ML20212F836

Evaluation for Effects of Dresden Unit 3 Polyurethane Fire

ML20212F836
Person / Time
Site:	Dresden
Issue date:	05/31/1986
From:	COMMONWEALTH EDISON CO., SARGENT & LUNDY, INC.
To:
Shared Package
ML20212F816	List: ML20212F812 ML20212F836
References
NUDOCS 8701120185
Download: ML20212F836 (111)

Text

__ - - . -

I .

!l ol Evaluation for the Effects of the Dresden Unit 3 lg Polyurethane Fire

!I by Commonwealth Edison Comoany and

,I Sargent & Lundy I

I I

i I

I

.I l

I u r 18a= n888 eoa6 l  % 86-389 o

l 1 l I l I

I EVALUATION FOR THE EFFECTS OF THE DRESDEN UNIT 3 POLYURETHANE FIRE I

I I

1 I

lI l

May 1986 1

'I BY COMMONWEALTH EDISON COMPANY I AND SARGENT & LUNDY CHICAGO, ILLINOIS I

l

b r

L TABLE OF CONTENTS PAGE Executive Summary 7 I.

r 1.1 General 7 1.2 Scope 8 1.3 Summary of Evaluation 9 1.4 Conclusion 11 II. Description of the Fire Event 12 2.1 General 12 2.2 Chronology of the Fire Event 13 2.3 Postfire Observations 17 I III. Scope of Evaluation 18 1

3.1 General 18 3.2 Itemized Tasks 18 I IV. Fire Temperature Definitions 22 I 22 i 4.1 General 4.2 Steel and Concrete Temperature During 22 i

the Fire 1

L 4.3 Radial Thermal Gradients 27 4 .' 4 Conclusions 28

v. Structural Assessment 29 5.1 General 29 5.2 Description of the Structure 29 5.3 Evaluation of the Ef fects of the Fire on the Structural Material Properties 30 5.4 Structural Analysis Model 33 5.5 Evaluation of Structural Integrity of the Steel Containment 34 5.6 Evaluation of the Structural Integrity

{

of the Concrete Shield 38 Conclusions 41 h 5.7

[ VI. Shielding Design Review 43 6.1 Introduction 43 6.2 Ef fect of High Temperature on Shielding Effectiveness 43 6.3 TemperTture Profile and Affected Area 44 6.4 Estimate of Reduced Shielding Effectiveness 45 45

( 6.5 Summary

[

[ 3 J

F VII. Fire Hazards Analysis 46 l 7.1 General 46 7.2 Fire Barrier Description 46 L 7.3 Safe Shutdown Cables 49 7.4 Combustible Materials 49 7.5 Fire Protection Features 51 7.6 Potential Health Effects Associated with the Fire 52

~

I 7.7 Design-Basis Fire 53 7.8 Conclusions and Corrective Action 59 j

i VIII. Conclusions and Corrective Actions 61 I

8.1 Conclusions 61 8.2 Corrective Actions 63 1

TX. References 64 l

Appendix 66 Tables I

Figures I

4

LIST OF TABLES MUMBER TITLE PAGE I 1 Concrete Temperature Profiles for Credible Fire Case 85 l

I I

2 Concrete Temperature Profiles for Hypothetical Fire Case 86 l

3 Results of Structural Analysis of Concrete Shield for l Credible 500 F Fire 87 4 Results of Structural Analysis of Concrete Shield for Hypothetical 1800 0 F Fire 88 5 Safe Shutdown Cables that Pass through Unit 2 Drywell Penetrations 89 6 Safe Shutdown Cables that Pass through Unit 3 Drywell Penetrations 91 7 Fire Zones Containing Polyurethane and/or Polyethelene Filler at the Top of Block Walls 93 8 Possible Effect of a Fire on Unit 2 Safe Shutdown Equipment 94 9 Possible Effects of a Fire on Unit 3 Safe Shutdown Equipment 97 5

L LIST OF EXHIBITS IIUMBER TITLE F

1 Typical Pipe Penetration Joint l

2 Post-Fire Conditions of Containment Coating l

3 Steel Containment Temperature Profiles l

l 4 Vertical Cross-Section of the Steel Containment and the Concrete Shield 5 Axisymmetric Shell Finite Element Model of the Steel Containment Vessel for Overall Analysis 6 Three-Dimensional Finite Element Model of a Local Area of the Steel Containment Vessel I

7 Meridional Stress Distribution in the Meridional Direction at 95 0 Azimuth - Local Analysis I 8 Meridional Stress Distribution in the Hoop Direction at Elevation 537 Feet - Local Analysis l

9 Typical Electrical Penetration Assembly Canister 10 Typical Mechanical Penetration Assemblies (Sheets 1 and 2)

I e

SECTION I EXECUTIVE

SUMMARY

f i 1.1 General l

l This report presents an evaluation of the effects of the j fire that occurred at Dresden Unit 3 on January 20, 1986. -

The plant was in a scheduled refueling outage. The I containment was deinerted with all fuel removed from the reactor vessel and the equipment hatch open. The fire was first detected by the presence of smoke in the reactor water cleanup pipeway at about 9:00 a.m., and was extinguished later the same Jay.

i The fire started when an air-arc cutting process used in the pipe replacement project to cu.t a pipe penetration through the steel containment vessel ignited the polyurethane foam installed in the expansion gap between the steel containment and the surrounding concrete shield. The polyurethane foam was used only to facilitate the pouring of the shielding concrete during initial construction and serves no other design function.

1 I

l l

1 7

I

1.2 Scope The scope of the evaluation report covers the following:

I

a. inspection of the postfire condition of the coating in the steel containment; l

1 l

b. assessment of credible fire temperature to which the containment was exposed, using analytical and experimental app 12ches; I

I

c. evaluation of the integrity of the affected structures, systems, and components during and after the fire;

d. discussion of the possible effects of the fire on the properties of steel and concrete; I

e. evaluation of the steel containment for the effects of potential hard spots that may have been created by the residue from the polyurethane and fiberglass skin exposed to the fire or hypothetical concrete spalls;

f. evaluation of the significance of the absence of the polyurethane foam in the fire-af fected areas of the expansion gap between the concrete shield and the steel containment; and B

g. ' evaluation of the radiation shielding function'of the concrete shield.

1 I h. evaluation of the fire hazards and the safe shutdown capability of the plant.

l 1 Specific responses to the Nuclear Regulatory Commission 's l

" Request for Information" are provided separately in the l Appendix to this evaluation report.

l l

1.3 Summary of Evaluation l

The findings of the postfire examination and evaluation of the affected structures, systems, and components are as i follows:

a. the containment coating has discolored and/or flaked in scattered locations in the fire-affected areas of the steel containment;

b. the maximum temperature experienced by the containment steel and shielding concrete during the fire was less than 500 0F; i

9

c. the fire had no adverse effects on the structural integrity of the steel containment, the concrete shield, or other affected systems and components;

d. the capability of the steel containment and the concrete shield to provide a barrier to control the release of fission products was not impaired;

e. the properties of steel and concrete remsin unaffected by the fire;

( f. the loss of polyurethane in the fire-affected areas of the expansion gap will not impair the intended design function of the steel containment or the concrete shield;

g. the presence of potential hard spots in the expansion gap will not affect the structural integrity of the steel containment or its capability to withstand the design pressure and temperature loads during normal operating and accident conditions;

h. the radiation shielding function of the concrete shield was maintained; 10

[

r

1. the fire cannot spread out of the expansion gap into l other fire areas, and fires started outside the gap will not propagate into the expansion gap; and l

j. safe shutdown of the plant can be achieved and 1

maintained for any fire in the expansion gap.

l 1.4 conclusion l

Based on the findings of this evaluation, it is concluded 1

that the overall integrity of the steel containment, the affected systems and components, and the safe shutdown capability of the plant were not adversely af fected by the fire and that the plant can be expected to continue serving its intended functions satisfactorily without repairs. In order to prevent future occurrences of this type of event, two changes to DMP 4100-1 will be implemented. First, the procedure will be made into a station administrative procedure to improve its implementation by all station working groups. Second, the fact that hot slag or other hot material cob'd travel along drywell pene trations to I reach the foam on the drywell exterior will be emphasized. A prefire plan will be developed to address this specific situation.

i I 11

s SECTION II

~ Di?SCRIPTION OF THE FIRE EVENT

~

2.1 General

_ Extensive construction work associated with the replacement

, of the recirculation system piping was in progress at Dresden Unit 3. The unit was in the refueling mode with

~

all fuel removed from the reactor. The primary containment was deinerted and the equipment hatch was open.

On January 20, 1986, a guard plate for the reactor water cleanup pipe penetration was being removed. This guard I plate was being cut from the drywell penetration area by means of an air-arc cutting process which melts metal through a combination of pressurized air and high electric current. During the cutting process, the polyurethane foam material sandwiched in the expansion gap between the steel containment and the outer concrete shield was ignited by hot slag which entered the annular space at the penetration. A typical pipe penetration detail is shown Exhibit 1.

I j

W The drywell consists of a steel containment and a concrete shield structure (Exhibit 4) . The steel containment is a I

12 l

L L

F pressure vessel with a spherical lower portion and a cylindrical upper portion fabricated of SA-516 Grade 70 L plate. The containment is enclosed by the reinforced concrete shield structure with concrete thickness from 4 r feet to 10 feet. At the foundation level, a sand pocket was used to " soften" the transition between the foundation and the containment vessel. Above the foundation l transition zone, the steel containment is separated all around from the reinforced concrete shield by a gap of l

approximately 2 inches. This annular space accommodated the thermal expansion of the steel containment and precludes any restrained thermal expansion load on the l steel containment or the concrete shield.

l 2.2 Chronology of the Fire Event l

l On January 20, 1986, at approximately 8:30 a.m., a contractor began removing the reactor water cleanup pipe as part of the Unit 3 pipe replacement project using an air-arc cutting process. Cutting was being done on the steel guard plates for penetration x-ll3. At 9:00 a.m., a possible fire was reported in the Unit 3 reactor water cleanup (RWCU) pipeway and a shift foreman (fire brigade leader) was immediately dispatched to the area.

