ML20235P376

Forwards Final Draft of Sandia Rept Sand 87-0750, Equipment Operability During Station Blackout Events. Info Provided for Consideration in Developing Guidelines for Coping W/ Station Blackout,As Appropriate

ML20235P376
Person / Time
Issue date:	07/15/1987
From:	Serkiz A NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES)
To:	Childers M NORTHEAST UTILITIES
References
NUDOCS 8707200466
Download: ML20235P376 (44)

Text

-O

'- l p s>" " ' co, y,' ,o, UNITr_D STATES e 3 #,f n NUCLE AR REGULATORY COMMISSION

& " .2 W ASHINGTON, D. C. 20$$5 M l 5 1987 e..,* ,

Michael L. Childers Northeast Utilities P. O. Box 270 Hartford, CN 06141-0270 l

SUBJECT:

TRANSMITTAL OF SAND 87-0750 DRAFT REPORT, " EQUIPMENT OPERABILITY DURING STATION BLACK 0UT EVENTS"

Dear Mr. Childers:

i l

Pursuant to our discussions with your NUGSB0 group on July 9,1987, enclosed 1s j l

a drcft copy of Sandia's report SAND 87-0750 dealing with possible analysis j methods for determining if equipment expected to operate during and after a j station blackout can be " reasonably" expectea to survive environments ]

generated as a result of station blackout. This infctmation is provided for your consideration in developing guidelines for coping with station blackout, l as appropriate.

l Yours truly, 1

i Aleck W. Serkiz, Sr. Task Manager  !

1 Reactor and Plant Safety Issues Branch  !

Division of Reactor and Plant Systems, RES 1

cc: W. Minners K. Kniel P. Norian l A. Rubin A. Serkiz l .PDR.'W/Enclosurf I

l t

l l

l 8707DOO466 07071D PDR 1DPRP EXISANL PM C

l .

'*' D Cs l w

l (c W SAND 87-0750

'XX tJM G l c f ~

[Qu lfV Eu T E N ' ENTAL OPERABILITY DURING' ,

STATION BLACKOUT EVENTS Mark J. Jacobus Vernon F. Nicolette Arthur C. Payne, Jr. .

Sandia National Laboratories Albuquerque, New Mexico 87185 Printed:

Sandia Project Monitor:

Nork performed under Sandia. Contract No.

Prepared for Sandia National Laboratories Albuquerque, NM 87185-Operated by Sandia Corporation for the US Department of Energy Prepared for Division of Reactor Syrter,0;f eti -- 6ad O'* i 621**f office.of Nuclear Regulatory Resea.rch US Nuclear Regulatory Commission Mashington, DC 20555 Under Memorandum of Underst.anding DOE 40-550-75 NRC FIN No.

L 1

L_ _ _ _ - . _ _ _ . _ _ _ _ _ _ . _ _ _ _ _

-l l

o ABSTRACT This report is an initial loox at a possible analysis method for determining if equipment expected to operate during and after a station blackout can be " reasonably" expected to  !

survive environments generated as a result of the station l blackout. The general conclusion is . that, given the current qualification levels of representative equipment, " reasonable" assurance of operability should. be demonstrable for most j equipment in most locations.  ;

l l

i l

i

.)

1 1 l j 1 ,

1- j l

I l

1 s

i 1

1 l

iii/iv  ;

~

1 i

i TABLE OF CONTENTS Section Pace ABSTRACT............................................. iii j EXECUTIVE

SUMMARY

.................................... 1 l

1. INTRODUCTION......................................... 3  !

2. DEFINITION OF PROBLEM................................ 3 ,

I

3. SYSTEMS AND COMPONENTS NECESSARY TO COPE WITH A STATION BLACKOUT..................................... 4 l

3.1 Pressurized Water Reactors (PWRs)............... 4 j i

3.1.1 Decay Heat Remova1....................... 4 3.1.2 Primary System Isolation................. 5 3.1.3 Reactor Suberiticality................... 6 3.1.4 Containment Integrity.................... 6 l l 3.1.5 Miscellaneous............................ 6 Systems Needed After Restoration of 3.1.6 AC Power................................. 6 1

3.2 Boiling Water Reactors (BWRs)................... 7 l

l 3.2.1 Decay Heat Remova1....................... 7 )

3.2.2 Primary System Isolation................. 7 j 3.2.3 Reactor Subcriticality................... 8 i 3.2.4 Containment Integrity.................... 8 j 3.2.5 Miscellaneous............................ 8 l i

3.2.6 Systems Needed after Restoration of l AC Power................................. B i i

4. EVALUATION OF ENVIRONMENTAL CONDITIONS............... B 1

l 4.1 Heat Sources.................................... 8 l 4.2 Heat Sinks...................................... 9 4.3 Heat Transfer Processes......................... 10 4.4 Doors and Vents................................. 12 4.5 Energy Balance................... .............. 12 l l 4.6 Analysis Summary................................ 13 i l 4.7 Sample Environment Calculation........... ...... 13 l 4.8 Othat Areas..................................... 15

5. SUGGESTED METHOD TO ASSESS EQUIPMENT OPERABILITY..... 15 5.1 Operability of Previously Qualified Equipment During a Station Blackout....................... 16  !

5.2 Example Determination of Equipment Operability 5.2.1 Specific Components...................... 16 L.2.2 Temperature Calculations for Equipment Locations...................... 17 Y

.-__-______-__-____a

i I

-TABLF OF CONTENTS (Cont.)

Section Pace j 5.2.3 Equipment Survival Temperatures.......... 18 5.2.4 Comparison of Calculated Environments with Survival Limits..................... 19 5.3 Design Characteristics of Equipment.......... . 20 5.4 Additional Information on Equipment Survivability................................... 21'  !

5.5 Factors for Testing'a' Component's Capability.... 22 i l

6. ESTIMATION OF COSTS ASSOCIATED WITH PERFORMING THE ENVIRONMENTAL ANALYSIS............................... 23 6.1 Estimation of Costs to Determine the Critical Components and Heat Loads....................... 23 6.2 Estimation of Costs for Calculating Environments......................... ......,... 24 6.3 Estimation of costs for Comparing Equipment Limits with Calculated Temperatures............. 24 -

6.4 Estimation of Costs to Test, Replace, Or Relocate Equipment.............................. 25 ,

6.5 Estimation of Costs to Modify Procedures........ 25 l 6.6 Estimation of Costs for Performing Cabinet.

Tests............................................ 25 6.7 Independent Review............................... 26

7. CONCLUSIONS AND LIMITATIONS.......................... 26 7.1 Conclusions..................................... 26 7.2 Ltaitations of Analysis......................... 27 REFERENCES............................................... 36 i

vi a J

<. LIST OF FIGURES Section p a q , ..

4 '.1 ' Estimated RCIC Air Temperature vs. Time............ 33 4.2 Estimated' Control Room Air Temperature vs. Time... 34 5.1 ' Proposed-Method for Environmental Analysis........

35 I' j i

c I e 1

I

.I vii

i LIST OF TABLES f:ection Page 4 4.1 Environment Definition Factors.................... 29 I 5.1 Equipment List for Example Plant RCIC System...... 30 I' 5.2 Summary of Equipment Test Conditions for Equipment Located in Harsh Environments.........~. 31 l

i i

l l

1 l

I l

l l

1 l

viii L__.----.---

EXECUTIVE

SUMMARY

This report is an initial look at a possible . analysis method for determining if equipment expected to operate during and l after a station blackout can be " reasonably" expected .to survive environments generated as a result of the station .

blackout. The general conclusion is that, given the current I qualification levels of repreter.tative equipment, "rFasonable" l I assurance of operability should be demonstrable for most equipment in most locations.

The method can be divided into five steps: 1) identify the .

equipment which must function, 2) calculate cabinet and room l heat loads, 3) calculate cabinet and room temperatures,

4) compare the calculations with the equipment qualification,  ;

and 5) test or reanalyze equipment whose qualification does not I meet the desired criteria. Since reasonable assurance' is j expected to be demonstrable for most components, the proposed j calculational techniques are straightforward and contain many j simplifying assumptions in order to allow an initial screening of the components. For the example plant considered, some equipment appears to be marginal in some environments and ,

slightly more refined applications using plant-specific data or  !

tests may be required. I The analysis begins by assuming that a station blackout. occurs at a nuclear power plant operating at 100% power and lasts for a specified period of time. Based on systems analyses, the systems and components necessary to mitigate the. station I blackout are identified. These systems include front-line and support systems for initial mitigation of the station blackout as well as the systems required after restoration of AC power. ,

Next, the temperatures at locations of the " critical" equipment i are estimated using simple heat transfer models for rooms' and l cabinets, assuming that heat is supplied by energized components and hot piping in the rooms. Next, the calculated l temperatures are compared to survival limits of the equipment base ( on manufacturer's specifications and test results, equipment qualification results, fire test data, and any other available data. In this process, the objective was to verify equipment survivability with " reasonable assurance," considered to be somewhat less demanding than the environmental qualification requirements of 10 CPR 550.49. The final step in the analysis, not performed as part of this study, is to test, relocate, or reanalyze equipment which cannot be shown to meet the " reasonable assurance" criteria in the initial screening.

