:.:.'.:.I:.:.:.:. :. I.':I -I.:i~:: ~ I:.:.':: .:IIi Figure 4.4 LTOPS Setpoint Factor, F(s), vs. LTOPS Setpoint 8821 ~:1d/060889 4.5

~ Cf?fJ CWfC V ~

C??

~ C* 'f

• L

~jff tf

'(v) = 1.435 - (0.725E"4)~(RCS Volume, cu.ft.)

= 0.53 for Cook unit 2. (Cook unit 2 cold RCS volume = 12509.2 cu.ft.)

F(s) = 1.810 - (0.00135)~(PORV Setpoint, psig)

F(z) = 0.200 + (0.267)~(PORY Opening Time, sec.)

The overpressures resulting from the application of these factor s are documented in tabular form below and through page 4.9, for the several PORV

'opening-'-times'elected for the analysis. A summary of the overpressures is given on page 4.9.

Mass In ection Ove ressures at PORV Onenin Time = 1.0 sec.

LTOPS Set oint si 400 500 600 700 Inj. Mass (ref Table 3.3)

(gpm) 439.2 429.7 420.2 410.7 (ibm/sec.) 60.96 59.64 58.32 57.01 Delta P(Ref) ... . . .,. 83.31 81.51 79.70 77.91 4'(v) 0 ~ ~ ~ I- ~ ~ 0. ~ ~. i ~ ..0.53 -. 0.53 0.53 0.53 F(s) o ~ ~ ~ ~ ~ ~ ~ ~ ~ 1.270 1.135 1.000 0.865 F(z) o o ~ ~ ~ ~ ~ ~ o ~ 0.467 0.467 0.467 0.467 Delta P (psig)..... 26.20 22.90 19.70 16.70 Overpressure (psig) 426.2 522.9 619.7 716.7 8921 e:1d/060889 4.6

S(E/E + EE)heal EE t I~QE $ II I I 'E ~

r~

li E5 lgI

, Mass In 'ection Overoressures at PORV Ooenin Time = 2.0 sec.

LTOPS Setpoint si 400 500 600 700 Inj. Mass (ref Table 3.3)

(gpm) ~ ~ ~ ~ e ~ 439.2 429.7 420.2 410.7 (ibm/sec.) . 60.96 59.64 58.32 57.01 Oelta P(Ref) . . . . 83.31 81.51 79.70 77.91 F(v) ~ ~ . ~ ~ ~ ~ ~ ~ 0.53 0.53 0.53 0.53 F(s) e ~ ~ o o ~ ~ ~ 1.270 1.135 1.000 0.865 F(x) o ~ ~ ~ ~ ~ ~ ~ 0.734 0.734 0,734 0.734

-,6=,-.'Oe1.ta, P",.(ps i g),...,. 41.20 , '36.00 31.00 26.20 Overpressure (psig) 441.2 536.0 631.0 726.2 Mass In'ection Overoressures at PORV enin Time = 4.0 sec.

LTOPS Set@oint si 400 500 600 700 Inj. Mass (ref Table 3.3)

(gPlll) o ~ ~ ~ ~ ~ 439.2 429.7 420.2 410.7 (ibm/sec.) . 60.96 59.64 58.32 57,01 Delta P(Ref) . '.. 83.31 81.51 79.70 77.91 F(V) ~ ~ ~ ~ ~ ~ ~ ~ 0.53 0.53 0.53 0.53 F(s) o ~ ~ ~ ~ ~ e ~ 1.270 1.135  ; 1.000 0.865 F(Z) ~ ~ ~ ~ ~ ~ ~ ~ 1.268 1.268. 1.268 1.268 Oel ta P (psig) 71.10 62.20 53.60 45.30 Overpressure (psig). 471.1 562.2 -"653.6 745.3 6921e:ld/060789 4.7

0 t~

~r 'v. s v v:

~h EL f

~ r I

I'r.)

.Mass In ection Over ressures at PORV Onenin Time = 6.0 sec.

LTOPS Setooint si 400 500 600 700 Inj. Mass (ref Table 3.3)

( gPlll ) ~ ~ ~ ~ ~, ~ 439.2 429.7 420.2 410.7 (ibm/sec.) 60.96 59.64 58.32 57.01 Delta P(Ref) 83,31 81.51 79.70 77.91 F(v) J ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 0J 53 0.53 0.53 0.53 f(s) J ~ ~ ~ ~ ~ ~ o. 1.270 1.135 1.000 0.865 F(z) i ~ ~ ~ ~ ~ ~ ~ 1.802 1.802 1.802 1.802

. ~ <<'Del,ta,.P. (psig) ......... ......., .. 101.0 ,...,88.40 76.10 64.40 Overpressure (psig). 501.0 588.4 676.1 764.4

~

Mass In ection Overoressures

~

at PORY enin Time = 8.0 sec.

LTOPS Setooint si 400 500 600 700 Inj. Mass (ref Table 3.3)

(gpm) J ~ ~ ~ ~ ~ 439.2 '429.7 420.2 410.7 (ibm/sec.) 60.96 59. 64. 58.32 57.01 Oelta P(Ref) 83.31 '1.51 79.70 77.91 F(v) a ~ ~ ~ ~ ~ ~ ~ 0.53 0.53 0.53 0.53 F(s) J ~ ~ ~ ~ ~ ~ ~ 1.270 1.135 1.000 0.865 f(I) ~ ~ ~ ~ ~ ~ ~ ~ 2.336 2.336 2.336 2.336 ta P (psig) .

'el 131.0 11'4.5 98.70 83.40 Overpressure (psig). 531.0 614.5 698.7 783.4 8921 e:1d/060189 4.8

~ ~

Mass In ection Ove ressures at PORV enin Time = 10.0 sec.

LTOPS Setooint si 400 500 500 700 Inj. Mass (ref Table 3.3)

(gpm) 0 ~ ~ ~ ~ ~ 439.2 429.7 420.2 410.7 (ibm/sec.) 60.96 59.64 58.32 57.01 Delta P(Ref) 83.31 81.51 79.70 77.91 F(Y) 0.53 0.53 0.53 0.53 F(s) i ~ ~ ~ ~ o ~ ~ 1.270 1.135 1.000 0.865 F(x) o ~ ~ ~ ~ ~ ~ ~ 2.870 2.870 2.870 2.870

.',Delta P,,(psig) . 160.9 140.7 121.2 102.5 Overpressure (psig). 560.9 640.7 721.2 802.5 RCS Overoressure Summar LTOPS Setpoint si PORV Opening Time sec 400 500 600 700

..1.0 426.2 ....., 522.9 619. 7 716.7 2.0 441.2 536.0 631.0 726.2 4.0 471.1 S62.2 653.6 745.3 6.0 501.0 588.4 676.1 764.4 8.0 531.0 614.5 698.7 783.4

10. 0 '560'.9 640.7 721.Z 805.5

&921e:1d/0807&9 4.9

C.

4

~ VI

~ 4 H.(

0

~ W1

~

~ riel 4 ~- ~ e P~t 4 H1 V M~4 I" ~ htlt+4 0 'I &t ~ '4 +4 % Es W 1 0

The summary is shown in Figure 4.5, illustrating the high degree of linearity I

of the peak system pressure as a function of setpoint pressure. The peak system pressure, resulting from implementation of the HOG methodology, can

,therefor be expressed as a linear function of setpoint pressure:

1) Peak System Press = A + B~(P MOS )

and the peak system pressure must be less than the Appendix G limit. The coefficients of the equation have been determined by performing a least-squares fit on the peak pressures (from the above table) as a function of setpoint pressure:

Coefficients PORV Opening Time sec 1.0 38.81 0.9683 2.0 61.10 0.9500 4.0 105.35 0.9140 6.0 149.63 0.8779 8.0 194.13 0.8414 10.0 234.21 0.8143 4.2 LOFTRAN/MOG Correlation The system overpressures determined from the LOFTRAN based analysis're also quite linear (reference Figures 3.10 through 3.15), and can be expressed as linear functions of setpoint pressure for each of the PORV opening times:

2) Peak System Press = C: ~ D"(PiLOFT) 89216:1d/060189 4.10

0 Qc

+l C

~ M l 4

18.8 8.8-6.8

~ PORV Opening Tine (sec) q 2.8 1.8 Figure 4.5 Maximum RCS Overpressrue Sased on "MOG" Methodology vs. LTOPS Setpoint Pressure 6921 ~ :1d/060769 4.11

$ f); >+

0 r a 1

L

a with the peak system pressures required to be'less than the Appendix G limit.

~ ~

A least-squares fit of the data in,Table 3.4 results in. the following coefficients:

Coefficients PORV Opening Time sec D 1.0 47.80 0.9790 2.0 55.50 0.9750 4.0 76.90 0.9570 6.0 96.50 0.9400 8.0 115.40 0.9270 10.0 132.00 0.9150 The peak system pressure determined by either the MOG algorithm or the LOFTRAN based analyses must be less than the Appendix G limit. Therefore, at the limit, the overpressure determined from equation 1 must be equal to the overpressure determined from equation 2:

Mos

+ D~(PLoFT~ ~ A + B~~P )

L0FT A

'- C' Mos w IJ 8921 e:) d/060889 4.12

w Ps 0

~ L A

%1 9 U Ok,

C The coefficients r'esulting from the combined equations (1) and (2) are as .

follows:

1 Coefficients PORV Opening A-C '

8 Time sec 1.0 "9.18 0.9891 2.0 5.74 0.9744 4.0 29.73 0.9551 6.0 56.52 0.9339 8.0 84.93 0.9077 10.0 111.70 0.8899 Using the above table of coefficients, the LOFTRAN equivalent LTOPS setpoints corresponding to a series of selected HOG setpoints is tabulated below:

Summar of LOFTRAN E uivalent LTOPS Setooints si LTOPS Set oint si Sased on MOG Al orithm PORY Opening

'ime: sec 300 400 500 600 700 800 1.0 287.6 386.5 485.4 584.3 683.2 782.1 2.0 298.1 395.5 492.9 590.4 687.8 785.3 4.0 316.3 411.8 507,3 602.8 698.3 793.8 6.0 336.7 430.1 523.5 616.9 710.3 803.6 8.0 357.2 . 448.0 -538.8 -629.6 ...720.3 .,811.1 .

10.0 378.7 467.7 556.7 645.6 734.6 823.6 8921e:1d/OB07B9 4.13

V<

f 0

~l A gO

~ t,

"%4*

k =I

4'he tabulation" is shown graphically in Figure 4.6, and plots the LOFTRAN

'derived LTOPS setpoint as a function oft PORV opening time; parametric with the WOG LTOPS setpoint. This figure is used to translate the LTOPS setpoint derived from application of the WOG methodology to the LOFTRAN equivalent value.

Utilization of the figure requires that the LTOPS setpoint first be determined using the WOG methodology, At the PORV opening time corresponding to that selected for the WOG calculation, determine the LOFTRAN analysis setpoint from the ordinate by linearly interpolating between the two curves bounding the WOG setpoint.

4.3 IMPACT OF STEAM GENERATOR TUBE PLUGGING Both the LOFTRAN and the WOG based analyses were performed assuming no tubes

'plugged"in the steam generators. The-impact of tube plugging is a small reduction in RCS volume; the consequence, of which, is slightly higher over-pressures as a result of mass injection events,'and reduced overpressures from heat input events. The heat input events are reduced in importance because of the reducti'on in heat transfer surface area of the steam generators.

The importance of tube plugging with respect to its impact on LTOPS setpoints is accounted for by the F(v) term in the WOG methodology (reference Figure 4.2), and is directly translatable to the LOFTRAN based analysis through the correlation developed in this chapter. As a worst case example,

~

reducing the number of steam generator tubes by 154 over all four steam generators, 'results in a reduction in RCS cold volume of 492..6 cu.ft. (based on an average tube length of 69.77 ft., a tube O.D. of 0.875 inches, and a wall thickness of 0.050 inches). This represents a fractional reduction in initial RCS volume of a bit less than 4A (i.eee 0.0394).. The impact on overpressure can be determined from the F(v) equation in: section 4.1. Without tube plugging (RCS:volume = 12509.2 cu.ft.), F(v) = 0.53. 'With 15K of the tubes plugged (RCS-volume = 12016.6 cu.ft.), F(v) increases to 0.56. This represents an increase in the delta overpressore of 1.068, almost a 7%

increase.

'9216:1d/060789 4.14

O

~ ~

MOS LTOPS Setpnt (yahoo)

Opening Time Figure 4.6 LOFTRAN/HOG Setpoint Correlation vs. PORV

4. 15 89216:1d/060789

l~ 4 0

Wkc 4't \ ~ 4,, tt- 4 W L 4 J4 + 4 I -4 '

Pt P$ l+~[h 1

'i

' ) r I'A \ I

'4 l 0-

~ h Fi OW 4'+ lip AO u S W SECTICN 4 AH.uc:0c~ ~ u op~ <<byBa" k C p~- o+s)

(KRATIN6 INSTRUCTINS IIITA ID TlGIHl&SH93TI%

fal NORMA4 MININNp ND MAXIMLMOPERATING VALUES 4.1.1 Pump

1. IIo. I Sea1 ITEM UNIT NORMAL MIN IMUM MAX IMUM NOTES Plow gpm See Figure 4-0 0.2 5.0 1, 2, 3 Temperature degrees F 100 - 190 60 235 Pressure psig 2250 325 2485 hP psi 2217 200 2470 fvl$4iHI A 4)4'1@Co't% ~ rl% ~

HOT OPERATIONAL RECOMMENOKO ALARM SETTINGS HIGH fLOW- 5.0GPM LOW FLOW- .TGPM O 3 C7 I

4l I

Qg 2 .::SAFE'PERA TING RA NGK.',

~ 7 I

I

~ 2 I 0

0 200 400 600 RAA IA>0 I."CIA l400 IRAA 1800 2000 2200 2400 I

'"'50 2'IOO Sgg. Dlml Nrmllnl I~HI p,>@Ill I!Ql.

1$ AOOORO I FIGURE 4-0 tlo. 1 Seal Performance Parameters 4-1 Rev. 1

0 C

NO-

~ 5P Ol

~.

0

• Nores Du'ring heatup or cooldown, vhen the system vacer pressure is 1000 'psig or below, leak~ff Elow may be insufficient to caol the bearing and seal compo-nents. When che leak-off flow is belov 1 gpm, the seal bypass valve should be-opened. This permits a limired flov to bypass the No. 1 seal through a non-ad)useable orifice block (external to the pump itself).

2~ If the leak-off flow is less than 0.2 gpm, ic is probable char. minor foreign macter is restricting che Elov ac the seal face inlet. The seal faces act like a filter. Foreign particles in the order af 25 - 40 microns can collect between the faces and restricr. the flow. Increasing the seal dif-ferential pressure (hP) may clear che seal. If ic does noc, decrease che seal differential pressure to 100 psi and turn che pump rocor by hand. Do not start the pump motor if the seal flow is belov che specified minimum.

3 ~ If a slow increase in che No. 1 seal leak-off Elav is observed, i.e. over a period of severaL weeks or more, Wescinghouse should be notified co provide guidance. During this period che pump may be operaced and the No. 1 seal Leak~ff valve should be Lefr. open. The seal leak-off Elov should not be permitted to exceed che limits of the safe operating range of Figure 4W.

If these conditions do noc exist, the procedures for emergency operation shown under the No. 2 seal should be followed.

4., This temperature is measured by a chermocouple ac, the outlet of che No. 1 seal. The maximum value shown should not be exceeded either in normal service or during a loss-of-injection condition.

The minimum loop pressure (325 psig) applies only during the filling and venting operation. The minimum loop pressure for subsequent operations is cancrolled by >the minimum 4P across che No. 1 seal (200 psi).

2. No. 2 Seal, MIHlt%M f"AXINN NOTES

" Flow gph .. Negligible Temperature degrees F N/A N/A N/A

%Let Pressur psi 33 15 75 psi 30 73 Flow 12 Temperature degrees F N/A N/A psi 2235 N/A if/A

• Notes Normal operation - No. 1 seal operacive.

2. The Ho. 2 seel temperature will vary vith the No. 1 seal temperacure and is noc considered i,nfozmntivc.

4-La M(v. I

~ ~

t

~'

3. Established by prevailing system conditions.

4; Emergency operation - No. l seal inoperative, with a primary pressure drop occurring across the No. 2 seal. The following action should be taken:

a. Close the No. 1 leak-off valve within five minutes.

b. Prepare for pump shutdown. The pump may be operated for a period not to exceed an additional 30 minutes. During this period, the reactor power should be ramped down to the N-1 allowable power level, where N is the number of operating RC pumps.

c. Secure the pump.

d. Do not restart the pump until the cause of the seal malfunction has been determined.

3. No. 3 Seal UNIT MINIMlN NXINN Plow cc/hr 100 . Negligible 200 Temperature degrees F N/A N/A N/A psi Flow gph N/A N/A Temperature degrees F -

N/A N/A N/A psi 15 N/A N/A

• Notes 1., Normal. operation No. 1 seal operative.

2. The No. 3 seal temperatures are not considered informative.

3. The sea1 differential pressure Ls established by the head tank.

4. Emergency operation - No. 1 seal, inoperative, with a primary pressure drop occurring across the No. 2 seal. Refer to note 4 of paragraph 4.1.1.1.

4. lnjectlon Water ITEM UNIT MINIMLN NXIMLH NOTES Plow 12 Temperature degrees F 130 60 1'50 Pressure N/A N/A 4-lb Rev. 1

"~

0

's 4~ -qp) i < q ~ ~~ k> ~ t.pi~ e t ~ e ~ @ ewa> ~

t Wc I '

~<

• Notes

..1....The, normal flow distribution is 5 gpm'into the system and 3 gpm for the seal supply.

2. If the in]ection ~ater is increased to 150'F, the reactor coolant tempera-

~

ture should be ad)usted to a temperature not to exceed 400'F.

3. Infection water pressure is ad)usted to obtain the required flow.

5. Thermal Barrier Cooi ing Water MINION /AXING NOTES Flow 40 35 60 Temperature degrees F 80 60 105 Pressure psi 150 N/A 200

• Notes

1. Water shall be supplied from the non-radioactive component cooling system.

Pressure shall be adequare to ensure the required flow.

6. Bearing Mater ITEM UN [T MININN Temperature degrees F 160 Amb ient 225

• Note

1. Refer to paragraph 4 1, items 1 and 2 (Loss of In]ection Water and High Temperature of Injection Water).

7. Seal Purge Water The No. 3 seal is fitted with a connection for purge water to minimize the buildup of boric acid crystals ac the top of the seal. If considered necessary, 2 to 4 gph of cool, clean, deminerali-ed, non-borated purge flow is water should be injected. The purge water drains out through the normal No. 3 seal leak-off line. When purge ~ater is used, the flows of the table in para-graph 4-1.1-3 do not apply.

8. Alarm Settings, 1 Alarm settings should be set in accordance with the maximum or minimum values own.on"the preceding charrs.

4-lc Rev. 1

'0 0

LICENSING REPORT FOR STORAGE DENSIFICATION OF D.C. COOK SPENT FUEL POOL INDIANAMICHIGAN POKER COMPTE&

by Holtec International AEPSC Contract No. C-7926

<< Holtec Project 00480 b

L

I

'I I

I I

HOLTEC INTERNATIONAL REVIEW AND CERTIFICATION LOG DOCUMENT NAME: LICENSING REPORT FOR STORAGE DENSIFICATION OF D.C. COOK SPENT FUEL POOL HOLTEC DOCUMENT I AD. NO. HZ-90488 HOLTEC PROJECT NO.

CUSTOMER/CLIENT: AMERICAN ELECTRIC POWER (INDIANA MZCHZCAN POWER CO.)

REVISION BLOCK ISSUE NO.

ORIGINAL REVISION 1 AUTHOR 57 DATE mC'~g((q@

Cg~

QS REVIEWER

& DATE

+res o arrv COfl r~

i I Q.A.

g

,5o(r l~

DATE MANAGER Cl V+

APPROVED BY & DATE M

+c+

~l Cc

~

(

(9'lV4+

REVISION 2 A c~y gl . c 9~8 't Cs 9/ >>1

..~~rc~

r'Fi REVISION 3 YHl 5'sl9 ~

~t$ 9( ggCi0<

r REVISION 4 lg g.ln f CM n.

REVISION 5 ~'/

Ylv @4 Rid ~~GO

~P PE REVISION 6 Irb]~ I C 8 So&

/g NOTE: Signatures and printed names are required in the review block.

Must be Project Manager or his Designee.

This document conforms to the requirement of the design specification and the applicable sections of the governing codes.

This document bears the ink stamp of the professional engineer who is certifying this document.

Pro esszonal Engineer SEAL r ggs}gg PAL SlHGH t",

1 2G906 E l J>g"3 P

I

SUMMARY

OF REVISZONS Revision 1 contains the following number of pages of text includin tables but exc udin fi ures Title Page 1 Review and Certification Log 1 Summary of Revisions Page 1 Table of Contents 4 List of Figures 3 Section 1 8 Section 2 17 Section 3 8 Section 4 (later)

Section 5 23 Section 6 48 Section 7 5 Section 8 13

.Section 9 (later)

Section 10 11 Revision 2 contains the

' following number of pages of text inc ud'n tables and es Title Page 1 Review and Certification Log 1 Summary of Revisions Page 1 Table of Contents 4 List of Figures 3 Section 1 8 Section 2 18 Section 3 14 Section 4 Not included in Revision 2 Section 5 33 Section 6 78 Section 7 5 Section 8 17 Section 9 Not included in Revision 2 Section 10 Not included in Revision 2

i)

~

SUMMARY

OF REVISIONS Revision contains the following number of pages of text 3

includ'n tables and fi ures Title Page 1 Review and Certification Log 1 Summary of Revisions Page 2 Table of Contents 4 List of Figures 3 Section 1 9 Section 2 19 Section 3 14 Section 4 34 Appendix A to Section 4 9 Section 5 33 Section 6 76 Section 7 5 Section Section Section 8

9 10 ll 17 Section ll 5 4

Revision 4 contains the same number of pages as Revision 3 with these exce tions:

List of Tables (added to Rev. 4): 3 Sections revised in Rev. 4 now contain the following number of pages:

Section 1 9 Section 2 18 Section 4 35 Section 34 Section 5

9 ll Individual pages revised and transmitted in Revision 4 are:

Pages 3-1, 3-3< 3-4, 6>>6, 6-15, 6-18, 6-28, 6-30, 7-4, 7-5, 7-6, 8,7 and 10-2.

SUMMARY

OF REVISIONS Holtec Report HZ-90488 Revision 5 The following is revised in Revision 5:

Pages 4-8 and 4-9 Section 9

,Appendix A Page v of Table of Contents Revision 6 The followin a es are revised in Revision 6:

List of Figures Table of Contents (page v) 2-1 4-15, 4-16 5 2I 5 3I 5 5I 5 6I 5 8~ 5 9I 5 10I 5 12I 5 15I 5 16I 5 17~ 5 5-20, 5-21, 5-24 through 5-38 18'-19, 6-3, 6-4, 6-35 7-2, 7-4 8-6, 8-7, 8-13 10-4 11-3

4 f

~

l I

~

~

~

~

l I

~

t

TABLE OF CONTENTS

1.0 INTRODUCTION

2.0 MODULE DATA 2-1 2.1 Synopsis of New Modules 2-1 2.2 Mixed Zone Two Region Storage (MZTR) 2-1 2.3 Material Considerations 2-4 2.3.1 Introduction 2-4 2.3.2 Structural Materials 2-4 2.3.3 Poison Material 2-4 2.3.4 Compatibility with Coolant 2-7 2.4 Existing Rack Modules and Proposed Reracking 2~7 Operation 3.0 CONSTRUCTION OF RACK MODULES 3-1 3.1 Fabrication Objective 3-1 3.2 Mixed Zone Two Region Storage 3-2 3.3 Anatomy of Rack Modules 3-2 3.4 Codes, Standards and Practices 3-5 for the D.C. Cook Spent Fuel Pool Racks 3.5 Materials of Construction 3-9 4.0 CRITICALITY SAFETY ANALYSES 4-1 4.1 Design Basis 4-1 4.2 Summary of Criticality Analyses 4-4 4.2.1 Normal Operating Conditions 4 4 4.2.2 Abnormal and Accident Conditions 4-6 4.3 Reference Fuel Storage Cells 4-8 4.3.1 Reference Fuel Assembly 4-8 4.3.2 High Density Fuel Storage Cells 4-9

~

~

~

f

TABLE OF CONTENTS (continued) 4.4 Analytical Methodology 4-10 4.4.1 Reference Design Calculations 4-10 4.4.2 Fuel Burnup Calculations and 4-12 Uncertainties 4.4.3 Effect of Axial Burnup 4-13 Distribution 4.5 Criticality Analyses and'olerances 4-15 4.5.1 Nominal Design 4-15 4.5.2 Uncertainties due to 4-15 Manufacturing Tolerances 4.5.2.1 Boron Loading Tolerances 4-15 4.5.2.2 Boral Width Tolerance 4-16 4.5.2.3 Tolerance in Cell Lattice Spacing 4-16 4.5.2.4 Stainless Steel Thickness 4-16 Tolerances 4.5.2.5 Fuel Enrichment and Density 4-16 Tolerances 4.5.3 Water-gap Spacing Between Modules 4-17 4.5.4 Eccentric Fuel Positioning 4-17

4. 6 Abnormal and Accident Conditions 4-17 4.6.1 Temperature and Water Density 4-17 Effects 4.6.2 Dropped Fuel Assembly 4-18 4.6.3 Lateral Rack Movement 4-18 4.6.4 Abnormal Location of a Fuel 4-19 Assembly 4.7 Existing Spent Fuel 4-19 4.8 References 4-21 5.0 THERMAL-HYDRAULIC CONSIDERATIONS 5.1 Introduction 5-1 5.2 Spent Fuel Cooling System Description 5-2 5.2.1 System Functions 5-2 5.2.2 System Description 5-3 5.2.3 Performance Requirements 5-4 5.3 Decay Heat Load Calculations 5-4

~

~

~

~

l I

TABLE OF CONTENTS (continued) 5.4 Discharge Scenarios 5-5 5.5 Bulk Pool Temperatures 5-6 5.6 Local Pool Water Temperature 5-11 5.6.1 Basis 5-11 5.6.2 Model Description 5-12 5.7 Cladding Temperature 5-13 5.8 Blocked Cell Analysis 5-16 5.9 References for Section 5 5-16 6.0 RACK STRUCTURAL CONSIDERATIONS 6-1 6.1 Introduction 6-1 6.2 Analysis Outline 6-2 6.3 Artificial Slab Motions 6-3 6.4 Outline of Single Rack 3-D Analysis 6-5 6.5 Dynamic Model for the Single Rack 6<<7 Analysis 6.5.1 Assumptions 6-9 6.5.2 Model Description 6-11 6.5.3 Fluid Coupling 6-12 6.5. 4 Damping 6-13 6.5. 5 Impact 6-13 6.6 Assembly of the Dynamic Model 6-14 6.7 Time Integration of the Equations 6-17 of Motion 6.7.1 Time History Analysis Using 6-17 Multi-Degree of Freedom Rack Model

~

6.7.2 Evaluation of Potential for 6-19 Inter-Rack Impact.

6.8 Structural Acceptance Criteria 6-19 6.9 Material Properties 6-21

c' l

TABLE OF CONTENTS (continued) 6.10 Stress Limits for Various Conditions 6-22 6.10.1 Normal and Upset Conditions 6-22 (Level A or Level B)

Level D Service Limits 6-25 6.11 Results for the Analysis of Spent Fuel 6-25 Racks Using a Single Rack Model and 3-D Seismic Motion 6.12 Impact Analyses 6-28 6.12.1 Impact Loading between Fuel 6-28 Assembly and Cell Wall 6.12.2 Impacts between Adjacent Racks 6-28 6.13 Weld Stresses 6-31 6.13.1 Baseplate to Rack Welds and 6-29 Cell-to-Cell Welds 6.13.2 Heating of an Isolated Cell 6-30 6.14 Whole Pool Multi-Rack Analysis 6-30 6.14.1 Multi-Rack Model 6-32 6.14.2 Results of Multi-Rack Analysis 6-34 6.15 Bearing Pad Analysis 6-36 6.16 References for Section 6 6-37 7.0 ACCIDENT ANALYSIS AND MISCELLANEOUS 7-1 STRUCTURAL EVALUATIONS 7.1 Introduction 7-1 7.2 Refueling Accidents 7-1 7.2.1 Dropped Fuel Assembly 7-1 7.3 Local Buckling of Fuel Cell Walls 7-2 7.4 Analysis of Welding Joints in Rack 7~3 due to Isolated Hot Cell 7.5 Crane Uplift Load of 3000 lbs. 7-4 7.6 References for Section 7 7-4

TABLE OF CONTENTS 8.0 STATIC AND DYNAMIC ANALYSES OF FUEL POOL STRUCTURE 8.1 Introduction 8-1 8.2 General Features of the Model 8-3 8.3 Loading Conditions 8-6 8.4 Results of Analyses 8-10 8.5 Pool Liner 8-11 8.6 Conclusions 8-11 8.7 References for Section 8 8-12 9.0 RADIOLOGICAL EVALUATION 9-1 9.1 Fuel Handling Accident 9-1 9.1.1 Assumptions and Source 9-1 Term Calculations 9.1-2 Results 9.2 Solid Radwaste 9-5 9.3 Gaseous Releases 9-5 9.4 Personnel Exposures 9-5 9.5 Anticipated Exposure during Reracking 9-6 9.6 References for Section 9 9-8 10.0 IN-SERVICE SURVEILLANCE PROGRAM 10-1 10.1 Purpose 10-1 10.2 Coupon Surveillance 10-2 10 '.1 10.2.2 Description of Test Benchmark Data Coupons 10-2 10-3 10.2.3 Coupon Reference Data 10-3 10.2.4 Accelerated Surveillance 10-4 10.2.5 Post-Irradiation Tests 10-4 10.2.6 Acceptance Criteria 10-4 10.3 References for Section 10 10-5

I

~

i1

~

g

~

~

~

~

TABLE OF CONTENTS 11.0 COST/BENEFIT ANALYSIS, 11.1 Introduction 11.2 Project Cost Assessment 11.3 Resource Commitment 11.4 Environment Assessment

5 I

LIST OF TABLES Table 1.1.1 Discharge Schedule Table 1.1.,2 Available Storage in the Donald C. Cook Pool Table 1.1.3 Rack Module Data, Existing and Proposed Racks Table 2.1.1 Module Data Table 2.1.2 Common Module Data Table 2.1.3 Module Data Table 2.3.1 Boral Experience List (Domestic and Foreign)

Table 2.3.2 1100 Alloy Aluminum Physical and Mechanical Properties Table 2.3.3 Chemical Composition (by weight) - Aluminum (1100 Alloy)

Table 2.3.4 Boron Carbide Chemical Composition, Weight  %

Boron Carbide Physical Properties Table 4.1 Summary of Criticality Safety Analyses Normal Storage Configurations Table 4.2 Summary of Criticality Safety Analyses Interim Checkerboard Loading Table 4.3 Reactivity Effects of Abnormal and Accident Conditions Table 4.4 Design Basis Fuel Assembly Specifications Table 4.5 Reactivity Effects of Manufacturing Tolerances Table 4.6 Effect of Temperature and Void on Calculated Reactivity of Storage Rack Table 5.4.1 Fuel Specific Power and Pool Capacity Data Table 5.4.2 Data for Scenarios 1 through 3

l

~

~

~

-l l

LIST OF TABLES (continued)

Table 5.4.3 Data for Scenarios 1 through 3 Table 5.5.1 Pool Bulk Temperature and Heat Generation Rate Data Table 5.5.2 Time-to-Boil for Various Discharge Scenarios Table 5.6.1 Peaking Factor Data Table 5.6.2 Data for Local Temperatures Table 5.7.1 Local and Cladding Temperature Output Data for the Maximum Pool Water Condition (Case 1)

Table 6.3.1 Correlation Coefficient Table 6.5.1 Degrees of Freedom Table 6.6.1 Numbering System for Gap Elements and Friction Elements Table 6.6.2 Typical Input Data for Rack Analyses (lb-inch units)

Table 6.9.1 Rack Material Data (200'F)

Support Material Data (200'F)

Table 6.11.1 Stress Factors and Rack-to-Fuel Impact Load Table 6.11.2 Rack Displacements and Support Loads Table 6.14.1 Rack Numbering and Weight Information Table 6.14.2 Maximum Displacements from WPMR Run MPl Table 6.14.3 Maximum Displacements from WPMR Run MP2 Table 6.14.4 Maximum Displacements from WPMR Run MP3 Table 6.14.5 Maximum Rack Displacements and Foot Load Table 8.4.1 Safety Factors for Bending of Pool Structure Regions ViL3.

I LIST OF TABLES (continued)

Inventories and Constants of Significant Fission Product Radionuclides Data and Assumptions for the Evaluation of the Fuel Handling Accident Typical Concentrations of Radionuclides in the Spent Fuel Pool Water Preliminary Estimate of Person-Rem Exposures During Reracking Donald C. Cook Nuclear Plant Worst Case Spent Fuel Inventory

I LIST OF FIGURES Figure 2. l. 1 Cook Spent Fuel Pool Layout (upper bound cell count 3616 cells)

Figure 3.3.1 Seam Welding Precision Formed Channels Figure 3.3.2 Composite Box Assembly Figure 3.3.3 Array of Cells for Non-Flux Trap Modules Figure 3.3.4 Adjustable Support Leg Figure 3.3.5 Elevation View of Rack Module Figure 4.1 Normal Storage Pattern (Mixed Three Zone)

Figure 4.2 Interim Storage Pattern (Checkerboard)

Figure 4.3 Acceptable Burnup Domain in Regions 2 & 3 Figure 4.4 Fuel Storage Cell Cross Section Figure 4.5 KENO Calculational Model Figure 4.6 Equivalent Enrichment for Spent Fuel at Various Burnups for Initial Enrichment of 4.95%

Figure 4.7 Effect of Water-Gap Spacing Between Modules on System Reactivity Figure 4.8 Acceptable Burnup Domain in Regions 2 6 3 Showing Existing Spent Fuel Assemblies .