13 I

A person was positioned in the room as ' fire-watch' to observe the cutting operation. At 9:00 a.m. smoke was f observed coming from the spare penetration x-114, about 20 inches from the guard plate being cut. The person assigned l as ' fire-watch' told the radiation chemistry technician

( RCT) assigned to this work to notify the control room.

l Meanwhile, a dry chemical fire extinguisher was discharged into the area from which the smoke was coming. Neither the l

' fire-watch' nor the air-arc operator saw any fire, glowing embers, or flame. However, they observed thick, black smoke, similar to that of a burning tire. The amount of smoke decreased as the fire extinguisher was being discharged. After the extinguisher was completely discharged, only a trace of smoke was observed coming from the spare area of penetration x-114. The ' fire-watch' and the air-arc operator left the room to await the fire brigade. When the fire brigade leader arrived, they all entered the RWCU room and observed that the smoke coming from the penetration had subsided, but, the Unit 3 reactor building started to fill with smoke from the RWCU room.

Since the drywell had its own ventilation for the pipe replacement work, the smoke travelled from the RWCU room through the reactor building and the drywell equipment hatch and into the drywell. Due to potential airborne radioactivity in these areas, the drywell and the drywell entrance were evacuated.

14 I

l l

l Gradually, the reactor building was clearing of smoke and l

the fire brigade leader considered the fire extinguished. I l Fire brigade members entered the drywell at 10:15 a.m. to check for smoke and found that the smoke had cleared. Air l

samples were taken to check the oxygen level and to check for any airborne radioactivity. No airborne contamination l

was found. The construction personnel then were allowed l back into the drywell at 11:30 a.m.

l l

CECO personnel who were conducting an unrelated walxdown in the drywell between 11:30 a.m. and 12:00 p.m. noticed elevated temperatures near elevation 537'-0" near azimuth 125 0. At 12:00 p.m. contractor personnel noticed i

discoloration of containment coating in the area between azimuths 95 and 125 . The drywell was again evacuated, the contractor informed the shift engineer of the evacuation, and CECO personnel confirmed that the control room had been notified of the problem. At approximately 12:20 p.m., a group of contractor personnel and the shift foreman entered the RWCU room again to assess the situation. At 12:30 p.m., water was applied through the annular gap around penetration x-ll3. CECO and contractor personnel detected hot spots inside the drywell near the RWCU penetration x-ll3. A temperature of 450 0P in the drywell at penetration x-143 was reported by contractor personnel. Subsequently, the calibration of the instrument used to measure the temperature was checked and found to be 15

- - - - - - - - - - - - - - - - - - - - - - - - - _ _ _ _ _ _ _ _ _ - - _ - _ - _ _ - - - _ _ __]

l l I

I accurate and the location of the measurement was verified. At 1:00 p.m., the fire brigade leader notified the Fire Marshall that the fire had probably not been extinguished and that water was being applied through penetration x-ll3. The odor in the RWCU room had increased, but no smoke could be seen. CECO personnel, at the direction of the Fire Marshall, prepared to add water through the annular gap around four additional penetrations x-122, x-143, x-144, and x-133.

I At the same time, the Unit 3 and recirculation pipe replacement (RPR) operating engineers, the Station Fire Marshall, and a Station Nuclear Engineering representative met to further investigate what was burning and to plan a course of action. They concluded that the polyurethane foam in the expansion gap between the steel containment and the concrete shield was probably smoldering, with the possibility of releasing toxic gases. The drywell and the Unit 3 reactor building were again evacuated due to the possibility of toxic particles in the air and CECO personnel informed the Fire Marshall that water was being applied through the annular gap around five penetrations.

At 3:00 p.m., CECO personnel and contractor personnel reentered the drywell to inspect and check penetration x-143 where the 450 F temperature was reported at noon. At 3:00 p.m., the temperature reading was recorded as 185 F.

I 16

I g At at 3:00 p.m., a meeting with the working groups from the outage was conducted. It was decided that the water was extinguishing the fire and that the station Technical Staff personnel should enter the drywell and take readings off the interior surface of the s*. eel containment using an optical pyrometer. Technical Staff personnel made a visual and pyrometer survey of the drywell from approximately 75 azimuth to 140 azimuth near elevations 515 feet, 0 inches and 537 feet, 0 inches and near the upper drywell spray header. A peak temperature of 140 F was recorded . 4-143, which later decreased to less than 120 0 F by 5:00 p.m.

Based on these observations, the fire was considered to have been extinguished by 5:00 p.m. A meeting was held at 5:15 p.m. and the afternoon shif t of contractor 's personnel were allowed to go to work. CECO and contractor personnel took temperature readings until 9:00 p.m. when all the ambient temperature readings in the drywell decreased to approximately 85 F. The water was shut off at 9:00 p.m.

I 2.3 Postfire Observations I

Postfire examination of the steel containment indicated that the coating had discolored or flaked away in several scattered areas inside the containment.

3 A field walkdown map of the affected area is shown in Exhibit 2.

17

l SECTION III l

I SCOPE OF EVALUATION 1

3.1 General Immediately after the fire, Commonwealth Edison Company (CECO) and Sargent & Lundy (S&L) drafted a project plan to evaluate the integrity of the affected structures, systems, and components during and after the fire and to assess fire hazards and the safe shutdown capacity of the plant. The plan was comprehensive in scope to address questions regarding the effects of the fire on the integrity of the containment and on the safe shutdown capability of the plant.

I 3.2 Itemized Tasks I The scope of work identified in the project plan is outlined in the following items. These tasks also encompass most of the NRC's request for information described in their February 26, 1986 letter to CECO (Re ference 1) .

I I

I 18

a. Perform a field walkdown to examine the postfire condition of the containment coating.

b. Perform tests on samples of the polyurethane foam and fi.3erglass-epoxy panels to:

I

1. determine the combustion characteristics of the foam and panels, I

2. determine the chemical composition of the residues remaining after the fire and their i

potential corrosive effect on containment steel, and I

3. determine the crushing strength and compressibility of the combustion residue of the foam.

c. Perform tests on samples of the water collected in the torus basement to determine their potential aggressive effects on steel and concrete.

I d. Perform assessments based on field and test data to:

I 1. determine the maximum credible temperature to which the containment was exposed, I 19 1

[

2. determine the temperature profile on the surface of the containment steel and through the thickness of the shielding concrete, and l

3. determine the temperature profile through the containment steel with and without the polyurethane for both normal operating and l accident conditions.

I

e. Determine the possible effects of the polyurethane fire and the water quenching on the material properties of the steel and concrete.

f. Perform structural analyses to evaluate:

I 1. the stress level in the steel containment during the fire,

2. the stress and strain levels of the concrete shield during the fire,

3. the effects of a wet sand pocket on the behavior of the steel containment, and

4. the adequacy of the steel containment to carry design accident pressure and temperature loads in the presence of potential "hard spots" created by I

- - -- o

E

[

{ the combustion residue of the foam in the expansion gap between the concrete shield and the I steel containment.

l

g. Perform fire hazard analyses to:

l

1. determine the effects of a fire in the expansion I gap on all essential and associated circuits that pass through the drywell expansion gap, l

I 2. determine whether a fire could spread from one fire area to another by way of the polyurethane gap,

3. evaluate the effects of the fire on the electrical penetrations located in the vicinity of the fire and the safe shutdown capability of the plant in the event of a fire during operations, I 4. determine the toxic gases that are by-products of a polyurethane fire and the potential ill effects on personnel upon exposure, and

5. recommend measures to prevent recurrence of the event.

21

l SECTION IV FIRE TEMPERATURE DEFINITIONS

(

{

( 4.1 General

[

This section presents the assessment of the containment

( steel and the shielding concrete temperatures that resulted from the polyurethane fire. The calculated temperatures are intended for use as input to the structural assessment of this event.-

( 4.2 Steel and Concrete Temperature During the Fire

[ a. Basis for the Fire Temperature Definition

[

The temperature assessment results are based upon

( bounding interpretations of the available field observations and subsequent test data that describe the effects of the subject fire. The physical evidence provided during and after the fire was used

[. .

to determine that the maximum temperature experienced by both the steel containment and the concrete shield

{

was less than 500 0F. The physical condition of the coating on the inside surface of the containment and

[ 22

[

temperature measurements taken on that surface during the fire event were used to establish the conservative

( bounding value of 500 F.

[ When coating systems burn, they undergo a pyrolysis process or thermal degradation. The coating is decomposed into volatile and flammable products and s

[ nonflammable char. The presence of the char is indicated by its flat black color, a unique condition which is not influenced by the pigments used in a given coating. Since no evidence of coating charring was reported from postfire walkdowns, it is concluded

[ that the coating was not heated to its ignition temperature. In general, the range of combustion temperatures associated with organics is 4000 F to 5000 F. Therefore, an upper bound temperature of 500 0 F is suggested.

[

This bounding temperature was corroborated by special tests conducted by the coating manufacturer (Reference

2) . In these tests, samples of the coatings used in the Dresden Unit 3 drywell were applied to instrumented steel plates and heated until the coating

{

ignited. A black char was observed to be produced at

[ temperatures ranging from 420 0F to 496 0F, These results verify the conservative 500 F used for the thermal analyses. Additionally, the steel sample 23

l l

plates with the coating char were subsequently taken l

into the drywell and compared to the affected portions of the steel containment. This acditional walkdown confirmed that a similar coating char was not present l on the containment steel surface.

l l

Furthermore, the 500 F temperature limit is consistent with the other physical evidence, namely, the temperature measurements taken during the fire. A general-use pyrometer was used by a member of the construction staff to measure the temperature in the vicinity of pipe penetration No. 143 on the inside of the steel containment between 12:30 p.m. and 1:15 p.m.

on the day of the fire. The highest reading recorded was about 440 F to 450 0F. Thus, the physical evidence provides a consistent set of data which supports the conservative 500 F temperature.

b. Steel Containment Temperature The 5000 F hot-spot temperature was used in the steady-state thermal model of the steel containment. A sufficiently large segment of the steel containment wall was included in the model such that the steel at locations away from the hot spot was at ambient -

temperature. The model included thermal conduction in the steel and convective and radiative heat transfer 24

s to the drywell atmosphere. Exhibit 3 presents the resulting temperature profiles in the steel containment wall. The profiles are isotherms centered at a hot spot produced by the fire. The hot spot is marked by an asterisk. The rings are at different c radial distances which are measured from the hot spot. The steel temperature for each ring is displayed in the figure.