The environments calculated for the sample plant analyzed in this report could resch as high as 250*F inside containment.

l 238'F in the Reactor Core Isolation Cooling (RCIC) room, and 156*F inside a control room cabinet with a 500W load. These temperatures are based on a number of assumptions of the heat transfer processes. room and cabinet dimensions, and heat 1

1

'q

, e' !

Ioads. Consequently they should. 'be interpreted with corresponding care. For most ' equipment - located ~ at the sample plant,. reasonable' assurance.. of equipment ' operability could - be shown. . The~ major exception ~' is - for some electronic equipment under some of the more severe assumed heat loads: f or - these cases, reasonable assurance of equipment operability is not demonstrated in this report. -

. .i i

}

1 1

1 l

l 1

['

I i

1 l

l 2

t

^ - - - - - --m---_-___ __m____-_b

1.. INTRODUCTION The Draft ' Regulatory Guide on Station Blackout [1] contains guidance that equipment needed to function both during and after a station blackout event should be able to operate in the environmental conditions that would be expected as a result of the blackout. Industry commented that this guide is not ~

specific enough in detining the criteria to be met and the types of analyses to be performed [2]. As a result, they are concerned that the requirements are open-ended and the analyses could be very costly if the level of detail in 10 CFR 550.49 reviews are required.

An attempt has been made to address these concerns in two ways: 1) industry has prepared a draft ANS 58.12 standard, sections of which address the environmental issues and describe acceptable criteria for meeting the expected environmental conditions [3), and 2) NRC has asked Sandia National Laboratories Albuquerque (SNLA) to provide: (1) additional information on specific factors which might need to be considered in determining the environmental conditions encountered as a result of the blackout, and equipment operability in such environments, and (2) estimates of costs for analyses and/or tests. Information and currently available data from work done in the fire / equipment qualification programs at SNLA would be used for this task.

The objective of this task is to suggest a method that utilities might use to show that necessary equipment expected to" operate during and after a station blackout can reasonably be expected to do so. The suggested method is applied to a i l sample nuclear power plant as a demonstration of the techniques l involved. This approach implies that the equipment could

" reasonably" be expected to operate for an extended period of I I

time in an environment that: 1) the same or similar equipment has been shown by tests or analyses to be operable in, or 2) for which operating limits on the equipment are not exceeded.

2. DEFINITION OF PROBLEM l It is assumed that a station blackout occurs at a nuclear power plant initially operating at 100% power, and the blackout lasts a specified time (either four or eight hours as determined in Table 1 of the Draft Regulatory Guide on Station Blackout

[1]). During this time, there is no AC power (either offsite or station diesel generators (DGs)), but there may be some AC equipment working if it has its own dedicated DG separate from the station DGs or from inverters powered by station batteries.

In order to mitigate the accident, DC powered equipment or '

independently-fed AC equipment must be available to shutdown the reactor, cool the core, maintain core coolant inventory (i.e., isolate the system so that leakage will not uncover the 3

core for the duration of the blackout or supply additional makeup), and isolate containment. This equipment must be able to operate without any plant ventilation and cooling unless supplied by its own DC or independent AC power source. If equipment operation is required for only a short time (several minutes) after the station blackout, and not later on, there may be no problem with operability since not enough_ time will have elapsed for the environment to degrade equipment.

However, some equipment must continue to operate as the room / cabinet / panel temperatures increase with time at a rate  ;

determined by external or internal heat sources. Unless additional secondary failures occur, we expect these environments to be dry heat outside of the primary containment. The environmental conditions at various locations around the plant have not necessarily been analyzed for this situation because they are in locations which would not be affected by a design basis event (they may have been analyzed i for some other event such as fire). Substantial P0rtions of I this equipment may not be explicitly qualified for the environments which way occur as a result of a station blackout, an event which is beyond the design basis because of multiple failures of safety systems which must occur.

When AC power is restored (either offsite or a station DG) then AC-powered systems and/or the DC or independent AC systems must operate for a sufficient period to bring the plant to a cold shutdown condition. Those non-operating systems which must now start should not have been damaged by the station blackout environment while in the deenergized mode. Also, the appropriate support systems (e.g., room or component cooling) must also be able to operate.

3. SYSTEMS AND COMPONENTS NECESSARY TO COPE WITH A STATION BLACKOUT 3.1 Pressurized Water Reactors (PWRs) 3.1.1 Decay Heat Removal The main system used to mitigate station blackout events in PWRs is the auxiliary / emergency feedwater turbine-driven pump.

This system takes water from a storage tank and injects it into the steam generators where it cools the primary coolant thereby removing the decay heat. The water boils off creating steam to drive the turbine pump and the rest is released to the atmosphere. In order to maintain circulation in the primary system and control the rate of cooldown, the rate of feedwater injection must be controlled and appropriate instrumentation available. Control of the system may be accomplished either by controlling injection flow, steam flow, or pump speed. The method varies at different plants. In any case, the pump and its controlling valves need to operate for the duration of the blackout.

4

In addition, various supporting equipment needs to function.

This includes: 1) DC power-battery, DC bus?s. switches, and cabling: 2) instrumentation - supply tank level: steam generator level and pressure; reactor temperature, pressure, and level: room temperatures, if important; pump ilow rate: and any relay or solid state circuitry necessary to process these signals and allow system control; and 3) any secondary support such as storage tank makeup or pump cooling, if needed.'

3.1.2 Primary System Isolation Reactor coolant inventory must be maintained in order to allow i natural circulation to cool the core and prevent core  !

uncovery. This is accomplished in one of two ways: 1) isolate the system sufficiently so that, even with expected leakege, the level will stay sufficiently high for the duration of the event; and 2) supply some AC-in3ependent makeup capability. In both of these, reactor pressure and level instrumentation would be necessary to monitor the inventory.

For the first method, three different cases can be identified:

1) isolation lines with active and/or passive valves: 2) power or safety relief valves; and 3) reactor coolant pump seals.

In the first case, the active valves are either AC or DC-powered. The AC valves should have failed closed or be manually operable locally within some specified time frame.

The DC valves should be remotely operable or locally operable.  ;

The valve position indication should work at least until the i time the valve is expected to be closed.

In the second case, since most block valves on PORVs are AC powered, if there is a stuck open PORV or SRV, it will not be able to be isolated for the duration of the blackout. While the operator can not directly do anything about this, he needs instrumentation to know if this has occurred so that he can roughly determine how much time he has to restore AC power and supply makeup. The instrumentation required would be down-stream temperature on the discharge lines or valve position.

In the third case, upon loss of AC power, reactor coolant pump seal cooling will be unavailable since most seal cooling is dependent on AC power (i.e., from charging, HPI, or component cooling water systems). Unless the seals have AC-independent cooling, the pump seals will degrade over some unknown period of time. As in the second group, there is not much the operator can do to stop the leakage if it occurs; however, the leak may vary from large to small and may take a long time to develop. The operator needs appropriate instrumentation to determine if a seal leak has occurred, such as: seal inventory presserc or bleedoff flow or temperature.

5

Currently in PWRs, high pressure AC-independent makeup systems are not available so method two can not be performed. However, if they were, then pump, valve, water supply, and instruments- I tion would be needed. l 3.1.3 Reactor Suberiticality  ;

In order to monitor reactor suberiticality, the source range monitors and their power supplies would need to work.

3.1.4 Containment Integrity i

As with the primary system, isolation position monitors on the containment isolation lines would need to work at least until j the lines can be shown to be isolated. In addition, the lines teus t be isolatable upon loss of AC power either remotely or locally. Containment pressure instrumentation is necessary to monitor this. 4 i

3.1.5 Miscellaneous Lighting - DC lighting should last the duration of the blackout.

Room / Cabinet Temperatures - If equipment in certain rooms is j going to be needed, temperature indication may be necessary I unless analysis shows it is not needed or a time-dependent action is prescribed in the procedures (e.g., open doorc or cabinets or establish alternate air or water flow paths on the basis of elapsed time to critical temperature). )

i I 3.1.6 Systems Needed After Restoration of AC Power After AC power is restored, either by recovering offsite power or by starting one or more of the diesel generators, then additional cystems will need to function. Reasons for this are: 1) extended loss of makeup and possible seal or relief valve leaks may require additional primary makeup in a short time frame, 2) provide containment heat removal in order to prevent system or containment failure, 3) provide redundant ways of maintaining the plant in a safe condition, and 4) the additional heat loads due to the energization of AC loads may require ventilation or other support systems to come on relatively soon after AC power is restored in order to prevent equipment damage.

For any (at least one of each type) of the systems required to be re-energized after the blackout, procedural actions should be defined and/or additional analysis should be performed to show: 1) that after the system is re-energized appropriate cooling systems will come on, if they are necessary, in time to prevent any subsequent equipment damage, and 2) that system components would not be damaged by any external heat sources during the blackout. This will be particularly important if 6

any high temperature sources are near temperature-sensitive equipment for which long-term blackout may create adverse environments. The utility should select one preferred method of removing decay heat, supplying primary makeup, and removing containment heat to analyze. This preferred method should be consistent with plant emergency procedures.