Figure 5.5.1 Pool Bulk Temperature Model Figure 5.5.2 Donald C. Cook SFP Normal Discharge, One Cooling Train, Case la Figure 5.5.3 Donald C. Cook SFP Normal Discharge, One Cooling Train, Case 1b Figure 5.5.4 Donald C. Cook SFP Normal Discharge, Two Cooling Trains, Case 2 Figure 5.5.5 Donald C. Cook SFP Full Core Offload Two Cooling Trains, Case 3 Figure 5.5.6 Donald C. Cook SFP Full Core Offload One Cooling Train, Case 4 Figure 5.5.7 Cook SFP Loss of Cooling Scenario, Case la

I l

LIST OF FIGURES (continued)

Figure 5. 5. 8 Cook SFP Loss of Cooling Scenario, Case 1b Figure 5.5.9 Cook SFP Loss of Cooling Scenario, Case 2 Figure 5.5.10 Cook SFP Loss of Cooling Scenario, Case 3 Figure 5.5.11 Cook SFP Loss of Cooling Scenario, Case 4 Figure 5.6.1 Idealization 'of Rack Assembly Figure 5.6.2 Thermal Chimney Flow Model Figure 5.6.3 Convection Currents in the Pool Figure 6.2.1 Pictorial View of Rack Structure Figure 6.3.1 DBE N-S Acceleration Time History Figure 6.3;2 DBE E-W Acceleration Time History Figure 6.3.3 DBE Vertical Acceleration Time History LIST OF FIGURES (continued)

Figure 6.3.4 Horizontal Design Spectrum and N-S Time History Spectrum (5% damping)

Figure 6.3.5 Horizontal Design Spectrum and E-W Time History Spectrum (5% damping)

Figure 6.3.6 Vertical Design and Time History Derived Spectra (5% damping)

Figure 6.3.7 OBE N-S Acceleration Time History Figure 6.3.8 OBE E-W Acceleration Time History Figure 6.3.9 OBE Vertical Acceleration Time History Figure 6.3.10 'Horizontal Design Spectrum and Time History Derived N-S Spectrum (2% damping)

Figure 6.3.11 Horizontal Design Spectrum and E-W Time History Derived Spectrum (2% damping)

Figure 6.3.12 Vertical Design and Time History Derived Spectra (2% damping)

LIST OF FIGURES (continued)

Figure 6.5.1 Schematic Model for DYNARACK Figure 6.5.2 Rack-to-Rack Impact Springs Figure 6.5.3 Impact Spring Arrangement at Node i Figure 6.5.4 Degrees of Freedom Modelling Rack Motion Figure 6.5.5 Rack Degrees of Freedom for X-Z Plane Bending Figure 6.5.6 Rack Degrees of Freedom for Y-2 Plane Bending Figure 6.6.1 2-D View of Rack Module Figure 6.14.1 Rack and Foot Pedestal Numbering for Cook Multi-Rack Model Figure 6.14.2 Cook Pool Multi-Rack Seismic Analysis, Run MP2 Rack 16 to Rack 17 South Corner Dynamic Gap at Rack Top Figure 6.14.3 Cook Pool Multi-Rack Seismic Analysis, Run MP2 Rack 16 to Rack 17 South Corner Dynamic Gap at Rack Top Figure 6.14.4 Cook Pool Multi-Rack Seismic Analysis, Run MP3 Rack 12 to Rack 18 WEst Corner Dynamic Gap at Rack Top Figure 6.14.5 Cook Pool Multi-Rack Seismic Analysis, Run MP3 Rack 12 to Rack 18 East Corner Dynamic Gap at Rack Top Figure 7.3.1 Loading on Rack Wall Figure 7.4.1 Welded Joint in Rack Figure 8.2.1 Isometric View of Cook Spent Fuel Pool Figure 8.2.2 Overall Finite Model of Cook Pool Top View Figure 8.2.3 Overall Finite Model of Cook Pool Bottom View Figure 8.3.1 Pedestal Load vs. Time

I

1.0 INTRODUCTION

Donald C. Cook is a twin unit pressurized water nuclear power reactor installation owned and operated, by Indiana Michigan Power Company. Donald C. Cook received its construction permit from the AEC in March, 1969, and its operating License in October, 1974 for Unit 1 and December 1977 for Unit 2. The two reactors went into commercial operation in August, 1975 (Unit 1) and July, 1978 (Unit 2), respectively. The Donald C. Cook fuel storage system is made up of a fuel pool 58'-3 1/8" long x 39'-1 9/16" wide with an integral cask laydown area. The pool presently contains 1367 spent fuel storage assemblies and 36 miscellaneous hardware items.

Thus, out of the total installed storage capacity of 2050 storage cells, 1403 storage cells are presently occupied. Since the full core has 193 fuel assemblies for both Donald C. Cook reactors, maintaining full core offload capability from one reactor implies that 1857 storage cells (2050 minus 193) are available for normal offload storage. Table 1.1.1 provides the data on previous and projected fuel assembly discharge in the Donald C. Cook spent fuel pool. Table 1.1.2, constructed from Table 1.1.1 data, indicates that Donald C. Cook will lose full core discharge capability (for one reactor) in 1995. This projected loss of full core discharge capability prompted the present undertaking to increase spent fuel storage capability in the Donald C. Cook pool.

I I

I l

The purpose of this licensing submittal is to rerack the Donald C.

Cook pool and equip it with new poisoned high density storage racks containing 3613 storage cells. The reracking also entails relocation of the thimble plug tool, spent fuel handling tool, Rod Cluster Control Assembly (RCCA) change tool, and Burnable Poison Rod Assembly (BPRA) tool brackets to the South wall adjacent to the cask pit.

Twenty three free-standing poisoned rack modules positioned with a prescribed and geometrically controlled gap between them will contain a total of 3613 storage cells (including 3 triangle cells located at the SW, NW and NE corners of the pool). Out of these cells, the peripheral cells located in each rack module are flux-trap cells*, and the interior ones are of the so-called non-flux trap type. The storage cells suitable for storing fresh fuel (up to 5% enrichment) are uniquely identified (see Section 4.0, Figures 4.1 and 4.2), and are surrounded by non-flux trap cells which have a burnup restriction on the fuel which they can store.

Consistent with the concept of two region storage, the placement of fuel with a given burnup in the allowable location is administratively controlled. No credit is taken for soluble boron in normal refueling and full core offload storage conditions.

A flux trap construction implies that there is a water gap between adjacent storage cells such that the neutrons emanating from a fuel assembly are thermalized before reaching an adjacent fuel assembly.

l It is noted that, the proposed reracking effort will increase the number of licensed storage locations to 3613 and,. as indicated in Table 1 '.2, will extend the date of loss of full core discharge, capability through the year 2008. Table 1.1 3 presents key

~

comparison data for existing and proposed rack modules for Donald'.

Cook.

The new spent fuel storage racks are free-standing and self supporting. The principal construction materials for the new racks are SA240-Type 304 stainless steel sheet and plate stock, and SA564-630 (precipitation hardened stainless steel) for the adjust- able support spindles. The only non-stainless material utilized in the rack is the neutron absorber material which is boron carbide and aluminum-composite sandwich available under the patented product name "Boral".

The new racks are designed and analyzed in accordance with Section III, Division 1, Subsection NF of the ASME Boiler and Pressure Vessel (B&PV) Code. The material procurement, analysis, and fabrication of the rack modules conform to 10CFR 50 Appendix B requirements.

This Licensing Report documents. the design and analyses performed to demonstrate that the new spent fuel racks satisfy all governing requirements of the applicable codes and standards, in particular/

"OT Position for Review and Acceptance of Spent Fuel Storage and Handling Applications", USNRC (1978) and 1979 Addendum thereto.

I I

N(

l

The safety assessment of the proposed rack modules involved demonstration of its thermal-hydraulic, criticality and structural adequacy. Hydrothermal adequacy requires that fuel cladding will not fail due to excessive thermal stress, and that the steady state pool bulk temperature will remain within the limits prescribed for the spent fuel pool to satisfy the pool structural strength constraints. Demonstration of structural adequacy primarily involves analysis showing that the free-standing rack modules will not impact. with each other or with the pool walls under the postulated Design Basis Earthquake (DBE) and Operating Basis Earthquake (OBE) events, and that the primary stresses in the rack module structure will remain below the ASME B&PV Code allowables. The structural qualification also includes analytical demonstration that the subcriticality of the stored fuel will be maintained under accident scenarios such as fuel assembly drop, accidental misplacement of fuel outside a rack, etc.

The criticality safety analysis shows that the neutron multiplication factor for the stored fuel array is bounded by the USNRC limit of 0.95 (OT Position Paper) under assumptions of 95%

probability and 95% confidence. Consequences of the inadvertent placement of a fuel assembly are also evaluated as part of the criticality analysis. The criticality analysis also sets the requirements on the length of the B-10 screen and the areal B-10 density.

This Licensing Report contains documentation of the analyses performed to demonstrate the large margins of safety with respect to all USNRC specified criteria. This report also contains the results of the analysis performed to demonstrate the integrity of the fuel pool reinforced concrete structure, and an appraisal of 1-4

I radiological considerations. A cost/benefit ana ysis demonstrating reracking as the most cost effective approach to increase the on-site storage capacity of the Donald C. Cook Nuclear Plant has also been performed and synopsized in this report.

All computer programs utilized in performing the analyses documented in this licensing report are identified in the appropriate sections. All computer codes are benchmarked and verified in accordance with Holtec International's nuclear Quality Programe The analyses presented herein clearly demonstrate that the rack module arrays possess wide margins o f safety from all three thermal-,hydraulic, criticality, and structural vantage points.

The No Significant Hazard Consideration evaluation submitted to the Commission along with this Licensing Report is based on the descriptions and analyses synopsized in the subsequent sections of this report.

1-5

I I

I I

I I

I

Table DISCHARGE SCHEDULE Number of Cumulative Month/ Assemblies Inventory

~Cele Year In the Pool lA+ 12/1976 65 65 2A 4/1978 64 129 3A 4/1979 64 193 1B** 10/1979 80 273 4A 5/1980 65 338 2B 5/1981 92 430 5A 5/1981 64 494 6A 7/1982 64 558 3B 11/1982 72 630 7A 7/1983 80 710 4B 3/1984 92 802 8A 4/1985 80 882 SB 2/1986 88 970 9A 6/1987 80 1050 6B 5/1988 80 1130 10A 3/1989 80 1210 7B 6/1990 77 1282 11A 10/1990 80 1362 8B 11/1991 76 1438 12A 2/1992 80 1518 9B 3/1993 80 1598 13A 6/1993 80 1678 10B 7/1994 80 1758 14A 10/1994 80 1838 11B 11/1995 80 1918 15A 4/1996 80 1998 12B 3/1997 80 2078 16A 8/1997 80 2158 13B 7/1998 80 2238

• • 'A- Reactor Unit 1 Reactor. Unit 2 1>>6

Table 1. 1. 1 (continued)

DISCHARGE SCHEDULE Number of Cumulative Month/ Assemblies Inventory

~Cele ~e D'schar e I n the Poo 17A* 12/1998 80 2318 14B** 1/2001 80 2398 18A 4/2000 80 2478 15B 5/2001. 80 2558 19A 8/2001 80 2638 16B 9/2002 80 2718 20A 12/2002 80 2798 17B 1/2004 80 2878 21A 6/2004 80 2958 18B 5/2005 80 3038 22A 10/2005 80 3118 19B 9/2006 80 3198 23A 2/2007 80 3278 20B 1/2008 80 3358 24A 7/2008 80 3438 21B 7/2009 80 3518

• • A - Reactor Unit 1 B - Reactor Unit 2 1-7

I I

I I

I I

I

Table 1.1.2 AVAILABLE STORAGE IN THE DONALD C COOK POOL NUMBER OF STORAGE LOCATIONS AVAILABLE With Present Month/ Licensed Capacity After Reracking

~Cc1 e Year 2050 Locations 3616 Locations 7B 6/1990 768 2334 11A 10/1990 688 2254 8B 11/1991 612 2178 12A 2/1992 532 2098 9B 3/1993 452. 2018 13A 6/1993 372 1938 10B 7/1994 292* 1858 14A 10/1994 212 1778 11B 11/1995 132** 1698 15A 4/1996 52*** 1618 12B 3/1997 1538 16A 8/1997 1458 13B 7/1998 1378 17A 12/1998 1298 14B 1/2000 1218 18A 4/2000 1138 15B 5/2001 1058 19A 8/2001 978 16B 9/2002 898 20A 12/2002 818 17B 1/2004 738 21A 6/2004 658 18B 5/2005 578 19B 9/2006 418 23A 2/2007 338*

20B 1/2008 258 24A 7/2008 178**

21B 7/2009 98 25A 11/2009 1 8***

• Date of loss of full core offload capability from both

• • reactors.

• • • Date of loss of full core offload capability for one reactor.

Date of loss of normal discharge capability 1>>8

I I

I I

I I

I

Table 1.1.3 RACK MODULE DATA'XISTING AND PROPOSED RACKS ITEM EXISTING RACKS PROPOSED RACKS Number of cells 2050 3616*

Number of modules 20 23 Neutron Absorber Boral Boral (Nom.) cell pitch, inch 10.5" 8 97n (Nom.) cell opening size, inch 8.884 + 0 '24 8.75" + 0.04 Include three triangular corner storage cells.

1-9

I I

I I

I I

I I

2.0 MODULE DATA 2.1 S no sis of New Modules The Donald C. Cook spent fuel pool consists of a 39'-1 9/16" x 58'-3 1/8" rectangular pit with a 10'-4" x 10'-6" space designated for cask handling operations. The pool is connected 'to the fuel transfer canal through a weir gate on the West wall. This gate is normally closed.

At the present time, the Donald C. Cook pool contains medium density racks with a 10. 5" nominal assembly center-to-center pitch. There is a total of 2050 storage cells in the pool. There are two sizes of modules, 10xl0 and 10xll. The 10x10 module weighs 33,800 lb. and the 10xll module weighs 37,200 lb.

Figure 2.1.1 shows the module layout for the Donald C. Cook pool after the proposed reracking campaign. As shown in Figure 2. 1. 1 and tabulated in Table 2.1.1, there are twenty-three racks containing a total of 3613 storage cells with a 8.97" nominal pitch.

The essential cell data for all storage cells is given in Table 2.1.2. The physical size and weight data on the modules may be found in Table 2. l. 3. In summary, the present reracking application will increase the licensed storage capacity of the Donald C. Cook pool from 2050 to 3613 cells.

2 ~2 Mixed Zone Three Re ion Stora e MZTR The high density spent fuel storage racks in the Donald C. Cook pool will provide storage locations for up to 3613 fuel assemblies and will be designed to maintain the stored fuel, having an initial enrichment of up= to 5 wt% U-235, in a safe, eoolable, and subcritical configuration during normal discharge and full core offload storages and postulated accident conditions.

2-1

I I

I I

I

All rack modules for Donald C. Cook spent fuel pool are of the so-called "free-standing" type such that the modules are not attached to the pool floor nor do they require any lateral braces or restraints. These rack modules will be placed in the pool in their designated locations using a specifically designed lifting device, and the support legs remotely leveled (using a telescopic removable handling tool) by an operator on the fuel handling bridge. The leveling operations are done when the support legs are lifted off the floor. Except for the crane, no additional lifting equipment is needed while leveling is being performed.

As described in detail in Section 3, all modules in the Donald C.

Cook pool are of "non-flux trap" construction. However, the module baseplates extend out by 7/8" (nominal), such that the nom'inal gap between the adjacent walls of two neighboring racks is 2" (nom.).

Thus, although there is a single screen of neutron absorber panel between two fuel assemblies stored in the same rack, there are two poison panels with a water flux trap (2" wide) between them for fuel assemblies located in cells in two facing modules. Out of these flux trap locations, and peripheral cell locations (cells adjacent to pool walls) a certain number of storage cells are designated for storing fresh fuel. In addition, as described in Section 4, a certain number of interior cells in each rack are designated for storing fresh fuel of 5% wt. U-235 (max.)

enrichment. In this manner, a sufficient number of locations without any burnup restriction (Region I cells) are identified to enable unrestricted full core offload of the Donald C. Cook reactor in the spent fuel pool. These so-called Region I cells are identified in Section 4 of this report. The remaining storage cells have enrichment/burnup restrictions. Appropriate restrictions on the enrichment/burnup of the stored fuel in Region II and Region IXI cells are presented in Section 4.

2~2

I Ol I

I

Each rack module is supported by at least four legs which are remotely adjustable. Thus, the racks can be made vertical and the top of the racks can easily be made co-planar with each other.

The rack module support legs are engineered to accommodate variations of the pool floor. The support legs also provide an under rack plenum 'for natural circulation of water through the storage cells. The placement of the racks in the spent fuel pool has been designed to preclude any support legs from being located over existing obstructions on the pool floor.

The Donald C. Cook racks are subjected to mandated seismic loadings per the plant UFSAR. The Design Basis Earthquake (DBE) and Operating Basis Earthquake (OBE) seismic response spectra are provided and synthetic time histories are generated. These acceleration time histories are applied as inertia loads (see Section 6.3).

Under these seismic events, the rack modules have four designated locations of potential impact:

(i) Support leg to bearing pad (ii) Storage cell to fuel assembly contact surfaces (iii) Baseplate edges (iv) Rack top corners The support leg to pool slab bearing pad impact would occur whenever the rack support foot lifts off the pool floor during a seismic event. The "rattling" of the fuel assemblies in the storage cell is a natural phenomenon associated with seismic conditions. The baseplate and rack top corners impacts would occur if the rack modules tend to slide or tilt towards each other during the postulated DBE or OBE seismic events. Section 6 of this report presents the analysis methodology and results for all three locations of impact, and establishes the structural integrity of the racks under the load combinations specified for plant conditions required by the NRC.

2~3

r I

l

~

~

i I

I I

A bearing pad, made of austenitic stainless steel, is interposed between the support foot and the liner such that the loads transmitted to the slab by the rack module under steady state as well as seismic conditions are diffused into the pool slab, and allowable local concrete surface pressures are not exceeded.

Section 8 of this report presents the detailed pool structure analysis.

2 ' Material Considerations 2.3.1 Introduction Safe storage of nuclear fuel in the Donald C. Cook spent fuel pool requires that the materials utilized in the fabrication of racks be of proven durability and be compatible with the pool water environment. This section provides the necessary information on this subject.

2.3.2 Structural Materials The following structural materials are utilized in the fabrication of the spent fuel racks:

a. ASME SA240-304 for all sheet metal stock.

b. Internally threaded support legs: ASME SA240-304.

c ~ Externally threaded support spindle: ASME SA564-630 precipitation hardened stainless steel.

d ~ Weld material - per the following ASME specification:

SPA 5.9 ER308.

2.3.3 oison Materia In addition to the structural and non-structural stainless material, the racks employ Boral, a patented product of AAR Brooks S Perkins, as the thermal neutron absorber material. A brief description of Boral, and its fuel pool experience list follows.

Boral is a thermal neutron absorbing material composed of boron carbide and 1100 alloy aluminum. Boron carbide is a compound 2-4

I I

having a high boron content in a physically stable and chemical inert form. The 1100 alloy aluminum is a light-weight metal with high tensile strength which is protected from corrosion by a highly zesistant oxide film. The two materials, boron carbide and aluminum, are chemically compatible and ideally suited for long-,

term use in the radiation, thermal and chemical environment of a spent fuel pool.

Boral's use in the spent fuel pool as the neutron absorbing material can be attributed to the following reasons:

The content and placement of boron carbide provides a very high removal cross section for thermal neutrons.

(ii) Boron carbide, in the form of fine particles, is homogenously dispersed throughout the central layer of the Boral.

(iii) The boron carbide and aluminum materials in Boral do not degrade as a result of long-term exposure to gamma radiation.

(iv) The thermal neutron absorbing central layer of Boral is clad with permanently bonded surfaces of aluminum.

(v) Boral is stable, strong, durable, and corrosion resistant.

The passivation process of Boral in an aqueous environment results in the generation of hydrogen gas. If the generation rate of hydrogen is too rapid, then swelling of Boral may occur.

Laboratory studies by Boral's supplier indicate that the rate of hydrogen generation is a strong function of the so-called impurities in, the chemical composition of the boron carbide powder, namely sodium hydroxide and boron oxide. AAR Brooks Perkins has instituted a strict program of monitoring of the chemistry of boron carbide used in the, manufacturing of Boral to ensure that no swelling of the panels will occur. Furthermore, 2-5

I I

randomly selected coupons of Boral panels from production runs are subjected to swelling test checks to preclude any possibility of swelling of Boral.

Boral is manufactured by AAR Brooks & Perkins under the control and surveillance of a computer-aided Quality Assurance/Quality Control Program that conf orms to the requirements of 10CFR50 Appendix B, "Quality Assurance Criteria for Nuclear Power Plants and Fuel Reprocessing Plants". As indicated in Table 2.3.1, Boral has been licensed by the USNRC for use in numerous BWR and PWR spent fuel storage racks and has been extensively used in overseas nuclear installations.

Boral Material Characteristics Aluminum: Aluminum is a silvery-white, ductile metallic element that is abundant in the earth's crust. The 1100 alloy aluminum is used extensively in heat exchangers, pressure and storage tanks, chemical equipment, reflectors and sheet, metal 'work.

It has high resistance to corrosion in industrial and marine atmospheres. Aluminum has atomic number o f 13, atomic weight o f 26.98, specific gravity of'.69 and valence of 3. The physical/

mechanical properties and chemical composition of the 1100 alloy aluminum are listed in Tables 2.3.2 and 2.3.3.

The excellent corrosion resistance of the 1100 alloy aluminum is provided by the protective oxide film that develops on its surface from exposure to the atmosphere or water. This film prevents the loss of metal from general corrosion or pitting corrosion and the film remains stable between a pH range of 4.5 to 8.5.

2-6

I I

Boron Carbide: The boron carbide contained in Boral is a fine granulated powder that conforms to ASTM C-750-80 nuclear grade Type III'he particles range in size between 60 and 200 mesh and the material conforms to the chemical composition and properties listed in Table 2.3.4.

2.3.4 Com atibilit with Coolant All materials used in the construction of the Donald C. Cook racks have an established history of in-pool usage. Their physical, chemical and radiological compatibility with the pool environment is an established fact at this time. As noted in Table 2.3.1, Boral has been used in both vented and unvented configurations in fuel pools with equal success. Consistent with the recent practice, the Donald C- Cook rack construction allows full venting of the Boral space. Austenitic stainless steel (304) is widely used in nuclear power plants.

2.4 Existin Rack Modules and Pro osed Rerackin 0 eration The Donald C. Cook fuel pool currently has medium density rack modules'ontaining a total of 2050 storage cells in twenty modules. At the time of the proposed reracking operation, approximately 1678 cells (between 6/1993 and 7/1994) out of 2050 locations will be occupied with spent fuel. There is sufficient number of open (unoccupied) cells in the pool to permit relocation of all fuel such that the existing modules can be emptied and removed from the pool, and new modules installed in a programmed manner.

A remotely engagable lift rig, which is designed to meet the criteria of NUREG-0612 "Control of Heavy Loads of Nuclear Power Plants", will be used to lift the empty modules. Auxiliary Building Cranes will be used for this purpose. A module change-out 2-7

I I

I I

I

scheme and procedure will be developed which ensures that all modules being handled are empty when the module is moving at a height which is more than 12" above the pool floor.

The Auxiliary Building has two overhead cranes which ride on rails that traverse the entire fuel handling area of the building. Each crane has a main hook rated at 150 tons. These hooks are single failure proof (SFP) (up to 60 tons). In addition there is an auxiliary hoist on the East Crane rated at 20 tons.

Pursuant to the defense-in-depth approach of NUREG-0612, the following additional measures of safety will be undertaken for the reracking operation.

The crane and hoist will be given a preventive maintenance checkup and inspection within 3 months of the beginning of the reracking operation.

(ii) The crane hook will be used to lift no more than 50% of its single failure proof capacity of 60 tons at any time during the reracking operation. (The maximum weight o f any module and its as s ociated handling tool is 24 tons).

(iii) The old fuel racks will be lifted no more than 6" above the pool floor and held in that elevation for approximately 10 minutes before beginning the vertical lift.

(iv) The rate of vertical minute.

lift will not exceed 6'er (v) The rate of horizontal movement will not exceed minute. 6'er (vi) Preliminary safe load paths have been developed.

"old" or "new" racks will not be carried over The any region of the pool containing fuel.

(vii) The rack upending or laying down will be carried out in an area which is not overlapping to any safety related component.

(viii) All crew members involved in the reracking operation will be given training in the use of the lifting and upending equipment. The training 2-8

I seminar will utilize videotaoes of the actual lifting and upending in the rigs on tyoical modules to crew member will be installed pools Every be required to pass a written examination in the use of lifting and upending apparatus administered by the rack designer.

( ix) Referring to Figure 2.1.1, fuel handling bridge crane it is noted that the cannot access storage cells facing the east wall and several locations in the southwest corner. Therefore, it necessary to load the inaccessible cells with fuel will be when the rack is staged' certain distance (approximately 20 inches) from the pool wall.

Having loaded these cells, the module will be lifted approximately 4 inches above the pool liner, and laterally transported to its final designated locations. A fuel shuffling and rack installation sequence has been developed to ensure that all heavy load handling criteria of NUREG-0612 are satisfied. The rack handling rig is designed with consideration of the rack module weight along with the contained fuel assembly mass.

The fuel racks 'will be brought directly into the Auxiliary Building through the access door which is at ground level This direct access to the building greatly facilitates (609'levation).

the rack removal and installation effort.

The "old" racks will be decontaminated to the extent practical on-site and approved for shipping per the requirements of 10 CFR71 and 49 CFR 171-178, be housed in shipping containers, and transported to a processing facility for volume reduction. Non-decontaminatable portions of the racks will be shipped to a licensed radioactive waste burial site or returned to site for storage if disposal access is unavailable. The volume reduction is expected to reduce the overall volume of the racks to about 1/10th of their original value.

All phases o f the reracking activity will be conducted in accordance with written procedures which will be reviewed and approved by X6M.

2-9

I Table 2 1.1 MODULE DATA Module Array Cell Total Cell Count I.D. ~nantit Size for this Module T e A** 5 13x14 910 B 4 12x14 672 C 4 13x12 624 D 2 12xl2 288 E 4 13xll 572 F 2 12xll 264 G 1 12x10 120 H* 1 13x14 (8x2) 166 Total 23 3616 Non-rectangular module.

• • Three of the A modules have one triangle cell to accommodate pool corner curvature.

2-10

I I

I I

Table 2.1.2 COMMON MODULE DATA Storage cell inside dimension: 8.75" + 0.04" Storage cell height (above the baseplate): 168 + 1/16" Baseplate thickness: 0.75" (nominal)

Support leg height: 5. 25" (nominal)

Support leg type: Remotely adjustable legs Number of support legs: 4 (minimum)

Remote lifting and handling provision: Yes Poison material: Boral Poison length: 144" Poison width: 7.5" Cell 'Pitch: 8. 97" (nominal) 2-11

Table 2. 1. 3 MODULE DATA Dimensions (inch)*

Shipping Weight Module I.D. East-West North-South ~ki s A 117-3/16 126-3/16 25.7 B 108-1/8 126-3/16 23.7 C 117-3/16 108-1/8 22.5 D 108-1/8 108-1/8 20.9 E 117-3/16 99-1/16 20.8 F 108-1/8 99-1/16 19.3 G 108-1/8 90-1/8 17 '

H 117-3/16 126-3/16 23.9

• All dimensions are bounding rectangular envelopes rounded to the nearest one sixteenth of an inch.

2-12

I I

I

Table 2.3.1 BORAL EXPERIENCE LIST (Domestic and Foreign)

Pressurized. Water Reactors Vented Construc- Mfg.

Plant Utility tion Year Bellefont 1, 2 Tennessee Valley Authority No 1981 Donald C. Cook Indiana & Michigan Electric No 1979 1, 2 Indian Point 3 NY Power Authority Yes 1987 Maine Yankee Maine Yankee Atomic Power Yes 1977 Salem 1, 2 Public Service Elec & Gas No 1980 Seabrook New Hampshire Yankee No Sequoyah 1,2 Tennessee Valley Authority No 1979 Yankee Rowe Yankee Atomic Power Yes 1964/1983 Zion 1,2 Commonwealth Edison Co. Yes 1980 Byron 1,2 Commonwealth Edison Co. Yes 1988 Braidwood 1,2 Commonwealth Edison Co. Yes 1988 Yankee Rowe Yankee Atomic Electric Yes 1988 Three Mile Island I GPU Nuclear Yes 1990 Boiling Water Reactors Browns Ferry 1,2,3 Tennessee Valley Authority Yes 1980 Brunswick 1,2 Carolina Power & Light Yes 1981 Clinton Illinois Power Yes 1981 Cooper Nebraska Public Power Yes 1979 Dresden 2,3 Commonwealth Edison Co. Yes 1981 Duane Arnold Iowa Elec. Light & Power No 1979 J.A. Fitzpatrick NY Power Authority No 1978 E.I. Hatch 1,2 Georgia Power Yes 1981 Hope Creek Public Service Elec & Gas Yes 1985 Humboldt Bay Pacific Gas & Electric Yes 1986 LaCrosse Dairyland Power Yes 1976 Limerick 1,2 Philadelphia Electric No 1980 Monticello Northern States Power Yes 1978 Peachbottom 2,3 Philadelphia Electric No 1980 Perry, 1,2 Cleveland Elec. Illuminating No 1979 Pilgrim Boston Edison No 1978 Shoreham Long Island Lighting Yes Susquehanna 1,2 Pennsylvania Power & Light No 1979 Vermont Yankee Vermont Yankee Atomic Power Yes 1978/1986 Hope Creek Public Service Elec & Gas Yes 1989 2-13

I I

I I

I I

I

Table 2.3.1 (continued)

Foreign Installations Using Boral France 12 PHR Plants Electricite de France South Africa Koeberg 1,2 ESCOM Switzerland Beznau 1,2 Nordostschweizerische Kraftwerke AG Gosgen Kernkraftwerk Gosgen-Daniken AG TaiWBIl Chin-Shan 1,2 Taiwan Power Company Kuosheng 1,2 Taiwan Power Company 2-14

I I

I I

I I

I I

I

Table 2.3.2 1100 ALLOY ALUMINUM PHYSICAL AND MECHANICAL PROPERTIES Density 0.098 lb/cu. in.

2.713 gm/cc Melting Range 1190-1215 deg. F 643-657 deg. C Thermal Conductivity 128 BTU/hr/sq ft/deg. F/ft (77 deg. F) 0.53 cal/sec/sq cm/deg. C/cm Coef. of Thermal 13.1 x 10 /deg. F Expansion 23.6 x 10 /deg. C (68-212 deg. F)

Specific heat 0.22 BTU/lb/deg. F (221 deg. F) 0.23 cal/gm/deg'. C Modulus of 10xl06 psi Elasticity Tensile Strength 13,000 psi annealed (75 deg. F) 18,000 psi as rolled Yield Strength 5,000 psi annealed (75 deg. F) 17,000 psi as rolled Elongation 35-45% annealed (75 deg. F) 9-20% as rolled Hardness (Brinell) 23 annealed 32 as rolled Annealing Temperature 650 deg. F 343 deg. C 2-15

I I

I I

I I

I

Table 2.3.3 CHEMICAL COMPOSITION (by weight.) - ALUMINUM (1100 Alloy) 99.00% min. Aluminum 1.00% max. Silicone and Iron 0.05-0.20% max. Copper

.05% max. Manganese

.10% max. Zinc

.15% max. others each 2-16

I I

I I

I I

I

Table 2.3.4 BORON CARBIDE CHEMICAL COMPOSITION Wei ht Total boron 70.0 min.

B isotopic content in 18.0 natural boron Boric oxide 3.0 max.

Iron 2.0 max.

Total boron plus 94.0 min.

total carbon BORON CARBIDE PHYSICAL PROPERTIES Chemical formula B4C Boron content (weight) 78.28%

Carbon content (weight) 21.72%

Crystal Structure rombohedral Density 2.51 gm./cc-0.0907 lb/cu. in.

Melting Point 2450 C (4442 F)

Boiling Point 3500 C (6332 F)

Microscopic thermal- 600 barn neutron cross-section 2-17

I I

I I

I t:

I I

I

l.e I

~1

~ I' I ~

W a

'e

~ ~

~:

~ ~

o s + a 0~ ~~ '

I I

I

3.0 CONSTRUCTION OF RACK MODULES The object of 'this section is to provide a description of rack module construction for the Donald C. Cook spent fuel pool to enable an independent appraisal of the adequacy of the design.

Similar rack structure designs have recently been used in previous licensing efforts for Kuosheng Units 1 & 2 (Taiwan Power Company);

J.A. FitzPatrick (New York Power Authority); Indian Point 2 (Consolidated Edison Company of New York, Inc.); Three Mile Island Unit 1 (GPU Nuclear); and Hope Creek 1 (Public Service Electric Gas Company). A list of applicable codes and standards is also presented.

3.1 Fabrication Ob 'ective The requirements in manufacturing the high density storage racks for the Donald C. Cook fuel pool may be stated in four interrelated points:

(1) The rack module will be fabricated in such a manner that there is no weld splatter on the storage cell surfaces which would come in contact with the fuel assembly.

(2) The storage locations will be constructed so that redundant flow paths for the coolant are available.

(3) The fabrication process involves operational sequences which permit immediate verification by the inspection staff.

(4) The storage cells are connected to each other by austenitic stainless steel corner welds which leads to a honeycomb lattice construction. The extent of welding is selected to "detune" the racks from the seismic input motion of the Operating Basis Earthquake (OBE) and Design Basis Earthquake (DBE).

3-1

I I

3.2 Mixed Zone Two Re ion Stora e All rack modules designed and fabricated for the Donald C. Cook spent fuel pool are of the so-called "non-flux trap" type. Zn the non-flux trap modules, a single screen of the poison panel is interposed between two fuel assemblies. The poison material utilized in this project is Boral, which does not require lateral support to prevent slumping due to the inherent stiffness.

However, accurate dimensional control of the poison location is essential for nuclear criticality and thermal-hydraulic.

considerations. The design and fabrication approach to, realize this objective is presented in the next sub-section.

3.3 Anatom of Rack Modules As stated earlier, the storage cell l'ocations have a single poison panel between adjacent austenitic stainless steel surfaces. The significant components of the rack module are: (1) the storage box subassembly (2) the baseplate, (3) the thermal neutron absorber material, (4) picture frame sheathing, and (5) support legs.

The rack module manufacturing begins with fabrication of the box. The "boxes" are fabricated from two precision formed channels by seam welding in a machine equipped with. copper chill bars and pneumatic clamps to minimize distortion due to welding heat input. Figure 3.3.1 shows the box.

The minimum weld penetration will be 80% of the -box metal gage which is 0.075" (14 gage). The boxes are manufactured to 8.75" X.D. (nominal inside dimension).

The design objective calls for installing Boral with minimal surface loading. This is accomplished by die forming a "picture frame sheathing" as shown in Figure 3-2

I 4

I I

3.3.2. This sheathing is 0.035" thick and is made to precise dimensions such that the offset is .010" to

.005" greater than the poison material thickness.