F L The extent of the fire's effects is illustrated in p Exhibit 2. This figure was produced by observations L

made during postfire walkdowns. The fire event

{ produced the pattern of hot spots shown ranging approximately from plant azimuth 90 0 through 135 and F plant Elevation 534 feet 10 inches through Elevation 547 feet 6 inches.

I

c. Concrete Shield Temperature l

l The concrete shield temperature profile assessment paralleled the steel containment assessment and used a l

hot spot temperature of 500 F. Due to the difference in the thermal properties of concrete ompared to those for steel, the concrete fire exposure duration was considered by performing transient analyses.

Thus, based upon the event duration described in I Section II of this report, the concrete is 25

~

conservatively assumed to be continuously exposed to the hot-spot temperature for at least 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> (9:00 a.m. to 11:00 a.m.) or for as long as 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> (9:00

- a.m. to 3:00 p.m.) The results of these two credible fire cases are presented in Table 1. The initial

~

concrete temperature was taken to be uniform at 80 F.

Two additional hypothetical fire scenarios were postulated as extremely conservative upper-bound conditions based upon the configuration of the polyurethane foam and the epoxy layer between the polyurethane foam and the concrete shield. These I hypothetical scenarios ignore the physical evidence and facts that the polyurethane was treated with a l

fire-retardant and located in a confined space which would not facilitate an open flame. An insufficient oxygen supply would produce incomplete burning of the polyurethane foam and thus leave portions of the foam unburned. The postfire observations recorded in Exhibit 2 display a spotted pattern of hot spots and i

not a continuous solid burn pattern. These data suggest that the nature of the fire was smoldering and slow-moving. Based upon the observed azimuthal extent of the fire damage of approximately 21 feet and a fire duration between 2 and 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br />, the estimated fire I

spreading rate ranges from 1.8 to 0.6 inches / minute.

The presence of the melted and unburned foam, as well 26

L

[

as the char produced by those portions of the

{

polyurethane foam that did burn, would serve to impede

( the spread of the fire and to insulate the concrete from the effects of the flame temperature.

[ Nevertheless, the transient temperature of the concrete for these two hypothetical scenarios was assessed by conservatively assuming the 1800 F flame l temperature was either directly in contact with the concrete or separated from the concrete by the half thickness (1/8 inch) of the epoxy layer. The 1800 F flame temperature is based on a conservative interpretation of the polyurethane flame temperatures l measured in the tents described in Reference 3. The results for these additional four cases are presented in Table 2.

4.3 Radial Thermal Gradients This section addresses the significance of the absence of portions of the polyurethane for nonfire conditions. For normal operation and accident conditions, the steady-state radial temperature profile has been calculated through the steel containment with and without the polyurethane present. The calculated results for normal operation show a change in temperature difference from 0.0llO F to 0.02 F when the polyurethane is removed. The temperature difference is taken as the surface temperature on the I

27 I

4

L I

L drywell side of the steel minus the surface temperature on the outboard (polyurethane) side of the steel. The

[ absolute value of this temperature difference is small and does not necessitate additional structural evaluation.

Similarly, the calculated results for the accident conditions show a change in temperature dif ference from l

0.043 F to 0.083 F. Again, the absolute value of the j temperature dif ference is small and does not necessitate additional structural evaluation.

4.4 Conclusions l The results described in this section indicate that the maximum temperature experienced by the containment steel and concrete shield during the fire was less than 500 0F.

~

I I

I I

I 28 I .

. -- I

SECTION V I STRUCTURAL ASSESSMENT 5.1 General This section presents a' general description of the structure, the structural analysis model, and the results of the evaluation of the integrity of the steel containment and the surrounding concrete shield during and after the

.I fire.

I 5.2 Description of the Structure Dresden Unit 3 is a General Electric Boiling Water Reactor (BWR) Mark I plant. The primary containment system consists of a drywell, a pressure suppression torus, and interconnecting vent pipes.

I The drywell consists of a steel containment and a concrete shield structure (Exhibit 4). The steel containment is a pressure vessel with a spherical lower portion and a cylindrical upper portion fabricated of SA-516 Grade 70 plate. The containment is enclosed by the reinforced concrete shield structure with concrete thickness varying 29 I

l from 4 feet to 10 feet. At the foundation level, a sand pocket was used to " soften" the transition between the j foundation and the containment vessel. Above the foundation transition zone, the steel containment is separated all around from the reinforced concrete shield by a gap of approximately 2 inches. This annular space l

accommodates the thermal expansion of the steel containment l and precludes any restrained thermal expansion load on the steel containment or the concrete shield.

l To facilitate the pouring of shielding concrete without I

reducing the required gap space during construction, prefabricated crushable polyurethane foam sheets were installed over the exterior surface of the steel containment. Epoxy impregnated fiberglass tape was applied over all joints in the polyurethane foam, and 1/4 to 3/8-inch thick fiberglass-epoxy pref abricated cover panels were then installed over the polyurethane foam. Once the concrete has hardened, these materials sandwiched in the annular space do not serve any design function and are no longer required. ,

I 5.3 Evaluation of the Effects of the Fire on the Structural Material Properties

a. Containment Steel i

I 30

-______ ---- - - - - - - - - - - _- --____-- -_-_-__-___--__- - -_- ------ ----- u

L r

L The steel containment material SA-516 is a carbon steel with microstructure of pearlite in a ferrite I matrix. The tensile strength of the steel reduces at l temperatures above 850 0F. Since the peak temperature attained during the fire was determined to be less than 500 F, it is concluded that there was no significant change in the material properties of the

{ steel.

I Thermal shock for steel at 5000 F temperature is not a concern for any cooling rate. The fact that the l

Dresden polyurethane fire was a one-time occurrence j and that a slow cooling was observed during the later stage of the fire event further lessen the concern.

l Sample tests on water solutions of the burnt foam l

residues (Reference 4) also indicated that no significant concentrations of chemicals hazardous to the steel were found. If acids had formed, they would l have been highly diluted by the quantity of water used to extinguish the fire. This is also supported by the water sample collected from the sand pocket drain lines. Corrosion of the containment steel is, therefore, considered to be minimal (Reference 16).

31

( ,

The volume of water supplied to extinguish the fire

{

was calculated from the plant records to be 34,000 (L

gallons. The volume of water processed by the l radwaste system was approximately 30,000 to 35,000

( > gallons.

(

b. Shielding Concrete The current ASME Section III, Division 2 Code for Concrete Reactor Vessels and Containments (Reference 5) allows concrete in a local area to reach a temperature of 650 F during a pipe rupture accident

( without a strength reduction. Based on this criterion, the concrete material properties are considered unaffected by the maximum credible fire temperature of 5000 F defined in Section IV of this report.

Also, analyses of the effects of chemical species in the water sample on concrete have indicated that chemical attack on the shielding concrete is highly unlikely and that it would take more than 200 years

{ for the concrete cover to deteriorate to the point where the reinforcing steel will start to corrode

( (Reference 17) .

[

[ 32

[

[' .

Based on the above evaluations, it is concluded that there is no change of design significance in the material properties of the containment steel and the shield concrete.

5.4 Structural Analysis Model To investigate the overall integrity of the steel containment at various design sections (Reference 6) , an axisymmetric shell finite element analysis model was constructed (Exhibit 5). This model is referred to as the overall analysis model in this report. The stiffness of the sand pocket described earlier was conservatively neglected in the model because the sand may be wet due to the water applied to extinguish the fire.

I To assess the structural distress around the fire-affected areas with localized effects, such as the highly concentrated fire temperature profile and the potential hard spots resulting from the fire in the expansion gap, a three-dimensional plate-shell finite element model with 780 nodes and 725 elements was used (Exhibit 6). This model is referred to as the local analysis model.

i 33

I I 5.5 Evaluation of the Structural Integrity of the Steel Containment

a. Structural Integrity of the Steel Ccutainment During the Fire The temperature profiles described in section IV and given in Exhibit 3 were considered in this evaluation. For the axisymmetric overall analysis model, the I

nonaxisymmetric fire temperature distribution in the circumferential direction was represented by Fourier harmonics and analyzed as harmonic loads. For the finite element local analysis model, the actual temperature profile was input to the finite elements. Thus, the peak temperature location was conservatively represented at one node in the overall analysis model, whereas in the local analysis model it was more realistically applied over an element.

I The stress patterns in the meridonal and hoop directions around the fire spot obtained from a linear elastic analysis of the local analysis model are shown in Exhibits 7 and 8, respectively. The results from the overall analysis are similar, but the peak stresses are slightly higher. The difference is due to more realistic modeling of the peak temperature in 34

s c

L c the local analysis model. These results show that the thermal stresses calculated are highly localized F around the peak temperature spot and attenuate very L

rapidly. The magnitude of the pesk stress can also be e

I shown to be comparable to that resulting from the thermal expansion of a fully restrained structural element.

In contrast to a general or overall thermal loading,

( thermal stress in a small hot spot is defined by the design basis code, namely 1965 ASME Section III

[ Subsection N-412 (Reference 7) , as local thermal stress. The allowable. thermal stress for general thermal effects is specified by Subsection N-414 to be

( 57.75 ksi. The allowable stress for local thermal stress is based on f atigue considerations. This value is determined in accordance with subsection N-415 to be 260 ksi preudo-elastic stress for the anticipated number of normal temperature and pressure cycles for

( the remainder of the life of the plant.

The peak local thermal stress obtained is greater than the yield strength of the material but is within the allowable value even for general thermal effects, and

{ is f ar less than the allowable local thermal stress.

[

35

l I

l The potential overlapping effects of the multiple hot I

spots shown in Exhibit 2 were also investigated.

j Because of the rapid attenuation of the peak thermal stresses as shown in Exhibits 7 and 8, the overlapped l stresses are also within the code allowable values.