~

3.2 Boiling Water Reactors (BWRs) 3.2.1 Decay Heat Removal There are three basic BWR designs in terms of decay heat removal methods during a station blackout. These are: {

1) Isolation Condensers, 2) High Pressure Coolant Injection (HPCI) and Reactor Core Isolation Cooling System (RCIC), and 3)

High Pressure Core Spray (HPCS)/RCIC.

For the isolation condenser, initially only DC-powered return

)

valves need to operate; however, after about two hours, additional secondary shell-side water is necessary. This water is usually supplied by a diesel-driven fire water pump. The return valves need to remain open and the fire pump and supply l

valves need to operate for the duration of the blackout.

l For the HPCI/RCIC case, both systems have a steam-driven turbine pump that is dependent on DC power. These systems take water from the CST tank or suppression pool and deliver it directly to the reactor coolant system. Either pump and its controlling valves need to operate for the duration of the blackout.

For the HPCS/RCIC case.-one AC powered system (HPCS) has its own dedicated diesel generator and support systems such as pump ,

and room cooling. The other system (RCIC) has a steam-driven j turbine pump, and its operation is dependent on DC power. '

Either system needs to operate for the duration of the blackout.

l Alce, the Automatic Depreccurization System (ADS) or the Safety l Relief Valves (SRVs) should be functional so that depressuriza- i tion can be performed in order to remain within the suppression  :

pool temperature limits for the longest time possible. '

In addition, various support equipment is needed such as:

1) DC power, and 2) instrumentation - reactor coolant system parameters, system control devices and equipment room ,

temperature, if important. 1 3.2.2 Primary System Isolation See Section 3.1.2. Primary system isolation is not as critical in BWRs as in PWRs since most decay heat removal systems inject directly into the primary and thus supply makeup for any lost inventory. However, the same types of actions apply in BWRs to 7

l

I T- i 1

I obtain system isolation. BWRs do not have PORVs, only SRVs, j and the SRVs can not be isolated if they stick open. If a '

recirculation pump seal leak occurs, it can possibly be isolated when AC power is restored.

3.2.3 Reactor Subcriticality In order to monitor reactor suberiticality, the source range monitors and their power supplies would need to work.

3.2.4 -Containment Integrity As with.the primary system, isolation position monitors on the containment isolation lines would need to work at least until the lines can be shown to be isolated. 'In addition, the lines must be isolatable upon loss of AC power either remotely or locally. Contain' ment pressure instrumentation is necessary to monitor this. -

3.2.5 Miscellaneous Lighting - DC lighting should last the. duration of the blackout.

Room / Cabinet Temperatures - If equipment in certain rooms is going to be needed, temperature indication may be necessary unless analysis shows it is not needed or a. time-dependent' action is prescribed in the procedures (e.g., open doors .or cabinets or establish alternate air or water flow paths on the basis of elapsed time to critical temperature).

3.2.6 Systems Needed after Restoration of AC Power See Section 3.1.6. Since some systems are known to be close to the suppression pool and other heat sources in some designs and since some systems will be drawing water from initially hot sources, the effects of long-term high temperature from nearby i heat sources and thermal shock from pumping potentially very hot water appears to be relatively more important in some BWR designs.  !

4. EVALUATION OF ENVIRONMENTAL CONDITIONS l 1

This section identifies the important parameters which are necessary to define a typical station blackout' environment.  ;

Ranges of tcaperature, heat flux, or other relevant thermal i effects are discussed. A brief review of an example BWR plant -l is conducted in Section 4.7 to provide rough estimates of these  ;

parameters. l 4.1 Heat Sources j i'

The two primary sources of heat during a station blackout scenario are electrical circuits and hot steam pipes.

l 8 l l

1

.i i

l l

Electrical circuits are present in virtually every room or compartment in the plant, but are largely con:entrated in cabinets in the control room. Hot steam pipes may also be l present in a room (such as RCIC) and may be insulated or bare. I The heat load due to electrical components can be estiniated based on the FSAR for the plant. During a s tation -blackout, j the load on the batteries is typically in the range of 0-100 kW '

for a typical room. The load for a particular compartment can be determined by adding up the power dissipated by all the DC i relays, lights, notors, switches, etc., which are present in l that compartment. A similar procedure can be used to estimate the power dissipated in a particular control console or cabinet. Typical DC power dissipation for cabinets in the control room is in the range of 0 .5 kW.

Residual heat remaining in electrical / mechanical components from normal operation is of minimal concern. Res.idual heat effects are only existent immediately following the station blackout, whereas DC loads remain relatively constant over time. Since the transient under study covers time periods up to eight hours, residual heat is not expected to be of any significance relative to the DC load.

If hot steam pipes are present in a room, they may add signifi- j cant energy to a stagnant environment. A bare (uninsulated) j steam pipe can conservatively be estimated to be at the temperature of the steam in the line. The associated heat flux by natural convection and thermal radiation from the bare pipe  !

to the compartment air and walls is in the range of 5-10 kW/m2 of pipe surface. If the pipe is reasonably insulated, a conservative estimate of the pipe outer surface temperature is 93*C (200*F). The associated heat flux from the insulated i pipe surface is less than 1 kM/m2 of pipe surface. ,These I heat fluxes are discussed in more detail in Section 4.3. j The heating up of adjacent rooms and subsequent conduction of heat through the walls into the compartment of interest chould not represent a significant source of energy. The Fourier 2

number (at/L ) for .6m (2 ft) thick concrete walls after an 8 hour9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> time period is approximately 0.2. For this Fourier number the Heisler charts [4] indicate very little penetration of any thermal effects.

4.2 Heat Sinks Energy dissipated by electrical components is transferred to the ambient air. Some of it is then transferred from the air to the walls, similarly, energy is convected from any hot steam pipes which are present to the ambient air, while some of it is transferred directly to the walls via thermal radiation, 9

For any compartment or room being analyzed, the walls represent a very large heat ~ sink due to their large surface arca and l thermal capacitance. .Using only the walls and ceiling as a ,

heat sink in an analysis of the thermal environment in a l compartment represents a conservative estimate of the actual I beat sink available. No credit is then taken for the floor or  !

any equipment which may be present in the room.

The important parameters are the wall thermal propertiec (thermal conductivity, density, and specific heat) and dimensions (surface area and thickness). For concrete walls, the product of density and specific heat, is approximately 2000 kJ/m 3 .K. The thick walls pep,large and surface area of a typical room (wall volume) multiplied by pep results in a substantial heat sink. A steel cabinet ,

has a pep of approximately -4000 kJ/m 3 *K. The thin walls j and relatively small surface area of a typical cabinet result in a volume of steel which provides a much smaller, heat sink than the concrete walls.

l 4.3 Heat Transfer Processes j l The relevant modes of heat transfer in a station blackout i include natural convection, thermal radiation, and conduction,  !

unless fans or blowers powered by inverters from the batteries are available to provide forced convection. Since we have assumed that no additional failures occur, the environment is assumed to be dry heat, and processes such as condensation and evaporation are neglected. Natural convection occurs as the electrical circuits and . hot steam pipes lose energy to the ambient air. The heated air in turn transports energy to tne walls of the compartment or cabinet being analyzed.

While natural convection is the heat transfer mechanism for transporting thermal energy from the electrical components to the compartment air, it is not feasible to model the temperature of each electrical component in a compartment. For analysis purposes it can be assumed that the electrical energy to the components is immediately transferred or dumped into the air of that particular compartment. The electrical components are in effect assumed to be at a steady state operating condition. This assumption does not affect the maximum compartment air temperature.

It is also appropriate to treat the compartment air as one lumped volume with uniform (although time-dependent) temperature. This is a conservative assumption for any  !

equipment which is not located near the ceiling where a hot l layer would be expected to develop or directly by the heat j source. Possible local hotspots are not accounted for with i t this method, but other areas of conservatism should generally offset this effect.

10 I

)

1 l

i The calculation of natural convection heat transfer rates to or from the walls and from the hot steam pipes requires an appropriate natural convection heat transfer coefficient.

Values of the natural convection heat transfer coefficient for vertical flat plates and cylinders in crossflow can be estimated using correlations from any heat transfer textbook (e.g., [4)). Fo: transient analyses where repeated evaluations of the Rayleigh number are necess,ary, the correlations ~ proposed  ;

by Churchill and Chu [5, 6) are convenient. Typical values of j the natural convection heat transfer coefficient in the l environment of interest are 2-10 W/m2.K.