As shown in Figure 3.3.1, each box has four lateral 1" diameter holes punched near its bottom edge to provide auxiliary flow holes. The edges of the sheathing and the box are welded together to form a smooth edge. The box, with integrally connected sheathing, is referred to as the "composite box".

The "composite boxes" are arranged in a checkerboard array to form an assemblage of storage cell locations (Figure 3.3.3). The 'nter-box welding and pitch adjustment are accomplished by small longitudinal connectors. Further details are given later in this section.

This assemblage of box assemblies is welded edge-to-edge as shown in Figure 3.3.3, resulting in a honeycomb structure with axial, flexural and torsional rigidity depending on the extent of intercell welding provided.

Zt can be seen from Figure 3.3.3 that the edges of each interior box are connected to the contiguous boxes resulting in a well defined path to resist shear.

h b pl horizontal surface for supporting the fuel assemblies.

The baseplate is attached to the cell assemblage by fillet welds. The baseplate in each storage cell has a 5" diameter flow hole. The baseplate is 3/4" thick to withstand accident fuel assembly drop loads postulated and discussed in Section 7 of this report.

(3) The thermal neutron absorber material: As mentioned in the preceding section, Boral is used as the thermal neutron absorber material.

(4) Picture Frame Sheathin : As described earlier, the sheathing serves as the locator and retainer of the poison material. Figure 3.3.2 shows a schematic of the sheathing.

3~3

I l

I I

The poison material is placed in the customized flat depression region of the sheathing, which is next laid on a side of the "box". The precision of the shape of the sheathing obtained by die forming guarantees that.

the poison sheet installed in surface compression.

it will not be subject to The flanges of the sheathing (on all four sides) are attached to the box using skip welds. The sheathing serves to locate and position the poison sheet accurately, and to preclude its movement under seismic conditions.

Su ort Le s: All support legs are the adjustable type (Figure 3.3.4). The top portion is made of austenitic steel material. The bottom part is made of SA564-630 stainless steel to avoid galling problems.

Each support leg is equipped with a readily accessible socket to enable remote leveling of the rack after its placement in the pool. Lateral holes in the support leg provide the requisite coolant flow path.

An elevation cross-section of the rack module shown in Figure 3.3.5 shows two box cells, and a developed cell in elevation. The Boral panels and their location are also indicated in this figure. The boral panels are positioned such that the entire enriched fu'el portion of the fuel assembly is enveloped by the thermal neutron absorber material.

The joint between the composite box arrays and the baseplate is made by single fillet welds which provide a minimum of 7" of connectivity between each cell wall and the baseplate surface.

As shown in Figure 3.3.4, the support leg is gusseted to provide an increased section for load transfer between the support legs and the cellular structure above the baseplate. Use of the gussets also minimizes heat input induced distortions of the support/baseplate contact regions 3-4

I I

I

3.4 Codes Standards and Practices for the Donald C. Cook S ent Fuel Pool Racks The fabrication of the rack modules for the Donald C. Cook spent fuel pool is performed under a strict quality assurance system suitable for manufacturing and complying with the provisions of 10CFR50 Appendix B.

The following codes, standards and practices will be used as applicable for the design, construction, and assembly of the spent fuel storage racks. Additional specific references related to detailed analyses are given in each section.

a~ Codes and Standards for Desi n and Testin (1) AZSC Manual of Steel Construction, 8th Edition, 1980.

(2) ANSI N210-1976, "Design Objectives for Light Water Reactor Spent Fuel Storage Facilities at Nuclear Power Stations".

(3) American Society of Mechanical Engineers (ASME),

Boiler and Pressure Vessel Code, Section Subsection NF, 1989.

III/

(4) ASNT-TC-1A, June, 1980 American Society for Nondestructive Testing (Recommended Practice for Personnel Qualifications).

(5) ASME Section V - Nondestructive Examination (6) ASME Section ZX - Welding and Brazing Qualifications (7) Building Code Requirements for Reinforced Concrete, ACI318-89/ACI318R-89.

3-5

I I

Ot I

I

(8) Code Requirements for Nuclear Safety Related Concrete Structures, ACI 349-85 and Commentary ACI 349R-85 (9) Reinforced Concrete Design for Thermal Effects on Nuclear Power Plant Structures, ACI 349.1R-80 (10) ACI Detailing Manual 1980 (11) ASME NQA-2, Part 2.7 "Quality Assurance Requirements of Computer Software for Nuclear Facility Applications (draft).

(12) ANSI/ASME, Qualification and Duties of Personnel Engaged in ASME Boiler and Pressure Vessel Code Section III, Div. 1, Certifying Activities, N626 1977.

Mate 'a Codes

( 1) American Society for Testing and Materials (ASTM)

Standards A-240.

American Society of Mechanical Engineers (2)

Boiler and Pressure Vessel Code, Section II (ASME),

Parts A and C, 1989.

ASME Boiler and Pressure Vessel Code, Section IX-Welding and Brazing Qualifications (1986) or latex issue accepted by USNRC.

ualit Assurance C ea liness Packa 'n Shi Receivin Stora e and Handlin Re uirements (1) ANSI N45.2.2 - Packaging, Shipping, Receiving, Storage and Handling of Items for Nuclear Power Plants.

(2) ANSI 45.2.1 - Cleaning of Fluid Systems and Associated Components during Construction Phase of Nuclear Power Plants."

3-6

I I

I I

I

(3) ASME Boiler and Pressure Vessel,Section V, Nondestructive Examination, 1983 Edition, including Summer and Winter Addenda, 1983.

(4) ANSI N16.1-75 Nuclear Criticality Safety Operations with Fissionable Materials Outside Reactors.

(5) ANSI N16.9-75 Validation of Calculation Methods for Nuclear Criticality Safety.

(6) ANSI N45.2.11, 1974 Quality Assurance Requirements for the Design of Nuclear Power Plants.

(7) ANSI 14.6-1978, "Special Lifting Devices for Shipping Containers weighing 10',000 lbs. or more for Nuclear Materials".

(8) ANSI N45.2 ',

Testing Personnel.

Qualification of Inspection and (9) ANSI N45.2.8, Installation, Inspection.

(10) ANSI N45.2.9, Records.

(ll) ANSI N45.2..10, Definitions.

(12) ANSI N45 ~ 2 12, QA Audits.

(13) ANSI N45.2.13, Procurement.

(14) ANSI 45.2.23, QA Audit Personnel.

Other References (In the references below, RG is NRC Regulatory Guide)

(1) RG 1.13 - Spent Fuel Storage Facility Design Basis, Rev. 2 (proposed).

(2) RG 1.123 - (endorses ANSI N45.2.13) Quality Assurance Requirements for Control of Procurement of Items and Services for Nuclear Power Plants.

(3) RG 1.124 Service Limits and Loading Combinations for Class 1 Linear Type Component Supports, Rev. 1.

3-7

I I

I

(4) RG 1.25 Assumptions Used for Evaluating the Potential Radiological Consequences of a Fuel Handling Accident in the Fuel Handling and" Storage Facility of Boiling and Pressurized Water Reactors.

(5) RG 1.28 (endorses ANSI N45.2) Quality Assurance Program Requirements, June'972.

(6) RG 1.29 - Seismic Design Classification, Rev. 3.

(7) RG 1.31 - Control of Ferrite Content in Stainless Steel Weld Metal,'ev. 3.

(8) RG 1.38 - (endorses ANSI N45.2.2) Quality Assurance Requirements for Packaging, Shipping, Receiving, Storage and Handling of Items for Water-Cooled Nuclear Power Plants, March, 1973.

(9) RG 1.44 Control of the Use of Sensitized Stainless Steel.

(10) RG.1.58 - (endorses ANSI N45.2.2) Qualification of Nuclear Power Plant Inspection, Examination, and Testing Personnel, Rev. 1, September, 1980.

(ll) Assurance RG 1.64 - (endorses ANSI N45..2.11) Quality Requirements for the Design of Nuclear Power Plants, October, 1973.

(12) RG 1.71 Welder Qualifications for Areas of Limited Accessibility.

(13) RG 1 '4 - (endorses ANSI N45.2.10)

Assurance Terms and Definitions, February, Quality 1974.

III,Materials Acceptability (14) RG 1.85 Code Case ASME Section Division 1.

(15) RG 1.88 (endorses ANSI N45.2.9) Collection, Storage and Maintenance of Nuclear Power Plant Quality Assurance Records, Rev. 2, October, 1976.

(16) RG 1.'92 - Combining Modal Responses and Spatial Components in Seismic Response Analysis.

3-8

I I

I I

I I

(17).RG 3.41 Validation of Calculation Methods for Nuclear Criticality Safety.

(18) General Design Criteria for Nuclear Power Plants, Code of Federal Regulations, Title 10, Part 50, Appendix A (GDC Nos. 1, 2, 61, 62, and 63).

(19) NUREG-0800, 3 ' '1 3.F 1(

Standard 3 ' '/ ' '(

3 Review Plan, 3 8 '

Sections 3.2.1, (20) "OT Position for Review and Acceptance of Spent Fuel Storage and Handling Applications," dated April 14, 1978, and the modifications to this document of January 18, 1979. (Note: OT stands for Office of Technology).

(21) NUREG-0612, "Control of Heavy Loads at Nuclear Power Plants".

(22) Regulatory Guide 8.8, "Znformation Relative to Ensuring that Occupational Radiation Exposure at Nuclear Power Plants will be as Low as Reasonably Achievable (ALtGQ.).

(23) 10CFR50 Appendix B, Quality Assurance Criteria for Nuclear Power Plants and Fuel Reprocessing Plants (24) 10CFR21 Reporting of Defects and Non-Compliance 3 ' Materials of Construct'on .

Storage Cell: ASME SA240-304 Baseplate: ASME SA240-304 Support Leg (female): ASME SA240-304 Support Leg (male): Ferritic stainless steel (anti-galling material) ASME SA564-630 Poison: Boral 3-9

I I

I I

Lateral flow holes

'.75" Weld Seam

.075" Figure 3.3.1 SEAM WELDING PRECISION FORMED CHANNELS 3-10

I l

I

Sheathing Figure 3.3.2 COMPOSITE BOX ASSEMBLY

I e ~

I I

Figure 3.3.3 ARRAY OF CELLS FOR NON-FLUX TRAP MODULES 3-12

I I

I I

Baseplate Gusset Figure 3. 3. 4 ADJUSTABLE SUPPORT I EG 3-13

I I

I

Box Cells Cell Pitch Boral panel Developed

)

cell Sheathing Baseplate One Inch Lateral Flow Hole (Typical)

Figure 3.3.5 ELEVATION VIEN OF RACK NODULE 3-14

I I

I l

I

4.0 CRITICALITY SAFETY ANALYSES 4.1 Desi Basis The high density spent fuel storage racks for Donald C. cook Nuclear Plant are designed to assure that the effective neutron multiplication factor (k~ff) is equal to or less than 0.95 with the racks fully loaded with fuel of the highest anticipated reactivity, and flooded with unborated water at the temperature within the operating range corresponding to the highest reactivity. The maximum calculated reactivity includes a margin for uncertainty in reactivity calculations including mechanical tolerances. All uncertainties are statistically combined, such that the final k,<<

will be equal to or less than 0.95 with a 954 probability at a 954 confidence level.

Applicable codes, standards, and regulations or pertinent sections thereof, include the following:

o General Design Criteria 62, Prevention of Criticality in Fuel Storage and Handling.

o USNRC Standard Review Plan, NUREG-0800, Section 9.1.2, Spent Fuel Storage, Rev. 3 - July 1981 o USNRC letter of April 14, 1978, to all Power Reactor Licensees - OT Position for Review and Acceptance of Spent Fuel Storage and Handling Applications, including modification letter dated January 18, 1979.

USNRC Regulatory Guide 1.13, Spent Fuel Storage Facility Design Basis, Rev. 2 (proposed), December 1981.

ANSI ANS-8.17-1984, Criticality Safety Criteria for the Handling, Storage and Transportation of LWR Fuel Outside Reactors.

4-1

I I

I

To assure the true reactivity will always be less than the calculated reactivity, the following conservative assumptions were made:

Moderator is assumed to be unborated water at a temperature within the operating range that results in the highest reactivity (determined to be 20 4C).

The effective multiplication factor of an infinite radial array of fuel assemblies was used (see section 4.4.1) except for the boundary storage cells where leakage is inherent.

Neutron absorption in minor structural members is neglected, i.e., spacer grids are analytically replaced by water.

The design basis fuel assembly is a 15 x 15 (Standard) Westinghouse containing UO> at a maximum initial enrichment of 4.95 + 0.05 wt%

U-235 by weight. For fuel assemblies with natural UO) blanketsg the enrichment is that of the central enriched zone. Calculations

'I confirmed that this reference design fuel assembly was the most.

reactive of the assembly types expected to be stored in the racks.

i Three separate storage regions are provided in the spent fuel storage pool, with independent criteria defining the highest.

potential reactivity in each of the two regions's follows:

Region 1 is designed to accommodate new fuel with a maximum enrichment of 4.95 + 0.05 wt% U-235, or spent fuel regardless of the discharge fuel burnup.

Region 2 is designed to accommodate fuel of 4.954 initial enrichment burned to at least 50,000 MWD/MtU (assembly average), or fuel of other enrichments with a burnup yielding an equivalent reactivity.

Region 3 is designed to accommodate fuel of 4.954 initial enrichment burned to at least 38,000 MWD/MtU (assembly average), or fuel of other enrichments with a burnup yielding an equivalent reactivity.

4-2

I I

I I

I

The water in the spent fuel storage pool normally contains soluble boron which would result in large subcriticality margins under actual operating conditions. However, the NRC guidelines, based upon the accident condition in which all soluble poison is assumed to have been lost, specify that the limiting kgff of 0.9S for normal storage be evaluated in the absence of soluble boron. The double contingency principle of ANSI N-16.1-1975 and of the April 1978 NRC letter allows credit for soluble boron under other abnormal or accident conditions since only a single independent accident need be considered at one time. Consecgxences of abnormal and accident conditions have also been evaluated, where "abnormal" refers to conditions which may reasonably be expected to occur during the lifetime of the plant and "accident" refers to conditions which are not expected to occur but nevertheless must be prot'ected against.

4-3

l I

I I

4.2 Summar of Criticalit Anal ses 4.2.1 Normal 0 eratin Conditions The design basis layout of storage cells for the three regions is shown in Figure 4.1. Xn this configuration, the fresh fuel cells (Region 1) are located alternately along the rack periphery (where neutron leakage reduces reactivity) or along the boundary between two storage modules (where the water gap provides a flux-trap which reduces reactivity). High burnup fuel in Region 2 affords a low-reactivity barrier between fresh fuel assemblies and Region 3 fuel of intermediate burnup. There are at the present time, an adequate number of spent fuel assemblies to nearly fill and "block off" the Region 2 barrier locations (see Section 4.7). Thus, the administrative controls required are comparable to a conventional two-region storage rack design.

Prior to approaching the reactor end-of-life, not all storage cells are needed for spent fuel. Therefore, an alternative configuration may be used in which the internal cells are loaded in a checkerboard pattern of fresh fuel (or fuel of any burnup) with

'mpty cells, as indicated in Figure 4.2. This configuration is intended primarily to facilitate a full core unload when needed, prior to the time the racks are beginning to fill up.

Figure 4.3 define the acceptable burnup domains and illustrates the limiting burnup for fuel of various initial enrichments for both Region 2 (upper curve) or Region 3 (lower curve), both of which assume that the fresh fuel (Region 1) is enriched to 4.954 U-235.

Criticality analyses show that the most reactive configuration occurs along the boundary between modules with the reactivity of 4-4

4$ 1 I

I the edge configuration being slightly lower . The bounding criticality analyses are summarized in Table 4.1 for the design basis storage condition (which assumes the single accident condition of the loss of all soluble boron) and in Table 4.2 for the interim checkerboard loading arrangement. The calculated maximum reactivity of 0.940 (same for both the normal storage condition and the interim checkerboard arrangement) is within the regulatory limit of a k,<< of 0.95. This maximum reactivity

. includes calculational uncertainties and manufacturing tolerances (954 probability at the 954 confidence level), an allowance for uncertainty in depletion calculations and the evaluated effect of the axial distribution in burnup. Fresh fuel of less than 4.954 enrichment would result in lower reactivities. As cooling time increases in long-term storage, decay 'of Pu-241 results in a continuous decrease in reactivity, which provides an increasing subcriticality margin with time. No credit is taken for this decrease in reactivity other than to indicate conservatism in the calculations.

The burnup criteria identified above (Figure 4-3) for acceptable storage in Region 2 and Region 3 can be implemented in appropriate administrative procedures to assure verified burnup as specified in the proposed Regulatory Guide 1.13, Revision 2. Administrative procedures will also be employed to confirm and assure the presence of soluble poison in the pool water during fuel handling operations, as a further margin of safety and as a precaution in the event of fuel misplacement during fuel handling operations.

The thick base-plate on the rack modules extend beyond the storage cells and provide assurance that the necessary water-gap between modules is maintained.

4-5

~

I

~

l

~

~

l

For convenience, the minimum (limiting) burnup data in Figure 4.3 for unrestricted storage may be described as a function of the initial enrichment, E, in weight percent U-235 by fitted polynomial expressions as follows; For Re ion 2 Stora e Minimum Burnup in MWD/MTU 22'70 + 22I 220 E 2I 260 E + 149 For Re ion 3 Stora e Minimum Burnup in MWD/MTU 26,745 + 18,746 E 1,631 E + 98.4 E 4.2.2 Abnormal and Accident Conditions Although credit for the soluble poison normally present in the spent fuel pool water is permitted under abnormal or accident conditions, most abnormal or accident conditions will not result in exceeding the limiting reactivity (k,<< of 0.95) even in the absence of soluble poison. The effects on reactivity of credible abnormal and accident conditions are discussed in Section 4.7 and summarized in Table 4.3. Of these abnormal or accident conditions, only one has the potential for a more than negligible positive reactivity effect.

4-6

~

t l

~

~

~

~

~

~

(

I ~

The inadvertent misplacement of a fresh fuel assembly has the potential for exceeding the limiting reactivity, should there be a concurrent and independent accident condition resulting in the loss of all soluble poison. Administrative procedures to assure the presence of soluble poison during fuel handling operations will preclude the possibility of the simultaneous occurrence of the two independent accident conditions. The largest reactivity increase

(+ 0.065 Sk) would occur if a new fuel assembly of 4.954 enrichment were to be positioned in a Region 2 location with the remainder of the rack fully loaded with fuel of the highest permissible reactivity. Under this accident condition, credit for the presence of soluble poison is permitted by NRC guidelines, and calculations indicate that 550 ppm soluble boron would be adecpxate to reduce the k,<< to the calculated k,<< (0.940) and approximately 450 ppm would be sufficient to assure that the limiting koff of 0.95 is not exceeded.

Double contingency principle of ANSI N16.1-1975, as specified in the April 14, 1978 NRC letter (Section 1.2) and implied in the proposed revision to Reg. Guide 1.13 (Section 1.4, Appendix A).

4-7

I l

~

~

~

~

I i

~

~

4 ' Reference Fuel Stora e Cells 4.3.1 Reference Fuel Assembl The design basis fuel assembly, described in Figure 4.4, is a 15 x 15 array of fuel rods with 21 rods replaced by 20 control rod guide tubes and 1 instrument thimble. Table 4.4 summarizes the fuel assembly design specifications and the expected range of significant manufacturing tolerances. As shown below, initial cell calculations with CASMO-3 indicated that the W 15 x 15 fuel exhibited a slightly higher reactivity in the storage rack cell than either the W 17 x 17 standard or optimized (OFA) fuel or the ANF fuel assembly designs.

Burnup Cell Fuel t e Enrichment: ~D/DKU ~k~

W 15 x 15 2.5 0 1.0261 W 15 x 15 2.5 10 0.9210 W 17 x 17 OFA 2.5 0 1. 0205 W 17 x 17 OFA 2.5 10 0. 9144 W 17 x 17 Stnd 2.5 0 1. 0217 W 17 x 17 Stnd 2.5 10 0.9188 ANF 15 x 15 2.5 0 1.0148 ANF 17 x 17 2.5 0 1.0126 W 15 x 15 4.95 0 1.1941 W 15 x 15 4.95 40 0.9204 W 17 x 17 OFA 4.95 0 1.1933 W 17 x 17 OFA 4.95 40 0.9149 W 17 x 17 Stnd 4.95 0 1. 1880 ANF 15 x 15 4.95 0 1. 1857 ANF 17 x 17 4.95 0 1.1883 Highest values 4-8

~

~

l

~

Based upon the calculations listed above, the Westinghouse 15 x 15 rod design was used as the basis for the criticality calculations.

4.3.2 Hi h Densit Fuel Stora e Cells The nominal spent fuel storage cell used for the criticality analyses of the Donald C. Cook spent fuel storage cells is shown in Figure 4.4. Each storage cell is composed of Boral absorber panels positioned between a 8.75-inch I.D., 0..075-inch thick inner stainless steel box, and a 0.035-inch outer stainless steel sheath which forms the wall of the adjacent cell. The fuel assemblies are normally located in the center of each storage cell on a nominal lattice spacing of 8.97 + 0.04 inches. The Boral absorber has a thickness of 0.101 + 0.005 inch and a nominal B-10 areal density of 0.0345 g/cm .

4-9

f 'll'l I

1 w

IK I

4 ' Anal ical Methodolo 4.4.1 Reference Desi n Calculations In the fuel rack analyses, the primary criticality analyses of the high density spent fuel storage racks were performed with the KENO-5a computer code package (1) ,'sing the 27-group SCALE cross-section

. ~

library and the NITAWL subroutine for U-238 resonance shielding effects (Nordheim integral treatment). Depletion analyses and determination of equivalent enrichments were made with the two-dimensional transport theory code, CASMO-3 . Benchmark

/

calculations, presented in Appendix A, indicate a bias of 0.0000 with an uncertainty of + 0.0024 for CASMO-3 and 0.0090 + 0.0021 (954/954) for NITAWL-KENO-Sa. In tracking long-term (30-year) reactivity effects of spent fuel stored 'in Region 2 of the fuel storage rack, previous CASMO calculations confirmed a continuous reduction in reactivity with time (after Xe decay) due primarily to Pu-241 decay and Am-241 growth.

KENO-Sa Monte Carlo calculations inherently include a statistical uncertainty due to the random nature of neutron tracking. To minimize the statistical uncertainty of the KENO-calculated reactivity, a minimum of 500,000 neutron histories in 1000 generations of 500 neutrons each, are accumulated in each calculation. For the design calculation for the racks, 1,250,000 histories were used to confirm convergence of the KENO-5a calculation.

Figure 4.5 represents the basic geometric model used in the KENO-5a calculations. This model effectively describes a repeating array of 10 storage cells in the X-direction separated by a 2-inch water "SCALE" is an acronym for Standardized Computer Analysis for licensing /valuation, a standard cross-section set developed by ORNL for the USNRC.

4-10

L I

I t

I I

gap between modules and an infinite array of cells in the Y-direction (periodic boundary conditions). In the axial (Z) direction, the full length 144-inch fuel assembly was described with a 30-cm water reflector. A similiar model was used for calculations of the rack peripheral cells where the calculations were made with both water and concrete reflectors (a concrete reflector gave a slightly higher reactivity by 0.004 Sk).

Larger models, encompassing an entire storage module (half of an 11 x 11 array, run for 1,250,000 neutron histories to assure convergence) confirmed results obtained with the smaller infinite array model. The larger model was also used to confirm the reactivity calculation for the checkerboard arrangement with fresh fuel and empty cells in Region 3 and in the investigation of the consequences of potential accident conditions with a misplaced fresh fuel assembly. In addition, the corner intersection was explicitly modeled and, as. expected, gave a lower reactivity than the reference design calculation.

In the CASMO-3 geometric model (cell), each fuel rod and its cladding were described explicitly and reflecting boundary conditions (zero neutron current) were used in the axial direction and at the centerline of the Boral and steel plates between storage cells. These boundary conditions have the effect of creating an infinite array of storage cells in all directions and provide a conservative estimate of the uncertainties in reactivity attributed to manufacturing tolerances.

Because NITAWL-KENO-5a does not have burnup capability, burned fuel was represented by fuel of equivalent enrichment as determined by CASMO-3 calculations in the storage cell (i.e. an enrichment which yields the same reactivity in the stora e cell as the burned fuel).

4-11

l I

I I

Figure 4.6 shows this equivalent enrichment. for fuel of 4.954 initial enrichment at various discharge burnups, evaluated in the storage cell.

4.4.2 Fuel Burnu Calculations and Uncertainties CASMO-3 was used for burnup calculations in the hot operating condition. CASMO-3 has been extensively benchmarked (Appendix A and Refs. 2 and 7) against critical experiments (including plutonium-bearing fuel). In addition to burnup calculations, CASMO-3 was used for evaluating the small reactivity increments (by differential calculations) associated with manufacturing tolerances.

Since there are no critical experiment data with spent fuel for determining the uncertainty in burnup-dependent reactivity calculations, an allowance for uncertainty in reactivity was assigned based upon the assumption of 54 uncertainty in burnup.

This is approximately equivalent to 54 of the total reactivity decrement. At the design basis burnups of 38 and 50 MWD/KgU, the uncertainties in burnup are + 1.9 and + 2.5=MWD/KgU respectively.

To evaluate the reactivity consequences of the uncertainties in burnup, independent calculations were made with fuel of 36,100 and 47,500 MWD/MtU burnup in Regions 2 and 3, and the incremental change from. the reference burnups assumed to represent the net uncertainties in reactivity. These calculations resulted in an incremental reactivity uncertainty of + 0.0047 Sk in Region 2 (isolation barrier at 50 MWD/KgU burnup) and + 0.0019 for Region 3 (at 38 MWD/KgU burnup). .In the racks, the fresh unburned fuel in Region 1 strongly dominate the reactivity which tends to minimize the reactivity consequences of uncertainties in burnup. The Only that portion of the uncertainty due to burnup. Other uncertainties are accounted for elsewhere.

4-12

I 4e

allowance for uncertainty in burnup calculations is' conservative estimate, particularly in view of the substantial reactivity decrease with time as the spent fuel ages.

4.4.3 Effect of Axial Burnu Distribution Initially, fuel loaded into the reactor will burn with a slightly skewed cosine power distribution. As burnup progresses, the burnup distribution will tend to flatten, becoming more highly burned in the central regions than in the upper and lower ends, as may be seen in the curves compiled in Ref. 4. At high burnup, the more reactive fuel near the ends of the fuel assembly (less than average burnup) occurs in regions of lower reactivity worth due to neutron leakage. Consequently, it would be expected that over most of the burnup history, distributed burnup fuel assemblies would exhibit a slightly lower reactivity than that calculated for the average burnup. As burnup progresses, the distribution, to some extent, tends to be self-regulating as controlled by the axial power distribution, precluding the existence of large regions of significantly reduced burnup. Among others, Turner has provided generic analytic results of the axial burnup effect based upon calculated and measured axial burnup distributions. These analyses confirm the minor and generally negative reactivity effect of the axially distributed burnup. The trends observed, however, suggest the possibility of a small positive reactivity effect at high burnup.

Calculations were made with KENO-5a in three dimensions, based upon the typical axial burnup distribution of spent fuel (that observed at the Surrey plant was taken as representative). Xn these calculations, the axial height of the burned fuel was divided into a number of axial zones (6-inch intervals near the more significant top of the fuel), each with an enrichment equivalent to the burnup 4-13

I Ol I

l

of that zone. These calculations resulted in an incremental reactivity increase of 0.0037 b'k for the reference design case.

Fuel of lower initial enrichments (and lower burnup) would have a smaller (or negative) reactivity effect as a result of the axial variation in burnup. These estimates are conservative since smaller axial increments in the calculations have been shown to result in lower incremental reactivities 4-14

I I

4.5 Criticalit Anal ses 'and Tolerances 4.5.1 Nominal Desi n For the nominal storage cell design, the NlTAWL-KENO-Sa calculation resulted in a bias-corrected k of 0.9250 + 0.0012 (9S%/95%),

which, when combined with all known uncertainties and the axial burnup effect, results in a kof 0.929 + 0.011 or a maximum kof 0.940 with a 954 probability at the 95: confidence level For the interim loading pattern of checkerboarded fuel and empty cells in Region 3, calculations resulted in essentially the same reactivity as the reference design within the normal KENO-5a statistics (maximum kof 0. 940, including all allowances and uncertainties, see Table 4.2).

4 ' ' Uncertainties Due to Manufacturin Tolerances The uncertainties due to manufacturing tolerances are summarized in Table 4-5 and discussed below.

4.5.2.1 Boron Loadin Tolerances The Boral absorber panels used in the storage cells are nominally 0.101 inch thick, 7.50-inch wide and 144-inch long, with a nominal B-10 areal density of 0.034S g/cm . The vendors manufacturing tolerance limit is + 0.004S g/cm in B-10 content which assures that at any point, the minimum B-10 areal density will not be less than 0.030 g/cm . Differential KENO-5a calculations for the reference design with the minimum tolerance B-10 loading results in an incremental reactivity of + 0.00614 Sk uncertainty.

4-15

I I

1 l

l

4.5.2.2 Boral Width Tolerance The reference storage cell design uses a Boral panel with an initial width of 7.50 + 0.06 inches. For the maximum tolerance of 0.06 inch, the differential CASMO-3 calculated reactivity uncertainty is + 0.0009 Sk.

4.5.2.3 Tolerances in Cell Lattice S acin The manufacturing tolerance on the inner box dimension, which directly affects the storage cell lattice spacing between fuel assemblies, is + 0.06 inches. This corresponds to an uncertainty in reactivity of + 0.0015 Sk determined by differential CASNO-3 calculations.

4.5.2.4 Stainless Steel Thickness Tolerances The nominal stainless steel thickness is 0.075 + 0.005 inch for the inner stainless steel box and 0.035 + 0.003 inch for the Boral cover plate. The maximum positive reactivity ef fect of the expected stainless steel thickness tolerances was calculated (CASMO-3) to be + 0.0009 Sk.

4.5.2.5 Fuel Enrichment and Densit Tolerances The design maximum enrichment is 4.95 + 0.05 wt% U-235. Separate CASMO-3 burnup calculations were made for fuel of the maximum enrichment (5.004) and for the maximum UO> density (10.50 g/cc).

Reactivities in the storage cell were then calculated using the restart capability in CASMO-3 and equivalent enrichments determined for the reference fuel burnups of 38 and 50 MWD/KgU. The incremental reactivities between these calculations and the reference CASMO-3 cases, were conservatively taken as the 4-16

l I

sensitivity'to small enrichment and density variations. For the tolerance on U-235 enrichment, the uncertainty in k is + 0.0034 6k and for fuel density is + 0.0035.

4.5.3 Water- a S acin Between Modules The water-gap between modules constitute a neutron flux-trap for the outer (peripheral) row of storage cells. Calculations with KENO-5a were made for various water-gap spacings (Figure 4.7).

From these data, it was determined that the incremental reactivity consequence (uncertainty) for the minimum water-gap tolerance of +

1/4 inch is + 0.0045 6k. The racks are sonstructed with the base plate extending beyond the edge of the cells. This assures that a minimum spacing of 1.75 inch between storage modules is maintained under all credible conditions.

4.5.4 ccentric Fuel Positionin The fuel assembly is assumed to be normally located in the center of the storage rack cell. Infinite array calculations were made using KENO-5a for a single cell with the fuel assemblies centered and with the assemblies assumed to be in the corner of the storage rack cell (four-assembly cluster at closest approach). These calculations indicated that the reactivity uncertainty could be as much as + 0.0019 6k.

4.6 Abnormal and Accident Conditions 4~6~1 Tem erature and Water Densit Effects The moderator temperature coefficient of reactivity is negative; a moderator temperature of 20'C (68'F) was assumed for the reference designs, which assures that the true reactivity will always be lower over the expected range of water temperatures. Temperature 4-17

I I

I V

I I

I

effects on reactivity have been calculated and the results are shown in Table 4.6. With soluble poison present, the temperature coefficients of reactivity would differ from those inferred from the data in Table 4.6. However, the reactivities would also be substantially lower at all temperatures with soluble boron present, and the data in Table 4.6 is pertinent to the higher-reactivity unborated case.

4.6 ' D o ed Fuel Assembl For a drop on top of the rack, the fuel assembly will come to rest horizontally on top of the rack with a minimum separation distance from the fuel in the rack of more than 12 inches, including 'the potential deformation under seismic or accident conditions. At this separation distance, the effect on reactivity is insignificant

(<0.0001 6'k). Furthermore, soluble boron in the pool water would substantially reduce the reactivity and assure that the true reactivity is always less than the limiting value for any conceivable dropped fuel accident.

4.6.3 ateral Rack Movement Lateral motion of the rack modules under seismic conditions could potentially alter the spacing between rack modules. However, the maximum rack movement has been determined to be less than the tolerance on the water-gap spacing. Furthermore, soluble poison would assure that a reactivity less than the design limitation is maintained under all accident or abnormal conditions.

4-18

I I

I I

I

4.6.4 Abnormal Location of a Fuel Assembl The abnormal location of a fresh unirradiated fuel assembly of 4.95 wt4 enrichment could, in the absence of soluble poison, result in exceeding the design reactivity limitation (k of 0.95). This could occur if a fresh fuel'ssembly of the highest permissible enrichment were to be either positioned outside and adjacent to a storage rack module or inadvertently loaded into either a Region 2 or Region 3 storage cell. Calculations (KENO-5a) showed that the highest reactivity, including uncertainties, for the worst case postulated accident, condition (fresh fuel assembly in Region 2) would exceed the limit on reactivity in the absence of soluble boron. Soluble boron in the spent fuel pool water, for which credit is permitted under these accident conditions, would assure that the reactivity is maintained substantially less than the design limitation. It is estimated that a soluble poison concentration of 550 ppm boron would be sufficient to maintain k at the reference design value of 0.940 under the maximum postulated accident condition. Approximately 450 ppm boron would be required to limit the maximum reactivity to a k~ff of 0.95.

4.7 Existin S ent Fuel As of May 1990, there were 1596 spent fuel assemblies in storage at the Donald C. Cook plant, including those now in the reactor and their projected burnups at discharge. Figure 4.8 superimposes the enrichment-burnup combination of these fuel assemblies on the curves defining the acceptable burnup domains. As may be seen in this figure, most of the spent fuel now in storage falls well into the acceptable domain for the barrier fuel (Region 2). The number of fuel assemblies meeting the enrichment-burnup criteria for storage in Region 2 is 1390 which will nearly fill the 1447 Region 2 storage locations. Twelve fuel assemblies (discharged 4-19

I I

I I

prematurely for various reasons) will need to be kept in a Region 1 storage location, and the remaining 194 assemblies may be s'tored in Region 3 locations. Future discharge batches may reasonably be expected to have a preponderance of highly burned fuel capable of being stored in Region 2 (or in Region 3 once Region 2 is filled).