I l

b. Postfire Structural Integrity of the Steel Containment to Carry the Design Basis Loads l Sample ftre tests of the polyure thane foam and fiberglass panels (Reference 4) showed that residues l

of the burnt foam and panels have very low crushing I strengths. Stress analyses of the shielding concrete discussed later in this section also showed that l concrete will not spall under the maximum credible fire temperature. Nevertheless, a hypothetical hard spot representing the combustion residue was conservatively assumed in the structural analysis for the postfire structural integrity of the steel con ta inmen t . The hard spot was simulated in the analysis model, first with a finite stiffness and later with infinite stiffness. The model was then subjected to the design pressure and the design basis accident temperature, respectively. The results of these analyses show that regardless of the presence of a potential hard spot of various stif fness, the I

36

I I stresses in the steel containment under these design loads will be within the code allowable values.

A structural analysis was also performed to investigate the hydrostatic pressure effect of the extinguishment water potentially trapped below the vent line after the fire. The results of the analysis indicate that the stresses induced in the steel containment by the hydrostatic pressure are negligible.

As discussed in the evaluation of the effects of the fire on the structural material properties, no significant concentrations of hazardous chemicals to steel are expected to exist in the potentially trapped water. In addition, the primer on the fire side of the steel containment is judged..to be intact based on the conditions of the primer on the drywell side.

Corrosive potential of the steel containment is, therefore, considered minimal and containment analyses for design basis conditions have shown that there are

, ample safety margins to accommodate the normal corrosion for the remainder of the plant life.

Based on the above, it is concluded that the structural integrity of the steel containment is not impaired and that 37

l l

l it can perform its functions for the remainder of its design life.

l 5.6 Evaluation of the Structural Integrity of the Concrete Shield l

~

l The structural integrity of the reinforced concrete shield was evaluated for conditions during the fire and for postftre service. Elevations 537 feet, 6 inches and 541 feet, 6 inches are considered to be most af fected by the fire. Stresses in the concrete and reinforcing steel were determined for the credible and upperbound temperature profiles and compared with the maximum allowable stresses in accordance with the ACI Building Code 318 ultimate strength design.

The structural capability to withstand postfire design basis loads was also evaluated in accordance with the design basis allowable stresses specified in FSAR Table 12.1.1.1.

a. Structural Integrity of Shield During the Credible Fire Structural analyses of the reinforced concrete shield were performed to determine the stresses and strains in concrete and reinforcing stool for the credible 38

I fire temperature profile discussed in Section IV. The

' temperature distribution through the thickness of the concrete shield for this case is given in Table 1.

The results of the structural evaluation along with the maximum allowable values per ACI 318 (Reference 8) are presented in Taole 3 and show that:

I

1. the maximum compressive strain in the concrete is within the allowable value, I

2. the maximum stress in the inner layer of reinforcing steel in both the meridional and hoop directions is within the allowable value, and 1

3. the maximum stress in the outer layer of l reinforcing steel in both the meridional and hoop directions is within the allowable value, i

I Dased on the above, it is concluded that the l

structural integrity of the shield was not impaired f during the fire.

I 1

b. Postfiro Structural Integrity I

Since the proporties of the concrete and steel were not af fected at the tempera ture level associated with I

I

l l

l l the credible temperature fire event and since the allowable stresses and strains in the concrete and l reinforcing steel were not exceeded during the fire, the structural integrity of the concrete shiold is l

intact to carry the design basis loads and load combinations without exceeding the allowable stresses specified in FSAR (Reference 9) Table 12.1.1.1.

I

c. Capability of the Concrete Shield to Withstand 18000 F Fire i

The structural integrity of the concrete shield during and after the hypothe tical upperbound fire given in Table 2 was also evaluated. The extremely conservative assumptions used in defining this hypothetical temperaturo profile are discussed in Section IV. In 'the structural ovaluation for this caso, portions of the concreto and stool in the shield wall, where the temperature magnitudo in the concreto exceeded the 6500F limits discussed earlier (Reference 5) , were conservatively assumed to be inoffectivo. Strossos and strains in the concreto and reinforcing stool for loading condition during and after the fire woro then dotormined and are presented in Table 4.

40

L 0

The results show that, in spite of the conservative assumptions, the stresses and strains in the of fective concrete and steel during this 1800 0F fire are within the allowable values, that the postfire structural

( integrity of the shield is maintained and that it is adequate to carry the design basis loads and combinations without exceeding the design allowable stresses specified in FSAR Table 12.1.1:1.

{

( 5.7 Conclusions

( a. The properties of the containment steel and the shielding concrete are not affacted by the fire.

{

( b. During the fire, the stresses in the steel containment were within the allowable values.

c. During the fire, the stresses and strains in the concrete shield were within the allowable values.

d. The postfire structural integrity of the steel containment is not impaired, and it is adequate to withstand the design basis loads.

e. The

{ postf tre structural integrity of the concrete shield is not impaired, and it is adequate to

( withstand the design basis loads.

41

r F

f. No repair or modification to the steel containment or f the concrete shield is recommended.

u L

E u

I L

I 1

I I

I I

I I

u g

e k

SECTION VI I SHIELDING DESIGN REVIEW L

b 6.1 Introduction This section presents an evaluation of the capability of the concreta shield to provide radiation shielding after the tire.

6.2 Ef fect of liigh Temperature on Shielding Ef fectiveness i volume II, Shielding Materials, of Engineering Compendium of Radiation Shielding (Reference 10) states that: "The maximum internal temperatures of a concrete neutron shield should be less than 200 0F. Gamma ray shields may be safely taken to 350 F..."

The reason for the temperature limit is that at elevated temperatures, concrete is more prone to dehydration, and hydrogen is a necessary ingredient for offective neutron shielding. Gamma shielding is a function of total density. As the total density of concrete is comprised mostly of other nonhydrogen elements, the gamma ray shielding is relatively unaffected by clovated temperatures.

I 43 I

. . . . . __ _ _____________________________J

(-

6.3 Temperature Profile and Affected Area The temperature distribution through the thickness of the

{

concrete shield for the credible fire case is given in

b. Table 1. This estimate was obtained conservatively from a one-d imens ional slab-geometry . calculation. Of concern is the concrete which reached temperatures above ~2000 F.

Table 1 shows that the maximum thickness from the fire

{

surface to the point where concrete reached this

( temperature is 9 inches.

[ Exhibit 2 shows the spotty nature of the fire affected areas. Based on this, one may conclude that the

{

temperatures calculated from the uniform slab geometry are

(. an upper bound estimate of temperatures experienced during the fire.

b The fire-affected area of the concrete shield requiring radiological evaluation is bounded between azimuth 90 0 and 135 0 and between elevations 534 feet, 0 inch and 547 feet,

[

6 inches. The concrete at this location is more than 10 feet thick (distance froin the interior wall to the reactor vessel, i.e., not a horizontal distance) , which is thicker than at all other locations in the drywell.

[

[

[ 44

l 6.4 Estimate of Reduced Shielding Effectiveness i

I The concrete shield at the reactor core midplane is 6 feet, 6 inches thick (the active core extends from elevation 548 feet, 8 1/4 inches to 560 feet, 8 1/4 inches). The 6 feet, 6 inch-concrete wall provides adequate protection at the core midplane elevation, the region of maximum neutron and gamma ray dose rate.

The fire-affected shield wall is below the active core and farther from it. Therefore, the incident neutron and gamma ray dose rates are much lower than at reactor midplane.

This coupled with the thicker (10 feet, O inch) shield wall leads to the conclusion that even if 1.h e equivalent of 9 inches of concrete shielding is lost completely (which will not be the case), there is still more shielding in place in the af fected area than at core midplane. Since the dose rates are acceptable at core midplane elevation, they are equally acceptable at the fire-affected elevations and thus the remaining concrete is more than adequate.

6.5 Summary Fire damage assessments have been reviewed and the impact of the fire on the shielding capability of the drywell wall has been qualitatively assessed. The judgment is that the shielding integrity of the wall has not been compromised.

45

[

[ SECTION VII FIRE HAZARDS ANALYSIS

[ 7.1 General This section includes descriptions of the fire barrier, the

( safe shutdown cables, and the combustible material. The fire protection criteria and measures and the design basis

[ fire are then examined.

7.2 Fire Barrier Description The 2-inch gap between the steel containment and the r

concrete shield is separated from the drywell fire area by I the steel containment shell. Separation of the expansion gap from the rest of the reactor building is provided by minimum 4 feet, O inch-thick structural concrete that is penetrated by mechanical and electrical penetrations. In the area where the fire took place, the concrete is a minimum 8 feet, 0 inch thick.

I In Unit 3, the concrete wall separates the expansion gap from two separate fire area groups. These fire areas were established based on the location of safe shutdown

I components including associated circuitry. One of these fire area groups consists of the TIP room, isolation condenser floor, and the isolation condenser pipe chase, and has been given the designation RB3-I. The other fire area is made of all the other parts of the reactor building not included in Fire Area RB3-I and is designated as Fire Area RB3-II.

In Unit 2, this concrete wall also separates the expansion gap from two separate fire area groups. As with Unit 3, I the fire areas were established based on the location of safe shutdown components including associated circuitry.

One of these fire area groups consists of the shutdown cooling pump room, isolation condenser area and the isolation condenser pipe chase and has been given the designation RB2-I. The other fire area is made of all the other parts of the reactor building not included in Fire Area RB2-I and is designated as Fire Area RB2-II.

Also providing a barrier to fire spread are the electrical cable penetration assemblies and the mechanical penetrations. There are three standard types of electrical penetration assemblies present; Low Voltage Power and Control Cable Penetration, High Voltage Power Cable Penetration, and the Shielded Cable Penetration. Each type of electrical penetration has the same basic configuration shown in Exhibit 9. An assembly is sized to be inserted in 47 l

l

I l

the penetration nozzles which are 12-inch schedule 80 steel pipe (wall thickness of 0.688 inches). They are furnished l as part of the containment structure and the design and fabrication of each assembly is in accordance with the requiretr e nts of the ASME Boiler and Pressure Code,Section III, Class B Vessel. The assembly extends approximately 1 foot beyond the drywell wall on both sides of the pene t rations and the drywell wall, in the vicinity -

l of the electrical penetrations, is at least 6 feet thick.

The mechanical pene trations are of two types; those which accommodate thermal movement (hot) and those which experience relatively little thermal stress (cold). The hot fluid line penetrations have a guard pipe between the hot line and the penetration nozzle in addition to a double-seal arrangement (see Figure Exhibit 10, sheet 1).