Thermal radiation may also play an important role in defining l the thermal environment, particularly if bare pipes are present l in a compartment. Since air is not a good absorber of thermal ,

radiation, most of the energy radiated from the steam pipes is l transferred directly to the wall surface if no steam is l present. This radiative exchange can readily be determined if j some estimate of the temperature of the outer surface of the  !

pipe is made. 1 For bare pipes the maximum temperature can be estimated as the  !

temperature of the steam inside. insulated pipes are consider-ably cooler than bare pipes. Assuming a 0.2m (8 in) diameter  ;

pipe with inside steam temperature of 280*C (550*F), and .05m I (2 in) of .04 W/;n K insulation, the maximum outer surface temperature of the insulation can conservatively be estimated i at 93*C (200*F). If the pipe or insulation emissivity and i viewfactor from the pipe to the walls are conservatively {

assumed to be unity, the radiative fluxes from the pipe to the l walls are about 5.0 kW/m2 of pipe surface for a bare pipe. i For an insulated pipe the corresponding radiative flux is less than 0.5 kW/m2 of pipe surface. This neglects any asymmetry l due to the heat source being near a particular wall, but this should not lead to large differences in the average room I conditions.

Because the walls represent a large heat sink, conduction through the walls should be properly modeled. As a conservative assumption, the floor can be treated as being insulated with zero thermal capacity. If a steel cabinet is being analyzed, the Biot number is generally small enough to permit treating the thermal response of the cabinet in a lumped fashion. On the other hand, thick concrete walls have a relatively large Biot nuber and require more effort to analyze. Since the problem is transient in nature with time and temperature dependent heat fluxes, it is probably easiest to model the concrete valin using a one-dimensional finite difference method. Because of the slow thermal response of the wall, the outside (far) boundary condition is not important.

The inside boundary condition should include the convective and radiative fluxes from the air and hot pipes respectively.

11

4.4 Doors and Vents ]

The effects of open doorways in rooms and vents in cabinets are  ;

difficult to estimate quantitatively and to model analytically.

Proper modeling of these effects necessitates involved calculations of the location of the neutral plane and hot layer characteristics. While transport codes (field models_and zone nodels) exist which can accommodate these influences, implementation of such a code usually entails significant time and money and may not be the most efficient or desirable approach.

In cases where a preliminary analysis of a closed compartment or closed cabinet indicates a potential problem, some  !

engineering judgement must be used. For example, if a closed I cabinet can be opened allowing virtually unrestricted air l circulation inside of it, then the electrical components inside j the cabinet are probably acceptable if they are qualified to within the maximum ambient or room temperatures: otherwise, a j test of the cabinet may be required. 1 In cases where opening a door to a room or a cabinet is deemed ,

necessary to protect equipment, such procedures should be l written into the standard operating procedure for a station j blackout. i l

l 4.5 Energy Balance l Once the various heat transfer mechanisms have been identified, l they can be incorporated into energy balances to estimate the '

i change in air or wall temperature. For example, assuming dry air, the rate of change of the air temperature can be analyzed j knowing the heat sources and heat sinks present: l i

l dT peCp *Ve p = Qin - Oout 1

l where p - air density cp - air specific heat V- room or cabinet volume T = air temperature t - time 12

Qin represents the rate of heat addition from electrical energy dissipation and hot steam pipes (convection- only).

Oout is the heat transfer rate to the walls (convection only).

A similar energy balance can be written for the room or cabinet walls. If bare steam pipes are present, radiation ,to the wall's should be included in the Qin term. _

4.6 Analysis Summary The factors which enter into the environment definition as discussed above are summarized in Table 4.1. A range for the relevant thermal properties using the example plant is also indicated. A qualitative estimate of the relative importance of each factor is given on a scale of 1 to 3, where 1 represents a very important factor, 2 a moderately important factor, and 3 represents a slightly important factor.

Summarizing Table 4.1, the power dissipated by elec'trical and mechanical components along with natural convection from hot (bare) steam pipes are the two main contributors to room air heat up. The large surface area, volume, and heat capacity of the concrete walls act as a very large heat sink.

The main weakness in the analysis is the inability to quantitatively estimate the effect of opening doors, opening cabinets, and/or vents on the results (see Section 4.4 for more discussion). However, it is recommended that a' . conservative approach (no open doors or vents) be used to obtain a first estimate of the environment. It is believed that the results of such an analysis will make consideration of the effects of open doors and vents unnecessary for the majority of rooms and cabinets (see the discussion concerning the results for the example plant, Section 5.2).

4.7 Sample Environment Calculation Several sample environments that have been roughly calculated using the example BWR plant will be discussed in this section.

Calculations were performed for the RCIC , room, the control room, and for a control cabinet in the control room. The method of analysis discussed above was used in the calculations. The results of these calculations for changes in temperatures will be used in Section 5.2.2 to determine final component temperature environments. .

Most of the assumptions have been discussed already but will be summarized here. All of the power dissipated'by the electrical components is immediately transported to the ambient air. The floors of the RCIC and control room (and control cabinet) are insulated and have no thermal capacity. There is no flow to or from adjacent rooms (no open doors) or out of the cabinet (no vents). RCIC steam pipes are assumed to be at 288'C (550*F) if 13

bare and at 93*C (2OO'F) if insulated. All air and wall properties are constant with the exception of air density. The validity and/or conservatism of most of these assumptions has been discussed previously.

The initial temperature for the RCIC room and control cabinet calculations was 49'C (120*F) which is a conservative _ estimate of their temperatures during normal operation. For the control cabinet calculations, the external ambient air (control room air) van also ascumed to be at a constant 49'C (120*F). This nay not be conservative due to heating up of the control room air with time, but was selected here for illustration. For the control room calculations, the initial temperature was assumed to be 27*C (80*F).

The following geometry is assumed for the example pla tit calculations. The RCIC room is shaped like a right isosceles triangle with legs of 15.4m (50.5 ft) and a vertical, height of 7.5m (24.7 ft). The_ wall and ceiling surface area (neglectina the is 514md (5530 ft2) and the volume is 892m5 f looftr (31,500 )3 ) .The walls are assumed to be 0.6m (2 ft) thick concrete.

Heat loads in the RCIC toom are provided by the dissipation of electrical power and by hot steam pipes. The electrical /

mechanical components which are sources of energy include the following: emergency bearing oil pump, emergency seal oil pump, RCIC turbine condenser vacuum pump, RCIC turbine condenser condensate return pump, motor-operated valves, emergency lighting and indicating lights. These components represent an estimated 63 kW of power dissipated during the station blackout, with most of the heat load coming from the two oil pumps.

Hot steam pipes 0.2 m (8 in) in diameter and of length equal to twice the length of the room (for conservatism) or 30.8 m (101 ft) were assumed present in the RCIC room. Bare and insulated pipes were modeled, with the insulated pipes measuring 0.3 m (1 ft) in diameter including insulation. The assumed outside pipe temperatures were 288*C (550*F) for the bare pipe and 93*C (200*F) for the insulated pipe as discussed previously.

The concrete walls were modeled using an explicit one-dimensional finite difference method. The concrete walls were represented by 100 uniformly spaced nodes in order to l obtain sufficient resolution of thermal gradients near the wall I

surface. A non-uniform grid could also have been used to save

' computer time and storage. The inside wall boundary condition represented heat transfer by natural convection from the BCIC air and heat transfer by radiation from the hot steam pipes.

The outer wall boundary condition was selected to be natural convection also. The correlations referenced earlier were used to calculate the natural convection coefficients.

14 l

l The results of the RCIC calculations are shown in Figure 4.1.

For the case of bare steam pipes, the maximum predicted RCIC air temperature change was 65C* (118F*) after 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br />. The insulated pipe case resulted in a prediction of a 44C' (80F*)

temperature change after 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />.

A similar calculation was performed for the control-room (no ,

steam pipes present). The control room of the sample plant has a volume of 3135 m3 and a surface area of 1130 m 2 (excluding floor and equipment), with a height of 5.6 m. The maximum 4 anticipated control room load on the batteries was estimated to be about 45 kW per unit. Since two units may share the same control room, a maximum anticipated control room load of 90 kW was used. Calculations were performed for 45 kW and 90 kW power sources, since 90 kW was believed to be a high estimate.

The results are shown in Figure 4.2 and predict maximum control room temperature increases of 16C* (30F*) and 29C* (53F*) for i the 45 and 90 kW loads, respectively, after 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />.

Calculations were also performed for a typical control console or cabinet with no ventilation. Cabinet dimensions of .6 mx ,

l 1.8 m x 2.4 m (2 ft x6 ft x8 ft) were selected along withu a {

wall thickness of 3.2 mm (.125 in). The thin steel walls vere i treated using a lumped capacitance approach (uniform t emp e r -- '

ature). Representative loads of 100, 300, and 500 W were selected. For these three respective loads, the roaximum changes in cabinet air temperature were 4C* (8F'). 11C* (19F*),

and 16C* (28F*), respectively.  !

l 4.8 Other Areas j Other rooms can be handled in a similar fashion to the ones i above. In large rooms other than the RCIC andLeontrol rooms.

the effects of electrical loads will probably not be sig-nificant (except possibly for local effects). Hot steam pipes (uhere present) may be the major source of heat to the room.

P.c at transfer from adjacent rooms (including the primary containment) via conduction through the thick concrete walls is relatively small. In light of these facts, bulk temperature increases in these other areas are expected to be small and not l to challenge electrical equipment therein. Simple calculations such as these discussed above can confirm this assumption.