An appreciable number of spent fuel assemblies have enrichment-burnup combinations well in excess of the design basis and, this provides further conservatism in the criticality safety of the spent fuel storage rack design.

4-20

I l

I I

4.8 References Green, Lucious, Petrie, Ford, White, and Wright, "PSR /NXTAWL-1 (code package) NITAWL Modular Code System For Generating Coupled Multigroup Neutron-Gamma Libraries from ENDF/B", ORNL-TM-3706, Oak Ridge National Laboratory, November 1975.

2 ~ R.M. Westfall et. al., "SCALE: A Modular System for Performing Standardized Computer Analysis for Licensing Evaluation", NUREG/CR-0200, 1979. Volume 2, Section F11, "KENO-5a An Improved Monte Carlo Criticality Program with Supergrouping".

3. A. Ahlin, M. Edenius, and H. Haggblom, "CASMO A Fuel Assembly Burnup Program", AE-RF-76-4158, Studsvik report.

A. Ahlin and M. Edenius, "CASMO A Fast Transport Theory Depletion Code for LWR Analysis", ANS Transactions, Vol.

26 ~ p 604 I 1977 ~

"CASMO-3 A Fuel Assembly Burnup Program, Users Manual",

Studsvik/NFA-87/7, Studsvik Energitechnik AB, November 1986 4 ~ H. Richings, Some Notes on PWR (W) Power Distribution Probabilities for LOCA Probabilistic Analyses, NRC Memorandum to P.S. Check, dated July 5, 1977.

5. S. E. Turner, "Uncertainty Analysis Burnup Distribu-tions", presented at the DOE/SANDIA Technical Meeting on Fuel Burnup Credit, Special Session, ANS/ENS Conference, Washington, D.C.,November 2, 1988

6. M.G. Natrella, Experimental Statistics National Bureau of Standards, Handbook 91, August 1963.

4-21

I I

I I

I I

Table 4.1

SUMMARY

OF CRITICALITY SAFETY ANALYSES NORMAL STORAGE CONFIGURATION Design Basis burnups at 4.954 0 in Region 1

+ 0.054 initial enrichment 50 in Region 2 38 in Region 3 Temperature for analysis 20 C (68 F)

Reference k (KENO-5a) 0.9160 Calculational bias, Sk 0.0090 Uncertainties Bias statistics (954/954) + 0.0021 KENO-5a statistics (954/95%) + 0.0012 Manufacturing Tolerances + 0.0064 Water-gap + 0.0045 Fuel enrichment + 0.0034 Fuel density + 0.0035 Burnup (38 MWD/KgU) + 0.0019 Burnup (50 MWD/KgU) + 0.0047 Eccentricity in position + 0.0019 Statistical combingion + 0.0110 of uncertainties Axial Burnup Effect 0.0037 Total 0.9287 + 0.0110 Maximum Reactivity (k ) 0.940 (2)

See Appendix A Square root of sum of squares.

4-22

I I

I

Table 4.2

SUMMARY

OF CRITICALITY SAFETY ANALYSES INTERIM CHECKERBOARD LOADING Design Basis burnups at 4.954 0 in Region 1

+ 0.054 initial enrichment 50 in Region 2 Region 3 -CHECKERBOARD (FRESH FUEL AND EMPTY)

Temperature for analysis 20~C (684F)

Reference k (KENO-5a) 0. 9168 Calculational bias, 6k 0.0090 Uncertainties (Assumed same as the reference case)

Bias statistics (954/954) +, 0.0021 KENO-5a statistics (954/954) + 0.0012 Manufacturing Tolerances + 0.0064' Water-gap 0.0045 Fuel enrichment + 0.0034 Fuel density + 0.0035 Burnup (38 MWD/KgU) NA Burnup (50 MWD/KgU) +, 0.0047 Eccentricity + 0.0019 Statistical combination

+ 0.0108 of uncertainties Axial Burnup Effect 0.0037 Total 0.9295 + 0.0108 Maximum Reactivity (k) 0.940 (2)

See Appendix A Square root of sum of squares.

4-23

I I

I I

I I

I I

Table 4.3 REACTIVITY EFFECTS OF ABNORMAL AND ACCIDENT CONDITIONS Accident/Abnormal Conditions Reactivity Effect Temperature increase (above 684F) Negative (Table 4.6)

Void (boiling) Negative (Table 4.6)

Assembly dropped on top of rack Negligible (<0. 0001 Sk)

Lateral rack module movement (Included in Tolerances)

Misplacement of a fuel assembly Positive (0.065 Max 6k)

(controlled by soluble poison) 4-24

I I

I I

Table 4.4 DESIGN BASIS FUEL ASSEMBLY SPECIFICATIONS FUEL ROD DATA Outside diameter, in. 0.422 Cladding thickness, in. 0.0243 Cladding inside diameter, in. 0.3734 Cladding material Zr-4 Pellet density, 4 T.D. 95.0 Stack density, g UO>/cc 10.29 + 0.20 Pellet diameter, in. 0.3659 Maximum enrichment, wt 4 U-235 4 95 + 0.05 FUEL ASSEMBLY DATA Fuel rod array 15 x 15 Number of fuel rods 204 Fuel rod pitch, in. 0 '63 Number of control rod guide and instrument thimbles Thimble O.D., in. (nominal) 0.533 Thimble I.D., in. (nominal) 0.499 4-25

I I

I I

I I

Table 4.5 Reactivity Effects of Manufacturing Tolerances Tolerance Incremental Reactivit Sk Boron-10 loading (+ 0.0045 g/cm ) 0.0061 Boral Width (+ 1/16 inch) 0.0009 Lattice spacing (+ 0.04 inch) 0.0015 Stainless Thickness (+ 0.005 inch) 0.0009 Total (statistical sum) + 0.0064 4-26

I Table 4.6 EFFECT OF TEMPERATURE AND VOID ON CALCULATED REACTIVITY OF STORAGE RACK Case Incremental Reactivity Change, Sk Region 1 Region 2 20oC (68oF) Reference Reference 40oC (104oF) -0.003 -0.002 66 C (150 F) -0.009 -0.005 90 C (194 F) -0 013 -0.010 122 C (252 F) -0.024 -0.015 122 C (252 F) + 204 void -0.071 -0.061 4-27

2g

rkrga'h1 I

~ gggs5gt,g ASSIS Ie aNas$

1%$ 888SII88 N0$$ )II 4Nra 8$ ?i))ma I aCSC gN)SN gII

~

Otftif?fNN$$ NS$ ftfN

~ SNPRESNSSNRft$

Of8lftmfskmf888%

~ attfftfNeftce@$$ NSS r r OSNR$ NNS

%85$ NRsfNR ols LSSI8s mk a cl8NSg 8$

OII88$ 8isf?I$ 886 lr Ij)5w NEIN%

msf fsmk NN$?IN Nrfftstfkr rkssmfftss SSRlSS ala BS S'$tstf NfflttfNRft$% k a %$$ $ N5f$$ NSS kk $ $ R?N$ $ 8tf$$$ ? Sr SINefsNu$ $ N NESS %%$ $ Nssfslkstf~QRSS akn5$IR$ Neisfefft ~ %Nssl$ $ NsSSNRQ f8% $$ NSS ssftfm@kk Qka

% f8NmssSM388%

~ kmSNSNQSNftkr OSRSNRHf8Smmr S PS

%$ $ fsmsf ISmssfsk omssss8mfssm 5S kmsfnmae4

~

N888fQ 8Sf RRO I8mf8 SNia$ $ $9%1 rkf88ffiff8t$$ 8$ PRN$ $

NQ NRSN O 4

~ algllkak)IIISI a%a NIIIImaa OS%SOS O gnggugetm5g

~ O O'O Iggg)5)ggg)L ~ O O tt$

~ Iaaa4karISI

~ OOOO%

O O O O O O O I5akmIarlakrEI ONR 'ciPNRSS

~ 4SIgi'5%Igiaam alaar %pirl)CI 'ING ka ala%I Earmark

'NNRQNIIaaakI

~ SQftwmsssmsssRSI I SWISS MpQ8%1 rsmsslsmss

%$ N'$$ NN NQmk$ I$ 5$$ fftmk NNg Ikfshtt"NamN88%I IS NEQ QS8tafsaI

- omsff88$

$ hmmsk if? 8$ 8Qtifk omml883immsslsl 88 Of8msf!88 86%

~ SNettfmc$ $ NK$$ kr r'5~NI ka aki$ t5fNlRWNN$$ kr rkNBSNR NQftf r $$ % I?$ SNIaoaa SSIPNifI8IN

~ SNSNQNNNSRSS kktftmfNt$IPSM4 Sk %%$ tfmssftfSNSNSS ~ SN$ $ NQft? ItWNSSS NQQNSS ~ SN@bmN

%8QR)%4lSSQI8% %88I888f8i38 ~ SfSSSS$ SS8KSSS iSSI

~ %Nst'tf'NR$$ NRQSa a%$ $ N $ $ N $$ tft ON$ $ fftNNf?INNtCIS a%%assr%ask%a ~ IS5gg5gggI ~ Sr%as%Is%Sr ksks1IS rgaRr

'SNwgssNQRSr I $$ $ ftm f8Rmssk tff$ $ 8tff SlQftfNO kr OSlmS$ SS8

'a SSNM ft$t ONNINN S O O S S ~ O O S O O amm  %%%ask asar% a O O O O LI SSSC 'I gN$ $ M<W$;5g CSC I 5)gg55g"5' ~ g)5g5ggga ~ agggft5ggg5I 'Ig~

'5 ala'5SCr Ils5cv85~5 aCasalmgCy'5l a s?$ 8sssmssh 8% r r'1158"LMM'r k NQE$ 8tssfsms$ I8%r a ssfsmss8tef8sa ~ iNNIS@luma ~

% $8PRmmlf"mmmk  %%f88$ 18$ lsO tf ~ emsseSI8emr Qg msfssmsfi88% Omf8SINBSSMS rsN$ $ eua'eweesa

%8SfsSISIS$ NNN$ %

rSRtf$ NANNA$I$ %r r% 0 Ntttf$$ NQ$ $ fft%

SsmssSmf8bmo OhlkRI88QI8%

r SEemLIPamNIIr

%$ hSS@8i $ SSQI8%

NRstfNRltfNSr SSNRlft$ NAftfNQft$%r O)$ Nf888Msmssk OES58 f888tl88S$ %

~ SssRmf8sttmf8I88%a assff8msf fsesf is%a rsBsssgmf8ismkr S RNNN NA$$ %

rI f8l8 8$ 88 8$ 888IS a assfss 8 8$ f88s$ %ms a akms3f88$ si88$ 8isgr N EW4N4 $ NS

~ 'Earl%ma%a%ask O ~ O ~ O ~

asar%%I%sask

~ %%%%%

rrasaklsaagr PS%SO %

as'""ISIFSNIr OC ~ OO ella OOOO tf a%a%a%a%a ~ Cakakafimkase OOOO

~  %% O O o a a

~ IrlaiaCrkrsr ~ krarka %%sr aCr5%%5 aa SI O

- kmsff88$5"38$ '?ls IkkaaaCS'LI am%Ca'5sksaCa OI88$ MII$58 mk asNst.st$ NiftfN&es ~ ssNRftfRRN$ $

oss8%888s)98% %8$ I88ISII8$8)ik

$$I II  %$ $ 4mmi85Ãc 8%

SQS$ MtIISNNs$ %

r oms$ 88$ 8I8 mka r SSQ$ 88MIS8% ak%)sgsf I88%lska sNstsNlR$ $ NRlsla SINNNNIeNIIIS a%8%m?8$ c'M%iftam %%8matQmf8%$ Sk a%f?f588 $ mI8$ tmNl88$ $ hgl NNgNN$tf ~ ra$ $ NstfftfNtttf$$ Naa am ANN(

rSNRS$

sssfsmssI88mt$ smlr 08Ksflms$ %8KSS

%Sf QRSf SNRISg rk5mf8$ mf8884 asmMm Ig a a

N gstf NlftflftQsg ssB @st 88RS 8$ $ Imm8emsslskl SSNlRstfNR$ $ SNfttkr

%%8@Nh@ssRms SSAI@t88$ $ %?8$ $ %%a

%$$ NStf$ $ NNftfiQO

%41INSf

%tarn ItfmmI ISS m5 O tfttftfN 4 Igiw ggg541

~ kraNSIIIakr oar%sara%as SI am%ha o ~ o o %%

~ ~

~ ~

g, o

Ji I

0 Sfy'5PSyCqCy'ES hagi

'~5 PaIIe SStCCS SSml88 iala)$ :i Ehaiaa g ~ giaiaa

~ S8NISWR ASA

'ifl 8MB ASS ~

)a iila a 55 NmlPaPSSS ll ~

EN'Ihi gI3 Q8 88I88S NIN8SSS NSCCNgS ISPSM FsPamre Sgl Q Pi Q Q

~ SNSSSSSSSSSS

~ 5 ~ ~ 5 0

~ ENESOa ~ L'$S@A45AS 0 ~ SS S ~

~ iaaagaga5 ~ I $ ASSPASAS4 SSSCCW)

'0 5 5 Alga SSSSSS

~

M<8~ ~

CS SS ~

IimII mimi, S ~

SNI

'>>i:hl ~

'ghee mS ~ 8@8 858 8@8 SI385 8Q I

AS Q Cd Q Q SA SS Sr SCM8551S48 A SS SS mRSA NNSSS SSQ Q Q Q QSS SS gS S.Q ~ S

~ WQAQQ.Q.Q.

Q Q QSS ~ X5$ IQN CIINgiANS SS SS Q Q Q g SS54I¹S¹Ã¹g¹gA SA S

S ~ i.i,,s,. ~ iagamaaa/aaa ASSS 0 S 0 S ~ S ~~ gSSS 0 5 lg gS 8 ASRSASQAASSSA Q M Q 0 l)C5~ENDSS CA

'a.ei N 5SS Capri )III S 5NNN SCP CSS S ~

SIICS 'II%!IIi:Ma AS.Q.Q QCd ~ ~ aag+paaa S Og8 ~ Sag88g388RSS I

e'88585 ggNNCC NNPRENSM4 SNINNSN88INS

~ a55 $ 8% SSaII388fI38SI38SS SAQQf3Q 5tS NICCN NICINCCWNN55NS SSAASASSSASSS S S S S S 0 BASSA iSMSS SSSSASA S SS S S SSS SSS S 5 S'SSS S 4SSSSSSSASS 5 0 0 0 0 5 SSSSSSSSSSSSS 0 0 5 0

~ S S S S ~

ASASASASASASS SSP A5)A%ASS SASS S S SS SllS8 5~IN S ' N~S S NSSSSNSS

~

hi S SPRS NamNSNA SSNNNI SSSS ASNBN SS Nia F$ IRSS SSNLCN$ 50SS SS5%85%85ISSS SRSNR8$ 88S SAND ~ CCRC

~ aii p <<SICIgNRNNIIII ~ ICIS NLNICINSI 8 m8im

)a ~ 58N ia SCa8 S

~ SSAXSASSSS S ~ ~ S 5 SS 0 0 SgNAS4

~ PSkmN 3AEASSS SSS4 S 0

4 n I

~

~

I

~

5QQOO 45000 hoaaa AC EPTA LE BUR UP D MAIN

~ 35000 a

> 30000 CL K 25aaa

~

CO

~ 2aaaa 4.

15000 NACC PTABL 10000 RNUP DOMAI (Requ res Ster e in Re 1an 1)

.0 1.5 2.0 28 3.0 3.5 h.o h.5 5.0 INITIAL ENRICHMENT, mU-235 F1g. 4 3 ACCEPTABLE BURNUP DOMAIN IN REGIONS 2 8c 3 4-30

7.50 + 0.06" BORAL IN 0.109" GAP 0.101 +. 0.005" THK 0.0345 9mB-10/crrP V4+g~p~

cQ FUEL OD 0.3659 IN.

DENSITY 10.29 g/cc oog NUMBER CLAD ID 204 (15 X 15) 0.3734 IN.

0.035" SS OOOO. SHEATH 00004 ROD I'ITCH OD 0.422 IN.

0.563 IN.

++ +++, THIMBLES OD 0,546 IN OOOOOO< ID 0.'512 IN.'0000004 ooooooooe 00ioooooo. 0.21 9" 0000000000E, 00000000000K 0000000000004

- 0.04" OOO 8.75 BOX I.D.

oo~ -'075" iIS~@@e I

SS BOX NOT TO SCALEI PITCH 8.969" NOMINAL FIG. 4-4 FUEL STORAGE CELL CROSS SECTION 4-31

I 2" WATER GAP WITH REFLECTING BOUNDARY CONDITIONS OR REFLECTOR (WATER OR CONCRETE) WITH ZERO FLUX BOUNDARY CONDITION PERIODIC BOUMDARY COHDITIOH C ~ C r

REGION 3 REGIOH 3 REGION 3 REGION 2 FRESH FUEL

.0:

x o'.x>>'o REGION 3 REGION 3 REGION 3 REGION 2 REGION 2 PERIODIC BOUMDARY CONDITION FIG. 4-5 KENO CALCULATIONALMODEL

I 5.00 4.75 4.50 4.25 4.00 3.75 I

3.50 8

3.25 3.'00 2.75 2.50 2.25 2.00 1.75 1.50 1.25 1.00 0.75 0 5 10 15 20 25 30 35 40 45 50 55 60 BURNUP, MWD/KgU Fig. 4-6 EQUIVALENT ENRICHMENT FOR SPENT FUEL AT VARIOUS BURNUPS FOR INITIAL ENRICHMENT OF 4.95%

4-33

l 0.940 0.935 0,930 0-g 0.925 hl Z 0.920

~~ 0,915 4

0.910 0.905 0.900 0 .5 10 15 20 25 3.0 WATER GAP BETWEEN MODULES, inches Fig. 4-7 EFFECT OF VIATER-GAP SPACING BETWEEN MODULES ON SYSTEM REACTIVITY'-34

l I

l

.I l

I

55000 saaoo 4SOOO 40000 AC EPTA LE BURN P DO AIN

+ 35000 CI

+ 30000 CL

~ 2500Q CQ

~ 2ooaa 4

tsaao NACC PTABL taaaa RNUP OOMAI (Requ res Stor e In Re Ion t) 5000 0 t.5 2.0 2D XO 3.5 4.0 4.5 5.0 INITIAL ENRICHMENT, mU-235 Fig. 4-8 ACCEPTABLE BURNUP DOMAIN IN REGIONS 2 8c 3 SHOWING EXISTING SPENT FUEL ASSEMBLIES 4-35

I 2CPPENDIX A BENCHMARK CALCULATIONS by Stanley E. Turner, PhD, PE HOLTEC INTERNATIONAL January 1991

I

1.0 INTRODUCTION

AND SUM~Y The .objective of this benchmarking study is to verify both the NITAWL-KENO-5a(~) methodology with the 27-group SCALE cross-section library and the CASMO-3 code() for use in criticality safety calculations of high density spent fuel storage racks. Both calculational methods are based upon transport theory and have been benchmarked against critical experiments that simulate typical spent. fuel storage rack designs as realistically as possible.

Results of these benchmark calculations with both methodologies are consistent with corresponding calculations reported in the literature.

Results of the benchmark calculations show that the 27-group (SCALE) NITAWL-KENO-5a calculations consistently under-predict the critical eigenvalue by 0.0090 + 0.0021 Sk (with a 95<

probability at a 954 confidence level) for critical experiments<)

that are as representative as possible of realistic spent fuel storage rack configurations and poison worths.

Extensive benchmarking calculations of critical experi-ments with CASMO-3 have also been reported<), giving a mean k,< of 1.0004 + 0.0011 for 37 cases. With a K-factor of 2.14<@ for 954 probability at a 954 confidence level, and conservatively neglect-ing the small overprediction, the CASMO-3 bias then becomes 0.0000

+ 0.0024. CASMO-3 and NITAWL-KENO-5a intercomparison calculations of infinite arrays of poisoned cell configurations (representative of typical spent fuel storage rack designs) show very good agreement, confirming that 0.0000 + 0.0024 is a reasonable bias and uncertainty for CASMO-3 calculations. Reference 5 also documents good agreement of heavy nuclide concentrations for the Yankee core isotopics, agreeing with the measured values within experimental A-1

l II I I

The benchmark calculations reported here confirm that either the 27-group (SCALE) NITAWL-KENO or CASMO-3 calculations are acceptable for criticality analysis of high-density spent fuel storage racks. Reference calculations for the rack designs should be performed with both code packages to provide independent verification.

2.0 NITAWL-KENO 5a BENCHMARK CALCULATIONS Analysis of a series of Babcock & Wilcox critical experiments(, including some with absorber panels typical of a poisoned spent fuel rack, is summarized in Table 1, as calculated with NITAWL-KENO-5a using the 27-group SCALE cross-section libra y and the Nordheim resonance integral treatment in NITAWL. Dancoff factors for input to NITAWL were calculated with the Oak Ridge SUPERDAN routine (from the SCALE() system of codes). The mean for these calculations is 0.9910 + 0.0033 (1 o standard deviation of the population). With a one-sided tolerance factor corresponding to 95< probability at a 954 confidence level(@, the calculational bias is + 0.0090 with an uncertainty of + 0.0021 for the sixteen critical experiments analyzed.

Similar calculational deviations have been reported by ORNL for some 54 critical experiments (mostly clean critical without strong absorbers), obtaining a mean bias of 0.0100 + 0.0013 (954/954). These published results are in good agreement with the results obtained in the present analysis and lend further credence to the validity of the 27-group NITAWL-KENO-5a calculational model for use in criticality analysis of high density spent fuel storage racks. No trends in k,z with intra-assembly water gap, with absorber panel reactivity worth, with enrichment or with poison concentration were identified.

A-2

)

Additional benchmarking calculations were also made for a series of French critical experiments ) at 4.754 enrichment and for several of the BNWL criticals with 4.264 enriched fuel.

Analysis of the French criticals (Table 2) showed a tendency to overpredict the reactivity, a result also obtained by ORNL . The calculated k,< values showed a trend toward higher values with decreasing core size. In the absence of a significant enrichment effect (see Section 3 below), this trend and the overprediction is attributed to a small inadequacy in NITAWL-KENO-5a in calculating neutron leakage from very small assemblies.

Similar overprediction was also observed for the BR<L series of critical experiments< >, which also are small assemblies (although significantly larger than the French criticals). In this case (Table 2), the overprediction appears to be small, giving a mean k<<of 0.9990 + 0.0037 (1 cr population standard deviation) .

Because of the small size of the BNWL critical experiments and the absence of any significant enrichment effect, the overprediction is also attributed to the failure of NITAWL-KENO-5a to adequately treat neutron leakage in very small assemblies.

Since the analysis of high-density spent fuel storage racks generally does not entail neutron leakage, the observed inadequacy of NITAWL-KENO-5a is not significant. Furthermore, omitting results of the French and BNWL critical experiment analyses from the determination of bias is conservative since any leakage that might enter into the analysis would tend to result in overprediction of the reactivity.

A-3

I

3. CASMO-3 BENCHMARK CALCULATIONS The CASMO-3 code is a multigroup transport theory code utilizing transmission probabilities to accomplish two-dimensional calculations of reactivity and depletion for BWR and PWR fuel assemblies. As such, CASMO-3 is well-suited to the criticality analysis of spent fuel storage racks, since general practice is to treat the racks as an infinite medium of storage cells, neglecting leakage effects.

CASMO-3 is a modification of the CASMO-2E code and has been extensively'benchmarked against both mixed oxide and hot and cold critical experiments by Studsvik Energiteknik ). Reported ana-lyses@ of 37 critical experiments indicate a mean k<< of 1.0004 +

0.0011 (le). To independently confirm the validity of CASHO-3 (and to investigate any effect of enrichment), a series of calculations were made with CASMO-3 and with NITAWL-KENO-5a on identical poisoned storage cells representative of high-density spent fuel storage racks. Results of these intercomparison calculations (shown in Table 3) are within the normal statistical variation of KENO calculations and confirm the bias of 0.0000 +

0.0024 (954/954) for CASM0-3.

Since two independent methods of analysis would not be expected to have the same error function with enrichment, results of the intercomparison analyses (Table 3) indicate that there is no significant effect of fuel enrichment over the range of enrich-ments involved in power reactor fuel. Furthermore, neglecting the French and BNWL critical benchmarking in the determination of bias is a conservative approach.

Intercomparison between, analytical methods is a technique endorsed by Reg. Guide 5.14, "Validation of Calculational Methods for Nuclear Criticality Safety".

A - 4

A E

I

REFERENCES TO APPENDIX A Green, Lucious, Petrie, Ford, White, and Wright, >>PSR /NITAWL-1 (code package) NITAWL Modular Code System For Generating Coupled Multigroup Neutron-GAmma Libraries from ENDF/B>>, ORNL-TM-3706, Oak Ridge National Laboratory, November 1975.

2. R.M. Westfall et. al., "SCALE: A Modular System for Perform-ing Standardized Computer Analysis for Licensing Evaluation",

NUREG/CR 0200'979 '

~ A. Ahlin, M. Edenius, and H. Haggblom, "CASMO A Fuel Assembly Burnup Program", AE-RF-76-4158, Studsvik report.

A. Ahlin and M. Edenius, >>CASMO - A Fast Transport Theory Depletion Code for LWR Analysis", ANS Transactions, Vol. 26,

p. 604, 1977.

>>CASMO-3 A Fuel Assembly Burnup Program, Users Manual>>,

Studsvik/NFA-87/7, Studsvik Energitechnik AB, November 1986

4. M.N. Baldwin et al., "Critical Experiments Supporting Close Proximity Water Storage of Power Reactor Fuel", BAW-1484-7, The Babcock & Wilcox Co., July 1979.

5. M. Edenius and A. Ahlin, >>CASMO-3: New Features, Benchmarking, and Advanced Applications", Nuclear Science and En ineerin 100/ 342-351/ (1988)

6. M.G. Natrella, E erimental Statistics, National Bureau of Standards, Handbook 91, August 1963.

7~ R.W. Westfall and J. H. Knight, "SCALE System Cross-section Validation with Shipping-cask Critical Experiments", QS Transactions, Vol. 33, p. 368, November 1979

8. S.E. Turner and M.K. Gurley, >>Evaluation of NITAWL-KENO Benchmark Calculations for High Density Spent Fuel Storage Racks", uclea Science and En ineerin p 80(2) 230 237, February 1982.

A-5

I

9. J.C. Manaranche, et. al., "Dissolution and Storage Exgeriment with 4.754 U-235 Enriched UOz Rods", Nuclear Technolo , Vol.

50, pp 148, September 1980

10. S.R. Bierman, et. al., "Critical Separation between Sub-critical Clusters of 4.29 Wt. 4 U Enriched UO> Rods in Water with Fixed Neutron Poisons", Batelle Pacific Northwest Labora-tories, NUREG/CR/0073, May 1978 (with August 1979 errata).

11. A.M. Hathout, et. al., "Validation of Three Cross-section Libraries Used with the SCALE System for Criticality Analy-sis", Oak Ridge National Laboratory, NUREG/CR-1917, 1981.

A-6

I l

Table 1 RESULTS OF 27-GROUP (SCALE) NZTAWL-KENO-5a CALCULATIONS OF B&W CRITICAL EXPERIMENTS Experiment Calculated Number kerr 0.9932 + 0.0016 1

0.9915 + 0.0015 ZZI 0.9916 + 0.0013

"'X 0.9918 + 0.0014 0.9923 + 0.0015 XI 0.9919 + 0.0014 XZZ 0.9961 + 0.0015 ZIZZ 0.9960 + 0.0015 XIV 0.9817 + 0.0015 0.9843 + 0.0014 0.9912 + 0.0015 0.9866 + 0.0013 0.9904 + 0.0014 XZX 0.9861 + 0.0013 0.9934 + 0.0013 0.9874 + 0.0014 Mean 0 '910 + 0.0014<'>

Bias 0.0090 + 0.0033@

Bias (954/954) 0.0090 + 0.0021 Calculated from individual standard deviations.

Calculated from k,< values and used as reference.

A - 7

I I

Table 2 RESULTS OF 27-GROUP (SCALE) NITAWL-KENO-5a CALCULATIONS OF FRENCH and BNWL CRITICAL EXPERIMENTS French Experiments Separation Critical Calculated Distance, cm Height, 'cm k,g 0 23.8 1.0231 + 0.0036 2.5 24.48 1.0252 + 0.0043 5.0 31.47 1.0073 + 0.0013 10.0 64.34 0.9944 + 0.0014 BNWL Experiments Calculated Case Expt. No. kea No Absorber 004/032 0.9964 + 0.0034 SS Plates (1.05 B) 009 0.9988 + 0.0038 SS Plates (1.62 B) 011 1.0032 + 0.0033 SS Plates (1.62 B) 012 0.9986 + 0.0036 SS Plates 013 0.9980 + 0.0038 SS Plates 0.9936 + 0.0036 Zr Plates 030 1.0044 + 0.0035 Mean 0.9990 + 0.0037 A-8

I l

Table 3 RESULTS OF CASMO-3 AND NITAWL-KENO-5a BENCEBGQK (INTERCOMPARISON) CALCULATIONS Enrichment+

Wt. c U-235 NITAWL-KENO-5a~ CASMO-3 iSkl 2.5 0.8385 + 0.0016 0.8379 0.0006 3.0 0.8808 + 0.0016 0.8776 0.0032 3.5 0.9074 + 0.0016 0.9090 0.0016 4.0 0.9311 + 0.0016 0.9346 0.0035 4.5 0.9546 + 0.0018 0.9559 0.0013 5.0 0 '743 + 0.0018 0.9741 0.0002 Mean 0.0017 Infinite array of assemblies typical of high-density spent: fuel storage racks.

k from NITAWL-KENO-5a corrected for bias of 0.0090 Sk.

A-9

I I

I I

l

5.0 THERMAL-HYDRAULIC CONSIDERATIONS 5.1 Introduction A primary objective in the design of the high density spent fuel storage racks for the Donald C. Cook spent fuel pool is to ensure adequate cooling of the fuel assembly cladding. In the following section a brief synopsis of the design basis, the method of analysis, and the numerical results is provided.

Similar methods of thermal-hydraulic analysis have been used in previous licensing efforts on high density spent fuel racks for Fermi 2 (Docket 50-341), Quad Cities 1 and 2 (Dockets 50-254 and 50-265)g Rancho Seco (Docket 50-312), Grand Gulf Unit 1 (Docket 50-416), Oyster Creek (Docket 50-219), Virgil C. Summer (Docket 50-395), Diablo Canyon 1 and 2 (Docket Nos. 50-275 and 50-323),

Byron Units 1 and 2 (Docket 50-454, 455), St. Lucie Unit One (Docket 50-335), Millstone Point I (50-245), Vogtle .Unit 2 (50-425), Kuosheng Units 1 & 2 (Taiwan Power Company), Ulchin Unit 2 (Korea Electric Power Company), J.A. FitzPatrick (New York Power Authority) and TMI Unit 1 (GPU Nuclear).

The analyses to be wed out for- the thermal-hydraulic ication o f the rack array may be broken down into the carr'ualif following categories:

(i) Pool decay heat evaluation and pool bulk temperature variation with time.

(ii) Determination of the maximum pool local temperature at the instant when the bulk temperature reaches its maximum value.

5-1

I I

I I

I

(iii) Evaluation of the maximum fuel cladding temperature to establish that bulk nucleate boiling at any location resulting in two phase conditions environment around the fuel does not occur.

( iv) Evaluation of the time-to-boil if all heat rejection paths through the cooling and cleanup are lost.

(v) Compute the effect of a blocked fuel cell opening on the local water and maximum cladding temperature.

The following sections present a synopsis of the methods employed to perform such analyses and a final summary of the results.

5.2 S ent Fuel Coolin S stem Descri tion The principal functions of the Spent Fuel Cooling System are the removal of decay heat from the spent fuel stored in the pool it serves and maintaining the clarity of, and a low activity level in, the water of the pool. Cleanup, of pool water is accomplished by diverting part of the flow, maintained for removal of decay heat, through filters and/or demineralizers as described in Section 9.4 of UFSAR.

5 2.1 S stem Functions The Spent Fuel Pool Cooling System is designed to remove from the spent fuel pool the heat generated by stored spent fuel elements .

The system serves the spent fuel pool which is shared between the two units.

5-2

I I

I I

I

The system design incorporates two separate cooling trains.

System piping is arranged so,that failure of any pipeline does not drain the spent fuel pool below the top of the stored fuel elements.

5.2.2 S stem Descri tion Each of the two cooling loops in the Spent Fuel Pool Cooling System consists of a pump, heat exchanger, strainer, piping, associated valves and instrumentation. The pump draws water from the pool, circulates it through the heat exchanger and returns it to the pool. Component cooling water cools the heat exchanger.

The clarity and purity of the spent fuel pool water is maintained by passing approximately 100 gpm of the cooling flow through a filter and demineralizer. Skimmers are provided to prevent dust and debris from accumulating on the surface of the water.

The refueling water purification pump and filter can be used separately or in conjunction with the spent fuel pool demineralizer to regain refueling water clarity after a crud burst in either unit. This can prevent loss of time during refueling due to poor visibility. The system is also used to maintain water quality in the Refueling Water Storage Tanks of both units.

The spent fuel pool filter/demineralizer is downstream of the spent fuel pool cooler. As a result, the pool purification components are subjected to water temperatures which correspond to the cooler outlets (less than 140'F). All elements of the purification system, including the resins, are qualified for 200 F design temperature.

5-3

I I

I I

I I

The spent fuel pool pump suction lines penetrate the spent fuel pool wall above the fuel assemblies stored in the pool to prevent loss of water as a result of a suction line rupture. The pool is initially filled with water at the same boron concentration (2400 ppm) as in the refueling water storage tank.

The spent fuel pool is located outside the reactor containment.

During refueling the water in the pool can be isolated from that in the re-fueling canal by a weir gate so that there is only a very small amount of interchange of water as fuel assemblies are transferred.

5.2.3 Performance Re uirements The first design basis of the system is based on the normal refueling operation with a normal batch of 80 assemblies being removed from the unit each time.

The second design basis for the system considers that it is possible to unload the reactor vessel (193 fuel assemblies) for maintenance or inspection at a time when a maximum of 3518 spent fuel assemblies are assumed already stored in the spent fuel pool.