This permits the penetration to be vented to the drywell should a rupture of the hot line occur within the penetration. The guard pipes are designed to the same pressure and temperature as the fluid line and is attached to a multiple flued head fitting, a one-piece forging wi,th integral flues or nozzles. This fitting was designed to the ASME Pressure Vessel Code, Section VIII. The penetration sleeve is welded to the drywell and extends through the biological shield where it is welded to a bellows which in turn is welded to the guard pipe. The l

bellows accommodates the the rmal expansion of the steam 48

I pipe and drywell relative to the steam pipe. A double bellows arrangement permits remote leak testing of the penetration seal. The lines have been constrained at each end of the penetration assembly to limit the movement of the line relative to the containment, yet will permit pipe movement parallel to the penetration.

The penetration details of cold piping lines are shown on Exhibit 10, Sheet 2. These penetrations have a double-seal arrangement, however, the guard pipe provided for the hot piping line penetrations is not provided.

7.3 Safe Shutdown Cables The only safe shutdown components located in the expansion gap are electrical conductors inside the electrical penetration assembly canisters. Tables 5 and 6 list the safe shutdown cables, for Units 2 and 3, respectively, the number of the penetrations to which the cable is connected, and the safe shutdown equipment associated with each cable.

7.4 Combustible Materials 1 The 2-inch gap is filled with polyurethane sheets that are covered with a fiberglass cover panel, as shown in Exhibit 1. Polyurethane is a polyester base material and '

the sheets used conform to the following requirements:

49

l

a. Base Specification - M/L-PPE-200F

b. Chemistry - Isocyanate foam formed by reaction of polyisocyanates with polyester polyols,

c. Density - 2 pcf i 0.10 pcf

d. Thermal Value - 0.26K factor l

e. Service Temperature - 285 F.

l

f. Physical Properties l 1. Tensile - 12 psi minimum

2. Elongation - 100%

3. Compressibility - 35% at 1.0 psi maximum

4. Compression Set - 10% at 50% compressibility 1

I l g. Sheet size 1/4 inches x 2 feet x 8 feet with tolerances as specified by M/L-C-26861.

l

h. Heat of Combustion - 12,000 Btu /lb.

i. Self-ignition Temperature - 1000 F.

I j. Ignition Temperature - 500 0F to 700 F.

, e0

l f

I I .

The expansion gap is not the only place where polyurethane l

1s used as a filler material. Both polyurethane and polyethylene have been used as a filler material at the top of block walls and polyurethane is used to seal penetrations. None of the block walls is considered a I

rated fire barrier and in those walls that use polyurethane as a penetration seal, either the wall is not a rated l

barrier or, if it is, the polyurethane has been replaced I with a fire rated or noncombustible material. Thus, the polyurethane or polyethylene cannot provide a path for fire spread from one fire area to another. Table 7 provides a i

list of those fire zones where polyurethane and polyethylene are used as a filler at the top of block walls and the percent increase in the fire loading of each zone resulting from this material. As can be seen, the percent change is very small which is why it was not included in past fire loading calculations. The amount of polyurethane used as penetration seals could not be easily determined.

However, even if this amounted to two or three times that used at the top of the walls, the percent change for any of the fire loading would still be small.

7.5 Fire Protection Features No fire protection feature or equipment is located in the expansion gap. Currently, fire detection is provided in 51 l

L

[

Unit 2 reactor building Fire Areas RB2-I and RB2-II and Unit 3 Fire Areas RB3-I and RB3-II above specific equipment. These detectors will detect the smoke combustion products of a fire originating in the gap. The smoke from the January 20, 1986, polyurethane fire was detected and alarmed by the detectors located above McC's in the reactor building. General area fire protection is now being installed throughout the reactor building as part

{

of the station 10 CFR 50 Appendix R modifications, f

Water system hose reels and portable fire extinguishers are provided throughout the reactor building. The hose reels can be used to reach all penetrations so that water can be applied to a fire 7.6 Potential Health Effects Associated with the Fire Pyrolysis of polyurethane yields the following combustion products (Reference ll) :

1. carbon monoxide,

2. carbon dioxide,

3. water vapor, and

4. hydrocyanic acid, The major hazard from the decomposition of rigid polyurethane is carbon monoxide. Other vapors emitted from 52

l l

a fire may cause irritation of the eyes, skin, and l

respiratory tract, however, they are generally acute l (short-term) with no chronic effects (Reference 15) .

l l

7.7 Design-Basis Fire l

There are two major concerns for a fire in the expansion gap.

1. Can a fire starting in the gap or in a specific fire area adjacent to a gap spread to another fire area where safe shutdown components of an alternate or redundant shutdown path are located?

2. Can the reactor be shutdown if a fire occurs inside the gap during operation?

A third concern that has to do with the January 20, 1986 fire was raised.

3. Did the January 20, 1986 fire affect any of the electrical penetrations?

Each of the concerns is addressed in this section.

Analysis of the data from the January 20, 1986 fire has 53 i

I l

l been used wherever possible to support the elimination of the concern.

1 Concern #1 l

As stated in Section 7.2, the barriers to the spread of a fire out of the expansion gap are the drywell wall and the electrical and mechanical penetrations. Since the concrete wall is a minimum 4 fee t , O inches thick, the discussion here is limited to justifying that the electrical and mechanical penetration assemblies provide a barrier to fire spread. Also, since the drywell is inerted during operation, the concern is limited to the spread of a fire from the gap to fire areas outside the drywell.

Since the polyurethane foam is located on the drywell side of the concrete wall, the only mechanism for fire spread from the expansion gap through electrical penetration assemblies to the fire areas outside the drywell or vice-versa is by conduction of sufficient heat through the penetration assembly to reach the autoignition temperature of the cables (600 0F) or the foam (1000 F) . This is an unlikely event due to the construction of the assembly. As can be seen in Exhibit 9, the penetration assembly is a metal canister into which a sleeve, two header plates and cable support plates have been inserted. Electrical conductors are contained within the sleeve and are passed 54

L

[

through the sleeve through openings in the header plates.

A potting compound has been applied at each end of the penetration to seal between the header plates and cable.

The highest temperatures of the drywell side of the steel shell during the January 20, 1986, molyurethane fire was 500 F. An analysis has shown that at a distance of 3 feet l

from a hot spot on the steel shell, the maximum temperature l was 94 F (see Exhibit 3). Thus, if the hot spot was next to a penetration, the temperature at the outside of the I penetration, which is about 6 feet away, would be very low. Furthermore, conservative calculations indicate that l

if an 1800 F exposure (i.e., flame impingement) were to occur in either the polyurethane foam or the reactor building, it would take at least 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> to conduct sufficient heat through the stainless steel penetration to threaten ignition of combustible materials on the opposite wall. (This calculation is suffici.ently conservative to also account for thermal radiation affects through the annulus around mechanical penetrations.) Therefore, a fire cannot spread through an electrical penetration assembly into another fire area or from the fire area into the gap.

I For mechanical penetrations, as is the case with the electrical penetration assemblies, the polyurethane foam is located on the drywell side of the concrete wall, thus, the only mechanism for fire spread from the expansion gap to the fire areas outside the drywell is by conduction through 55

l l

l the penetration. Since there are no combustible materials in contact with the pipes in the vicinity of the I penetrations, a fire cannot spread out of the gap. This conclusion is supported by the January 20, 1986, fire. As l

can be seen in Exhibit 2, the fire involved several mechanical penetrations, however, the fire did not spread out of the expansion gap. Analysis of the January 20, l 1986, fire shows that on the steel shell at a distance of 3

~

feet from any hot spot, the temperature was 94 0F.

I Therefore, at the outside end of the penetrations, the temperatures of the penetration could not have been high enough to ignite any combustible material in the vicinity l of the penetration.

l The concrete wall and the electrical and mechanical penetration also act as a barrier to the spread of fire from the reactor building into the expansion gap. The justification for this statement is given in the discussion that was given above for why a fire cannot spread from the expansion gap into another fire area.

Concern #2 As discussed above, a fire in the expansion gap cannot spread into bounding Unit 2 Fire Areas RB2-I or RB2-II or into bounding Unit 3 Fire Areas RB3-I or RB3-II via the electrical and mechanical penetrations. A fire in the 56

r .

L.

[ drywell is not considered a credible event during operation since they are inerted.

H y As discussed above, the construction of the assembly and i

the analysis of the January 20, 1986, fire makes it unlikely that a fire in the expansion gap could prevent the penetration assembly from performing its function.

Iloweve r , Tables 8 and 9 conservatively address the situation where the assemblies are affected. As can be I seen in the tables, shutdown is not affected because of the following:

I l 1. the safe shutdown valve is normally in the proper safe shutdown position and a fault in the cable will not l

change that position, l

2. the mechanical function of the Target Rock valve and the safety valves is not af fected by an expansion gap i fire, thus RPV pressure control will remain available,

3. Instruments are available to monitor reactor vess,el level that have their essential and associated circuits routed independent of the expansion gap, and

4. manual actions can be performed to open valves required for cold shutdown or close valves in lines that are not used as fluid paths for hot shutdown.

57

_ _ _ _ _ _ _ _ _ _ _ _ __________J

J J

The results of the analysis of these possible effects indicate that a fire that spreads throughout the entire gap area would not prevent achieving a safe shutdown. For a fire in the expansion gap hot shutdown can be achieved and maintained using the HPCI system for reactor water makeup and the LPCI/CCSW system for suppression pool cooling.

Cold shutdown can be achieved using the shutdown cooling system in conjunction with the reactor building closed F

i cooling water system and the service water system.

E Concern #3 F

l None of the electrical penetrations was involved in the I

l January 20, 1986 fire. The closest any hot spot (identified in Exhibit 2) came within an electrical i

penetration was about 3 fee t . As shown in Figure 3, the maximum temperature 3 feet from a hot spot was 94 F. As stated in the FSAR (Subsection 5. 2. 2) , the electrical l penetrations are capable during normal operations of continuous operation at temperatures of 1500 F inside the drywell and 125 0 F outside the drywell. Thus, the fire could not have af fected any of the electrical penetrations.