5. SUGCESTED MSTHOD TO ASSESS EQUIPMENT OPERABILITY this discussion is to propose a method that t

The intent of utilities can apply to provide reasonable. a s r,u r a n c tJ that components required to mitigate a station bluskout can operate in the environmental conditions associated with the event. The major parameter requiring consideration is increased temperature levels associated with the loss.. of AC-powered

.; cooling and ventilation systems. The general ' steps of the process are shown in Figure 5.1. The essence of Figure 5.1 is 15

that the utility must identify the required equipment along with heat loads in rooms and cabinets (or other confined areas) that could affect the required components. Using some combination of test and analysis of both environmental conditions and component survival temperatures, the utility )

would have to demonstrate with reasonable assurance that the required equipment will continue to function during and after a station blackout. It is expected that this process may, in some cases, involve iterations on the testing and analysis until the reasonably demonstrated survival temperatures of the equipment exceed the expected temperature conditions. For many of the required components, it is expected that only the first '

three steps of the process will be required, siuch as, for example, previously qualified components inside containment.

5.1 Operability of Previously Qualified Equipment During a Station Blackout l

Much of the instrumentation required to mitigate a station blackout will likely be located where it could be subjected to design basis accidents (i.e., HELB or LOCA) and will usually be equipment required to be environmentally qualified.

Consequently, the environmental qualification of any such instrumentation in many cases can provide reasonable assurance that the equipment will continue to operate during and after a station blackout. The environmental qualification temperature inside conta#nment typically ranges around 149'C (300*F), above the maximum temperature which might be expected during a station blackout. For example, calculations of the containment  !

temperature for BWRs give typical values of peak temperature of about 121*C (250*F) during an 8-hour station blackout. The qualified equipment inside the containment at the example plant is generally tested to above 149'C (300*F). Although the duration of the environmental qualification might not be quite long enough to envelop the station blackout time in all cases, it does provide reasonable assorance of equipment functionality, in part because of the significant temperature margin in the qualification tests. Based on the above arguments, it should be relatively easy for utilities to give reasonable assurance that previously qualified equipment inside containment will remain functional daring and after a station blackout. For other equipment, both inside and outside containment, environmental qualification may be used in combination with other information to estimate survival temperatures.

5.2 Uxample Determination of Equipment Operability 5.2.1 Specific Components As an example of the methodology involved, primarily in the first three steps of Figure 5.1, the components of the RCIC rystem at the example plant have been considered. From the 16

plant master equipment list, those components which would likely be required f or. mitigation of a station blackout.using the RCIC system were determined. All components listed for the RCIC system were assumed to be . required for system functionality. In. addition to the components on this list, the DC distribution system components and certain instrumentation will be required as outlined in Section 3.2.1. However, only

~

j the RCIC system is considered for this example. Similar analyses can . be performed for other required equipment. The list of equipment. in the RCIC system at the example . plant is shown in Table 5.1. Obviously, cables, terminal blocks, and other interconnecting devices are used to connect the electrical equipment of- interest. Based on the equipment function, postulated failure modes. (open circuits, short circuits, or ' low insulation resistance), and typical survival temperatures [7,8,9), these types of equipment are not_ believed to be vulnerable to a station blackout environment (i.e., dry heat). .

5.2.2 Temperature Calculations for Equipment Locations .

Inside Containment As stated in Section 5.1, analyses of the containment temperature for BWRs give a peak temperature of about 121*C (250*F) during an 8-hour station blackout. These calcu-lations were performed using the Long-Term Accident i' Simulation (LTAS) code [10,11].

RCIC Room The calculated temperature peak for the.RCIC room as described in Section 4.7 is 114*C (238'F)'with uninsulated pipes (49'C (120*F) initial temperature + 65C' (118F*)

increase) and 93*C (200*P) with insulated pipes (49'C (12C'F) initial temperature + 44C' (80F') increase).

Control Room The general control room area temperature .was based on an initial temperature of 24*C- (75'F). This value is considered more representative than the 27'C (80*F) used for the base calculations in Section 4.7, yet still  ;

conservative (typical range is 21-24*C (70-75'F)). The- '

temperature increases calculated -in Section 4.7 are- still appropriate since the heat transfer coefficients are essentially constant' over the . range of interest and the j concrete walls act as a very large heat sink at the initial i temperature. A later calculation with an initial l temperature of 24*C (75'F) verified that indeed the I temperature rise was the same, within better than 0.06*C (0.20*F) for the control room. Thus, temperature increases may. be directly added to the assumed initial temperature to 17 I

._- _. _ . . . . _ . . _ _ _ _ _ . . _ _ _ . _ _ . ._ __ _______m___ _ _

get the final temperature. The same methodology is used below for the cabinet located in the control room with the final temperature taken as the initial control room temperature plus the calculated control room temperature rise plus the calculated cabinet temperature rise. The resulting peak temperatures in the control room are 41*C (105*F) for the 45 kW load (24*C (75*F) initial temperature

+ 16C' (30F*) increase) and 53*C (128'F) for the 9D kW load (24*C (75*F) initial temperature + 29C* (53F*) increase).

Cabinet in Control Room The peak temperature inside a control room cabinet is assumed to be the peak control room temperature plus the cabinet temperature rise (4C' (BF*) for 100 W, llc * (19F*)

for 300 W, and 16C* (28F') for 500 W) as described above.

The various results are summarized as follows:

l l Control Room Load Cabinet Load 45 kW 90 kW 100 W 45*C (ll3*F) 58*C (136*F) 300 W 51*C (124*F) 65*C (147'F) 500 W 56*C (133*F) 69'C (156*F) 5.2.3 Equipment Survival Temperatures l

Equipment Located in Harsh Environments in this discussion, harsh environment is taken to be the severe temperature / pressure / steam conditions associated with design l basis accidents (i.e., LOCA and HELB). Mild environment refers to environments which essentially do not change during a design i basis accident (or change very little). The equipment listed l in the Table 5.1 harsh environment section is all equipment which has been qualified for harsh environments. The peak temperatures to which the equipment was exposed in the cited reference with durations either specified (or estimated if possible) is given in Table 5.2. In general, where the duration is not specified, the information could be obtained from the testing organization. The table does not give all known test information, but it does give a basis from which a determination of survivability with reasonable assurance could be made if additional test information were obtained.

Equipment Located in Mild Environments For this example, only equipment located in the control room will be considered. Similar procedures can be followed for equipment located in other areas. The equipment located in the control room is given in Table 5.1 along with known equipment survival temperatures in Table 5.2. Although the references 18

from the Equipment Qualification Data Bank [12) were not checked to insure applicability (i.e., complete information such as model number, temperature, duration, etc. are not always specified), doing so would be fairly straightforward.

5.2.4 Comparison of Calculated Environments with Survival Limits _

i Because precise heat loads were not obtained for the example j plant, different assumptions give different temperatures in ,

various locations. The equipment located inside containment is assumed qualified above the maximum station blackout tempera- ]'

ture of 121*C (250*F). This could easily be verified by the utility checking their equipment qualification files.

The equipment listed in Table 5.1 for harsh environments is assumed to all be located in the RCIC room. Again, utilities know precise locations. Of the equipment assumed'in the RCIC room, the equipment survival temperatures clearly exceed the calculated environments for the solenoid valves, limit i switches, and the flow transmitter. Additionally, the given l information for the pump motor, the valve operators, and the l S-O-R Vacuum Switch model 6N-AA21, along with some additional 1 information from the manufacturers should provide reasonable assurance of functionality for these items.

The S-O-R flow switches of unknown model number would obviously have to be identified and information obtained for them. '

Functionality of the remaining items, the Barton flow switches and the S-O-R vacuum switch model 54N6-B118-NX, could probably be demonstrated with the given data (and some more detail on the tests) for the case where the temperature in the RCIC room reaches 93*C (200*F) (i.e., insulated pipes). For uninsulated pipes, the analysis predicts a peak temperature of 114*C (238'F); therefore, either additional information on the 1 I

equipment capability would be required (either from literature or testing) or one of the alternatives in Figure 5.1 would have j to be used.  !

For equipment in the control room, all three types of switches have data to demonstrate reasonable assurance of equipment functionality up to the maximum postulated cabinet temperature  :'

of 69*C (156*F).

The available data indicate that meters would not suffer permanent damage from the effects of the maximum cabinet ,

temperature. There is some possibility that a temporary loss i of accuracy (during the high temperature exposure) would cause increased errors. However, this is not expected to be significant because of the design of the meters (i.e., D' Arsonval movement with no electronics). Based on the Sandia test which exposed a meter to temperatures above 70'C (159'F) for 50 minutes (100*C peak (212*F)), reasonable assurance of 19

l 1

l equipment operability at 70*C is indicated. Therefore, for all l

! cabinet conditions including the case of the most conservative I heat loads for both the control room and the cabinet (69'C I

(156*F)), reasonable assurance of equipment operability is indicated.