5.3 Deca Heat Load Calculations The decay heat load calculation is performed in accordance with the provisions of USNRC Branch Technical Position ASB9-2, "Residual Decay Energy for Light Water Reactors for Long Term Cooling", Rev. 2, July, 1981. For purposes of this licensing

I I

I I

application, it is assumed that the pool contains an inventory accumulated through scheduled discharges from 1975 to 2009 (Table 1.1.1). Further, since the decay heat load increases monotonically with reactor exposure time, an upper bound of 1260 full power operation days (approximately 3.5 years) is assumed for all stored fuel. The cumulative decay heat load is computed for the instance of hypothetical normal discharge (21B in Table 1.1.1) in the year 2009. As shown in Table 5.4.1, the ratio of this decay heat load due to the inventory of previously stored fuel to the average assembly operating power P is 0.3303.

This decay heat load is assumed to remain invariant for the duration of the pool temperature evaluations performed in the wake of normal and full core offloads discussed below.

5.4 Dischar e Scenarios The following discharge scenarios are examined:

Case 1: Normal discharge A normal batch of 80 assemblies with 1260 days of reactor exposure time at full power is discharged in the pool at the end of a normal 18 month operating cycle. There are 43 previously discharged batches in the pool. As described later, the normal discharge is assumed to occur at the rate of approximately 4 assemblies per hour a fter 16 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> o f decay in the reactor. One fuel pool cooling train is active and running. One cooling train contains one heat exchanger and one fuel pool pump.

This case is run with the design fuel pool water flow rate (2300 gpm) and actual available flow rate (2800 gpm) . These two cases are labelled as Case la and Case 1b, respectively.

Case 2: Normal discharge Same as Case 1 except two fuel pool cooling trains are operating.

5-5

I I

I I

I

Case 3: Back-to-Back Full core offload The full core offload condition corresponds to the emergency reactor offload condition wherein the shutdown of a reactor occurs 30 days after the other reactor shutdown for a normal refueling. Two cooling trains are assumed to be operating in parallel after the shutdown. The decay time of the core in the reactor and the rate of discharge to the pool are. the same as in Case l.

Case 4:

Same as Case 3 except only one cooling train is in operation.

This case is listed for reference only; it is not a design basis case by the Donald C. Cook Technical Specification or the USNRC guidelines (NUREG-0800).

Detailed data on the three cases are given in Table 5.4.1 to 5.4.3.

5.5 Bulk Pool Tem erature In order to perform the analysis conservatively, the heat exchangers are assumed to be fouled to their design maximum. Thus, the temperature effectiveness, p, for the heat exchanger utilized in the analysis is the lowest postulated value calculated from heat exchanger thermal hydraulic codes. p is assumed constant in the calculation.

The mathematical formulation can be explained with reference to the simplified heat exchanger alignment of Figure 5.5.1.

Referring to the spent fuel pool/cooler system, the governing differential equation can be written by utilizing conservation of energy:

dT

= QL - QHx (5-1)

QL = Pcons + Q (r) QEV (T, t'a) 5-6

I I

I I

I I

I

where:

C: Thermal capacitance of the pool (net water volume times water density and times heat capacity), Btu/'F.

Qz,- Heat load to the heat exchanger, Btu/hr.

Heat generation rate from recently discharged fuel, which is a specified function of time, z, Btu/hr.

Pcons = P Po'eat generation rate from "old" fuel, Btu/hr. (Po = average assembly operating power, Btu/hr.)

QHX: Heat removal rate by the heat exchangerI Btu/hr.

QEy (T(ta)'eat loss to the surroundings, function of pool temperature which T

is a and ambient temperature ta, Btu/hr.

QHX is a temperature non-linear function of time effectiveness is if constant we assume during the the p

calculation. QHX can, however, be written in terms of effectiveness p as follows:

QHX +t Ct p (T ti) (5-2) where:

Wt. Coolant flow rate, lb./hr.

Ct. Coolant specific heat, Btu/lb.'F.

p: Temperature effectiveness of heat exchanger.

5-7

I I

I I

T: Pool water temperature, 'F ti: Coolant inlet temperature, 'F to. Coolant outlet temperatureI F.

p is obtained by rating the heat exchanger on a Holtec proprietary thermal/hydraulic computer code. Q(r) is specified according to the provisions of "USNRC Branch Technical Position ASB9-2, "Residual Decay Energy for Light Water Reactors for Long Term Cooling", Rev. 2, July, 1981. Q(r) is a function of decay time, number of assemblies, and in-core exposure time. During the fuel transfer, the heat load in the pool will increase with respect to the rate of fuel transfer and equals to Q(r) after the fuel transfer.

QEV is a non-linear function of pool temperature and ambient temperature. QEV contains the heat evaporation loss through the pool surface, natural convection from the pool surface and heat conduction through the pool walls and slab. Experiments show that the heat conduction takes only about 4% of the total heat loss I'5.5.1], therefore, can be neglected. The evaporation heat and nature convection heat loss can be expressed as:

QEV m 1's + hC As 8 (5-~)

where:

m: evaporation rateI lb /hr ft.

As'ass Latent heat of pool water, Btu/lb.

Pool surface area, ft..2 hc- Convection heat transfer coefficient at pool surface, Btu/ft. ,hr. 'F 5-8

-I

~

~

8 = T-ta. The temperature difference between pool water and ambient air, 'F The mass evaporation rate m can be obtained as a non-linear function of 8. We, therefore, have m = hD (8) (Wps Was) (5-4) where:

Wps ~ Humidity ratio of saturated moist air at pool water surface temperature T.

Was. Humidity ratio of saturated moist air at ambient temperature ta hD(8): Diffusion coefficient at pool water surface. hD is a non-linear function of 8, lb./hr. ft. 'F The non-linear single order differential equation (5-1) is solved using Holtec's Q.A. validated numerical integration code "ONEPOOL".

Figures 5.5.2 through 5.5.6 provide the bulk pool temperature profiles for the normal discharge, and full core offload scenarios respectively. Table 5.5.1 gives the peak water temperature,

'I coincident time, and coincident heat load to the cooler and coincident heat loss to the ambient for three cases.

The next step in the analysis is to determine the temperature rise profile of the pool water if all forced indirect cooling modes are suddenly lost. Make-up water is provided with a fire hose.

Clearly, the most critical instant of loss-of-cooling is when pool water has reached its maximum value. Et is assumed that cooling water is added through a fire hose at the rate of G lb./hr. The 5-9

I

~

~

~

~

~

~

~

I L

cooling water is at temperature, tcool. The .governing enthalpy balance equation for this condition can be written as dT

[C + G(Ct)(z zo)] = Pcons + Q(~ + rins) + G (Ct) (tcool T) d'r

- ')EV where water is assumed. to have specific heat of unity, and the time coordinate z is measured from the instant maximum pool water temperature is reached. -zo is the time coordinate when the direct addition (fire hose) cooling water application is begun. tins is the time coordinate measured from the instant of reactor shutdown to when maximum pool water temperature is reached. T is the dependent variable (pool water temperature). For conservatism, QE~

is assumed to remain constant after pool water temperature reaches and rises above 170'F.

A Q.A. validated. numerical quadrature code is used to integrate the foregoing equation. The pool water heat up rate, time-to-boil, and subsequent water evaporation-time profile are generated and compiled for safety evaluation.

Assuming no make-up water (G = 0), the time-to-boil output results are presented in Table 5.5.2. Figures 5.5.6 through 5.5.10 show the plot of the inventory of water in the pool after loss-of-coolant-to-the-pool condition begins.

Zt is seen from Table 5.5.2 that sufficient time to introduce manual cooling measures exists and the available time is consistent with other PWR reactor installations.

5-10

I 1

~

i

~

5.6 Local Pool Water Tem erature Zn this section, a summary of the methodology, calculations and results for local pool water temperature is presented.

5.6.1 Basis Zn order to determine an upper bound on the maximum fuel cladding temperature, a series of conservative assumptions are made. The most important assumptions are listed below:

0 The fuel pool will contain spent fuel with varying time-after-shutdown (rs). Since the heat emission falls off rapidly with increasing rs, it is conservative to assume that all fuel assemblies are from the latest batch discharged simultaneously in the shortest possible time and they all have had the maximum postulated years of operating time in the reactor. The heat emission rate of each fuel assembly is assumed to be equal and'aximum.

0 As shown in the pool layout drawings, the modules occupy an irregular floor space in the pool. For analysis, a circle circumscribing the the'ydrothermal actual rack floor space is drawn (Fig. 5. 6. 1) . Zt is further assumed that the cylinder with this circle as its base is packed with fuel assemblies at the nominal layout pitch.

0 The actual downcomer space around the rack module group varies. The nominal downcomer gap available in the pool is assumed to be the total gap available around the idealized cylindrical rack; thus, the maximum resistance to downward flow is incorporated into the analysis (Figs. 5.6.2 and 5.6.3) (i.e. minimum gap between the pool wall and rack module, including seismic kinematic effect).

0 No downcomer flow is assumed to exist between the rack modules.

5-11

~

~

~

I

~

~

~

~

I'

~

~

I

0 The ANF 17x17 fuel assembly has been used in the analysis which, from the thermal-hydraulic standpoint, bounds all types of fuel bundles utilized in the Donald C. Cook reactor.

0 No heat transfer is assumed to occur between pool water and the surroundings (wall, etc.)

5.6.2 Model Descri tion Xn this manner, a conservative idealized model for the rack assemblage is obtained. The water flow is axisymmetric about the vertical axis of the circular rack assemblage, and thus, the flow is two-dimensional (axisymmetric three-dimensional) . Fig. 5. 6. 2 shows a typical "flow chimney" rendering of the thermal hydraulics model. The governing equation to characterize the flow field in the pool can now be written. The resulting integral equation can be solved for the lower plenum velocity field (in the radial direction) and axial velocity (in-cell velocity field), by using the method of collocation. The hydrodynamic loss coefficients which enter into the formulation of the integral equation are also taken from well-recognized sources (Ref. 5.6.1) and wherever discrepancies in reported values exist, the conservative values are consistently used. Reference 5.6.2 gives the details of mathematical analysis used in this solution process.

After the axial velocity field is evaluated, it is a straight-forward matter to compute the fuel assembly cladding temperature.

The knowledge of the overall flow field enables pinpointing of the storage location with the minimum axial flow (i.e, maximum water outlet temperatures). This is called the most "choked" location.

Zn order to find an upper bound on the temperature in a typical cell, it, is assumed that it is located at the most choked location. Knowing the global plenum velocity field, the revised 5-12

axial flow through this choked cell can be calculated by solving the Bernoulli's equation for the flow circuit through this cell.

Thus, an absolute upper bound on the water exit temperature and maximum fuel cladding temperature is obtained. In view of these aforementioned assumptions, the temperatures calculated in this manner overestimate the temperature rise that will actually occur in the pool. Holtec's computer code THERPOOL*, based on the theory of Ref. 5.6.2, automates this calculation. r The analysis procedure embodied in THERPOOL has been accepted by the Nuclear Regulatory Commission on several dockets. The Code THERPOOL for local temperature analyses includes the calculation of void generations. The effect of void on the conservation equation, crud layer in the clad, flux trap temperature due to gamma heating, and the clad stress calculation when a void exists, are all incorporated in THERPOOL. The peaking factors are given in Table 5.6.1.

5.7 Claddin Tem erature The maximum specific power of a fuel array ~ can be given by:

M= q Fxy where:

Fxy = radial peaking factor q = average fuel assembly specific power

• THERPOOL has been used in qualifying the spent fuel pools for Enrico Fermi Unit 2 (1980), Quad Cities I and II (1981)g Oyster Creek (1984), V.C. Summer (1984), Rancho Seco (1983)f Grand Gulf I (1985), Diablo Canyon I and II (1986), among others.

5-13

1l The maximum temperature rise of pool water in the most disadvantageously placed fuel assembly is computed for all loading cases. Having determined the maximum local water temperature in the pool, it is now possible to determine the maximum fuel cladding temperature. A fuel rod can produce Fz times the average heat emission rate over a small length, where Fz is the axial rod peaking factor. The axial heat distribution in a rod is generally a maximum in the central region, and tapers off at its two extremities.

Zt can be shown that the power distribution corresponding to the chopped cosine power emission rate is given by n (a + x) q(x) = gA sin h+ 2a where:

h: active fuel length a: chopped length at both extremities in the power curve x: axial coordinate with origin at the bottom of the active fuel region The value of a is given by h z a

1 - 2z where:

1/2 1 1 2

+

m Fz +2F2 nFz z2 5-14

l where Fz is the axial peaking

)

factor.

The cladding temperature Tc is governed by a third order differential equation which has the form of d3 T d x

+ al d2 T d x2 a2 f dT dx

= (x) where al, a2 and f(x) are functions of x, and fuel assembly geometric properties. The solution of this differential equation with appropriate boundary conditions provides the fuel cladding temperature and local water temperature profile.

In order to introduce some additional conservatism in the analysis, we assume that the fuel cladding has a crud deposit which results in .005 F-sq.ft.-hr/Btu of crud resistance, which covers the entire surface.

Table 5.6.2 provides the .key input data for local temperature analysis. The results of maximum local pool water temperature and minimum local fuel cladding temperature are presented in Table 5.7.1.

The local boiling temperature of water is approximately 242'F at 26'elow the free water surface and higher at lower elevations.

The location where the local water temperature reaches its maximum value is deeper than 26'elow the free water surface, where the coincident boiling temperature of water is greater than 242'P.

It is shown that the local pool water temperature is lower than the local boiling point and therefore, nucleate boiling will not occurs 5-15

Finally, it is noted that the fuel cladding temperature is considerably lower than the temperatures to which the cladding is subjected inside the reactor. Therefore, it is concluded that there is sufficient margin against fuel cladding failure in the spent fuel pool.

5.8 Blocked Cell Anal sis Calculations are also performed assuming that 50% of the top opening in the thermally limiting storage cell is blocked due to a horizontally placed (misplaced) fuel assembly. The corresponding maximum local pool water temperature and local fuel cladding temperature data are also presented in Table 5.7.1.

There is also no incidence of localized nucleate boiling of the pool water or potential for fuel cladding damage.

5.9 References for Section 5 5.6.1 General Electric Corporation, R&D Data Books, "Heat Transfer and Fluid Flow", 1974 and updates.

5.6.2 Singh, K.P. et al., "Method for Computing the Maximum Water Temperature in a Fuel Pool Containing Spent Nuclear Fuel", Heat Transfer Engineering, Vol. 7, No. 1-2I pp. 72-82 (1986) ~

5-16

I l

I

Table 5.4.1 =-

FUEL SPECIFIC POWER AND POOL CAPACITY DATA Total water volume of Pool: 635645 gallons Specific Operating Power of a Fuel Assembly: 60.3E+06 Btu/hr.

Dimensionless decay power of "old" discharges: 0.3303 5-17

Table 5.4.2 DATA FOR SCENARXOS 1 through 4 CASE NO. Ia Ib Pool thermal capacity 4.241 4.241 4.241 4.241 4.241 Cxl0, Btu/F No. of Cooling Trains No. of Discharges considered for the Analysis Time between 720 720 shutdowns, hr.

Cooler Inlet Temp., 108.4 108.4 103.9 101.4 104 '

OF coolant Flow Rate/ 1.49 1 49 l. 49 1. 49 l. 49 cooler, 10 lb./hr.

Fuel Pool Water l. 14 1.40 l. 14 l. 14 l. 14 Flow Rate, 10 lb. /hr.

Temperature 0.3970 0 43 0.3975 0.3979 '.3987 Effectiveness/

cooler, p 5-18

I I

Table 5.4.3 DATA FOR SCENARIOS 1 THROUGH 4 Time After shutdown when Offload Expo.

Case Discherge No. of Transfer Begins T~e T'me No. lD Assemblies (hrs) (hrs) (hrs) la Discharge 1 80 168 19.07 30240 or 1b Discharge 1 80 168 19.07 30240 Discharge 1 80 168 19.07 30240 3 (

or Discharge 2 80 168 46.00 10080 4 (Full core) 113 30240 5-19

I I

I I

Table 5.5.1 POOL BULK TEMPERATURE AND HEAT LOAD DATA Time Coincident Tmax o Coincident Coincident Max. Pool (after Evaporation Case Cooler Duty Bulk Temp., reactor Hept Loss, No. 106 Btu/hr. Op shutdown) 10~ Btu/hr.

la 30.241 159.54 207 3.00 lb 30.69 156 '1 206 2.578 32.787 131.57 198 0.689 50.690 143 '4 222 1.395 45.04 176.91 225 6.887 5-20

I Table 5.5.2 TIME-TO-BOIL FOR VARIOUS DISCHARGE SCENARIOS Time-to-Boil (hours)

Case Number G = 0 GPM la 7.82 1b 8.27 11.52 5.74 3.02 5-21

I I

I I

Table 5.6.1 PEAKING FACTOR DATA Radial Bundle Peaking Factor 1.65 Total peaking factor 2.40 5-22

I I

I I

Table 5.6.2 DATA FOR LOCAL TEMPERATURE Type of Fuel Assembly Fuel Cladding Outer Diameter, inches 0.36 Fuel Cladding Inside Diameter, inches 0.31 Storage Cell inside Dimension, inches 8.75 Active fuel length, inches 144 No. of Fuel Rods/Assembly 264 Operating Power per Fuel Assembly 60.3 Po x 6p Btu/hr Cell pitch, inches 8.97 Cell height, inches 168 Plenum radius, feet 29.3 Min. Bottom height, inches 4.75 Min. gap between pool wall 1.5 and outer rack periphery, inches 5-23

I I

I I

I

Table 5.7.1 LOCAL AND CLADDING TEMPERATURE OUTPUT DATA FOR THE MAXIMUM POOL WATER CONDITION (Case a)

Condition Water Tem . 'F Tem . 'F No blockage 168.0 212.9 50% blockage 219.2 246.9 5-24

l I

I I

SPERT W,T FUE'QQL.

Ã,CP l,, 7

~,T j 4 ~

FIGURE 5.5.1 Pool Bulk Temperature Model 5-25

I I

I I

165 163 DONALD C. COOK SFP NORMAL DISCHARGE, ONE COOLING TRAIN, CASE 1a 160 158

~ 155 IJJ 153 0

150 ca 148 145 140 138 50 100 150 200 250 300 350 400 450 500 TIME AFTER FIRST SHUTDOWN OF THE REACTOR, I.IR FIGURE 5.5.2

I I

I I

165 163 DONALD C. COOK SFP NORMAL DISCHARGE, ONE COOLING TRAIN, CASE 1b 160 158 155

a. 153

~

o. 150 O

148 g

145 143 140 138 50 100 150 200 250 300 350 400 450 500 TIME AFTER FIRST SHUTDOWN OF THE REACTOR, HR FIGURE 5.5.3

I I

I

135 DONALD C. COOK SFP NORMAL DISCHARGE, TWO COOLING TRAINS, CASE2 133 130

~ 128 O

O 125 123 120 118 50 100 150 200 250 300 350 400 450 500 TIME AFfER FIRST SHUTDOWN OF THE REACTOR, HR FIGURE 5.5.4

150 DONALD C. COOK SFP FULL CORE OFFLOAD,TWO COOLING TRAINS, CASE3 145 140 135 LB I

UJ 130 00

~ 125 120 115 110 100 200 300 400 500 600 700 800 900 1000 1100 1200 TIME AFTER FIRST SHUTDOWN OF THE REACTOR. HR FIGURE 5.5.5

I I

I I

I I

I

185 180 DONALD C. COOK SFP FULL CORE OFFLOAD, ONE COOLING TRAIN, CASE4 175 170 165 o 160 O ~ 155 O

Q.

150 g

145 140 135 130 125 100 200 300 400 500 600 700 800 900 1000 1100 1200 TIME AFTER FIRST SHUTDOWN OF THE REACTOR, HR FIGURE 5.5.6

I I

5.50E+005 5.00E+ 005 COOK SFP LOSS OF COOLING SCENARIO, CASE 1a 4,50E+ 005

~ 4.00E+005 c

c 3.50E+005 0

E

~ 3.00E+005 I

L

< 2.50E+005 0

2,00E+ 005 E

o 1 50E+005 0

~0 1.00E+005 5.00E+ 004.

O,OOE+ 000 0 40 80 120 160 Time After Max, temp. has been reached, hr F IGURE 5. 5. 7

I 5.50E+005 5.00E+ 005 COOK SFP LOSS OF COOLING SCENARIO, CASE 1b I

4.50E+005 D 4.OOE+OO5 C

c 3.50E+005 o

E

~ 3.00E+005 Vl I

O 2.50E+005 0

2,00E+005 o 1.50E+005 0

o 1.00E+005 5.00E+ 004 O.OOE+000 0 40 80 120 160 Time After Max. temp. has been reached, hr FIGURE 5.5.8

I I

I I

I I

I

5.50E+005 5.00E+ 005 COOK SFP LOSS OF COOLING SCENARIO, CASE 2 4.50E+005

> 4.00E+005 c 3.50E+005 0

E

~ 3.00E+005 I

4J

< 2.50E+005 0

2.00E+005 0 1.50E+005 0

~0 1.00E+005 5.00E+004 O.OOE+000 0 40 80 120 160 Time After Max. temp. has been reached, hr FIGURE 5.5.9

I I

I I

I

5.50E+005 5.00E+005 COOK SFP LOSS OF COOLING SCENARIO, CASE 3 4.50E+005

> 4.00E+005 c

c 3.50E+005 0

E

~ 3.00E+005

> 2.50E+005 0

2.00E+005 o 1.50E+005 0

~() 1.00E+005 I

5.00E+004 O.OOE+000 0 20 40 60 80 Time After Max. temp. has been reached, hr FIGURE 5.5.j.o

I I

I

~

~

I

5.50E+ 005 5.00E+005 COOK SFP LOSS OF COOLING SCENARIO, CASE 4 4.50E+ 005 u 4.OOE+OO5 C

c 3.50E+005 0

E CL 3.00E+005 0 2.50E+005 0

2.00E+ 005 0 1.50E+005 0

~~ 1.00E+005 I-5.00E+ 004.

O,OOE+000 0 20 40 60 .80 Time After Max. temp. has been reached, hr FIGURE 5.5.11

I I

I ' I I

OUi P()

~t ~

t Z

~

D P

+I ~

0 0

0 0HEAiADDtiIGN ZI 0 V Q

tN 5-37 THERMALCHIMNEY FLOW MODEL FlGLlRE 5~6~2

I I

I

i$

~ 4.

~

4 ~

r

~

~

.4

~ ~

~ 0 ~

~ ~

. 4 RACK 0 ~

~ ~

e e

~ ~ 4

~ g

~ ~

rd

~ ~

r ~ ~

oawN COMER h~ r ~

~

cf ~

V ~

go

~ ~

~

~mgrrm~

r' o

~

S ~ rr r ~

5-38 CONVECTION CVR RENTS IN THE POOL FIGURE 5.6. 3

I I

I I

I I

I

6.0 STATIC AND DYNAMIC ANALYSIS 0 RACK STRUCTURE 6.1 Introduction The purpose of this section is to present analyses which demonstrate the structural adequacy of the Donald C. Cook spent fuel high density rack design under normal storage and the postulated accident loading conditions as defined by and following the guidelines of the USNRC Standard Review Plan (Ref. 6.1.1). The method of analysis presented uses a time-history integration method similar to that previously used in the licensing reports on high density spent fuel racks for Enrico Fermi Unit 2 (USNRC Docket No. 50-341), Quad Cities 1 and 2 (USNRC Docket Nos. 50-254 and 50-265), Rancho Seco (USNRC Docket No. 50-312), Grand Gulf Unit 1 (USNRC Docket No. 50-416), Oyster 'Creek (USNRC Docket No.

50-219), V.C. Summer (USNRC Docket No. 50-395), Diablo Canyon Units 1 and 2 (USNRC Docket Nos. 50-275 and 50-323), Vogtle Unit 2 (USNRC Docket No. 50-425) and Millstone Point Unit 1 (USNRC Docket No. 50-245 ) . The analyses carried out for the Donald C. Cook racks are considerably more elaborate and exhaustive in scope and substance than those performed in the aforementioned dockets, and reflect advances in 3-D fuel rack simulation technology in the past two years. The details are presented later in this section, after the essential elements of the dynamic model are fully explained.

The results show that the high density spent fuel racks are structurally adequate to resist the postulated stress combinations associated with level A, B, C, and D conditions as defined in Refs. 6.1.1, 6.1.2, and 6.1.3.

6-1

I 6.2 Anal sis Outline The principal steps in performing the seismic analysis of Donald C. Cook racks are summarized below:

a0 Develop statistically independent synthetic time histories for three orthogonal directions which satisfy USNRC SRP3.8.4. Two time histories are considered to be statistically independent if correlation coefficient is less than 0.15.

their normalized

b. Prepare a three-dimensional dynamic model of the fuel rack which embodies all elastostatic characteristics and structural nonlinearities of the Donald C. Cook rack modules.

c~ Perform a series of 3-D dynamic analyses on a limiting module geometry type from those listed in Tables 2.1.1 and 2 '.3 and for varying physical conditions (such as coefficient of friction, extent. of cells containing fuel assemblies, and proximity of other racks).

d. Perform stress analysis for the critical case from the dynamic analysis runs made in the foregoing steps.

Demonstrate compliance with ASME Code Section III, sub-section NF (Ref. 6.1.2) limits.

e. Carry out a degree-of-freedom (DOF) reduction procedure on the single rack 3-D model such that the kinematic responses calculated by the reduced DOF (model RDOFM) are in agreement with the baseline model of step (b) above. This reduced DOF model is also truly three-,

dimensional.

Prepare a whole pool multi-rack dynamic model by compiling the RDOFM's of ~al rack modules in the pool, and by including all fluid coupling interactions among them, as well as those between the racks and pool walls.

This 3-D multi-module simulation is re ferred to as a Whole Pool Multi-Rack (WPMR) model.

6-2

l I

l

go Perform a 3-D Whole Pool Multi-Rack (WPMR) analysis to demonstrate that all kinematic criteria for Donald C.

Cook rack modules are satisfied (see Section 6.8), and that pedestal compressive loads are comparable to the loads used for structural qualification per item d above. Section 6.8 gives the criteria which need to be checked.

For the Donald C. Cook racks, the principal kinematic criteria are (i) no rack to pool wall impact, and ( ii) no rack-to-rack impact in the cellular region of the racks.

Figure 6.2.1 shows a pictorial view of the rack module. It is noted that the baseplate extends beyond the cellular region, envelope, thus ensuring that the inter-rack impact, if any, would first occur at the baseplate elevation. The baseplate of the rack modules is structurally qualifiable to withstand large in-plane impact loads.

We describe each of the above analysis steps in some detail, in the following sub-sections with special emphasis on the baseline 3-D dynamic model which is the building block for all subs equent analyses. We also present the results of the analysis in the concluding sub-section.

6.3 Artificial Slab Motions The UFSAR provides a single response spectrum in the horizontal direction and a single response spectrum in the vertical direction (2/3 of the horizontal) for the Design Basis Earthquake (DBE). A corresponding pair of spectra are provided for the Operating Basis Earthquake (OBE).

6-3

I I

I I

I

Holtec's Q.A. validated time history generation code GENEQ [6.3.1]

was used to generate three synthetic statistically independent time histories for the North-South, East-West and vertical directions, respectively, from the two response spectra. 5%

damping is used for the DBE condition. Figures 6.3.1 6.3.3 show the DBE time history plots. Response spectra corresponding to these time histories were also generated and are shown overlaid on the design spectra in Figures 6.3.4 6.3.6.

The normalized correlation coefficients pij between time histories i and j (1 = N-S, 2 = E-W, 3 = vertical) are provided in Table 6.3.1.

The above analyses were repeated for the OBE spectra using 2%

damping. Figures 6. 3. 7 - 6. 3. 9 present the time history plots, and Figures 6.3.10 -6.3.12 show the comparison between the design spectra and the derived spectra. Table 6.3.1 also provides pij for the OBE time histories. Xt is noted that the enveloping requirement on the derived spectra and statistical non-coherence of artificial motions are unconditional'ly satisfied.

6-4

I I

I

6.4 Outline of Sin e Rack 3-D Anal sis The spent fuel storage racks are Seismic Class I equipment. They are required to remain functional during and after a Design Basis Earthquake (Ref. 6.1.3). These racks are neither anchored to the pool floor nor attached to the sidewalls. The individual rack modules are not interconnected. Furthermore, a particular rack may be completely loaded with fuel assemblies (which corresponds to greatest rack inertia), or it may be completely empty. The coefficient of friction, p, between the supports and pool floor is another indeterminate factor. According to Rabinowicz (Ref.

6.4.1), the results of 199 tests performed on austenitic stainless steel plates submerged in water show a mean value of p to be 0.503 with a standard deviation of 0.125. The upper and lower bounds (based on twice the standard deviation) are thus 0.753 and 0.253, respectively. Analyses are therefore performed for single rack simulations using values of the coefficient of friction equal to 0.2 (lower limit) and 0.8 (upper limit), respectively. The bounding values of p ~ 0.2 and 0.8 have been found to bracket the upper limit of the module response in previous rerack projects.

A single rack 3-D analysis requires another key modelling assumption. This relates to the location and relative motion of neighboring racks. The gap between a peripheral rack and an adjacent pool wall is known, and the motion of the pool wall is prescribed without any ambiguity. However, another rack adjacent to the rack being analyzed is also free-standing and subject to motion during a seismic event. To conduct the seismic analysis of a given rack its physical interface with neighboring modules must be specified. The standard procedure in the single rack analysis is to specify that the neighboring. racks move 180'ut-of-phase in 6-5

I I

4 I

I

relation to the subject rack. Thus, the available gap before inter-rack impact occurs is one half of the physical gap. This "opposed phase motion" assumption increases the likelihood of predicting intra-rack impacts and is thus a conservative assumption. However, it also increases the relative contribution of fluid coupling terms, which depend on fluid gaps and relative movements of bodies, making the outright conservatism a less certain assertion. In fact, 3-D Whole Pool Multi-rack analyses carried out for Taiwan Power Company's Chin Shan Station, and for GPU Nuclear's Oyster Creek Nuclear Station show that the single rack simulations predict smaller rack displacement during seismic responses. Nevertheless, single rack analyses permit detailed evaluation of stress fields, and serve as a benchmark check for the much more involved, WPMR analysis results. In order to predict the limiting conditions of rack module seismic response within the framework of single rack analysis, module A4 (13x14) is analyzed. This is typical of the largest module, and is also a corner module. The corner module has larger rack-to-wall gaps which will minimize the fluid coupling.

The rack is considered fully loaded or half loaded, with limiting coefficients of friction; these simulations identify the worst case response for rack movement and for rack structural integrity.

After completion of reracking, the gaps between the rack modules and those between the racks and walls will be in the manner of I

Figure 2.1.1. We show in this report that all single rack 3-D simulations predict that no rack-to-rack or rack-to-wall impacts will occur in the cellular region of the racks.

The seismic analyses were performed utilizing the time-history method. Pool slab acceleration data presented in the preceding sub-section was used.

6-6

I l

W S I

I I

The objective of the seismic analysis of single racks is to determine the structural response (stresses, deformation, rigid body motion, etc.) due to simultaneous application of the three statistically independent, orthogonal seismic excitations. Thus, recourse to approximate statistical summation techniques such as the "Square-Root-of-the-Sum-of-the-Squares" method (Ref. 6.4-2) is avoided. For nonlinear analysis, the only practical method is simultaneous application of the seismic loading to a nonlinear model of the structure.

The seismic analysis of a single rack is performed in three steps, namely:

Development of a nonlinear dynamic model consisting of inertial mass elements, spring, gap, and friction elements.

2 ~ Generation of the equations of motion and inertial coupling and solution of the equations using the "component element method" (Refs. 6.4.3 and 6.4.4) to determine nodal forces and displacements. The Holtec computer code DYNARACK is used to solve the system of equations [6.4.5].

3 ~ Computation of the detailed stress field in the rack just above the baseplate and in the support legs is made using the nodal forces calculated in the previous step. These stresses are checked against the design limits given in a later sub-section.

A brief description of the dynamic model follows.

6.5 D namic Mode for The S' Rac Anal s's Since the racks are not anchored to the pool slab or attached to the pool walls or to each other, they can execute a wide variety of motions. For example, the rack may slide on the pool floor 6-7

I I

I I

I I

I

(so-called "sliding condition" ); one'r more legs may momentarily lose contact with the liner ( "tipping condition" ); or the rack may experience a combination of sliding and tipping conditions. The structural model should permit s imulat ion o f these kinematic events with inherent built-in conservatisms. Since the modules are designed to preclude the incidence of inter-rack impact in the cellular region, it is also necessary to include the potential for inter-rack impact phenomena in the analysis to demonstrate that such impacts do not occur. Lift-off of the support legs and subsequent liner impacts must be modelled using appropriate impact (gap) elements, and Coulomb friction between the rack and the pool liner must be simulated by appropriate piecewise linear springs.

The 'elasticity of the rack structure, relative to the base, must also be included in the model even though the rack may be nearly rigid. These special attributes of rack dynamics require a strong emphasis on the modeling of the linear and nonlinear springs, dampers, and compression only stop elements. The term non-linear spring is the generic term to denote the mathematical element representing the situation where the restoring force exerted by the element is not linearly proportional to the displacement. In the fuel rack simulation the Coulomb friction interface between the rack support leg and the liner is a typical example of a non-linear spring. The model outline in the remainder of this sub-section, and the model description in the following sub-section, describe the detailed modeling technique to simulate these effects, with considerable emphasis placed on the nonlinearity of the rack seismic response.

6-8

I I

I I

I I

I

6.5.1 ssum t'ons a." The fuel rack structure is a folded metal plate assemblage welded to a baseplate and supported on four legs. An odd-shaped module may have more than four legs. The rack structure itself is a very rigid structure. Dynamic analysis of typical multi-cell racks has shown that the motion of the structure is captured almost completely by modelling the rack as a twelve degree-of-freedom structure, where the movement of the rack cross-section at any height is described in terms of six degrees-of-freedom of the rack base and six degrees of freedom defined at the rack top. The rattling fuel is modelled by five lumped masses located at H, . 75H, . 5Hg

.25H, and at the rack base, where H is the rack height as measured from the base.

b. The seismic motion of a fuel rack is characterized by random rattling of fuel assemblies in their individual storage locations. Assuming a certain statistical coherence (i.e. assuming that all fuel elements move in-phase withinthea rack) in the vibration of the. fuel assemblies exaggerates computed dynamic loading on the rack structure. This assumption, however, greatly reduces the required degrees-of-freedom needed to model the fuel assemblies which are represented by five lumped masses located at different levels of the rack. The centroid of each fuel assembly mass can be located, relative to the rack structure centroid at that level, so as to simulate a partially loaded rack.

c. The local flexibility of the pedestal is modelled so as to account for floor elasticity, and local rack elasticity just above'he pedestal.

d. The rack base support may slide or lift-offthe pool floor.

e. The pool floor has a specified time-history of seismic accelerations along the three orthogonal directions.

i

f. Fluid coupling between simulated rack and fuel assemblies, and between rack and wall, is by introducing appropriate inertial coupling into the system kinetic energy. Inclusion of these effects uses the methods of Refs. 6.5.1 and 6.5.2 for rack/assembly coupling and for rack/rack coupling.