I I

1

<__ . - - _ _ - - _ -_----------_-----_-----------a

I

(

7.8 Conclusions and corrective Action In conclusion,-the analysis has shown that in the event of I

a fire during operation . safe shutdown can be achieved and  ;

( maintained. An analysis of the January 20, 1986 fire has shown'that the fire cannot spread out of the expansion gap and into the drywell during operation since the drywell is

( inerted. The fire, also, cannot spread into the rest of the reactor building due to the existence of substantial barriers to fire spread (the concrete wall and the electrical and mechanical penetrations). These -same barriers prevent a fire from spreading from any of the fire

( areas bounded by the expansion gap into the expansion l gap. The r e fore , safe shutdown capability cannot be affected because of the inability of a fire in the expansion gap to spread to other fire areas and the inability of a fire in other fire areas to spread into the expansion gap.

In order to prevent future occurrences of this type of event, two changes to DMP 4100-1 will be implemented.

First, the procedure will be made into a station administrative procedure to improve its implementation by all station working groups. Second, the fact that hot slag or other hot material could travel along drywell penetrations to reach the foam on the drywell exterior will be emphasized.

59 l

/

e L

r Promptly following this event, the cause was thoroughly discusaed with station construction and maintenance personnel. It is stressed that all such gap areas must be stuf fed with fire retardant sheeting prior to the start of curing or welding. The Station Prefire Plans will be enhanced to address this type of event.

F~

k Since an alternative shutdown method is used for a fire in the expansion gap, an exemption from the requ ir ements of Appendix R to 10 CFR 50, Section III.G.3 (requirement that fire detection and fixed fire suppression be installed) l will be requested.

I 1

E 1

I I

I I

I B

B

~

B

L p SECTION VIII CONCLUSIONS AND CORRECTIVE ACTIONS I

8.1 Conclusions

(

Based - on the evaluation of the fire-affected structures, systems, and components, the following conclusions are in order:

a. The structural integrity of the containment is not adversely af fected by the fire and it can be expected to continue serving its intended function.

b. For the normal operating and accident conditions, the steady state radial temperature profiles through the steel containment with and without the polyurethane are insignificant and do not warrant structural evaluation.

c. The properties of the containment steel are judged to remain unchanged as a result of the fire.

d. The stresses in the steel containment during fire are within the design code allowable values.

61

h r

e. No significant concentrations of chemicals hazardous L to the steel and concrete were introduced into the drywell expansion gap as a result of the fire and extinguishment.

I The short term and long term effects of the

( extinguishment water on the structural integrity of the containment steel, shielding concrete, and

[. electrical and pipe penetrations are minimal.

[

f. The presence of any potential "hard spots" will not

( affect the structural- integrity of the steel containment or its capability to wit 1. stand pressure and temperature loads during normal operating and accident conditions.

g. The fire did not affect the capability of the steel containment to provide a barrier and to effectively control the release of fission products.

h. The fire did not affect the structural and shielding 1

functions of the concrete.

1. A fire cannot spread from one fire area into the expansion gap and then into another fire area via the electrical and mechanical penetrations.

62

b

[.

j. A fire that spreads throughout the expansion gap area

( would not affect safe shutdown capability during normal operation.

8.2 Correc+ive Actions

( No repair or modification is recommended since the integrity of the steel containment, the concrete shield electrical and pipe penetrations, and other affected components is intact after the fire.

[

In order to prevent future occurrences of this type of event, two further changes to DMP 4100-1 will be implemented. First, the procedure to improve its implementation by all station working groups. Second, the fact that hot slag or other hot material could travel along drywell penetrations to reach the foam on the drywell exterior will be emphasized.

63

l SECTION IX

[ REFERENCES

(

l. USNRC Inspection Report No. 50-249/86006 (DRSS), dated February 25, 1986.

2. Carboline Test Report, " Thermal Degration of Rustbond f

s 6-C/Polyclad 933-1 Coating System," March 14, 1986.

3. Foam Rubber and Polyurethane Foam Fire Tests, " Automatic"

( Sprinkler Corporation Report No. 1676, September 2, 1985.

4. " Combustion Testing of Drywell Expansion Gap Filler Materials," by SMAD of Commonwealth Edison Company.

5. 1983 ASME Code Section III, Division 2.

6. CBI Stress Report No. 9-4646 for Dresden III Containment Vessel.

7. 1965 ASME Code Sections III and VIII.

8. ACI Code 381-71.

9. Dresden Final Safety Analysis report (FSAR).

64

10. Engineering Compendium on Radiation Shielding, volume II, Shielding Materials, Springer-Verlog, 1975, page 92.

1

11. Harris, John C., et al, " Toxicology of Area Formaldehyde and Polyurethane Foam Insulation." Journal of American Medical Association 245: 243-246, January 16, 1981.

12. Harrington, Robert, Rubber Age 82461 (Dec. 1967).

13. Schollenberger, C. S., et al., "Polyurethane Gamma

[ Radiation Resistance," B. F. Goodrich Co. Research Center, f Nobs - 72419 (June 1, 1959).

I

14. Kercher, J. F. and Bowman, R. E., "Ef fects of Radiation on Materials and Components".

15. Proctor, Nick and Hughes, James P. " Chemical Hazards in the r

Work Place," 1978.

16. Corrosion Handbook, Ahlig, 1966, John Wiley and Sons.

I

17. Clear, K. C., " Fire to Corrosion of Reinforced Steel in Concrete Slabs, V.3: Performance After 830 Daily Salt Applications," Report No. FHWA-RD-76-70, Federal Highway I Administration, Washington, D.C. 1976.

65 i

F APPENDIX RESPONSE TO REQUEST FOR INFORMATION l

, This appendix is prepared in response to NRC request for L

information identified in their Inspection Report No.

[ 50-249/86006 (DRSS), dated February 26, 1986.

I h In the following, each of the issues identified in paragraph 3 of the NRC report is restated and response is provided thereaf ter.

I B

l l

\

\

66 f

L

[ 1. Detailed Chronology of the Fire Event

. Provide a detailed chronology of the January 20, 1986, j l

Unit 3 drywell expansion gap fire occurrence.  !

Response: On January 20, 1986, around 9:00 a.m., a possible fire was reported in Unit 3 Reactor Water Cleanup I pipeway. As part of the Unit 3 pipe replacement project, contractors were removing reactor water cleanup pipe using an air / arc cutting torch. The flued head, bellows, and cleanup pipe had been removed from penetration x-ll3 and cutting was

( being done on steel guard plates. Cutting work started at approximately 8:30 a.m. A person was positioned in the lowest dose area in the room as

" fire-watch" to observe the cutting operation.

At 9:00 a.m.*, smoke was observed coming from the spare penetration x-ll4 about 20 inches from the guard plate being cut.

Additional information is provided in Section II of the report.

[

. Describe the sequence of events that led to the decision that offsite fire department assistance was not needed.

67

s

- Response: The station fire marshall determined offsite fire department assistance was not needed. This decision was based on the type of material smolde r ing (polyarethane foam) and the method L used to extinguish the fire.

H l

Polyurethane foam is a Class A type combustible which is extinguished most effectively by using water. Once it' was determined that the foam material was smoldering, water was applied through the annular gap around the penetrations into the annular space between the steel f containment and the concrete shield. Additional off-site resources were, therefore, deemed unnecessary.

l

2. Duration and Intensity of the Fire l

Determine the duration, physical extent, and intensity of l the fire and include, in this assessment, the highest metal and concrete temperatures reached during the fire. If no i

systematic approach was taken to record actual temperatures reached during the fire, determine the highest temperature that the steel and concrete structures may have been exposed to based on published (i.e., Underwriters Laboratories Inc., Factory Mutual Laboratories, Inc.) free 68

[

(;

-burning polyurethane foam calorific heat values for a fire of this duration.

Response: The steel containment and the concrete shield structures were exposed to fire for 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> to a maximum of 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br />.

( The pattern of hot spots produced by fire were observed approximately from

( plant azimuth 90 through 135 and plant elevation 534 feet, 10 inches through elevation 547 feet, 6 inches. The maximum temperature experienced by the containment steel and concrete

{

shield during the fire was less than 500 0F.

! . Provide an estimate of what changes occurred in the material properties of the steel, concrete, electrical and pipe penetrations, drywell penetration welds and other affected equipment or components.

Response: Changes of design significance do not occur in the material properties of the steel, concrete, electrical and pipe penetrations, containment penetration welds and other affected equipment or component at a temperature of 500 0F.

l

. For the normal operating and accident condition, determine the temperature profile through the drywell steel liner 69

{

(

with and without polyurethane present in order to show any changes in drywell expansion from the original design.

Response: The temperature profile through the steel l containment with and without polyurethane present has been calculated for the normal operating and accident conditions. The calculated results for

( normal operation shows a change in temperature gradient from 0.011 F to 0.20 F when the polyurethane is removed. Similarly, there is a change in temperature gradient for the accident condition is from 0.043 0 F to 0.083 0F. Since the absolute value of the temperature difference is so small, no further evaluation has been performed.

. Perform a structural analysis which evaluates the state of l

stress in the drywell steel liner during the fire and compare this with the yield strengths of the material.

Response A structural analysis of the steel containment has been performed for the conservative fire spot temperature of 500 0F. The stresses induced in the steel containment are very localized and attenuate rapidly away from the hot spot. The stress in a small hot spot is defined as local thermal stress in the design basis code, 1965 70

~

ASME Section III, Subsection N-414. The peak I j

local thermal stress obtained at the hot spot is greater than the yield strength of the material but is within the allowable value even for l general thermal effects which is far less than the allowable local the rmal stress specified by the ASME code.

I L

3. Corrosive Species Introduced into the Drywell Expansion Gap I

L

. Determine the type and quantity of corrosives that were introduced into the drywell expansion gap as a result of f the fire and its extinguishment.

E Response: Analyses of water sample collected at two sandbank drainline locations show the following 1

types and quantities of chemical species:

l Type At 450 Azimuth At 135 Azimuth l

Chloride 28 ppm 24 ppm i

I PH 8.09 7.92 Sulfite 3 ppm 1 ppm Sulfate 0.902 ppm 7.086 ppm 71 5

[.