For the remaining equipment, the flow controller and the signal convertor, the equipment is specified for continuous operation at 49'C (120*F) and storage to 82*C (180*F) [13]. Certainly these items would continue to function at temperatures above 49'C (120*F), but the accuracy and temperature limit are not easy to ascertain. Both items contain a significant amount of

, electronic components that could experience temperature drift j l and affect overall accuracy. The 701 uses about 6 watts of power and the 750 uses about 4 watts of pcwer [13]. Both these values are sufficient to cause localized hot spots on the order of the storage temperature limits. Consequently, using the storage temperature limit as an upper limit may be non-conser-vative. The 49*C (120*F) value would obviously be conservative, with reasonable aFsurance indicated for temperatures 6-17'C (10-30*F) higher. Using a middle value of 11*C (20*F) gives a limit of 60*C (140*F) ambient for continued operation during a station blackout. Therefore, these items would be considered .

acceptable for all cases except the 90 kW control room load with a cabinet load of 300 W or 500 W in the sample cases. If either of these loadings vera the correct case, additional work would have to be done as outlined in Figure 5.1.

5.3 Design Characteristics of Equipment ,

Equipment design characteristics can vary widely among different manufacturers and hence little can be said in general terms about specific design characteristics of equipment.

However, many equipment failure modes will be similar, although little can be said definitively about the precise failure temperatures of equipment by different manufacturers (except where test data exists). This fact is emphasized by recently completed Sandia tests [14] in which two Agastat relays failed in the temperature range of 160-210*C (320-410*F), but a General Electric relay survived to a temperature of about 350*C (662*F). Specific relay design characteristics, primarily different materials, caused the different failure temperatures. Consequently, establishing actual failure temperatures is virtually impossible for generic component types. However, survival temperatures may be estimated to give an indication where some designs of a component type might be expected to start showing indications of failure. This I estimation is provided in the next section. The potential I

failure modes of electrical equipment exposed to only high temperatures may be summarized by the following:

l 20

F

a. Melting leading to ' binding, . warpage, open circuits, or other effects.'which could. cause a loss of operation, indication, or accuracy from an instrument.- This failure mode would be primarily a function of materials of construction.

b. Leakage currents or shorts : leading to problems such_as spurious operation of equipment,- incoYrect instrument readings, fuse failures or breakers tripping. This failure mode would be dependent primarily on the capability of insulation materials used for isolation.

i

c. Electronic failures (including drift) leading to L incorrect readings,. temporary component i malfunction, fuse failure.. or- the: eventual complete failure of the device. This. failure mode.

would be' primarily a function.of the capabilities of individual electronic components and their' interrelationship in a circuit.

Of these failure modes, electronics- f ailures in . circuits is probably the greatest unknown because of the potential variability of component usage in circuits. Additionally, this failure mode would be expected to dominate . failures of components which use electronics. As an example of how circuit usage 'may affect failures, consider the characteristics of. a transistor which might be used in a circuit.. The material used in the transistor, typically silicon, has a relatively high survival temperature. The characteristics of .the transistor might be known to a temperature significantly above the temperature range of interest. Two of the many ' paramet ers which would be affected by the increased temperature would be current gain and base-emitter leakage current. In one circuit, the effects of these changes may be neglected if the circuit design allowed for wide tolerances in the transistor parameters. In another circuit, the changed parameters may lead to undesired secondary effects which may be very difficult to-predict.

5.4 Additional Information on Equipment' Survivability.

The following list . gives approximate survival temperature for various types of equipment.. No references are provided because this is only intended as a fi'rst approximation . which .may be used in a ger,eric f ashion together with the sample calculations of environments to gain a feel for the scope of potential effects of a station blackout on equipment.- Consequently, only ,

generic types of equipment are given. This list is pn .

expected to meet the reasonable assurance criteria. because: of the amount of estimation used to develop the table and *,he wide variations between - dif f erent - types of equipment. The values provided in the table are generally estimates of the level-21

_l

'd where the components might ' f ail (if this can be estimated)-or of survival temperatures (where failure temperatures cannot' be estimated or where. survival temperatures significantly exceed the worst case' expected temperatures during a station blackout). The list was' developed assuming industrial grade equipment for use outside containment. This table is 'in contrast to the specific example described earlier which was intended to provide reasonable assurance. Consequently, some of the' temperatures given here are higher. than those for reasonable assurance.

Estimated Failure.or Survival Temperature Range For. Durations l . Type of Component of Approximately 8 Hours Cables >300*F Solenoid Valves 250*F Temperature Switches 200-250*F Thermocouple /RTDs >300*F Motor Control Centers: 200-250*F .

Transformers 200-250*F Batteries 200*F Battery Chargers / Inverters 150*F Distribution Panels 200-250*F Recorders 150-200*F Controllers 150-200*F Power Supplies 150-200*F Logic Equipment 150-200*F Indicating Lights 200-250*F Meters 250*F Switches 250*F Electromechanical Relays 300*F Solid State Relays 150-200*F Limit Switches 250*F Valve Operators 250-300*F Transmitters 150-2OO*F Pressure Switches 250-300*F Motors 200-250*F Terminal Blocks >300*F 5.5 Factors for Testing a Component's capability The factors in testing a components's capability involve-primarily two factors. The. first is the creation of appropriate test conditions.. In general,. either steady state testing at the peak postulated temperature for the time required or transient. testing using the actual postulated profile would be required. Temperature rates of change are expected to be relatively slow, making steady state testing an acceptable choice. The second factor in testing is appropriate functional testing of the components. Similar to qualification testing, the functional tasts run during testing should be adequate to demonstrate that the component will perform its 22 1

l l

_-___----____-_a ---_ __-- __ __--____ _ -_ _ - - _ . _ _ _ - _ _ _ _ _ _ _ _ _ _ - - - . _ . _ _ _ _ _ - - _ _ _ _ - _ _ _ _ _ _ _ _ - - - -

n intended function during a station blackout. The functional tests should include appropriate accuracy requirements where  ;

this is a consideration. Any degradation associated with the station batteries should also be considered in functional testing, such as reduced (or increased) voltage levels associated with battery room temperature and battery depletion. As an example of appropriate functional t.ests, the function of a relay is to open or close contacts in response to an input signal. Therefore, functional tests to be considered would include contact resistances, insulation resistances, pickup voltage, and dropout voltage. Contact resistance would best be monitored by the capability of the relay to carry its normal load or range of loads. Insulation resistance can be checked by monitoring for current leakage at various terminals. Pickup and dropout voltage could be checked several i times during a test using a variable voltage power source.

The testing requirements for other equipment can be similarly I established.

l

6. ESTIMATION OF COSTS ASSOCIATED WITH PERFORMING THE ENVIRONMENTAL ANALYSIS 6.1 Estimation of Costs to Determine the Critical Components i and Heat Loads The problem of determining which systems and components are necessary for responding to a station blackout event is part of the utilities overall response to the station blackout rule.

There should be no incremental cost since the information i

should be directly applicable here.

The determination of heat loads falls into two parts: external l and internal. The external heat loads from steam pipes and '

other nearby heat sources should already be documented as part of the original environmental and accident analysis. These would need to be checked to see if the station blackout event would result in a decreased load.. Possibly, some long-term containment heat-up calculations for the station blackout event would need to be run to determine heat sources for nearby rooms. Use of a cheap, fast code would be preferred (such as LTAS [10] or comparable industry codes).

For internal loading, general loading information should be available from the heating and ventilation (HVAC) sizing calculations done by the Architect Engineer (AE). These loads, however, are most likely with AC power available and some analysis will lihely be necessary to estimate the DC or vital AC loading of each cabinet. A simple scoping calculation is recommended for the first pass, being slightly conservative, since it is expected that for most equipment " reasonable" assurance should be demonstrable with these values.

23 l

L-________--_____

4 It is estimated that it would take two to three man-months to:

1) obtain the initial loading information, 2) perform the analysis to modify these for the station blackout event, and i

3) possibly run a containment heat-up calculation. This estimate will depend on how detailed the initial information basis is. Because of the various licensing requirements, many of the required calculations may already be done to various levels of detail. The computer cost for the heat-up calcula-tion will vary with the code used. Also, some utilities may have already run analyses for this event.

6.2 Estimation of Costs for Calculating Environments If a general method of analysis is developed, it is only a matter of varying the input parameters to obtain the thermal environments in a variety of compartments. The estimated manpower for the development and implementation of a suitable analytical model is about 2 man-months for the initial model.

Subsequent mode); can be made by modifying the initial model and would take about one man-week to to set up, run, and analyze per room. This manpower estimate does not include the time necessa.y to identify critical equipment and what it is presently qualified to (Section 6.1), nor does it include any testing of equipment which may be required if the analysis predicts a potential problem (Section 6.4). The effort needed i to include the opening of doors and cabinets in the standard operating procedure for station blackouts is also not included (Section 6.5).

If a simple computer program is developed and used, the total computer costs should be minimal (<$500). Such a computer program should easily fit and run on a personal computer.