6-9

g. Potential impacts between rack and fuel assemblies are accounted for by appropriate "compression only" gap elements between masses involved.

h. Fluid damping due to viscous effects between rack and assemblies, and between rack and adjacent rack, is conservatively neglected.

i. The supports are modeled as "compression only" elements the vertical direction and as "rigid links" for transferring for horizontal stress. The bottom of a support leg is attached to a frictional spring as described in sub-section 6.6. The cross-section inertial properties of the support legs are computed and used in the final computations to determine support leg stresses.

j. The effect of sloshing is negligible at the level of the top of the rack and is hence neglected.

k. The possible incidence of rack-to-wall or rack-to-rack impact is simulated by gap elements at the top and bottom of the rack in the two horizontal directions . The bottom elements are located at the baseplate elevation.

1. Rattling of fuel assemblies inside the storage locations causes the "gap" between the fuel assemblies and the cell wall to change from a maximum of twice the nominal gap to a theoretical zero gap. Fluid coupling coefficients are based on the nominal gap.

m. The form drag due to motion of the fuel assembly in the storage cell, or that due to movement of a rack in the pool, has been neglected in this analysis for added conservatism.

n. The fluid coupling terms are based on opposed phase motion of adjacent modules.

Figure 6.5.1 shows a schematic of the model. Twelve degrees of freedom are used to track the motion'f the rack structure.

Figures 6.5.2 and 6.5.3, respectively, show the inter-rack impact springs (to track the potential for impact between racks or between rack and wall) and fuel assembly/storage cell impact 6-10

I I

I I

I I

springs at a particular level. Si (i = 1, 4) represent support locations, pi represent absolute degrees-of-freedom, and qi represent degrees-of-freedom relative to the slab. H is the height of the rack above the baseplate.

As shown in Figure 6.5.1, the model for simulating fuel assembly motion incorporates five rattling lumped masses. The five rattling masses are located at the baseplate, at quarter height, at half height, at three quarter height, and at the top of the rack. Two degrees-of-freedom are used to track the motion of each rattling mass in the horizontal plane. The vertical motion of each rattling mass is assumed to be the same as the rack base. Figures 6.5.4, 6.5.'5, and 6.5.6 show the modelling scheme for including rack elasticity and the degrees of freedom associated with rack elasticity. In each plane of bending a shear and a bending spring are used to simulate elastic effects in accordance with Ref.

6.5.1. Table 6.6.2 gives spring constants for these bending springs as well as corresponding constants for extensional and torsional rack elasticity.

6.5.2 Mode Descri tio The absolute degrees-of-freedom associated with each of the mass locations are identified in Figure 6.5.1 and in Table 6.5.1. The rattling masses (nodes 1*, 2*, 3*, 4*, 5*) are described by translational degrees-of-freedom q7-q16.

Ui(t) is the pool floor slab displacement seismic time-history.

Thus, there are twenty-two degrees of freedom in the system. Not shown in Fig. 6.5.1 are the gap elements used to model the support legs and the impacts with adjacent racks.

6-11

I I

I 1,.y I

I I

6.5.3 Fluid Cou lin An effect of some significance requiring careful modeling is the "fluid coupling effect" (Refs. 6.5.1 and 6.5.2). If one body of mass (m1) vibrates adjacent to another body (mass m2), and both bodies are submerged in a frictionless fluid medium, then Newton's equations of motion for the two bodies have the form:

N N (ml + Mll) Xl + M12 X2 applied forces on mass ml + 0 (xl )

M21 Xl + (m2 + M22) X2 = applied forces on mass m2 + 0 (x2 )

X1, X2 denote absolute accelerations of masses m1 and m2, respectively and the notation 0(x ) denotes non-linear terms which arise in the derivation.

M11~ M12 M21, and M22 are fluid coupling coefficients which depend on the shape of the two bodies, their relative disposition, etc. Fritz (Ref. 6.5.2) gives data for Mij for various body shapes and arrangements. The above equations indicate that the effect of the fluid is to add a certain amount of mass to the body (M11 to body 1), and an external force which is proportional to the acceleration of the adjacent body (mass m2) . Thus, the acceleration of one body affects the force field on another. This force is a strong function of the interbody gap, reaching large values for very small gaps. This inertial coupling is called fluid coupling. It has an important effect in rack dynamics. The lateral motion of a fuel assembly inside the storage location will encounter this effect. So will the motion of a rack adjacent to another rack if the racks are closely spaced. These effects are included in the equations of motion. For example, the fluid 6-12

I coupling is between nodes 2 and 2* in Figure 6.5.1. Furthermore, the rack equations contain coupling terms which model the effect of fluid in the gaps between adjacent racks. The coupling terms modeling the effects of fluid flowing between adjacent racks are computed assuming that all adjacent racks are vibrating 180 out of phase from the rack being analyzed. Therefore, only one rack is considered surrounded by a hydrodynamic mass computed as if there were a plane of symmetry located in the middle of the gap region.

Finally, fluid virtual mass is included in the vertical direction vibration equations of the rack; virtual inertia is also added to the governing equation corresponding to the rotational degree of freedom, q6(t) and q22(t).

6.5a4 ~Dam in Zn reality, damping ( Re f. 6. 5. 3 ) of the rack motion arises from material hysteresis (material damping), relative intercomponent motion in structures (structural damping), and fluid viscous effects (fluid damping). Zn the analysis, a maximum of 1%

structural damping is imposed on elements of the rack structure during OBE and DBE simulations. Material and fluid damping due to fluid viscosity are conservatively neglected. The dynamic model has the provision to incorporate form drag effects; however, no form drag has been used for this analysis.

6 ~ 5.5 ~Im act Any fuel assembly node (e.g., 2*) may impact the corresponding structural mass node 2. To simulate this impact, four compression-only gap elements around each rattling fuel assembly 6-13

I I

L node are provided (see Figure 6.5.3) . The compressive loads developed in these springs provide the necessary data to evaluate the integrity of the cell wall structure and stored array during the seismic event. Figure 6.5.2 shows the location of the impact L springs used to simulate any potential for inter-rack or rack-to-wall impacts. Sub-section 6.6 gives more details on these additional impact springs. Since there are five rattling masses, a total of 20 impact springs are used to model fuel assembly-cell wall impact.

6.6 Assembl o the D namic Model The .cartesian coordinate system associated with the rack has the following nomenclature:

x ~ Horizontal coordinate along the short direction of rack rectangular planform y = Horizontal coordinate along the long direction of the rack rectangular planform z = Vertical coordinate upward from the rack base Table 6.6.1 lists all spring elements used in the 3-D single rack analysis.

L If the simulation model is restricted to two dimensions (one L horizontal motion plus vertical motion, for example) gor the ur oses of od c ar'f'cat'o onl then a descriptive model of the simulated structure which includes gap and friction elements is shown in Figure 6.6.1.

The impacts between fuel assemblies and rack show up in the gap elements, having local stiffness KI, in Figure 6.6.1. In Table 6.6.1, gap elements 5 through 8 are for the vibrating mass at the 6-14

'top of the rack. The support leg spring rates KS are modeled by elements 1 through 4 in Table 6.6.1. Note that the local compliance of the concrete floor is included in KS. To simulate sliding potential, friction elements 2 plus 8 and 4 plus 6 (Table 6.6.1) are shown in Figure 6.6.1. The friction of the support/liner interface is modeled by a piecewise linear spring with a suitably large stiffness Kf up to the limiting lateral load, pN, where N is the current compression load at the interface between support and liner. At every time step during the transient analysis, the current value of N .(either zero for'ift-off condition, or a compressive finite value) is computed. Finally, the support rotational friction springs KR reflect any rotational restraint that may be offered by the foundation. This spring rate is calculated using a modified Bousinesq equation and is included to simulate the resistive moment of the support to counteract rotation of the rack leg in a vertical plane. This rotation spring is also nonlinear, with a zero spring constant value assigned after a certain limiting condition of slab moment loading is reached.

The nonlinearity of these springs (friction elements 9, 11, 13i and 15 in Table 6.6.1) reflects the edging limitation imposed on the base of the rack support legs and the shifts in the centroid of load application as the rack rotates. If this effect is neglected, any support leg bending, induced by liner/baseplate friction forces, is resisted by the leg acting as a beam cantilevered from the rack baseplate. This leads to higher predicted loads at the support leg baseplate junction than if the moment resisting capacity due to floor elasticity at the floor is included in the model.

The spring rate KS, modeling the effective compression stiffness of the structure in the vicinity of the support, is computed from the equation:

6-15

I I

W I

I

The spring rate KS, modeling the effective compression stiffness of the structure in the vicinity of the support, is computed from the equation:

1

+ +

1 1 1 KS K1 K2 K3 where:

K1 = spring rate of, the support leg treated as a tension-compression member K2 = local spring rate of pool slab K3 = spring rate of folded plate cell structure above support leg As described in the preceding section, the rack, along with the base, supports, and stored fuel assemblies, is modeled for the general three-dimensional (3-D) motion simulation by a twenty-two degree of freedom model. To simulate the impact and sliding phenomena expected, up to 64 nonlinear gap elements and 16 nonlinear friction elements are used. Gap and friction elements, with their connectivity and purpose, are also presented in Table 6.6.1. Table 6.6.2 lists representative values for a module used in the single rack dynamic simulations.

For the 3-D simulation of a single rack, all support elements (described in Table 6.6.1) are included in the model. Coupling between the two horizontal seismic motions is provided both by any offset of the fuel assembly group centroid which causes the rotation of the entire rack and/or by the possibility of lift-off of one or more support legs. The potential exists for the rack to be supported on one or more support legs during any instant of a complex 3-D seismic event.'ll of these potential events may be 6-16

I I

I I

I I

simulated during a 3-D motion so that a mechanism exists in the model to simulate the real behavior.

6.7 'me'e at'on of the t ons o Mot o 6.7.1 e- 'sto Ana sis Us'n u O'-De ree of Freedo Rack Mode Having 'assembled the structural model, the dynamic equations of motion corresponding to each degree of freedom are written by using Lagrange's Pormulation. The system kinetic energy can be constructed including contributions from the solid structures and from the trapped and surrounding fluid. A single rack is modelled in detail. The system of equations can be represented in matrix notation as:

[Ml (q") - (Q) + (G) where:

[M] total mass matrix; (q) the nodal displacement vector relative to the pool slab displacement; double prime stands for secondary derivations; a vector dependent on the given ground acceleration; a vector dependent on the spring forces (linear and non-linear) and the coupling between masses.

The equation can be rewritten as (q<<) ~ [M~ 1 (Q) + [M)-1 (G)

As noted earlier, in the numerical simulations run to verify structural integrity during a seismic event, the rattling fuel assemblies are assumed to move in phase. This will provide maximum impact force level, and induce additional conservatism in the time-history analysis.

6-17

I I

~ \

I I

I

This equation set is mass uncoupled, displacement coupled at each instant in time, and is ideally suited for numerical solution using a central difference scheme. The proprietary, USNRC accepted, computer program "DYNARACK"* is utilized for this purpose.

Stresses in various portions of the structure are computed from known element forces at each instant of time and the maximum value of critical stresses over the entire simulation is reported in summary form at the end of. each run.

In summary, dynamic analysis of typical multi-cell racks has shown that the motion of the structure is captured almost completely by the behavior of a twenty-two degree of freedom structure; therefore, in this analysis model, the movement of the rack cross-section at any height is described in terms of the rack degrees of freedom (ql(t),...q6(t) and q17-q22(t)). The remaining degrees of freedom are associated with horizontal movements of the fuel assembly masses. In'this dynamic model, five rattling masses are used to represent fuel assembly movement in the horizontal This code has been previously utilized in licensing of similar racks for Enrico Fermi Unit 2 (USNRC Docket No. 50-341),

Quad Cities 1 and 2 (USNRC Docket Nos. 50-254 and 265), Rancho Seco (USNRC Docket No. 50-312), Oyster Creek (USNRC Docket No.

50-219), V.C. Summer (USNRC Docket No. 50-395), and Diablo Canyon 1 and 2 (USNRC Docket Nos. 50-275 and 50-323), St. Lucie Unit I (USNRC Docket No. 50-335), Byron Units I and II (USNRC Docket Nos.

50-454, 50-455), Vogtle 2 (USNRC Docket 50-425), and Millstone Unit 1 (USNRC Docket 50-245), Indian Point Unit 2 (USNRC Docket No. 50-247), among others.

6-18

I I

I l

I

plane. Therefore, the final dynamic model consists of twelve degrees of freedom for the rack plus ten additional mass degrees of freedom for the five rattling masses. The totality of fuel mass is included in the simulation and is distributed among the five rattling masses.

6.7.2 Evaluation of Potential for Inter-Rack Im act Since racks are usually closely spaced, the simulation includes impact springs to model the potential for inter-rack impact. To account for this potential, yet still retain the simplicity of simulating only a single rack, gap elements are located on the rack at the top and at the baseplate level. Fig. 6.5.2 shows the location of these gap elements. The baseplate location is a designated potential impact region, and the impact springs located in this region are expected to register impact loads. However, the impact is disallowed in the cellular region of the racks.

Therefore, the impact springs located at the top must not indicate any loads at any time during the seismic event.

6.8 Structural Acce tance C iteria There are two sets of criteria to be satisfied by the rack modules:

a0 Kinematic C iter'o This'riterion seeks to ensure that the rack is a physically stable structure. The racks are designed to preclude inter-rack impacts in the cellular region.

Therefore, physical stability of the rack is considered along with the criterion that inter-rack impact or rack-to-wall impacts in the cellular region do not occur.

6-19

I l

I

b. Stress Limits The stress limits of the ASME Code,Section III, Subsection NFg 1989 Edition are used. The following loading combinations are applicable (Ref. 6.1.2) and are consistent with the plant UFSAR commitments.

Loadin Combinatio Stress Lim't D + L Level A service limits D + L + To D+ L+ To+

D+ L+ Ta+ E Leve B service mats D + L + To + Pf D + L + Ta + Level D service lmxts E'+L+Fd The functional capability of the fuel racks should be demonstrated.

The abbreviations in the table are'hose used in Section 3.8.4 of the Standard Review Plan and the "Review and Acceptance of Spent Fuel Storage and Handling Applications":

D = Dead weight-induced internal moments (including fuel assembly weight)

L = Live Load (not applicable for .the fuel rack, since there are no moving objects in the rack load path).

Fd Force caused by the accidental drop of the heaviest load from the maximum possible height.

Pf ~ Upward force on the racks caused by postulated stuck fuel assembly E ~ Operating Basis Earthquake (OBE)

E' Design Basis Earthquake (DBE) 6-20

I I

I I

I

Differential temperature induced loads (normal operating or shutdown condition based on the most critical transient or steady state condition).

Ta ~ Differential temperature induced loads (the highest temperature associated with the postulated abnormal design conditions).

The conditions Ta and To cause local thermal stresses to be produced. For fuel rack analysis, only one scenario need be examined. The worst situation will be obtained when an isolated storage location has a fuel assembly which is generating heat at the maximum postulated rate. The surrounding storage locations are assumed to contain no fuel. The heated water makes unobstructed contact with the inside of the storage walls, thereby producing the maximum possible temperature difference between the adjacent cells. The secondary stresses thus produced are limited to the body of the rack; that is, the support legs do not experience the secondary (thermal) stresses. For rack qualification, To, Ta are the same.

6.9 Material Pro erties The data on the physical properties of the rack and support materials, obtained from the ASME Boiler 6 Pressure Vessel Code, Section ZZZ, appendices, are listed= in. Table 6.9.1. Since the maximum pool bulk temperature is less than 200 F, this is used as the reference design temperature for evaluation of material properties.

6-21

I I

I I

I

6.10 Stress Limits for Various Conditions The following stress limits are derived from the guidelines of the ASME Code, Section IZZ, Subsection NF [6.1.2], in conjunction with the material properties data of the preceding section. All parameters and terminology are in accordance with the Code.

6.10.1 No a and U set Cond't'ons Leve A or Le~el B a~ Allowable stress in tension on a net section Ft ~ 0 ' Sy (Sy ~ yield stress at temperature)

Ft = (0 6) (25I000) = 15~000 psi (rack material)

~

Ft = is equivalent to primary membrane stresses Ft = (.6) (25,000) = 15,000 psi (upper part of support feet)

( ~ 6) ( 106/300) 63@780 psi (lower part of support feet)

b. On the gross section, allowable stress in shear iso Fv = .4 Sy

(.4) (25,000) ,10,000 psi (main rack body)

Ft = (.4) (25,000) = ~

10,000 psi (upper part of support feet)

(.4) (106,300) ~ 42,520 psi (lower part of support feet) 6-22

I I

I I

I I

I

c. Allowable stress in compressionI Fa.

[1 -

(kl)2 r2

/2Cc 2

Sy F

E( )+[3(/<<)-[(

5 3

kl r )

kl r )

3

/8Cc 3

where:

(2gz2 E) 1/2 Sy l = unsupported length of component k ~ length coefficient which gives influence of boundary conditions; e.g.

k I 1 (simple support both ends)

= 1/2 (cantilever beam)

= 2 (clamped at both ends)

E ~ Young's Modulus r = radius of gyration of component rack body is based on the full kl/r for the main section height and cross of the honeycomb region.

Substituting numbers, we obtain, for both support leg and honeycomb region:

Fa = 15,000 psi (main rack body)

Fa ~ 15,000 psi (upper part of support feet)

~ 63,780 psi (lower part of support feet) d0 Maximum allowable bending stress at the outermost fiber due to flexure about one plane of symmetry:

Fb ~ 0.60 Sy = 15,000 Psi (rack body)

Fb ~ 15,000 psi (upper part of support feet)

~ 63,780 psi (lower part of support feet) 6-23

I I

I I

I I

I "n" I I

e. Combined flexure and compression:

fa Cmx fbx Cmyfby

+ + < 1 Fa DxFbx DyFby where:

fa Direct compressive stress in the section fbx Maximum flexural stress along x-axis fby Maximum flexural stress along y-axis Cmy = 0.85 fa Dx 1 F'ex fa Dy~ 1 F'ey 12 n2 F'exiey =

23 (

kl

)

x,y and the subscripts x,y reflect the particular bending plane of interest.

f. Combined flexure and compression (or tension):

fa 0.6Sy

+

fbx Fbx

+

fby Fby 1 0 6-24

I I

I I

I

The above requirement should be met for both the direct tension or compression case.

6.'10.2 Leve D Serv ce 'm'ts Section F-1370 (ASME Section III, Appendix F), states that the limits for the Level D condition are the minimum of 1.2 (Sy/Ft) or (0.7Su/Ft) times the corresponding limits for Level A condition. Su is the ultimate tensile 'stress at 200'F per Table 6.9.1. Since 1.2 Sy is greater than 0.7 Su for the lower part of the support feet, the limit is 1.54 for the lower section under DBE conditions. The limit for the upper portion of the support foot is 2.0 under DBE conditions.

Instead of tabulating the results of the different stresses as dimensioned values, they are presented in a dimensionless form.

These dimensionless stress factors are defined as the ratio of the actual developed stress to its specified limiting value. With this definition, the limiting value of each stress factor is 1.0 for the OBE and 2.0 (or 1.54) for the DBE condition.

6.11 Results for the Analysis of Spent Fuel Racks Usin a Sin le Rack Model and 3-D Seismic Motion A complete synopsis of the analysis of the single rack, subject to the postulated earthquake motions, is presented in a summary Table 6.11.1 which gives the bounding values of stress factors Ri (i l... 7). The stress factors are defined as:

R1 = Ratio of direct tensile or compressive stress on a net section to its allowable value (note support feet only support compression)

R2 Ratio of gross shear on a net section in the x-direction to its allowable value 6-25

I I

I I

I I

R3 Ratio of maximum bending stress due to bending about the x-axis to its allowable value for the section R4 Ratio of maximum bending stress due to bending about the y-axis to its allowable value R5 Combined flexure and compressive factor (as defined in 6.10.le above)

R6 Combined flexure and tension (or compression) factor (as defined in 6.10.1f)

R7 Ratio of gross shear on a net section in the y-direction to its allowable value.

As stated before, the allowable value of Ri (i =1,2,3,4g5g6g7) is 1 for the OBE condition and 2 for t e DBE exce t fo the lower section of the su ort where the factor is 1.54 The dynamic analysis gives the maximax (maximum in time and in space) values of the stress factors at critical locations in the rack module. Ualues are also obtained for maximum rack displacements and for critical impact loads. Table 6.11.1 presents critical results for the stress factors, and for rack-to-fuel impact load. Table 6.11.2 presents maximum results for horizontal displacements at the top and bottom of the rack in the x and y direction. For single rack simulations "x" is always the short direction of the rack. In Table 6.11.2, for each run, both the maximum value of the sum of all support foot loadings (4 supports ) as well as the maximum value on any single foot is reported. The table also gives values for the maximum vertical load and the corresponding net shear force at the liner at essentially the same time instant, and for the maximum net shear load and the corresponding vertical force at a support foot at essentially the same time instant.

6-26

I I

I

The results presented in Tables 6.11.1 and 6.11.2 represent the totality of single rack runs carried out. The critical case for structural integrity calculations is included. Displacements at the baseplate level are minimal.

The single rack analysis for run A04 gave the highest stress factors for subsequent structural integrity calculations.

Subsequent to the detailed analysis, pedestals adjacent to the pool walls were relocated from the corner cell to new locations 2 cells inboard from the edge. Since this relocation could affect the conclusions concerning rack structural integiity, the critical case of run A04 was re-considered using the new pedestal locations. The results of that re-analysis are presented in the tables as run A94. The detailed structural integrity computations reported herein are based on the critical case for the loading scenario investigated. Subsequent Whole Pool Multi-Rack analyses are also based on the final pedestal locations.

The results corresponding to DBE give the highest load factors.

The critical load factors reported for the support feet are all for the upper segment of the foot for DBE simulations and are to be compared with the limiting value of 2 '. Results for the lower portion of the support foot are not critical and are not reported in the tables.

Analyses show that significant margins of safety exist against local deformation of the fuel storage cell due to rattling impact of fuel assemblies.

6-27

I I

Overturning has also been considered. This has been done by assuming a multiplier of 1.5 on the DBE horizontal earthquakes k

(more conservative than required by the USNRC Standard Review Plan) and checking predicted displacements. The horizontal displacements do not grow to such an extent as to imply any possibility for overturning.

It is noted that the analyses of the Donald C. Cook plant fuel racks have included some asymetrically loaded racks. The results of these studies can be used as bounding analyses for the case when a rack module is picked up and relocated when loaded asymmetrically with fuel assemblies. The results presented herein indicate that twisting or deformation that would cause loss of function or violation of safety margins will not occur during a planned rack relocation.

6. 12 Im act Anal ses 6.12.1 Im act Loadin Between Fuel Assemb and Ce Wall The local stress in a cell wall is conservatively estimated from the peak impact loads obtained from the dynamic simulations.

Plastic analysis is used to obtain the limiting impact load. The limit load is calculated as 3125 lbs. per cell which is much greater than the loads obtained from any of the simulations.

6.12.2 Im acts Between Ad'acent Racks All of the dynamic analyses assume, conservatively, that the racks are isolated. However, the displacements obtained from the dynamic analyses are less than 50% of the rack-to-rack spacing or rack-to-wall spacing if the pool is assumed fully populated.

6-28

l

~

I

Therefore, we conclude that no impacts between racks or between racks and walls occur during the DBE event. This has been further proven by the Whole Pool Multi-Rack Analysis discussed in Section 6 '4.

6.13 Weld Stresses Critical weld locations under seismic loading are at the bottom of the rack at the baseplate connection and at the welds on the support legs. Results from the dynamic analysis using the simulation codes are surveyed and the maximum loading is used to qualify the welds on these locations.

6.13.1 Base late to Rack Welds and Cell-to-Cell Welds Ref. [6.1.2] (ASME Code Section III, Subsection NF) permits, for the DBE condition, an allowable weld stress t = .42 Su 29 /820 psi. Based on the worst case of all runs reported/ the maximum weld stress for the baseplate to'ack welds is 7605 psi for DBE conditions.

The weld between baseplate and support leg is checked using limit analysis techniques. The structural weld at that location is considered safe if the interaction curve between net force and moment is such that a derived function of F/Fy and M/My is below a limiting value of 1.0.

6-29

,/

FyI My are the limit load and moment under direct load only and direct moment only. F, M are the absolute values of the actual force and moments applied to the weld section. The calculated value is .637 ( 1.0 based'n the instantaneous peak loading. This value conservatively neglects any gussets in place to increase pedestal area and inertia.

The critical area that must be considered for cell-to-cell welds is the weld between the cells. This weld is discontinuous as we proceed along the cell length.

Stresses in the storage cell to storage cell welds develop along the length of each storage cell due to fuel'ssembly impact with the cell wall. This occurs if fuel assemblies in adjacent cells are moving out of phase with one another so that impact loads in two adjacent cells are in opposite directions which would tend to separate the channel from the cell at the weld. The critical load that can be transferred in this weld region for the DBE condition is calculated as 5273 lbs. at every fuel cell connection to adjacent, cells. An upper bound to the load required to be transferred is 593 lbs. Where we have used a maximum impact load of 210 lbs. (obtained from Table 6.11.1), we have assumed two impact locations are supported by each weld region, and we have increased the load by V'2 to account for 3-D effects.

6.13.2 Heatin of an Isolated Cell Weld stresses due to heating of an isolated hot cell are also computed. The assumption used is that a single cell is heated, over its entire length, to a temperature above the value associated with all surrounding cells. No thermal gradient in the 6-30

I l

L

~

~

~

~

vertical direction is assumed so that the results are conservative. Using the temperatures associated with this unit, analysis shows that the weld stresses along the entire cell length do not exceed the allowable value for a thermal loading condition.

Section 7 reports the value for this thermal stress.

6. 14 Whole Poo Mu ti-Rack WPMR Anal s's The single rack 3-D simulations presented in the preceding sections demonstrate the structural integrity, physical stability, and kinematic compliance (no rack-to-rack impact in the cellular region) of the rack modules. However, as noted before, prescribing the motion of the racks adjacent to the module being analyzed introduces an assumption of unpredictable import in the single rack modules. For closely spaced racks, it is possible to demonstrate kinematic compliance only by modelling all rack modules in one comprehensive simulation which is referred to as Whole Pool Multi-Rack (WPMR) model. In the WPMR analysis, DBE seismic load is applied (Ref. 6.1.3) -and all racks are assumed fully loaded with fuel assemblies. The primary intent of the analysis is to confirm structural integrity conclusions from 3-D single rack analysis and to ensure that hydrodynamic effects not able to be modelled in a single rack analysis do not cause unanticipated structural impacts.

The cross coupling effects due to the movement of fluid from one interstitial (inter-rack) space to the adjacent one is modelled using classical potential flow theory and Kelvin's circulation theorem. This formulation has been reviewed and approved by the Nuclear Regulatory Commission, during the post-licensing multi-rack analysis for Diablo Canyon Unit I and II reracking project.

The coupling coefficients are based on a consistent modelling of 6-31

the fluid flow. While updating of the fluid flow coefficients, based on the current gap, is permitted in the algorithm, the analyses here are conservatively carried out using the constant nominal gaps that exist at the start of the seismic event.

Such a comprehensive WPMR model was prepared for the racks shown in the module layout drawing (Fig. 6.4.1). Computer code DYNARACK was used to perform the simulations.

Zn order to eliminate the last significant element of uncertainty in rack dynamic analyses, the friction coef ficient was also ascribed to the support leg/pool bearing pad interface in a manner consistent with Rabinowicz's experimental data [6.4.1]. A set of friction coefficients were developed by a random number generator with Gaussian normal distribution characteristics. These random derived coefficients are imposed on each pedestal of each rack in the pool. The assigned values are then held constant during the entire simulation in order that the results are reproducible.

6.14.1 Multi-Rack Model Figure 6.14.1 shows a planform view of the Donald C.'ook spent fuel pool. A rack and pedestal numbering scheme is set up in the figure. We set up a global x axis towards the East. Table 6.14.1 gives information on the number of cells per rack, and on the rack and fuel weights. All racks are assumed loaded with regular fuel.

There are twenty-three racks in the pool. The cask area in the pool is modelled as a fictitious rack (Rack f24 in Figure 6.14.1).

As noted previously, the presence of a fluid moving in the narrow gaps between racks and between racks and pool walls causes fluid coupling effects which cannot be modelled with a simulation using 6-32

only a single rack. Very simply, a single rack simulation can effectively include only the hydrodynamic effects due to contiguous racks when a certain set of assumptions is used for the motion of contiguous racks. In a multi-rack analysis the far field fluid coupling effects of all racks is accounted for using an appropriate model of the pool-rack fluid mechanics. For Donald C. Cook, the cask area was modelled assuming very large fluid gaps between racks 18 and 24 and between racks 23 and 24.

In the Whole Pool Multi-Rack analysis, used to investigate the interaction effects of all racks, we employ a reduced degree-of-freedom (RDOF) set for each rack plus its contained fuel. The purpose of the whole pool dynamic analysis, including the complete set of racks in the pool, is to determine whether effects, not able to be considered in a single rack analysis, alter any of the conclusions that are based on the results of the 22 DOF single rack analysis. In particular, the multi-rack analysis focusses on displacement excursions of each rack and on pedestal compressive loads. The Whole Pool Multi-Rack analysis is also utilized to investigate the possibility of impacts between racks or between racks and pool walls.

The reduced degree-of-freedom structural model for each rack is developed in a systematic way so that the important kinematic results from a dynamic analysis are in agreement with similar results from a solution obtained using the 22 DOF single rack model. The external hydrodynamic mass due to the presence of walls or ad jacent racks is computed in a manner cons istent with fundamental fluid mechanics principles and the use of a reduced 6-33

I DOF fuel rack model [6.14.1]. The fluid flow model, used to obtain the whole pool hydrodynamic effect is site specific and reflects actual gaps and rack locations.

The whole pool multi-rack model includes many non-linear compression only gap elements. 'here are gap elements representing compression only pedestals (normally four pedestals are assumed for each rack), gap elements describing the impact potential of the fuel assembly-fuel rack interface, and gap elements tracking rack-to-rack or rack-to-wall impact potential at, the top and bottom corners of the rack cell structure. Zn addition to the compression only gap elements,, each pedestal has two friction springs associated with the compression spring. As noted previously, a random number generator is used to establish a friction coefficient, for each pedestal at each instant when the pedestal is in contact, with the liner.

The seismic excitation directions X and Y are shown in Figure 6.14.1. The critical DBE event that governs the behavior of the single rack analysis is applied to the 3-D multi-rack model in the appropriate directions. Three simulations have been carried out using coefficients of friction assumed to be 0.2, to be random with a mean of 0.5 at all pedestals, and to be 0.8, respectively.

6.14.2-- Results of Mult'-Rack Anal sis Tables 6.14.2 6.14.4 show the maximum corner absolute displacements at both the top and bottom of each rack in x and y directions from three multi-rack runs. Zn Table 6. 14. 5, the maximum displacements obtained from the three multi-rack simulations are compared with a single rack analysis. Zn all of 6-34

~

~

5 l

r

these tables, the results for fuel rack 24 can be ignored sine there is no real rack at that Location.

"The absolute displacement values are higher than those obtained from single rack analysis. Thus, it appears essential to perform Whole Pool Multi-Rack analyses to verify that racks do not impact or hit the wall. Pigures 6.14.2 - 6.14.5 show the time history oz rack-to-rack gaps for the critical racks. Zt is shown that the rack-to-rack dynamic gaps are greater than 1. 65 " during a 15 second earthcpxake. Detailed examination of the rack-to-rack dynamic gaps show that the racks primar'y move in-phase in all three simulations. That is, the entire assemblage oz racks tends to move and minimize changes in rack-to-rack gaps.

Table 6.14.5 also presents peak pedestal compressive loads of all pedestals on the twenty-three real racks. l:n addition to a report of maximum pedestal loads, the time history of each pedestal Load for each rack is archived for use in the structural evaluation of the fuel pool slab and, the enveloping walls oz the fuel pool,.

Zt is noted that predicted, maximum pedestal zorce from the multi-rack simulation giving the Lazgest pedestal load (Run MP3 in TabLe 6-14. 5 ) is lower than the result obtained from single rack analysis. The macimum instantaneous vertical foot load obtained from single rack analysis is 183300 Lbs. Prom the Whole Pool Multi-Rack Run MP3, we find a peak single pedestal Load oz 180900 lbs. Because, detailed rack st ess calculations are based on the single rack analysis results, zo new structure concerns are identified by the scoping Whole Pool analysis and the overall structural integrity conclusions are conf~ed.

6-35

6. 15 Bearin Pad Anal sis To protect, the slab from high localized dynamic loadings, bearing pads are placed between the pedestal base and the slab. Fuel rack pedestals impact on these bearing pads during a seismic event and the vertical pedestal loading is transferred to the liner. The bearing pad dimensions are set to ensure that the average pressure impacted to the slab surface due to a static load plus a dynamic impact load does not exceed the American Concrete Institute

[6.15.1] limit on bearing pressures.

The time history results from the dynamic simulations for each pedestal are used to generate appropriate static and dynamic pedestal loads which are used to develop the bearing pad size.

From the whole pool multi-rack analysis, the worst case loading on a pedestal (instantaneous peak load) is 183,300 lbs. (see Table 6.14.5). For a 12" x 12" pad, this gives an average instanteous pressure pa = 1273 psi.

Section 10.15 of [6.15 1] gives the design bearing strength as F

fb = P (.85 fc') C where P ~ .7 and fc'3500 psi for Donald C. Cook. 6 = 1 except when the supporting surface is wider on all sides than the loaded area. In that case, C ~ (A2/A1)' but not more than 2. A1 is the actual loaded area, and A2 is an area greater than A1 which is defined pictorially in the ACI commentary on Section 10.15. For Donald C. Cook, 1 ~ 6 ~ 2; if we conservatively use 6 = 1, then fb 2083 psi which is in excess of the calculated pressure pa.