(

. Determine the short and long-term effects of these

.corrcsive species on the structural integrity of the drywell steel liner, structural and shielding concrete, electrical and pipe penetrations, drywell penetration welds, and other affected equipment and components.

Response: Laboratory tests (Reference 4) on water solution

~

( of the burnt polyurethane foam and fiberglass panel residues have been performed and the results indicate that no significant concentration of chemical spacies hazardous to the steel was found. If acids had formed, they would have been highly diluted by the large quantity of water used to extinguish the fire.

Also, the primer on the fireside is judged to be intact based on the condition of the primer on the drywell side. Corrosion of the containment steel is, therefore, considered to be minimal and structural analyses performed for the design basis conditions indicate that the steel containment has ample safety margin to accommodate the normal corrosion expected for the

{ remainder of the plant life.

i Analyses of the effects of chemical species in the water samples collected from the sand pocket drain lines also indicate that chemical attack on 72

the shielding concrete is highly unlikely and that it would take more than 200 years for the concrete cover to deteriorate to the point where the reinforcing steel would start to corrode.

4. Effects of Spalling Concrete and Polyurethane Residue Remaining Inside the Drywell Expansion Gap r

. Determine the effects of polyurethane and fiberglass residue as well as "hard spots" that may have been created by spalling concrete into the drywell expansion gap.

Determine the ef fects of potential "hard spots" on the drywell steel liner under pressure and temperature loads during normal operating and accident conditions.

I Response: Tests involving burning of polyurethane foam and fiberglass panel samples have been perfor=cd in the laboratory. The results show that the residuals are easily crushed by finger pressure. Analyses of the shielding concrete also show that concrete will not spall under the maximum credible fire temperature. The formation of "hard spots" is therefore unlikely. However, to verify the postfire structural integrity of the steel containment, a "hard spot" was, nevertheless, conservatively assumed. The stress analysis results show that regardless of the 73 E

L stiffness of the "hard spot," the steel containment stresses are within code allowable I

values for design pressure load and the design basis accident tempe rature load , respectively.

t i . Determine the compressive strength these "hard spots" must have to be of concern.

I Response: As indicated in the preceding response, formation l

of "hard spots" of any design consequence is l

unlikely, however, a hypothetical "hard spot" can i

have a stiffness range from zero to infinitive l without affecting the design functions of the structures.

4 1 5. Amount of Water Applied to the Drywell Expansion Gap to Extinguish the Fire

. Determine any thermal shock that may have occurred to the drywell steel liner.

Response: The highest temperature attained by the steel containment during the fire has been determined to be 500 F. Thermal shock in general is not a concern at this temperature, even less so for a one-time occurrence. In addition, the actual te...perature reading taken in the field during the 74

1 I

I event, showed a slow rate of cooling of the l

containment steel.

l

. Determine the amount of water used to extinguish the drywell expansion gap fire; how much of this water was removed; how much remains unaccounted for and what actions will be taken to remove any remaining moisture in the drywell expansion gap or in the surrounding structural and shielding concrete.

I Response: Water was supplied to the expansion gap to extinguish the fire. Black rubber hoses and thin wall conduit were used to deliver the water. The fire water system supplied water to the penetration in the RWCU room starting at 12:30 p.m. until 9:00 p.m. The clean-up demineralized water system supplied water to four additional penetration annular gaps from 2:00 p.m. until 9:00 p.m. Water was supplied through the annular gap around a total of five penetrations and was allowed to run through the expansion gap and .to collect o'n the floor in the torus basement. The collected water was processed by the radwaste system. The volume of water processed by the radwaste system was approximately 30,000 to 3 5,000 gallons. The volume of water supplied to extinguish the fire was calculated from the above 75 1

.. ___ -____ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __ _ _ _ _ _ ____________..______a

[

information and determined to be 34,000 gallons. ^This is in close agreement with the

(- volume of water processed by the radwaste system and acconnts for the water supplied in response to the fire. If any moisture remains - -in the expansion gap, it can escape by evaporation through the several penetrations in the shield

( wall.

6. Basic Drywell Liner and Structural and Shielding Concrete Design Functions

[.

. Determine to what extent (if any) the fire may have otherwise degraded the drywell steel liner's ability to provide a barrier which controls the release of fission products to the secondary containment.

[.

I Response: It has been determined that no change . of design

{

significance has occurred to the material

[ properties of the containment steel and that thermal shock is not a concern for the temperature experienced. Stress analyses of the steel containment for conditions during and after

{

the fire further verified the ability of the

( steel containment to ef fectively provide a

, barrier to control the release of fission .

products.

[ 76

l l

l . Determine to what extent, if any, the fire may have l

otherwise degraded drywell electrical or pipe penetrations l and the structural and shielding concrete design functions.

I Response: Fire damage assessments have been reviewed and the impact of the fire on the shielding capacity of the drywell wall has been qualitatively assessed. The conclusion is that even if the portions of concrete which reached temperatures above 2000 F (approximately 9 inches from the fire surface) were assumed to be inef fective, there is still more shielding in place in the affected area than at core midplane elevation where the neutron and gamma ray dose rate is the maximum.

The shielding integrity of the wall, therefore, has not been compromised.

7. Compliance with the Safe Shutdown Requirements of Appendix R to 10 CFR Part 50:

. Determine the effects of a fire of this nature on safe sh utdown capacity as prescribed in Section III G.2 of Appendix R to 10 CPR 50. During normal operation, this section requires redundant cables, including nonsafety circuits that could adversely affect safe shutdown capability th at are located in the same fire area outside of the primary containment, to be separated by a 3-hour 77

I I

fire barrier; be encased in a 1-hour fire barrier with automatic fire detection and suppr ession installed in the fire area; or be separated by a distance of more than 20 feet with no intervening combustible or fire hazards with automatic fire detection and suppression installed in the fire area. For normal operation of both Dresden Units 2 and 3, explain how such electrical cables and circuits passing through the drywell expansion gap are in compliance with the requirements of Appendix R so that a fire of this nature will not af fect safe shutdown capability during normal operations.

I Response: The effects of a polyurethane fire on safe shutdown capability were examined. It was found that safe shutdown can be achieved and maintained for any fire in the expansion gap. The fire itself cannot spread out of the expansion gap into other fire areas due to the existence of substantial barriers (the concrete wall and the electrical and mechanical penetrations). Fire starting outside the gap would not propagate into the gap, again due to the existence of the concrete wall and the mechanical and electrical penetrations. Ther e fore , safe shutdown capability will not be affected because of the inability of a fire in the expans ion gap to spread to neighboring fire areas, and the 78

inability of a fire to spread from neighboring fire areas into the expansion gap.

Since an alternative shutdown method is used for l

a fire in the expansion gap, an exemption from the requirements of Appendix R to 10 CFR 50 l

Section III.G.3 (requirements that fire detection and fixed fire suppression be installed) will be requested.

l I

l l

8. Potential Repairs Needed I

l . Determine the need for repairs (if any) to the drywell steel liner, structural and shielding concrete, electrical and pipe penetrations, or other affected equipment as a I

l result of the fire. Include in this assessment a time frame for completion and the impact of such repairs on l normal reactor operations.

I Response: The integrity of the steel containment, the concrete shield, electrical and pipe 1

penetrations, and other affected components has l been verified by detailed analyses to be intact after the fire. No repairs are, therefore, needed or recommended.

I I "

I I 9. Results of Water and Polyurethane Residue Samples I

. Provide the results of any and all extinguishing water and fire residue samples collected as a result of the fire for NRC review.

Response: Analyses of water sample collected at two sandbank drainline locations show the following types and quantities of chemical species:

I Type At 45 0 Azimuth At 135 0 Azimuth I Chloride 28 ppm 24 ppm PH 8.09 7.92

.I Sulfite 3 ppm 1 ppm l Sulfate 0.902 ppm 7.086 ppm

10. Corrective Actions to Prevent Recurrence:

I . Describe in detail the corrective actions that will be taken to prevent fires involving polyurethane material in the drywell expansion gap, including interim measures currently in place.

l I

• Response: Cutting and welding operations are governed by l

Procedure DMP4100-1, " Fire Prevention Procedure I for Use of Heat Sources in the Plant." Although this proced ure did caution against allowing hot slag to be ducted through nearby openings, the penetration gap area was apparently not covered l

with noncombustible protective material because l the welding personnel were unaware of the polyurethane foam. As part of an ongoing fire protection review, DMP4100-1 was revised February 23, 1986, to further clarify the use of l

fire watches. In order to prevent future l occurrences of this type of event, two further changes to DMP 4100-1 will be implemented.

l l First, the procedure will be made into a station administrative procedure to improve its l

implementation by all station working groups.

Second, the fact that hot slag or other hot l

material could travel along drywell penetrations l to reach the foam on the drywell exterior will be empr.asized. .

~

l

\

I Promptly following this event, the cause was thoroughly discussed with station construction and maintenance personnel. It was stressed that all such gap areas must be stuf fed with fire retardant sheeting prior to the start of cutting l

81

h

(;

or welding. The Station Prefire Plans will be j

{

enhanced to address this type of event.

11. Assessment of the Radiolyt..c and Thermal Decomposition of

[ Gap Materials

(

. Provide an assessment of the extent and results of the b radiolytic and thermal decomposition of materials in the drywell expansion gap in Unit 2 and an estimate of the

( ef fects of such decomposition on fire potential and containment structural integrity.

{

h Response: Pyrolysis of polyurethane yields the following combustion products;

a. carbon monoxide,

b. carbon dioxide,

[ c. water vapor,

d. hydrocyanic acid.

[

In its position outside the drywell, the polyurethane foam will be exposed to a maximum radiation exposure of x 10 7 based on 40

( 2.5 RADS, full years of reactor operation. Radiation data (Reference 12, 13 and 14) show the gamma radiation damage threshold to be between 8 x 10 6 and 4 X 10 7 RADS for polyurethane elastomers. Polyurethane foam samples, similar to that used in the gap, were

[ 82

~

{ irradiated at various levels from 10 7 to 109 RADS: there was no detectable change in resilience below 108 RADS, thus

. amply confirming the published data. Although the normal

- in-service temperature will be only 135 0F, the polyurethane which was used has a temperature rating of 2800 F (see FSAR

~

Subsection 5.2.3.6).

u

12. Other Areas with Similar Fire Potential L

. Provide a list of other plant locations where polyurethane or other combustible foam materials are installed in I concealed spaces. Identify whether these materials were explicitly addressed as part of fire hazards analysis.