6.3 Estimation of costs for Comparing Equipment Limits with Calculated Temperatures Regulations require that detailed equipment qualification records be kept by the utility. For equipment which is not i environmentally qualified, other sources of data such as l manufacturer's information, the equipment qualification data s bank, or available test results, would need to be obtained. l These records would need to be reviewed and the circumstances '

of the qualification tests would need to be compared to the calculated environments and expected component operating l condition. Some judgment would be required to determine if the j qualification program enveloped the expected station blackout j environment. It is estimated that, given the list of equipment  ;

one to two man-months would be required to obtain the ,

qualification records, review them, and decide which equipment l would need further analysis. i j

ll l

24 l l

E_______ . _ _

6.4 Estimation of Costs to Test, Replace, Or Relocate Equipment Where component testing is . deteririned to be' the- best (or only reasonable) alternative 'to demonstrate equipment capability, thel cost of' tests will be a. function of the type of equipment to be tested, the ' extent to which functional parameters 'are monitored, the organization running 'the test, etc. A simple oven test of a small meter with a visual calibration ~run every two hours should not cost auch.nore than about $1K. A more complicated ' test - of' a . motor control center, for example, would involve a' larger oven, somewhat more ~ difficult functional

testing, and a higher. cost -test specimen, say $5-10K plus the cost of a test' specimen. Still more complicated tests o f "a large motor operated valve, . f or ' example ..could reach $25K. or more, depending on test specimen cost.- Testing of equipment could either . be done by individual . utilities or by a utility :

group, as the utilities see fit. Much - equipment . is common-among many utilities, and hence, cost savings might be possible if a utility group were to conduct the tests. One possibility would be for the utility group to contract with manuf acturers ,

to perform the tests. Many manufacturers are equipped with facilities. to perform the type- of testing .which -might .be necessary since the only environment required is dry heat.

Estimating costs for replacing or rsloesting equipment is very difficult due to large variations in: 1) the types and number of equipment that may need to be changed: 2) characteristics of the plant layout: 3) reanalysis which might need to be done: and 4) the other regulatory considerations which must be addressed. These are obviously less preferable methods for solving the problem than modifying procedures or showing.that the equipment will survive the environment. Therefore, cost estimates for these methods will not be given.

6.5 Estimation of Costs to Modify procedures Because k.i s analysis should be performed as part of the utilities overall response to the station blackout rule. .any proposed procedural actions resulting from the analysis should be available at the time the station blackout procedure is to i

be written. These actions could, therefore, be ' part of ' the i original writeup and should. add only a small amount - to the overall cost. A separate cost estimate will not be made here.

6,6 Estimation of-Costs for performing Ca' binet Tests should cabinet . testing- be determined to be a ' desirable or necessary part of the analysis, the test costs are estimated in I

this section. If individual utilities conduct their. own testing with plant-specific parameters f or' each' component of interest, the costs would be unnecessarily large for meeting the desired " reasonable assurance" criterion. A selection of representative configurations and cabinet loadings could be 25 l

conducted by some central organization with individual utilities interpolating / extrapolating the results to their ,

particular plant-specific conditions. Tests could include  !

cabinets with both open and closed doors with both door configurations tested in the same experiment (i.e., run with closed doors to steady state, then open the doors and test to a second steady state). Lecause the temperature at 8_ hours is the desired value, test time would be minimal. The major expense would be cabinet procurement and test set-up. Cabinet costs are estimated to range from about $2-20K, based on recent l Sandia cabinet procurement, depending on the particular cabinet. With an appropriate test facility (a room with controllable temperature as a function of time would be most desirable), the total test cost is estimated to be $50-100K for several different tests on a single cabinet. Additional tests on different cabinets would be somewhat lower, probably in the  !

vicinity of $25-50K. The cabinets would have to be well l instrumented for temperature and use various heat loadings ~

j based on typical power plant cabinets. l 6.7 Independent Review In order to assure that the analysis assumptions, basic data, 1 and calculations are appropriate, an independent engineering i review should be performed. This review should be done by the I utility using people knowledgeable in the various areas but not directly involved in the project. It is not intended that this process be equivalent to a full quality assurance program.

Hence, only " reasonable" assurance is required here. This is strictly intended to be a review of the work done in order to assure that best engineering judgment has been used and that the calculations are correct based on the assumptions used. i Based on Sandia's past experience, independent review I constitutes about 10% of the resources for a project. This varies with how familiar the reviewers are with the subject and the overall cost of the project. Review tends to take a larger percentage on smaller projects due to increased ratio of j spin-up to actual review. In this case, the review is in-house i and done by individuals familiar with the subject areas so '

costs could be somewhat less.

j 7. CONCLUSIONS AND LIMITATIONS 7.1 Conclusions I J

l The major conclusions from the generic and plant specific I information collected are as follows:

I

a. The analytical method proposed is reasonably '

straightforward, and the costs are not excessive i and seem ira line with the desired level of effort.

i 26

)

b. The environments in some locations in a nuclear power plant during a station blackout may exceed the temperatures for which the equipment is qualified.

c. The expected failure temperatures of equipment which does not contain electronics appears to be above the typical calculated environments in many i cases. ]

d. The expected failure temperatures of equipment which does contain electronics may be marginal in some cases using tise reasonable assurance criteria.

e. Any testing done: should account for the expected operational state of the component (i.e.,

powered, cycling, standby, etc.). ,

7.2 Limitations of Analysis.

l This analysis was a first look at the problem. of long-term temperature effects during a station blackout event. Some obvious limitations of the analysis are as follows:

a. The maia weakness in the room and cabinet heat-up analysis is the inability to quantitatively esti- I mate the effects of open doors, open cabinets, and vents on the results. There exist very complica-ted and expensive models which could be used; but, for a reasonably conservative first cut, j neglecting these openings is recommended. This may result in some borderline cases and either l more detailed analyses or tests would then be

! required.

i

b. The proposed method is for . dry heat only. Steam i sources outside containment are not expected unless some additional failure has occurred.  !

These additional failures are not accounted for in this analysis. ,

i

c. For the example plant considered herein, precise I external and internal heat loads, sinks, and qualification information were not obtained.

Ranges of values or first ' approximations were obtained using easily available information from the FSAR, equipment qualification- lists, plant layout and electrical drawings, and information from the fire / equipment qualification program at Sandia. These values are believed to be generally I conservative and should bound the results.

l l

l 27 1

i

, .. j

d. A detailed' cost analysis was not performed for..the i

v. cost or manpower estimates.- The test costs are. 1 based. on.'similar . tests done at Sandia National Laboratories, and the manpower estimates: are. based on recent experience doing;similar tasks.

W l

]

1 1

l q

1 I

i i

1 28 .

L--___-___L__--_-_-____________________-___-____ _ _ _ _ _

Table 4.1 Environment Definition Factors Environment Definition Typical Thermal Potential Thermal Factors Range (For Importance*

Example Plant)

Heat Sources Elect./ Mech. Components 0-100 kW 1 Hot pipes to ambient air (natura1' convection) 4 a) bare pipes 0-3 kW/m2 og pipe i b) insulated pipes 0 .5 kW/m2 og pipe ,

3 Hot pipes to wall (thermal radiation) a) bare pipes 0-5 kW/m2 of pipe 2 b) insulated' pipes 0 .5 kW/m2 og pipe 3 Adjacent Rooms ~0 3 Heat Sinks Concrete Walls - 2000 kJ/K per m 3 of concrete 1 Steel Cabinet - 4000 kJ/K per m3 l of steel 2 Open Doors  ? 2 Open Cabinets and Vents  ? 1 i

""1" indicates most important. "2" moderately important, 'and "3" 1 indicates slightly important.

i i

I 29  ;

_. _ _ _ . .____.__.______________o

Table 5.1 Equipment List for Example Plant RCIC System Equipment in Marsh Environments Quantity Equipment Item Manufacturer /Model

~

1 Pump Motor Reliance 184T,DP 4 Scienoid Valve ASCO WPHV-206-381 4 Limit Switch Namco EA180 5 Valve Operator Limitorque SMB-000 8 Valve Operator Limitorque SMB-00 1 Valve Operator Limitorque SMB-0 2 Valve Operator Limitorque SMB-1 1 Flow Switch Barton 289A 2 S-0-R (Unknown model) 1 Vacuum Switch S-0-R 6N-AA21 1 5-0-R 54N6-B118-NX 1 Flow Transmitter Rosemount 1153DB6 Equipment in Mild Environments Quantity Equipment Item Manufacturer /Model Location 5 Meter GE 120 1H13-P601 in Centrol Room 14 Control Relay GE HFA 1H13-P621 in Aux Bldg.

13 GE HMA 1513-P621 in Aux Bldg.

4 CE HFA 1H13-P618 in Aux Bldg. ]

11 GE HMA 1H13-P611 in Aux Bldg. I 1 Delay Reimy Agastat 7012 1E13-P6f,1 in Aux Bldg.

11 Control Switch GE CR2940 1H13-?601 in Control Room 23 GE SBM 1H13-P601 in Control Room 1 Switch cutler-Hanner 10250T 1H13-P601 in Control Room 1 Aux. Steam 4E 21A9201AV Unknown Turbine Drives I 1 Power Supply GE 9T66Y987 1H13-P612 in Aux Bldg.