Thus, significant margin is provided by the bearing pads.

6-36

I References for Section 6 USNRC Standard Review Plan, NUREG-0800 (1981).

ASME Boiler & Pressure Vessel Code, Section III, Subsection NF, appendices (1989).

USNRC Regulatory .Guide 1.29, "Seismic Design Classification," Rev. 3, 1978.

Holtec Proprietary Report - Verification and User's Manual, Report HI-89364, January, 1990.

"Friction Coefficients of Water Lubricated Stainless Steels for a Spent Fuel Rack Facility," Prof. Ernest Rabinowicz, MIT, a report for Boston Edison Company, 1976.

USNRC Regulatory Guide 1.92, "Combining Modal Responses and Spatial Components in Seismic Response Analysis,"

Rev. 1, February, 1976.

"The Component Element Method in Dynamics with Application to Earthquake and Vehicle Engineeringg S.

Levy and J.P.D. Wilkinson, McGraw Hill, 1976.

"Dynamics of Structures," R.W. Clough and J. Penzieng McGraw Hill (1975).

Holtec Proprietary Reports: User's Manual, Report HI-89343, Revision 0; Theory, Reports HI-87162, Revision 1, and HI-90439, Revision 0; Verification, Report HI-87161, Revision 2.

"Dynamic Coupling in a Closely Spaced Two-Body System Vibrating in Liquid Medium: The Case of Fuel Racks,"

K.P. Singh and A.I. Soler, 3rd International Conference on Nuclear Power Safety, Keswick, England, May 1982.

R.J. Fritz, "The Effects of Liquids on the Dynamic Motions of Immersed Solids," Journal of Engineering for Industry, Trans. of the ASME, February 1972, pp 167-172.

USNRC Regulatory Guide 1.61, "Damping Values for Seismic Design of Nuclear Power Plants," 1973.

6-37

I "Fluid Coupling in Fuel Racks: Correlation of Theory and Experiment", by B. Paul, Holtec Report HI-88243.

ACI 318-89, ACI 318R-89, Building Code Requirements for Reinforced Concrete, American Concrete Instituteg Detroit, Michigan, 1989.

6<<38

I I

~

~

4 I

l

'4 I

Table 6.3.1 CORRELATION COEFFICIENT Time Histo Grou DBB OBE N-S "and E-W (1,2) 0.0146 0.1056 N-S to Vertical (1,3) 0.1269 0.0956 E-W to Vertical (2,3) 0.01016 0.1060 6-39

Table 6.5. 1 DEGREES OF FREEDOM Dz.sp acement Rotation Location Ux Uy Uz ex ey ez (Node)

Pl P2 P3 q4 q5 q6 P17 P18 P19 q20 q21 q22.

Point 2 is assumed attached to rigid rack at the top most po'int.

P7 P8 P9 P10 P11 P12 P13 P14 P15 P16 where:

Pi '

qi(t) + U1(t) i ~ 1,7,9,11,13,15,17 qi(t) + U2(t) i 2,8,10,12g14I16g18

~

qi(t) + U3(t) i 3,19

~

Ui(t) are the 3 known earthquake displacements.

6-40

I I

I I

I Ol I

Table 6.6.1 NUMBERING SYSTEM FOR GAP ELEMENTS AND FRICTION ELEMENTS Z. Nonlinear S in s Ga E ements (64 Total)

Number Node Locatio Descr'ion Support S1 Z compression only element Support S2 Z compression only element Support S3 Z compression only element Support S4 Z compression only element 2I2 X rack/fuel assembly impact element 2I2 X 'rack/fuel assembly impact element 2I2 Y rack/ fuel assembly impact element 212 Y rack/ fuel assembly impact element 9-24 Other rattling masses for nodes 1*i 3*i 4* and 5*

25 Bottom cross- Inter-rack impact elements section of rack (around edge)

Inter-rack impact elements Inter-rack impact elements Inter-rack impact elements Inter-rack impact elements Inter-rack impact elements Inter-rack impact elements 44 Inter-rack impact elements 45 Top cross-section Inter-rack impact elements of rack Inter-rack impact elements (around edge) Inter-rack impact elements Inter-rack impact elements Inter-rack impact elements Inter-rack-" impact elements Inter-rack impact elements 64 Inter-rack impact elements 6-41

I I

Table 6.6.1 (continued)

NUMBERING SYSTEM FOR GAP ELEMENTS AND FRICTION ELEMENTS II. Fr'ction Elements (16 total)

Number Node Locat'on Descri tion 1 Support S1 X direction friction 2 Support Sl Y direction friction 3 Support S2 X direction friction

'4 ~

Support S2 Y direction friction 5 Support. S3 X direction friction 6 Support S3 Y direction friction 7 Support S4 X direction friction 8 Support S4 Y direction friction 9 Sl X Slab moment, 10 S1 Y Slab moment 11 S2 X Slab moment 12 S2 Y Slab moment 13 S3 X Slab moment 14 S3 Y Slab moment 15 S4 X Slab moment 16 S4 Y Slab moment 6-42

I I

l I

Table 6.6.2 TYPICAL INPUT DATA FOR RACK ANALYSES (lb-inch units)

Support Foot Spring 4.91 x 106 Constant Ks (0/in.)

Frictional Spring 1.837 x 109 Constant Kf (0/in.)

Rack to Fuel Assembly 1.38 x 105 (x-direction)

Impact Spring Constant (0/in.) 1.61 x 10 (y-direction)

Elastic Shear Spring for 5.986 x 10 (x-direction)

Rack (4/in-) 4.866 x 10 (y-direction)

Elastic Bending Spring 5.458 x 10 (x-z plane) for Rack (0-in/in.) 4.71 x 1010 (y-z plane)

Elastic Extensional Spring 4.074 x 107 (5/in.)

Elastic Torsional Spring 1.322 x 109 (g-in./in.) I Gaps (in.) (for hydrodynamic calculations) 6-43

I I

I I

Table 6. 9. 1 RACK MATERIAL DATA (200'F)

Young's Yield Ultimate Modulus Strength Strength Material E (psi) Sy (psi) Su (psi) 304 S.S. 27.9 x 106 25000 71000 Section IXI Table Table Table Reference X-6.0 I-2.2 X-3.2 SUPPORT MATERXAL DATA (200 F)

Material 1 SA-240, Type 304 27.9 x 106 . 25,000 71000 (upper part of support Psi PSi PSi feet) 2 SA-564-630 27 9 x 106 106,300 140,000 (age hardened at PSl PSi PSl 1100'F) 6-44

I I

k

Table 6.11.1 STRESS FACTORS AND RACK-TO-FUEL IMPACT LOAD Rack/Fuel Impact Load Per Cell at Worst Location Along Height Critical Location Run Remarks ~lbs R 1 R2 R3 R5 RT a03 DBE 180.2 ~ 018 .023 .159 . 166 . 198 ~ 231 . 027*

p = o2 182 cells loaded with .274 .074 . 167 .161 . 417 .442 . 078*

reg. fuel a04 DBE 179. 6 .018 .025 . 172 . 178 . 204 .239 . 033 p = o8 182 cells .284 .079 . 214 . 172 .431 . 460 . 095 loaded with reg. fuel a30 p = 0.2 190 . 012 .012 . 090 .073 .109 . 127 . 013 91 cells loaded with . 181 .046 .118 .106 .281 .299 .052 reg. fuel a32 p = 0.2 209.8 .012 .011 . 094 . 090 .109 .127 .015 91 cells loaded with . 176 .049 ~ 112 .110 .271 .289 . 049 reg. fuel a94 Same as a04 174.8 . 018 . 018 .168 ~ 121 .187 . 219 . 032 with reloca-ted pedestals .325 . 056 .250 . 118 .483 .522 .113 Upper values are for rack cell cross-section just above baseplate.

Lower values are for support foot female cross-section just below attachment to baseplate.

I I

I Ii I

I I

Table 6.11.2 Rack Displacements and Support Loads (all loads are in lbs.)

FLOOR LOAD MAXIMUM MAXIMUM SHEAR (sum of all VERTICAL LOAD AND support feet) LOAD COINCIDENT in a rack (1 foot) VERTICAL DX DY**

~lbs. ~lbs. LOAD ~in. ~in.

a03 Full load 3.510x10 1.549x10 30212 (1.511x10 ) ~ 0609 .0562 p = 0' .0084 .0105 DBE, Reg. Fuel a04 Full load 3.510x10 1.605x10 35832(9.791x10 ) .0679 .0583 p = 0.8 .0015 .0012

~

DBE, Reg. Fuel a30 Half load in 1.883x10 1.021x10 20108(1.005x10 ) .0520 . 0450 Pos. x .0010 .0008 p = 0.2 DBE, Reg. Fuel a32 Half load in 1.883x10 9.973x10 19389(9.71x10 ) .0482 .0515 Pos. y .0055 .0080 p = 0' DBE, Reg. Fuel a94 Same as a04 3.508x10 1.833x10 44406 (1. 4829x10 ) . 0678 . 0778 with reloca- . 0014 .0018 pedestals The value in parenthesis is the vertical load at the insta t when the shear load is maximum. The maximum vertical and shear loads generally do not occur at the same instant.

Upper values are top movements; lower values are baseplate movements (not necessarily at the same time).

I I

I

Table 6. 14. 1 RACK NUMBERING AND WEIGHT INFORMATION Rack No. of Weight of Weight of No. ~Ce s Fuel Assembl b.

1 182 25700 1550 2 168 23700 1550 3 168 23700 1550 4 182 25700 1550 5 182 25700 1550 6 182 25700 1550 7 156 22500 1550 8 144 20900 1550 9 144 20900 1550 10 156 22500 1550 11 156 22500 1550 12 156 '22500 1550 13 143 20800 1550 14 132 19300 1550 15 132 19300 1550 16 143 20800 1550 17 143 20800 1550 18 143 20800 1550 19 182 25700 1550 20 168 23700 1550 21 168 23700 1550 22 166 23900 1550 23 120 17700 1550 24* 0 0 0 fictitious 6-47

I I

I

Table 6.14.2 MAXIMUM DISPLACEMENTS FROM WPMR RUN MP1 (Friction Coefficient = 0.2) rack uxt uyt uxb uyb 1 .7004E-01 .7756E-01 .6235E-01 .7303E-01 2 .7506E-01 .5227E-01 .6494E-01 .3936E-01 3 .8464E-01 .7521E-01 .6897E-01 .6619E-01

.5943E-01 ~ 5218E-01 .4960E-01 .3597E-01 5 .5131E-01 .5306E-01 .4290E-01 .4496E-01 6 .6793E-01 .9512E-01 .5135E-01 .9095E-01 7 .4783E-01 .8928E-01 .3978E-01 .7830E-01 8 .4856E-Ol .7065E-01 .3607E-01 .5917E-01 9 .4533E-01 .6377E-01 .3196E-01 .5192E-01 10 .3830E-01 .5754E-01 .2848E-01 . .4354E-01 11 .4224E-01 .5336E-01 .3659E-01 .4329E-01 12 .6411E-01 .9620E-01 .4885E-01 .8429E-01 13 .7253E-01 .1079E+00 .6568E-01 .9505E-01 14 .4602E-01 .1114E+00 .3650E-01 .9847E-01 15 .3557E-01 .1079E+00 .2634E-01 .9325E-01 16 ~ 3467E-01 ~ 9211E-01 ~ 2817E-01 .8608E-01 17 .5755E-Ol .4429E-01 .5326E-01 .3140E-01 18 .1011E+00 .1301E+00 .8596E-01 .9693E-01 19 ~ 6980E-01 ~ 1125E+00 ~ 6341E-01 .8575E-01 20 .8202E-01 .8680E-01 .6878E-01 .6048E-01 21 .8404E-01 1455E+00 ~ 6800E-01 .1229E+00 22 .8173E-01 .1057E+00 .7111E-'01 .9050E-01 23 ,.5647E-01 .6598E-01 ~ 4812E-01 .6156E-01 uxt=absolute value of maximum rack corner displacement in x-direction at rack top; uyt=absolute value of maximum rack corner displacement in y-direction at rack top; uxb=absolute value of maximum rack corner displacement, in x-direction at rack baseplate; uyb=absolute value of maximum rack corner displacement in y-direction at rack baseplate.

6- 48

I I

I I

I

Table 6. 14. 3 MAXIMUM DISPLACEMENTS FROM WPMR RUN MP2

'(Random Friction Coefficient) rack uxt uyt 'uxb uyb 1 .6524E-01 .4772E-01 .3373E-01 .2303E-01 2 .1423E+00 .5829E-01 .1442E+00 ~ 4598E-01 3 .1247E+00 .4122E-01 .1161E+00 .2566E-01 4 .1860E+00 .6628E-01 .1859E+00 .3161E-01 5 .1106E+00 .6379E-Ol .1091E+00 .2673E-01 6 .9642E-01 .7250E-01 .8330E-01 .6348E-01 7 .4742E-01 .6267E-01 .3334E-01 .5443E-01 8 .1801E+00 .5755E-01 .1819E+00 .4534E-01 9 .1275E+00 .3974E-01 .1207E+00 .2115E-01 10 .2336E+00 .7640E-01 .2336E+00 .5527E-01 11 .1710E+00 .8644E-01 .1712E+00 .6245E-01 12 .4015E-Ol .4740E-01 .2869E-01 .2678E-01 13 .1088E+00 .1034E+00 .1030E+00 .1040E+00

~

14 .1439E+00 .4029E-01 .1282E+00 .1865E-01 15 .6218E-01 .5620E-01 .6029E-01 .3386E-01 16 .3322E+00 .5413E-01 .3374E+00 .3677E-01 17 .1727E+00 .5385E-01 .1727E+00 .4896E-01 18 .1269E+00 .1958E+00 .1223E+00 .1913E+00 19 .8411E-01 .8106E-01 .6365E-01 .7508E-01 20 ~ 8402E-01 .6480E-01 .5976E-01 .4419E-01 21 .1280E+00 .4742E-01 .1281E+00 .3530E-01 22 8427E-01 .4951E-01 .7430E-01 .2335E-01 23 .2389E+00 .6471E-01 .2388E+00 .5758E-01 uxt=absolute value of maximum rack corner displacement in x-direction at rack top; uyt=absolute value of maximum rack corner displacement in y-direction at rack top; uxb=absolute value of maximum rack corner displacement in x-direction at rack baseplate; uyb=absolute value of maximum rack corner displacement in y-direction at rack baseplate.

6-49

I I

I I

Table 6.14.4 MAXXMUM DZSPLACEMENTS FROM WPMR RUN MP3 (Friction Coefficient=0.8}

rack uxt uyt uxb uyb 1 .2035E+00 .1702E+00 .1987E+00 . 1774E+00 2 .2751E+00 .5173E-01 -2732E+00 .1658E-01 3 .2637E+00 .5740E-01 -2638E+00 .4010E-01 4 .1363E+00 .5449E-01 .1321E+00 .2788E-01 5 .1333E+00 .8237E-01 1273E+00 .6876E-01 6 .1720E+00 .1514E+00 .1609E+00 .1617E+00 7 .2425E+00 .8747E-01 .2461E+00 .8782E-01 8 .1785E+00 .6039E-01 .1784E+00 .4260E-01 9 .1519E+00 .4434E;01 .1506E+00 .3129E-01 10 .8112E-01 .5007E-01 .7887E-01 .2883E-01 11 .1146E+00 .7975E-01 .1117E+00 .5071E-01 12 .1005E+00 .1602E+00 ~ 9143E-01 .1601E+00 13 .1604E+00 .1310E+00 .1633E+00 .1073E+00 14 .7786E-01 .7618E-01 .7823E-01 .5953E-01 15 .8616E-01 ~ 5521E-01 .8214E-01 .3148E-01 16 .9843E-01 ~ 4 /80E-01 .1024E+00 .2903E-01 17 .8975E-01 .7115E-01 .9063E-01 .7056E-01 18 .1418E+00 ~ 44 16E+00 .1089E+00 .4526E+00 19 .1959E+00 .1720E+00 .1922E+00 .1806E+00 20 .2741E+00 .5563E-01 .2727E+00 .3118E-01 21 .2117E+00 .5159E-01 .2120E+00 -2287E-01 22 .2361E+00 .6081E-01 .2242E+00 .3986E-01 23 .1016E+00 .7703E-01 .1033E+00 .7364E-01 uxt=absolute value of maximum rack corner displacement in x-direction at rack top; uyt=absolute value of maximum rack corner displacement in y-direction at rack top; uxb=absolute value of maximum rack corner displacement in x-direction at rack baseplate; uyb=absolute value of maximum rack corner displacement in y-direction at rack baseplate.

6-50

I I

Table 6.14.5 MAXIMUM RACK DISPLACEMENT AND FOOT LOAD Maximum Maximum Rack Corner Foot Displacement Pedestal Run Remarks inch Force lbs.

a94 Single Rack 0.0778 183,300 Analysis WPMR, p ~ 0.2 0. 1455 (Rack $ 21 in y) 157,400 (Rack $ 19, Foot 4)

WPMR, Random p 0.3322 (Rack i16 in x) 170,900 (Rack $ 19 I Foot 4 )

WPMR, p = 0 ' 0.4416 (Rack $ 18 in y) 180,900 (Rack $ 5, Foot 2) p friction coefficient 6-51

I I

TYPICAL CELL WALLS BASEPLATE BASEPLATE BEARING PAD Figure 6.2.1 Pictorial View of Rack Structure 6-52

I I

I

0.20 0.10 R

Q 0.00 l

Ld 'l LaJ I

>I C3 II C3 0.10 O.

100 100 300 500 700 900 1 100 1300 1500 TIME

• 0.0l sec.

F IGURE 6.3.l DOE t~-S ACCELERATION TIME-flISTORY

I

~

~

~

~

I

0.20 0.10 O. 00 Ld Ld C3 C3

0. 1 0 0

OO

'1 100 300 500 700 90 1 100 1300 1500 TIME

• 0.01 sec.

FIGURE 6.3.2 DBE E-W ACCELERATION TIME HISTORY

I I

0.20 0.10 o 0.00 Ul C3 0.10 O.

100 100 300 500 700 900 1 100 1300 1 500 TIME

• 0.0l sec.

FIGURE 6.3.3 DBE VERTICAL ACCELERATION TIME HISTORY

r~

(~

~

l

0.80 03 O. 60 e O I

CL 0.40 C3

~ 0.20 O.

10 10 FREQUENCY, IIz FIGU RE 6. 3. 4 HORIZONTAI'. DESIGN SPECTRUM AND N-S TIME HISTORY SPECTRUM (5% damping)

~

i

~

~

I 4

0.80

-~ 0.60 I

oI

~ 0.40 LLI C3 C3 0.20 0.

10 10 FREQUENCY, Hz FIGURE 6.3..5 HORIZOtlTAI', DESIGN SPECTRUM AND E-W TIME HISTORY SPECTRUM (5%, damping)

~

~

~

~

~

~

~

~

I

~

L

~

~

I

O. 60 0.40 CO LLI

~ 0.20 C3 0.

10 10 FR EQUENCY, Hz Figure 6.3.6 VERTICAL DESIGN AND TINE HISTORY DERIVED SPECTRA

( 5> damping)

I 0.08.

oi 0.03 Ch I

LJI Z,

~~ O. 02 LLJ C3 C3 0 07 O.

100 100 300 500 700 900 1100 1300 1500 TIME

• 0.01 sec.

FIGORE 6.3.7 OBE N-'CCELERATION TIME HISTORY

I 0.08 cri 0.03 Ch O

w~ 0.02 LLJ 0.07 O.

100 100 300 500 700 900 1100 1300 1500 TI&1E

• G.Ol sec.

FIGURE 6.3.8 OBE E-N ACCELERATION TINE HISTORY

1 4

0.05 0.05 100 300 500 700 900 1100 1300 1500 TItlE

• 0.0l sec.

FIGURE 6.3.9 OBE VERTICAL ACCELERATION TIME HISTORY

0. 60 10 FREQUENCY, Hz Figure 6.3.10 HORIZONTAL DESIGN SPECTRUM AND TIME HISTORY DERIVED N-S SPECTRUM (2% damping)

l

0. 60 Cf)

C3

~" 0.40 I

Ct Ld Ld

~ 0.20 0.

10 10 FRE(FLUENCY, Hz Figure 6.3.ll HORIZONTAL DESIGN SPECTRUM AND E-W TIME HISTORY'S DERIVED SPECTRUM (2% damping)

I O. 50 0.40

(/)

0.30 C)

~ 0.20 LLI 0.10 cn p

10 FREQUENCY, Hz Figure 6.3.12 VERTICAL DESIGN AND TIME HISTORY DERIVED SPECTRA (2% damping)

B,ack Pie Geometric Centerline piz H/q p,Y P16 Long Direction Ps Support f'I Typical Friction Element Figure 6.5.l SCHEMATXC MODEL FOR DYNARACK 6-65

)

Typical Top Impact Element 0

I I

I Rack Structure Typical Bottom Impact Element Figure 6.5.2 RACK-TO-RACK IMPACT SPRINGS 6-66

I

~ CFLL WALL tt>~~~K I xs FU"='SS:-!)'=LY!C"='

IMPACT h."-RlNG I

Ys FIGURE 6.5.3 IMPACT SPRING ARRANGEMENT AT NODE i 6-67

l I

17 20 L

Figure 6.5.4 DEGREES OF PREEDOH HODELLXHG RACK HOTXON

I L

2 FIGURE 6.5.5 RACK DEGREES OF FREEDOM FOR X-Z PLANE BENDING 6-69

I I

18 L

2 L

2 FIGURE 6.5.6 RACK DEGREES OF FREEDOM FOR Y-Z PLANE BENDING 6-70

l l

I

FiJEL ASSY/C-->>

llFACT 'SPRING y,

~

I 0.25H PJ2 XE P 2%g z::c<

,C.G.

H/2 T'PIC):

P i~aT.'iG EL>.o."-z t:-"

FR CTZO.'f S? REYG, .K IH x rZZ.-:C"-

SPRZ,'tG, K f

FOUNDATION ROTATIONAL CRK.EAt>C-SPREiVG, K FIGURE 6.6.1 2-D VIEW OF RACK MODULE 6-71

I I

2 2 2 1 RACK 24 RACK 6 RACK 12 RACK 16 (NOT REAL) 3 3 4 4 2 1 2 2 1 2 1 RACK 5 RACK ll RACK 17 RACK 23 4 3 2 1 2 1 2 2 1 RACK 4 RACK 10 RACK 16 RACK 22 4

2 1 2 2 1 2 RACK ~ RACK 9 RACK 15 RACK 21 3 4 3 2 1 2 2 'I 2 1 RACK 2 RACK 8 RACK 14 RACK 20 4

2 1 2 2 1 2 RACK 1 RACK 7 RACK IS RACK 19 4

FIGURE 6.14.l RACK AND FOOT PEDESDAL NUMBERING FOR D. C. COOK MULTI-RACKMODEL 6-72

I I

l l

l

2.03-COOK POOL MULTI-RACK SEISMIC ANALYSIS, RUN MP2 (Random Cof.)

2.00 RACK 16 TO RACK 17 NORTH CORNER DYNAMIC GAP AT RACK TOP 1.98 x'

1 95 o 1.93 CI 1.90 O

D 188 1.85 1.83 1.80 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 TIME, SEC.

FIGURE 6.14.2

I I

2.03 COOK POOL MULTI-RACK SEISMIC ANALYSIS, RUN MP2 (Random Cof.)

2.00 RACK 16 TO RACK 17 SOUTH CORNER DYNAMIC GAP AT RACK TOP 1.98 1 95 9 1.93 z

O 1.90 O

o188

~

I 1.85 1.83 1.80 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 TIME, SEC.

FIGURE 6.14.3

l' I

I L

t

~li l

i

2.05 COOK POOL MULTI-RACK SEISMIC ANALYSIS, RUN MP3 (Cof.=O.S) 2.00 RACK 12 TO RACK 18 WEST CORNER DYNAMIC GAP AT RACK TOP 1.95 1.90 u 1.85 1.80 o 17 I

1.70 1.65 1.60 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 TIME. SEC.

F IGURE 6.14.4

I I

I I

I I

I I

I I

I

2.05 COOK POOL MULTI-RACK SEISMIC ANALYSIS, RUN MP3 (Cof.=O.S) 2.00 RACK 12 TO RACK 18 EAST CORNER DYNAMIC GAP AT RACK TOP 1.95 xO 1 90

~ 1.85 D

1.80 C3 o 1.75 1.70 1.65 1.60 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 TIME, SEC.

FIGURE 6.14.5

I I

I

7.0 CCIDENT ANALYSIS AND MISCELLANEOUS STRUCTURAL EVALUATIONS 7.1 Introduction This section provides results of accident analyses performed to demonstrate regulatory compliance of the new fuel racks.

There are several types of accidents which could potentially affect the spent fuel storage pool. Installation of the proposed high density racks will enable the storage of increased amounts of spent fuel in the Donald C. Cook spent fuel pool. Accordingly, accidents involving the spent fuel pool have been evaluated to ensure that the proposed spent fuel pool modification does not change the present degree of assurance to public health and safety. The following accidents and miscellaneous structural evaluations have been considered:

Refueling accident Dropped Fuel Local Cell Wall Buckling Analysis of Welded Joints due to Isolated Hot Cell Crane Uplift Load 7~2 efuelin Accidents This section considers three (3) accidents associated with the handling of fuel assemblies.

7~2~1 Dro ed Fuel Assembl The consequences being moved over of dropping a new or spent fuel assembly stored fuel is discussed below.

as it is a~ D o ed Fue Assemb Acc'dent A fuel assembly is dropped from 36" above the top of a storage location and impacts the base of the module.

Local failure of the baseplate is acceptable; however, the rack design should ensure that gross structural failure does not occur and the subcriticality of the adjacent fuel assemblies is not violated. Calculated 7-1

I I

I I

I I

results show that there will be no change in the spacing between cells. Local deformation of the baseplate in the neighborhood of the impact will occur, but the dropped assembly will be contained and not impact the liner. We show that the maximum movement of the baseplate toward the liner after the impact is less than 1.52". The load transmitted to the liner through the support by such an accident is well below that caused by seismic loads.

b. Droa ed Fuel Assembl Accident ll One fuel assembly is (assumed dry weight. = 1550 lbs. )

dropped from 36" above the top of the rack and impacts the top of the rack. This is a more severe condition than the currently postulated drop of 1616 lbs. from a height of 15" above the top of the rack. Permanent deformation of the rack is acceptable, but is required to be limited to the top region such that the rack cross-sectional geometry at the level of the top of the active fuel (and below) is not altered. Analysis shows that although local deformation occurs, it is confined to a region above the active fuel area. The region of permanent deformation is to a depth 5.34" below. the top of the rack.

C ~ Droaaed Fuel Assembly Accident XIX.

This postulated accident is identical to (b) above except that the fuel assembly is assumed to drop in an inclined manner on top of the rack. Analyses show that the straight drop case (case b above) bounds the results.

7.3 Local Buckl'n of Fuel Cel Walls This subsection and the next one presents details on the secondary stresses produced by buckling and by temperature effects.

The allowable local buckling'tresses in the fuel cell walls are obtained by using classical plate buckling analys is . The following formula for the critical stress has been used based on a width of cell "b": (See Figure 7.3.1.)

pn2Et2 12 b2 (1 - p2) 7-2

I I

I

where E = 27.9 x 10 psi, p ~ 0.3, (Poison's ratio), t ~ 075 g b = 8.75". The factor p is suggested in (Ref. 7.3.1) to be 4.0 for a long panel.

For the given'data crcr ~ 7411 psi It should be noted that this stability calculation is based on the applied stress being uniform along the entire length of the cell wall.- In the actual fuel rack, the compressive stress comes from consideration of overall bending of the rack structures during a seismic event and as such is negligible at the rack top and maximum at the rack bottom. It is conservative to apply the above equation to the rack cell wall if we compare crcr with the maximum compressive stress anywhere in the cell wall. As shown in Section 6, the local buckling stress limit of 7411 psi is not violated anywhere in the body o f the rack modules, since the maximum compressive stress in the outermost cell is a 3585 psi. (From Table 6.11.1 for R6 ~ .239, the stress at the base of the rack under combined direct plus bending loads is cr ~ R6 x allowable stress).

7.4 Anal sis o Welded Joints in Rack due to Isolated Hot Cell In this subsection, in-rack welded joints are examined under the loading conditions arising from thermal effects due to an isolated hot cell.

A thermal gradient between cells will develop when an isolated storage location contains a fuel assembly emitting maximum postulated heat, while the surrounding locations are empty. We can obtain a conservative estimate of weld stresses along the length of an isolated hot cell by considering a beam strip (a cell wall) uniformly heated and restrained from growth along one long 7~3

I I

I I

I I

'I edge. The strip is subject to a uniform temperature rise hT 59.66 F. The temperature rise has been calculated from the difference of the maximum local water temperature and bulk water temperature in the spent fuel pool. (see Tables 5.5.1 and 5.7.1).

Then, using a shear beam theory, we can calculate an estimate of the maximum value of the average shear stress in the strip (see Figure 7.4.1).

The final result for wall maximum shear stress, under conservative restraint assumptions is given as t

EaT rmax =

.931 where a = 9.5 x 10 6 in/in 'F Therefore, 'we obtain an estimate of maximum weld shear stress in an isolated hot cell as 16984 '

Since this is a secondary thermal stress, it is appropriate to compare this to the allowable weld shear stress for a faulted event r < . 42Su = 29820 psi. In the fuel rack, this maximum stress occurs near the top of the rack and does not interact with any other critical stress.

7.5 Crane U lift Load of 3000 lb.

A local uplift load of 3000 lb. (UFSAR limit is 2950 lb.) will not induce any uplift stresses in the rack which are more severe than the limiting conditions discussed in the foregoing. This choice of load should be an upper bound load on the maximum load that can be applied to a struck fuel assembly during removal.

7-4

I I

I l

I t

~

-I

7.6 References for Section 7 7.3.1 "Strength of Materials", S.P. Timoshenko, 3rd Edition, Part IZ, pp 194-197 (1956).

7-5

I a

b

) cf, FIGURE 7 ~ 3. 1 LOADING ON RACK WALL Heated Cell Wall KA WdF&WWW&WWWA'W e ld Lin e

'FIGURE 7.4.1 Vf ELDED JOINT IN RACK'-6

I 8 ' STATIC AND DYNAMIC ANALYSIS OF FUEL POOL STRUCTURE 8.1 Int oduction The Donald C. Cook spent fuel pool is a safety related, seismic category I, reinforced concrete structure. In this section an abstract of the analysis to demonstrate the structural adequacy of the pool structure is presented. The object of the analysis is to demonstrate the compliance of the pool slab and confining walls to the applicable design codes and to NRC regulations for the condition of increased loadings due to high density fuel storage.

The loading on the pool structure is produced by the following discrete components:

a) Static Load'n

1) Dead weight of pool structure plus pool water (including hydraulic pressure on the pool walls).

2) Dead weight of the rack modules and fuel assemblies stored therein.

b) .D namic Loadin

1) Vertical loads transmitted by the rack support pedestals to the slab during a DBE or OBE event.

2) Inertia loads due to the slab, pool walls and contained water mass which arise during a DBE or OBE event.

c) e al Load'n

1) Mean temperature rise and temperature gradient across the pool slab and the pool walls due to temperature differential between the pool, water and the atmosphere external to the slab and walls.

8-1

I I

The spent fuel pool is analyzed using the finite element method.

The results for the above load components are combined using factored load combinations mandated by NUREG-0800, the Standard Review Plan (SRP), Section 3.8.4 (Ref. 8.1.1). It is demonstrated that for the critical factored load combinations, structural integrity is maintained when the fuel pool is assumed to be fully loaded with high density fuel racks with all storage locations occupied by fuel assemblies. The general purpose finite element code ANSYS (Ref. 8.1.2) is utilized to perform the analysis.

The critical regions examined are the fuel pool slab and the most critical wall sections adjoining the pool slab. Both moment and shear capacities of the critical regions are checked for structural integrity. Also evaluated is local punching integrity in the vicinity of a fuel rack bearing pad. Structural capacity evaluations are carried out in accordance with the requirements of the American Concrete Institute (ACI) (Refs. 8.1.3 and 8.1.4). In this analysis, the load factors of SRP Section 3.8.4 have been used together with the allowable concrete and reinforcement loads as called for by the American Concrete Institute. This constitutes the most conservative approach to the structural qualification of the pool structure based on a static load qualification method.

8.2 Gene al Features o the Mode The fuel pool model is constructed using information from design basis Donald C. Cook auxiliary building structural drawings. A description of the portion of the pool modelled for analysis is given in the following.

8-2

I I

~

~

~

I

~

I L

~

~

~

The fuel pool slab is a 5'-2 1/2" thick reinforced concrete slab with inside dimensions 39'-1 9/16" wide and 58'-3 1/8" long. The slab is located at elevation 600'-605'-2 1/2" and its long direction is aligned along the plant East-West direction. The East edge of the slab has a 5'-2" thick vertical reinforced wall which extends above the slab and is modeled to level 650'. The West edge of the slab has a 6'hick wall from level,605'-2 1/2" to level 650'. The West wall separates the fuel pool from the fuel transfer canal which is not modelled; however, the discontinuity in the wall structure in the center of the West wall is included.

All wall modeling is done to level 650', and we assume free edges at this level. The North wall is a 6'hick wall extending from the slab to level 650'. The South edge of the slab has a 5'hick wall extending up to level 650'. It is clear from the above description that the South wall has the largest length to thickness ratio, and therefore, may represent a limiting condition of structural strength. The foundation mat is at elevation 584'nd the pool slab and upper walls are supported on the foundation mat by walls and columns around the periphery. The North edge of the slab is supported by a continuous 3'-0" thick wall, while the East edge is partially supported along its length by a 2'-6" thick wall. There are three vertical columns located at the Southeast and Southwest corner of the slab, and intermediate along the South edge; There is also a portion of a wall below the South edge at one location. The floor slab has interior vertical support provided by a 2'-0" thick 'ertical wall providing vertical restraint in both the North-South and East-West direction over a substantial length of slab. In addition, there is a 25'pan standard W14 x 158 wide flange beam from the slab North edge supporting wall to give additional pool slab support. This propped beam is skewed toward the East 16'rom the North edge.

8-3

l I

The entire beam (the straight part plus the skewed part) is supported vertically by four TS10" x 10" square tubes. Each tubular column, has also been stiffened by four 8" x 3/8" plates.

Figure 8.2.1 shows a schematic of the above geometry.