1 Signal Convertor Bailey 750 1H13-P601 in Control Room 1 inverter DC-AC Topaz 1H13-P612 in Aux Blds.

N250-GWRS-125-60

$ Pressure Switch S-O-R AN6-E45-NK 1H22-P017 in Reactor Blds.

2 S-O-R AN6-E45-NX 1H22-P029 in' Reactor Bldg.

2 S-O-R 6N6-B2-NX 1H22-P017 in Reactor Bldg.

2 G-0-R 6N6-B2-NX 1H22-P029 in Reactor Bldg.

1 S-O-R 6N6-B5-NK 1H22-P017.in Reactor Bldg.

3 S-O-R 9N6-E45-NX 1H22-P017 in Reactor Bldg.

1 Level Switch Magnetrol 5.0-751 Locally mounted-turbine Bldg.

2 Magnetrol 3.5-751 Locally mounted-turbine Bldg.

1 Flow Controller Bailey 701 1H13-P601 in control Room 30

--m=___--_____--_____.___________.

i Table 5.2 1 (a) Summary of Equipment Test. Conditions for Equipment- i Located in Harsh Environments I

Component Qualification Summary j i

' Reliance. Pump Motor 184T.DP Same model motor qualified to 355' and to 608'F- peaks for  !

unknown times [12)..

ASCO Solenoid Valve Generically tested by ASCO to-WPHV-206-381 340*F minimum for 9 hours1.041667e-4 days <br />0.0025 hours <br />1.488095e-5 weeks <br />3.4245e-6 months <br /> [15).

1' l Namco' Limit Switch EA180 -Generically: tested ' by Namco to 346*F for 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> [16).

Limitorque Valve operators' 'All are typically qualified to SMB-000,-SMB-00 SMB-0 SMB-1 at least 300*F [12) f' or.. inside containment.- Outside contain-ment qualified to 250*F after j aging at 165'C for- 200 hour0.00231 days <br />0.0556 hours <br />3.306878e-4 weeks <br />7.61e-5 months <br />s- '

[17). . Unqualified- versions also ava.ilable with same model numbers.

Barton Flow Switch 289A Qualified by Barton to 200*F for continuous operation [18).

Qual. to 212*F for unknown time.

S-O-R Flow Switch Unknowr. qualification status-Unknown Model Number because of unknown model number.

S-O-R Vacuum Switch 6N-AA21 Qualified up to 290*F 1[13).

Estimated time <1 hr. Also qualified to 212*F' [13) for unknown time. s S-O-R Vacuum Switch Likely' qualified at example 54N6-Bil8-NX plant t o212

• F - [12 ) _ f or unknown time.

Rosemount Flow Transmitter Qualified by Rosemount to 303*F ll53DB6 for 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> [19).

-ij l

l 31 l

i 4

Table 5.2 (Continued)

(b) Summary of Equipment Test Conditions for Equipment )

Located in Mild Environments  !

)

Component Qualification'n Summary j GE Meter 180 Qualified to 9 7 F. a t- example plant (13), Sandia tests on GE type 100 meters (no operation .

during test but checked ~

afterward) had no failure at l 200*C peak -(above '70'C for 50 -I min.). Failed (melted) with 200*C' peak [14.]. j i

GE Switch CR2940 Qutlified to 267'F [13] for {

unknown time. Is. similar to l LBM switches tested iat Sandia 1 (see below). I GE Switch SBM Qualified- to 239'F' [13) for.

unknown time. Sandia tests of same type switch (no operation during test but checked I afterward) had no. failure. at l 200*C peak (above 150*C for 10 i min..) [14).

1 l Cutler-Hammer Switch Qualified to 400*F peak for l 10250T unknown time--continuous operation at 255*F [13).

Bailey Signal Convertor 750 Both of these devices are rated l and for a normal ambient up to 120"F .i Bailey Flow Controller'701 and storage up to 180*F [12].

i

.l

-)

l 32 l

_ _j

k 1

. i h

tn

. u

@ I O -

j o.

H ra W.

k

)

l c

kn m.

E < ,_j La C a r

m 'j g _ a z z H H

[ r g

~u

\ I H

. \ _

g I-

. \ . g

\

- \ . g

.s l

1 l n N. , --

a R R R s a f 8 C.d.) 3Hn1983dW31 NIB DID8 Figure 4.3 Estimated RCIC Air Temperature.vs. Time 33

R l 3'

g a o E C 1 -

t S 3 -

, 1 1

1 A l

, m e

R  !

n w

i.m u 0

Z H

g 5 w

I

\

ap i

8  !

B l

\. _ g N%

l i 1 1 - +-

O E B B E B B B -

(J.) 3Bn1BB3dW31 dIB WOON 1081NO3 -

Figure 4.2 Estimated Control Room Air Temperature vs. Time 34

l 1

I IDENTIFY COMPONENTS AND HEAT LOADS v -

l i

CALCULATE CABINET AND ROOM TEMPERATURES IMPROVE TEMP. CALCS.

n 4k

COMPARE TEMPERATURES TO ,

EQUIPMENT LIMITS FROM

QUALIFICATION, MANUFACTURERS' -*l OK l DATA FIRE TESTING, ETC.  ;

INOT OKl v

I TEST, RELOCATE, AND/OR MODIFY PROCEDURES TEST CABINETS ]

REPLACE EQUIPMENT SC TEMP. TO OPEN CABINETS / FOR ACTUAL

]

REQUIREMENTS ARE MET DOORS TO PROVIDE TEMPERATURES  !

COOLING l v l OK I 1

l i

CALCULATE NEW I TEMPERATURES - I SEE DISCUSSION IN TEXT l

Figure 5.1 Proposed Method for Environmental Analysis j 35 i i 1

REFERENCES

1. A. M. Rubin, et al., Draft Regulatory Guide . . and Value/

Impact Statement on Station Blackout. USNRC, March 1986.

2. Comments on the ' Proposed NRC Rule and Draft Regulatory Guide on Station Blackout. NUMARC, June 1986. -

3. ANS-58.12 Criteria f or . . Evaluation of Response Capability for Loss of all AC Power- (Station Blackout) at Light Water Reactor Nuclear Power Plants, ANS, November 1985.

4. F. P. Incropera and D. P. DeWitt, Fundamentals of Heat ,

Transfer,.-Wiley and Sons, p. 192'(1981).

5. Ibid, p.-442.

6. Ibid, p. 447. .

7. Chavez, J. M., " Steady-State Environment Cable Damage Testing," Quick Look Test Report, Sandia National Laboratories, July, 1984.

8. Wheelis, W. T., " Transient Cable . Damageability Results, Phase I," Quick Look Report, Sandia National Laboratories.

July, 1985.

9. Craft, C. M., " Screening Tests of Terminal Block Performance in a Simulated ~LOCA Environment,"

NUREG/CR-3418, SAND 83-1617, Sandia Nationa.1 Laboratories, August, 1984.

]

10. R. M. Harrington and L. C. Fuller, "BWR-LTAS: A Boiling Water Reactor Long-Term Accident Simulation Code,"

NUREG/CR-3764 ORNL-TM-9163, Oak. Ridge National Laboratory, February 1985.

11. Preliminary Results of LTAS Runs for the RMIEP Analysis'of-the .LaSalle NPP, A. C.- Payne, Jr., Sandia National Laboratories, October 1986.

S., A. Hodge, Station Blackout Calculations for Peach '

Bottta. Proceedings of the Thirteenth Water Reactor Safety 'l Research Information Meeting, Gaithersburg. . MD, October i 1985.  !

L. J. Ott, C. F. Weber, and C. R. Hyman, Station Blackout Calculations for Browns Ferry, Proceedings of the Thirteenth Water- Reactor Safety- Research Information Meeting, Gaithersburg, MD, October l1985. j

12. Ayotte, B. D., Eauipment Qualification,_ Data Bank, Quarterly Report, Electric Power Research' Institute, October, 1985.

36 i

- - E' - - _ _ _ _ _ _ _ _ _ _ _ ____J

,>s REFERENCES (Continued)

13. Bailey Controls General Catalog, Babcock and Wilcox, 1982.

14. Jacobus, M. J., " Screening Tests of Representative Nuclear Power Plant Components Exposed to Secondary Environments Created by Fires," NUREG/CR-4596. SAND 86-0394, Sandia National Laboratories, June, 1986. -

15. Sample Certificate of Compliance for ASCO 206 Series (and other) Nuclear Qualified Valves. Automatic Switch Co. Form VE 3239R1.

16. " Qualification of Namco Controls Limit Switch Model EAlb0 to IEEE Standards 344 ('75), 323 ('74), and 382 ('72),"

Revision 1. Acme-Cleveland Development Company, September 5, 1978.

17. "IEEE323 and IEEE382 Nuclear Qualification Data.for Safety Related Service," Limitorque Corporation, November, 1977.

18. Product Bulletin 288A/289A-3 Indicating Switches, ITT Barton , 1980.

19. "Qualifiestion Report for Pressure Transmitters Rosemount Model 1153 Series B," Revision B, Rosemount Repott 108025, February, 1981.

l 37

_ _ _ _ _ _ _ _