The pool slab is assumed to be loaded with 23 high density fuel racks having a total of 3616 cells. For analysis purposes, each cell is assumed to contain a 1550 lb. weight fuel assembly. As noted previously, all, fuel pool walls above the pool slab are assumed to have a free edge at level 650'. Lateral restraint is provided to the vertical walls at certain locations above the This restraint simulates the effect of adjacent structure 605'evel.

which is not included in the modelled envelope. Figures 8.2.2 and 8.2.3 show layouts of the entire 3-D finite element model. The gridwork in different regions shows the totality of elements used.

Shell elements are used to model the slab and walls, while beam elements are used to model the columns.

The finite element model is constructed using the ANSYS classical shell element STIF63 and the beam element STIF44 of the ANSYS finite element code. The shell element thickness in the various regions of the structure is the actual thickness of the structure at the location. The finite element model is prepared for the analysis of both mechanical load and thermal load. The effects of the reinforced concrete (cracked or uncracked) are accounted for in the finite element model by establishing an appropriate effective modulus for each shell element and effective inertias for the column elements. Effective moduli are defined for each local in-plane axis for the shell elements. The different moduli reflect the fact that different reinforcement geometries may be used in perpendicular directions of the plate-like sections when 8-4

I

! ~

I

the different concrete section assumptions (cracked or uncracked) are applied to the slab and walls. Only major reinforcement which affects the plate and shell-like behavior of the structure is incorporated into the definition of the effective moduli; additional, local reinforcement in various areas of the pool structure are neglected in the defining of the effective moduli.

However, such local reinforcement is accounted for in the strength evaluation after results are obtained. The non-homogeneous nature of the reinforcement is taken into account by defining different material types as necessary to reflect the varying values of effective moduli in different regions. The concrete section assumptions (cracked or uncracked) are fully in accordance with the 'equirements of American Concrete Institute (Refs. 8.1.3 and 8.1.4). In accordance with Ref. 8.1.4, we assume uncracked section properties for the mechanical load analyses (including load factors). For the thermal analyses, it is shown that the thermal gradients will always yield a cracked section if the uncracked stiffness is used; therefore, an iterative solution is used to show that cracked section properties should be used for the thermal analyses.

The effective properties for the elements used in the finite element model are calculated using standard procedures for reinforced concrete sections to define equivalent effective homogeneous materials having the appropriate stiffness and strength.

8-5

l 8.3 Loadin Conditions Zn order to evaluate the response due to the different load mechanisms outlined in Section 8.1, the following finite element analyses are carried out. Six loading cases are defined below which enable us to obtain the moments and shears for factored loadings by linear combination.

1. Dead loading from concrete, reinforcement and hydrostatic head.

40'f The loading is applied as a 1.0g vertical gravitational load for the structure and a surface pressure on the slab and walls for the hydrostatic head.

2. Dead loading due to weights of rack plus full fuel load.

These loads are applied as a uniform static pressure applied to the slab.

3. Seismic vertical loading due to racks plus fuel load applied as an effective sustained pressure on the floor slab pedestals. The loading applied is obtained from the

'-D whole pool multi-rack analysis described in Section 6 of this report. From the results of that analysis, we

.take the stored time history of each pedestal load and define an effective sustained pool pressure load which yields the same total impulse over the time duration of the seismic event. The details of developing this effective sustained pressure load are presented later.

We develop effective sustained vertical pressure loads for both OBE and DBE events and then perform appropriate finite element analyses.

4 ~ Seismic horizontal loading due to structure weight (including reinforcement). The loading is applied as a 1g horizontal and vertical acceleration applied to the structure plus a hydrodynamic pressure equivalent to an acceleration of all of the water mass against the weakest wall. The acceleration level is obtained from the applicable response spectra and is taken as the peak g level on the spectra at frequencies above the lowest natural frequency for the structure. A separate ANSYS frequency analysis simulation is carried out to establish the dynamic characteristics of the structure.

'8-6

er II I

Seismic horizontal load due to shear loads from each of the pedestals. This loading is obtained by using the static + effective dynamic loads developed for case 3 above and assuming a coefficient of friction = .8. The direction of these loads is set so as to develop stresses that maximize the load combinations necessary to satisfy structural integrity requirements discussed below. Zn this load case we also impose a lateral pressure on the weakest pool wall to simulate hydrodynamic effects from fluid coupling due to rack motion relative to the wall.

6. A mean temperature rise plus a thermal gradient is applied across the walls and floor slab to simulate the heating effect of the water in the pool. This gradient is calculated based on the maximum wall temperature deduced from the pool bulk temperature calculations for the licensing basis scenarios presented in Section 5 of this report.

For subsequent discussion of structural integrity checks using various mandated load combinations, we refer to the above individual finite element load cases as "case 1-6", respectively.

As noted above, in addition to the static analyses using the developed finite element model, we also perform a frequency analysis of the pool structure assuming that all contained fluid is attached to the pool slab. Uncracked section properties are used here. This frequency analysis is used to determine the lowest pool structural frequencies so as to establish appropriate seismic amplifiers to apply to load cases 1 and 4. These seismic 8-7

l I

I

amplifiers are obtained from the response spectra of the seismic event and multiply the results of load cases 1 and 4 when forming the mandated load combinations.

As noted above, the case 3 loading involves the determination of an effective pressure load to represent the seismic load on the slab due to the racks plus fuel. The method of determination of this effective pressure is described below.

As noted previously, the Holtec 3<<D dynamic simulation code DYNARACK is used to simulate the seismic response of the entire fuel pool containing multiple racks. The vertical load time history from each pedestal on each rack is saved in an archival file. For the pool slab structural analysis, which is based on static analyses, we compute an effective static load increment based on averaging of the time history. Figure 8.3.1 is used to illustrate the concept where the total pedestal load is considered as the static load (Fs in Figure 8.3.1 plus a time varying component). Note that in Figure 8.3.1 a zero load during a portion of the time means that the pedestal has lifted off. We define an effective static load for the purposes of pool static analysis and structural qualification as follows:

a0 Prom the archival pedestal load time history we may, at each point in time, determine the total pool load FT by summing the total loads for each pedestal.

b. At each point in time i, we can define the dynamic load increment for the pool as FT FS = DFi where FS now represents the total static load on the slab. We keep i

track of the number of time points where DFi ) 0.

C ~ An equivalent. static pool load (seismic adder to the static pool load) is defined as SEISMIC ADDER SUMDFI/SUMNIi 8-8

where SUMDFZ is the sum of all of the non zero DF1 and SUMNi is the total number of points in the time history where the dynamic pool load increment is greater than zero.

d~ Zn forming the appropriate load combinations mandated for structural integrity checks, the calculated "seismic adder" divided by the pool area, is used as the effective seismic pressure on the slab.

Of all loading conditions mandated in Ref. 8.1.1, the factored loads which apply to this structure and are deemed critical are:

A. 1 'D + 1 'E B. ~ 75 (1.4D + 1.9E + 1.7To)

CD D + E' To where:

D ~ Dead load E' Design Basis Earthquake (DBE)

E ~ Operating Basis Earthquake (OBE)

To = Steady State Thermal Load The appropriate load cases are formed from the individual finite element analyses as follows:

D ~ case 1 + case 2 E' DBE amplifier x case 1 + DBE amplifier x case 4 + case 3 (for DBE) + case 5 (for DBE)

E ~ OBE amplifier x case 1 + OBE amplifier + case 4 + case 3 (for OBE) + case 5 (for OBE)

To case 6 Load combinations are formed using absolute values where necessary so as to maximize critical stress resultants.

8-9

I I

As noted above, for analysis of fuel pool structural integrity, the seismic amplifiers are based on the peak g level responses at the lowest resonant frequency that are obtained from the plant acceleration response spectrum. We show that this is conservative.

8.4 Results o A al ses The ANSYS postprocessing capability is used to form the appropriate load combinations identified above and to establish the critical bending moments in various sections of the pool structure. The ultimate moments for each section are computed using allowable limit strength levels as described in Ref. 8.1.3.

For Donald C. Cook, the following limit strengths for concrete and for reinforcement are used in the computation of limit (ultimate) moments.

concrete o'c ~ 3500 psi (compression) reinforcement ~ cry ~ 40000 psi (tension/compression)

Zn each section, we define the safety margin for bending as the ultimate bending moment divided by the calculated bending moment (from the ANSYS postprocessing of the required load cases). Table

8. 4 1 summarizes the results obtained from the finite element

~

analyses and shows minimum safety margins on each section of the structure. Note that these are safety margins based on the factored load conditions as mandated in Ref. 8.1.1 and need only satisfy a limit ~ 1.0.

8-10

I l

The floor slab perimeter is also checked against gross shear failure under factored load conditions. Local bearing strength and punching shear calculations are performed in accordance with (Reft 8.1.3) ~

The pool liner is subject to in-plane strains due to movement of the rack support feet during the seismic event. Calculations are made to establish that the liner will not fail due to cyclic straining caused by the rack foot loading. An ANSYS analysis of a liner plate section subjected to vertical and horizontal static pedestal loading is carried out. The time history result for the pedestal loading is then used to evaluate the number of stress cycles to be expected in the liner for each event. The cumulative damage factor (CDF) is computed and shown to be less than 1.0 in critical regions of the liner and attachment locations. The

@umber of stress cycles'sed in the CDF evaluation is based on 1 DBE and 20 OBE events.

Critical regions affected by loading the fuel pool completely with high density racks are examined for structural integrity under bending and shearing action. It is determined that adequate safety factors exist. assuming that all racks are fully loaded with normal (unconsolidated) fuel and that the factored load combinations are'hecked against the appropriate structural design strengths. It is also shown that local frictional loading on the liner results in in-plane stresses that are low enough so that liner fatigue is not a concern.

8-11

a I

8.7 References fo Section 8 8.F 1 NUREG-0800/ SRP for Review of Safety Analysis Reports for Nuclear Power Plants, Section 3.8.4, July 1981.

8.1.2 ANSYS User's Manual, Swanson Analysis Rev. 4.3, 1987.

8.1.3 ACI 318-89, ACI 318R-89, Building Code Requirements for Reinforced Concrete, American Concrete Institute, Detroit, Michigan.

F 1.4 ACI349.1R-80, Reinforced Concrete Design for Thermal Effects on Nuclear. Power Plant Structures, 1981.

8>>12

I I

I I

I I

Table 8.4.1 SAFETY FACTORS FOR BENDING OF POOL STRUCTURE REGIONS REGION FACTOR OF SAFETY Slab 1.23 North Wall East Wall 1.08 South Wall 1.26 West Wall 1.05 The factors of safety have been obtained using conservative assumptions on mechanical and thermal load distribution. They represent factors of safety over the values required by NUREG-0800.

8-13

I I

I

EAST WALL NOR ALL SOUT ALL I

WEST WALL I I

I

~ ~

~ I

~ ~

~ ~

~ ~

I

~ I

~ ~I

~

rI r rr

~e t.

~ ~ r r rt r

~ ~ ~

r r I ~

~

~

I r >r

~

~

r rr I I

~ s

~

I

~

~

I I ~

I r ~ ~

~

~

I r ~ r rt

~ ~

~

I r sr '

~

I r I I ~ ~

~

~

~

~

I I

~

~

r ~

~

~ rr r ~

rt ~

I srt ~

~

rI V

~ ~ ~

r ret aeseeespasoSesptag

~,Srrr I I

~ ~

~

~ ~ I ~ ~ ~ ~

r ~

~

~ ~

~ s

~ r FIGURE 8.2.1 ISOMETRIC VIEW OF COOK SPENT FUEL POOL 8-14

I I

I I

I I

I

FIGURE 8.2.2 OVERALL FINITE MODEL OF COOK POOL TOP VIEN 8-15

I I

I

FIGURE 8.2.3 OVERALL FINITE MODEL OF COOK POOL BOTTOM VIEN 8-16

I I

I I

FIGURE 8.3.1 PEDESTAL LOAD VS. TIME (Positive Load Means Pedestal in Contact with Liner) 8- l7

I I

9.0 RADIOLOGICAL EVALUATION 9.1 Fuel Handlin Accident 9.1.1 Assum tions and Source Term Calculations An evaluation of the consequences of a fuel handling accident has been made for fuel of 5.0 wt4 initial- enrichment burned to 60,000 MWD/MTUg with the reactor conservatively assumed to have been operating at 3411 MW thermal power (38.8 MWD/KgU specific power) prior to reactor shutdown. Except for the fuel enrichment and discharge burnup, the assumptions used in the evaluation are the same as those previously reviewed and accepted by the USNRC.

r,,

As in the, previous evaluation, the fuel handling accident was conservatively assumed to result in the release of the gaseous fission products contained in the fuel-rod gaps of all the rods in the peak-power fuel assembly at the time of the accident. Gap inventories of fission products available for release were estimated using both the assumptions identified in Regulatory Guide 1.25+ and those in NUREG/CR-5009(). NUREG/CR-5009 "'has confirmed that the Reg Guide 1.25 assumptions remain conservative for extended burnup except for I-131, for which the release fraction was reported to be 20% higher.

Most of the gaseous fission products having a significant impact on the off-site doses are the short-lived nuclides of Iodine and Xenon which reach saturation inventori'es during in-core operation.

These inventories depend primarily on the fuel specific power over f

the ew months immediately preceding reactor shutdown. In the highest power assembly, the specific power and hence the inventory of Iodine and Xenon will be directly related to the peaking factor (assumed to be 1.65 per Reg. Guide 1.25).

9-1

I 1

l I

I

The inventory of long-lived Kr-85 (10.73 year half-life), however, is nearly proportional to the accumulated fuel discharge burnup and hence is independent of the peaking factor. Because Kr-85 is a weak beta emitter, it has only a minor impact on off-site doses, primarily affecting the whole-body beta dose. The off-site radiological consequences are dominated by the short-lived radionuclides (which are at saturation concentration independent of fuel burnup). In the present analysis, the calculated doses are higher and more coservative than those of the previous evaluation because (1) the analyses reported here use higher gap inventories based on Reg Guide 1.25 assumptions and (2) the use of the up-dated ORIGEN-2 code<> for calculating the fission product inventories.

Results of the evaluation confirm that the off-site doses remain within the regulatory limits.

The present evaluation uses values for the 2-hour atmospheric dispersion factor (X/Q) and filter efficiencies that have previously been reviewed and accepted. Core inventories of fission products were estimated with the ORIGEN-2 code based upon a reactor power of 3411 MWt and fuel with an initial enrichment of 5.0: U-235 burned to 60,000 NWD/MTU. Calculations were made for 100 hours0.00116 days <br />0.0278 hours <br />1.653439e-4 weeks <br />3.805e-5 months <br /> cooling time as the source term for the fuel handling accident. The release fraction of the core inventories assumed to be in the gap by both the Reg Guide 1.25 and NUREG/CR5009 assumptions are listed in Table 9.1.

The following equation, from -Reg Guide 1.25, was used to calculate the thyroid dose (D) from the inhalation of radioiodine, D= Z FgI;FPBR;(x/Q)

Rads DFp DFg summed over all Iodine radionuclides.

9-2.

~

~

~

I

~

~

~

I

FS fraction of fuel rod Breathing rate =

Iodine inventory in gap 3.47 x 10 cubic space meters per second core Iodine radio-nu-clide inventory at time Dose conversion of the accident (cu- factor (rads/curie) ries) from Reg. Guide 1.25 fraction of core dam-aged so as to release (X/Q) = atmospheric Iodine in the rod gap diffusion factor (1/193) (3. 15 x 10 sec/m )

Core peaking factor (1.65) DFp = effective Iodine decontamination DF< = effective Iodine factor for pool decontamination factor water (= 150) for filters (= 10)

The gap inventories listed in Table 9-1 are the product of I; (core inventory) and Fz (the fraction existing in the gap).

The function used to calculate the external whole body dose from beta (D,) or gamma (D<) radiation in the cloud uses many of the terms defined above and is given by:

D~ = Z 0.23 (x/Q) F P G< E~ and Df Z 0 ~ 25 (X/Q) F P Gi E where G< is the gap inventory of the gaseous radionuclides of Xe and Kr and the functions above are summed over all the noble gases.

E> and E< are the average energies of decay (beta and gamma radiation respectively) for the various radionuclides. These functions assume the noble gas decontamination factors in water and the charcoal filters are 1.0. The gap inventories of radioiodine 9-3

C

~

l

~

~

~

I

make a negligible contribution to the whole body doses, D~ or D<,

because of the large decontamination factors appropriate to the iodines.

9.1.2 Results A summary of the assumptions used to evaluate the fuel handling accident is given in Table 9-2. The minimum time after shutdown when fuel assemblies would be moved was conservatively assumed to be 100 hours0.00116 days <br />0.0278 hours <br />1.653439e-4 weeks <br />3.805e-5 months <br /> as identified in the Technical Specifications. At 100 hours0.00116 days <br />0.0278 hours <br />1.653439e-4 weeks <br />3.805e-5 months <br /> after shutdown, the two-hour dose at the site boundary, for a fuel handling accident releasing all of the gaseous fission product radioactivity in the gaps of all rods in the highest power assembly, are as follows:

Two-Hour Site Boundar Dose NUREG/CR-5009 Reg. Guide Previous Method 1.25 A~nal sis Inhalation thyroid dose = 7.07 Rads 5.97 Rads 2.15 Whole body beta dose, Dp = 0.36 Rads 0.70 Rads Whole body gamma dose, D< = 0.31 Rads 0.58 Rads 0.51 These doses are well within the limits of 10 CFR Part 100 in conformance with the acceptance criteria of SRP 15.7.4. (Rev.1, July 1981) 9-4

9.2 Solid Radwaste The necessity for resin replacement is determined primarily by the requirement for water clarity and the resin is normally changed about once a year. No significant increase in the volume of solid radioactive wastes is expected with the expanded storage capacity.

During reracking operations, a certain amount of additional resins may be generated by the pool cleanup system on a one-time basis (perhaps 10 to 30 cubic feet).,

9.3 Gaseous Releases Gaseous releases from the fuel storage area of the auxiliary building are combined with other plant exhausts. Normally, the contribution from the fuel storage area of the auxiliary building is negligible compared to the other releases and no significant increases are expected as a result of the expanded storage capacity.

9.4 Personnel Ex osures During normal operations, personnel working in the fuel storage area may be exposed to radiation from the spent fuel pool.

Operating experience has shown that the area radiation dose rates, which originate primarily from radionuclides in the pool water, are generally less than 1 mrem/hr but may temporarily increase to 2.5 3 mrem/hr during refueling operations. No evidence has been observed of any crud deposition around the edges of the pool that might cause local areas of high radiation.

9-5

Radiation levels in zones surrounding the pool are not expected to be significantly affected. Existing shielding around the pool (water depth and concrete walls) provide more than adequate protec-tion, despite the slightly closer approach to the walls of the pool+

Typical concentrations of radionuclides in the pool water are shown in Table 9.3. During fuel reload operations, the concentrations will increase due .to crud deposits spalling from spent fuel assemblies and to activities carried into the pool from the primary system. While these effects may increase the concentrations (as much as a factor of 10), the pool cleanup system soon reduces the concentrations to the normal operating range. No evidence has been seen of any significantly higher radiation doses near the edge of the pool that might suggest the accumulation of crud deposits.

Operating experience has shown that there have been negligible concentrations of airborne radioactivity and no increases are expected as a result of the expanded storage capacity. Area monitors for airborne activities are available in the immediate vicinity of the spent fuel pool.

No increase in radiation exposure to operating personnel is expected and, therefore neither the current health physics program nor the area monitoring systems need to be modified.

9.5 Antici ated E osure Durin Rer ackin Total occupational exposure for the reracking operation is estimated to be between 6 and 11 person-rem, as indicated in Table 9.4. While individual task efforts and exposures may differ from those in Table 9.4, the total is believed to be a reasonable estimate for planning purposes. Divers will be necessary to remove 9-6

l 1,

J

~ .

certain underwater appurtenances. These appurtenances are well removed for the stored fuel which minimizes the radiation dose rate to the divers. Careful monitoring and adherence to pre-prepared procedures will assure that the radiation dose to the divers will be maintained ALARA. All of the reracking operation will utilize detailed procedures prepared with full consideration of ALARA principles. Similar operations have been performed in a number of facilities in the past and there is every reason to believe that reracking can be safely and efficiently accomplished at the Donald C. Cook Nuclear Plant, with minimum radiation exposure to personnel.

The existing radiation protection program at the Cook Nuclear Plant is adequate for the reracking operations. Where there is a potential for significant airborne activity, continuous air samplers will be in operation. Personnel wear protective clothing and, if necessary, respiratory protective equipment. Activities are governed by a Radiation Work Permit and personnel monitoring equipment will be assigned to each individual. As a minimum, this includes thermoluminescent dosimeters and pocket dosimeters.

~ Additional personnel monitoring equipment (i.e., extremity badges or alarming dosimeters may be utilized as required. Work, personnel traffic, and the movement of equipment will be monitored and controlled to minimize contamination and to assure that exposures are maintained ALARA.

In reracking, the existing storage racks will be removed, decon-taminated as much as possible by washing and wipe-downs, packaged and shipped to a licensed processing/disposal facility. Shipping containers and procedures will conform to Federal DOT regulations and the requirements of any State DOT office through which the shipment may pass.

9-7

9.6 References for Section 9 Reg. Guide 1.25 (AEC Safety Guide 25), "Assumptions used for evaluating the potential radiological consequences of a fuel handling accident in the fuel handling and storage facility for boiling and pressurized water reactors".

2 ~ C.E. Beyer, et al., "Assessment of the Use of Extended Burnup Fuel in Light Water Power Reactors", NUREG/CR-5009, Pacific Northwest Laboratory (PNL-6258).

3. A.G. Croff, "A User's Manual for the ORIGEN2 Computer Code",

ORNL/TM-7175, July 1980 (ORIGEN = ORNL Isotope Generation and Depletion)

Section 15.7.4, "Radiological Consequences of Fuel Handling Accidents" NUREG-0800, Section 15.7.4, Rev. 1 July 1981 9-8

I I

I

Table 9-1 INVENTORIES AND CONSTANTS OF SIGNIFICANT FISSION PRODUCT RADIONUCLIDES TOTAL GAP INVENTORY, CURIES SHUTDOWN DECAY NUREG/CR-5009 Reg. Guide 1.25 DOSE CORE NUCLIDE CONST. CONVERSION E (MEV) E (MEV)

INVENTORY P 7 CURIES X, 1/HRS 100 hrs 100 hrs Ri l-131 9.0 E+7 3.591E-3 7.5 E+6 6.3 E+6 1.48E+6 0.186 0.389 l-132 1.3 B.8 3.013E-1 Negligible ~ Negligible 5.35E+4 I-133 1.8 E+8 3.332E-2 6.3 B5 6.3 B5 4 0 E+5 0.419 0.597 I-134 1.9 E+8 7.905E-1 Negligible Negligible 2.5 E+4 I-135 1.7 E+8 1.048 E-1 Negligible Negligible 1.24E+5 0.394 1.456 Kr-85M 1.9 E+7 1.547E-1 Negligible Negligible Kr-85 1.4 E+6 7.376E-6 2.0 E+5 4.2 E+5 0.251 0.002 Kr-87 3.6 E+7 5.451 E-1 Negligible Negligible Kr-88 5.0 E+7 2.442 E-1 Negligible Negligible Xe-131M 1.0 E+6 2.427E-3 7.9 E+4 7.9 E+4 0.163 Xe-133M 5.6 E+6 1.319 E-2 1.5 E+5 1.5 E+5 0.233 Xe-133 1.8 E+8 5.506E-3 5.1 E+6 1.0 E+7 0.102 0.081 Xe-135 3.9 E+7 7.626E-2 Negligible Negligible 0.309 0.262

'O RELEASE FRACTION GIVEN ASSUMED SAME AS REG. GUIDE 1.25

I I

Table 9.2 DATA AND ASSUMPTIONS FOR THE EVALUATION OF THE FUEL HANDLING ACCIDENT Source Term Assum tions VALUES Core power level, MWT 3411 Fuel burnup, MHD/MTU 60,000 Analytical method ORIGEN Release Assum tions Number of failed fuel all rods in 1 rods of 193 assemblies Fraction of core Re . Guide 1. 25 inventory released to gap (NUREG/CR-5009 4 of the Iodine 10 release of Iodine-131 4 of the Xenon 10 is reported to be 20% 4 of Kr-85 30 higher)

Assumed power peaking 1. 65 factor Inventory in gap Table 9.1 available for release Pool decontamination factors For Iodines 150 For noble gases 1 Filter decontamination factors For Iodines 10 For noble gases 1 Atmospheric Dispersion, 3. 15 x 10 sec/m (x/Q)

Breathing rate 3.47 x 10 m /sec 9-10

I l

I

Table 9.3 Typical Concentrations of Radionuclides in the Spent Fuel Pool Water Concentration Nuclide ~C;~ml Ag-110M 4.6 x 10 Co-58 1' x 10 Co-60 4.4 x 10 Cs-134 3.2 x 10 Cs-137 6.4 x 10 9-11

I-I I

Table 9.4 PRELIMINARY ESTIMATE OF PERSON-REM EXPOSURES DURING RERACKING Number of Estimated

~Ste Personnel Hours Ex osure< >

Remove empty racks 40 0.5 to 1.0 Wash and Decon racks 10 0.08 to 0.2 Clean and Vacuum Pool 25 0.3 to 0.6 Remove underwater 0.4 to 0.8 appurtences Partial installation 20 0.25 to 0.5 of new rack modules Move fuel to new racks 150 0.8 to 1.5 Remove remaining racks 120 1.5 to 3.0 Wash and Decon racks 30 0.2 to 0.4 Install remaining new 35 0.4 to 0.8 rack modules Prepare old racks for 80 1.0 to 2.0<'i shipment Total Exposure, person-rem 6 to 12 Assumes minimum dose rate of 2 1/2 mR/hr (expected) to a maximum of 5 mR/hr, except for pool vacuuming operations which assumes 4 to 8 mR/hr and diving operations which assume 20 to 40 mR/hr.

Maximum expected exposure, although details of preparation and packaging of old racks for shipment have not yet been deter-mined.

9-12

l I

10.0 IN-SERVICE SURVEILLANCE PROGRAM 10.1 ~Pur ose This section describes the programmatic commitments made by Indiana Michigan Power Company (I&M) for in-service surveillance of the Boral neutron absorption material to comply with the provisions of Section IV (8) of the OT Position Paper (Ref.

10.1.1).

All material used within a storage system for spent nuclear fuel are qualified to a level of performance predicated upon calculated worst case environmental conditions and are based on accelerated testing of the materials to levels of service life corresponding to that environment. Because such environmental compatibility testing in the laboratory conditions is accelerated, it is prudent that each of the system components be monitored to some extent throughout the service life to assure that the actual in-service performance remains within acceptable parameters as defined by the accelerated testing. For many of the materials, monitoring throughout the service life is relatively easy, however, the neutron absorbing material is encased in a stainless steel jacket precluding a direct visual or physical examination during the in-service condition.

The coupon surveillance program presented herein is intended to provide a definitive assessment of the present physical integrity of the neutron absorber', as well as inferential information to detect future degradation.

10-1

i The coupon surveillance procedure consists of preparing twelve neutron absorber coupons carefully, encased in a stainless steel metal jacket, and suspending them from a "coupon tree".

The coupon tree is placed in the center of a group of freshly discharged fuel assemblies each time a new batch is discharged to the pool. The group of assemblies surrounding the coupon tree shall be those which have the above-average values of radial peaking factor. The object, of course, is to subject this "tree" to the maximum radiation exposure in the fuel pool in the minimum amount of time.

Further details are provided in the following.

10.2 Cou on Surveillance 10.2.1 Descri tion of Test Cou ons The neutron absorber used in the surveillance program shall be representative of the material used within the storage system. It shall be of the same composition, produced by the same method, and certified to the same criteria as the production lot neutron absorber. The sample coupon shall be the same thickness as the neutron absorber used within the storage system and shall meet the referenced Holtec drawing dimensional requirements. Each neutron absorber specimen shall be encased in a stainless steel jacket of an alloy identical to that used in the storage system, formed so as to encase the neutron absorbing material and fix it in a position and with tolerances similar to that for the storage racks. The jacket would be similar to that for the storage racks.

10-2

I l'

l I

~

~

I 1

The jacket would be closed by quick disconnect clamps or screws with lock nuts in such a manner as to retain its form throughout the use period and also allow rapid and easy opening without

. contributing mechanical damage to the neutron absorber specimen contained therein.

Consistent. with the USNRC OT Position Paper [reference 10.1.1],

requirements of a statistically acceptable sample size, a total of twelve jacketed neutron absorber specimens, shall be used.

10.2.2 Benchmark Dat The following benchmark tests shall be performed on test coupons derived from the same production run as the actual neutron absorber panels.

(i) Length, width, thickness and weight. measurements (ii) Wet chemistry (iii) Neutron attenuation measurement (optional) 10.2.3 Cou on Reference Data Prior to encasing the coupons, each coupon shall be carefully calibrated. Their width, thickness, length and weight shall be carefully measured and recorded. The wet chemistry will be performed on a strip taken from the same Boral plates from which the coupons are made to provide a benchmark B-10 loading data.

Three points on each coupon will be designated for neutron attenuation measurement. Neutron attenuation measurements at those three points will be made and recorded.

10-3

I I

I i

I I

I

10.2.4 Accelerated Surveillance At the time of the first off-load of spent fuel, the coupon tree is surrounded by storage cells containing fuel assemblies from the peak power region of the reactor core. At the time of the second off-load of the fuel assemblies, the tree is withdrawn from the fuel pool and one coupon is taken for evaluation. The specimen strip is replaced in the fuel pool in a new location, where it is again surrounded by peak power region fuel assemblies. The storage cell that was vacated may now be used to store a fuel assembly. This arrangement is repeated at the first two off-loads of fuel and after that, every third outage. By evaluation of the specimens, an accelerated monitor of environmental ef fects on the neutron absorber will be obtained.

10. 2. 5 Post-Erradiation Tests Coupons removed from, the pool will be t'es ted for dimensional, neutron attenuation, and wet chemistry changes using the same procedures which were used in initial benchmarking to minimize the potential for instrument errors.

10.2.6 Acce tance Criteria A plant procedure will be developed to execute the commitments made in this licensing submittal. Equipment requirements, step-by-step instructions for executing inspections and acceptance criteria will be described in that procedure for use by plant personnel.

10-4

I I

5 l

References for Sect on 10 OT Position for Review and Acceptance of Spent Fuel Storage and Handling Applications", by Brian K. Grimes, USNRC, April 14, 1978, and Revision dated January 18, 1979.

10-5

I I

I OI

11.0 ENVIRONMENTAL COST BENEFIT ASSESSMENT 11 ~ 1 Intr'oduction The specific need to increase the existing storage capacity of the spent fuel pool at the Donald C. Cook Nuclear Plant is based on the continually-increasing inventory in the pool, the prudent requirement to maintain full-core offload capability, and a lack of viable economic alternatives.

The inventory increase can be inferred from the fuel assembly discharge schedule contained in Table- 11.1.

The proposed project contemplates the reracking of spent fuel pool with free-standing, high density, poisoned spent fuel racks. The engineering design and licensing will be completed for a full reracking of the pool, which is currently only partially racked.

Engineering and design will also be completed to accommodate consolidated fuel. The licensing effort for consolidated fuel will, however, be pursued at a later date if consolidation is chosen to accommodate future storage needs.

11 ' Pro ect Cost Assessme t The total capital cost for the rerack project. is estimated to be approximately $ 14.1 million.

Many alternatives were considered prior to proceeding with reracking, which is not the only technical option available to increase on-site storage capacity. Reracking does, however, enjoy a cost advantage over other technologies, as shown:

11-1

I I

Capital Costs T e of Sto a e K U Rerack g2 0(1)

Fuel consolidation S20 - 34(')

Dry cask storage $ 45 110(')

Storage vault $ 40 90( )

New pool $ 115(')

There are no acceptable alternatives to develop off-site spent fuel storag~ capacity for the Cook Nuclear Plant. First, there are no commercial independent spent fuel storage facilities operating in the U.S. Second, the adoption of the Nuclear Waste Policy Act (NWPA) created a de facto throw-away nuclear fuel cycle. Since the cost of spent fuel reprocessing is not offset by the salvage value of the residual uranium, reprocessing represents an added cost for the nuclear fuel cycle which already includes the NWPA Nuclear Waste Fund fees. In any event, there are no domestic reprocessing facilities. Third, I&M has no other operating power plant; therefore, shipment of spent fuel from the Cook Nuclear Plant to other system nuclear power plants is not possible. Fourth, at

$ 600,000 per day replacement power cost, shutting down the Cook Nuclear Plant is many times more expensive than simply reracking the existing spent fuel pools.

From EPRI NF-3580, May 1984 From DOE RW-0220, "Final Version Dry Cask Storage Study,"

February 1989 Actual estimated cost per KgU of storage space gained for this project 11-2

I I

I I

I

Resource Commitment The expansion of the spent fuel pool capacity is expected to require the following primary resources:

Stainless steel 360 tons.

Boral Neutron Absorber 30 tons, of which 30 tons are Boron Carbide Powder and 20 tons are aluminum.

The requirements for stainless steel and aluminum represent a small fraction of total world output of these metals (less than .0001.).

Although the fraction of world production of Boron Carbide required for the fabrication is somewhat higher than that o'f stainless steel or aluminum, it is unlikely that the commitment of Boron Carbide to this project will affect other alternatives. Experience has shown that the production of Boron Carbide is highly variable and depends upon need, and can easily be expanded to accommodate worldwide needs.

11.4 Environment Assessment Due to the additional heat-load arising from increased spent fuel pool inventory, the anticipated maximum bulk pool temperature increases from a previously-licensed 140'F to approximately 160'F, as detailed in the calculations described in Section 5.0 of this report. The resultant total heat-load (worst, case) is 35.5 million BTU/HR, which is less than 0.54 of the total plant heat loss to the environment.

The net result of the increased heat loss and water vapor emission (due to increased evaporation) to the environment is negligible.

11-3

I I

I I

I I

I

Table 11.1 DONALD C. COOK NUCLEAR PLANT HORST CASE SPENT FUEL INVENTORY ASSEMBLIES ItLssTORAB ',

1991 1362 1992 1518 1993 1678 1994 1838 1995 1918 Lose Ml core discharge capability with current capacity 1996 1998 1997 2158 Lose normal discharge capability with current capacity 1998 2318 1999 2318 2000 2478 2001 2638 2002 2798 2003 2798 2004 2958 2005 3118 2006 3198 2007 3278 2008 3438 Lose full core discharge capability with proposed rerack 2009 3598 2010 3678 Lose normal discharge capability with proposed rerack 2011 3758 2012 3918 2013 4078 2014 4158 2015 4351 2016 4431 2017 4624 11-4

I I

I I

I I

I I