I I Response: The expansion gap is not the only place where polyurethane is used as a filler material. Both l polyurethane and polyethylene have been used as a filler material at the top of block walls and l

polyurethane is used to seal penetrations. None of the block walls are considered rated fire 1

barriers and in those walls that use polyurethane as a penetration seal, either the wall is not a rated barrier or, if it is, the polyurethane has been replaced with a fire rated or noncombustible material. Thus, the polyurethane or polyethylene cannot provide a path for fire spread from one g

I *'

l I fire area to another. Table 7 provides a list of those fire zones where polyurethane and polyethylene are used as a filler at the top of block '.; alls and the percent increase in the fire loading of each zone resulting from this material. As can be seen, the percent change is very small, which is why it was not included in past. fire loading calculations. The amount of polyurethane used as penetration seals could not be easily determined. However, even if this amounted to two or three times that used at the top of walls, the percent change for any of the fire loading would still be small.

I I

I I

I I

I 84 I

F TABLE 1 L

CONCRETE TEMPERATURE PROFILES FOR CREDIBLE FIRE CASE TEMPERATURE, O F Concrete and Char Distance From

~ Surface of L Concrete (inch) 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> 0 500 500 E 3 303 382 L 6 173 279 9 111 199 m 12 88 145 L 15 82 112 18 80 95 21 80 86

{ 24 27 80 80 82 81 30 80 80 r 33 80 80 L 36 80 80 39 80 80 E

F l

1 l

1 l

l l

85

TABLE 2 CONCRETE TEMPERATURE PROFILE FOR HYPOTHETICAL FIRE CASE TEMPERATURE, OF Concrete Only Concrete and Epoxy I Distance From Surface of Concrete (inch) 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> 0 1800 1800 1596 1677 3 994 1317 836 1205 6 459 894 372 813 I 9 12 15 206 114 88 568 346 212 170 103 85 503 305 189 I 18 21 24 82 80 80 140 105 90 81 80 80 128 100 87 27 I

l 30 33 36 80 80 80 80 83 81 80 80 80 80 80 80 83 81 80 80 g 39 80 80 80 80 I

I I

I I

I I

I I

I ee au u i

w v r- r- r,_ t- v m v v r r r r - -

TABLE 3 RESULTS OF STRUCTURAL _ ANALYSIS OF CONCRETE SiiIELD FOR CREDI_BLE,500*F FI,RE, I

Maximum Value Loading Condition Load Combination Acceptance Criteria Ec in/in s ksi f

2-hour fire 0.0024 - 32.3 D+L+T ACI 318 USD f

6-hour fire 0.0021 - 37.2 D+L+T 6 e d 0.003 in/in f

8-hour fire 0.0020 - 37.2 D+L+T f f 54.0 ksi f

3 -

Unit Width.

e_

Interior Unit Width Face y f' = 4000 psi f = 60 ksi A

S = 2.03"2/ Ft A S

= 9.36"2j Ft y 1 1 91" 91" 1

Ag = 3.74"2jFt A S = 9.36"2j NOTATION: -

D = Dead Load H TION L = Live Load T = Fire Temperature Load f

Ag = Area of Reinforcinn Steel g = Compressive Strain in Concrete f = Stress in Steel, + for tension, - for compression s

'm m p m m m - w - - - - - - -

TABLE 4 RESULTS OF STRUCTURAL ANALYSIS OF CONCRETE SIIIELD FOR IlYPOTIIETICAL 1R00*P FIRE M xi Values Governing Loading Condition . Acceptance Criteria j Load Combination i b e in/in fy ksi During Fire ACI 318 USD 2 - hour fire 0.00083 +17.1 D+L+T f bc d 0.003 in/in I 6 - hour fire 0.00276 +33.8 D+L+T g fs s 54.0 ksi Postfire FSAR Table 12.1.1:1 Design Basis 0.0004 +26.6 D+L+Eg +Tg f 1800 psi (f $ 1500 psi) c C f e 30 kri m 3 -

m _

f' = 4000 psi Unit Width Interior Face Unit Width j= z' b r' f = 60 ksi Y .. --

0 Ag = 3.12"2jFt 82" 82" NOTATION:

Ag = 3.74"2/ 7 A = 9.36"27y D = Dead Load 2 2 L = Live Load I- U-l E O

= Operating Basis Earthquake MERICIONAL SECTION IiOOP SECTION I

T = Operating Temperature Load T = Fire Temperature Load A

8

= Area of Reinforcina Steel l G = Compressive Strain in Concrete fg = Compressive Stress in Reinforcing Steel, + for tension, - for compression f

C

= Stress in Concreto I

H L

TABLE 5 SAFE SHUTDOWN CABLES THAT PASS THROUGH UNIT 2 DRYWELL PENETRATIONS Cable Penetration Np. No. Safe Shutdown Equipment 22555 X-205E MO2-1301-1 22556 X-205E MO2-1301-1 1 22557 X-205E MO2-1301-1 l 22S58 X-205E MO2-1301-1 l 22572 X-200B MO2-1301-4 3 22573 X-200B MO2-1301-4 l 22574 X-200B MO2-1301-4 g 22575 X-200B MO2-1301-4 g 23938 X-200B 2-203-3B 23939 X-200B 2-203-3B 25069 X-2008 2-203-3B I 23946 23941 23942 X-200B X-205E X-205E 2-203-3B 2-203-3C l 2-203-3C I

l 23905 25068 23944 X-205E X-205E X-205E 2-203-3C 2-203-3C 2-203-3D 23945 X-205E 2-203-3D I

l 23906 25051 23947 X-205E X-205E X-205E 2-203-3D 2-203-3D 2-203-3E I 23948 23940 25070 X-205E X-205E X-205E 2-203-3E 2-203-3E 2-203-3E g 23935 X-200B 2-203-3A g 25061 X-200B 2-203-3A 25060 X-2003 2-203-3A 22655 X-205E LI2-263-106 A&B, LI2-263-ll6 & 117 I 22656 20687 X-204S X-204S LI2-263-106 MO2-1301-1, A&B, LI2-263-ll6 & 117 MO2-1301-2, MO2-1301-3 MO2-1031-4, A02-1301-17, A02-1301-20 20693 X-204S MO2-1301-1, MO2-1301-2, MO2-1301-3, I 26342 X-200B MO2-1301-4, A02-1301-17, A02-1301-20 A02-1301-17, A02-1301-20 26343 X-200B A02-1301-17, A02-1301-20 26353 X-200B A02-1301-17, A02-1301-20 26456 X-200B A02-1301-17, AO2-1301-20 26360 X-200B A02-1301-17, A02-1301-20 26370 X-205E A02-1301-17, A02-1301-20 26371 X-205E A02-1301-17, A02-1301-20 26385 X-205E A02-1301-17, A02-1301-20 26384 X-205E I 26352 X-200B A02-1301-17, A02-1301-20 A02-1301-17, A02-1301-20 Il ee

k.

l 1

TABLE 5 (Cont'd)

Cable Penetration No. No. Safe Shutdown Equipment 26346 X-200B A02-1301-17, A02-1301-20 26357 X-200B A02-1301-17, A02-1301-20 26359 X-200B A02-1301-17, A02-1301-20 26374 X-205E A02-1301-17, A02-1301-20 26373 X-205E A02-1301-17, A02-1301-20

( 26388 26387 X-205E A02-1301-17, A02-1301-20 X-205E A02-1301-17, A02-1301-20 22486 X-200B MO2-1001-1A 22464 X-200B MO2-1001-1A

{ 22485 X-200B MO2-1001-1A 22483 X-200B MO2-1001-1A 22846 X-205E MO2-1001-1B 22844 X-205E MO2-1001-1B 22845 X-205E MO2-1001-1B 22843 X-205E MO2-1001-1B 22498 X-205E MO2-0202-4A 22499 X-205E MO2-0202-4A 22500 X-205E MO2-0202-4A 22497 X-205E MO2-0202-4A E 22858 X-204S MO2-0202-4B 22859 X-204S MO2-0202-4B 22860 x-204S MO2-0202-4B 22857 X-204S MO2-0202-4B

[

[

[

[

(

90

i TABLE 6 SAFE SHUTDOWN CABLES THAT PASS THROUGH UNIT 3 DRYWELL PENETRATIONS Cable Penetration No. No. Safe Shutdown Equipment 32555 X-204M MO3-1301-1 f 32556 X-204M MO3-1301-1 L 32557 X-204M MO3-1301-1 32588 X-204M MO3-1301-1 32572 X-204S MO3-1301-4 32573 X-204S MO3-1301-4 32574 X-204S MO3-1301-4 32575 X-204S MO3-1301-4 32627 X-200C MO3-2301-4 32629 X-200C MO3-2301-4 32629 X-200C MO3-2301-4 32630 X-200C MO3-2301-4 33668 X-204S 3-203-3A 33935 X-204S 3-203-3A 33936 X-204S 3-203-3A 35060 X-204S 3-203-3A 33939 X-204S 3-203-3B 33938 X-204S 3-203-3B 33669 X-204S 3-203-3B 33946 X-204S 3-203-3B 33670 X-204M 3-203-3C 33941 X-204M 3-203-3C 33942 X-204M 3-203-3C 30568 X-204M 3,203-3C 32498 X-204M MO3-0202-4A

{ 32497 32499 X-204M X-204M MO3-0202-4A MO3-0202-4A 32500 X-204M MO3-0202-4A

(- 32957 X-200C MO3-0202-4B L 32858 X-200C MO3-0202-4B 32859 X-200C MO3-0202-4B 32680 X-200C MO3-0202-4B

( 33945 35051 X-204M X-204M 3-203-3D 3-203-3D 33671 X-204M 3-203-3D 33944 X-204M 3-203-3D

{ 35070 X-204M 3-203-3E 33948 X-204M 3-203-3E 33947 X-204M 3-203-3E 33672 X-204S 3-203-3E 36342 X-204S A03-1301-17, A03-1301-20 36343 X-204S A03-1301-17, A03-1301-20 36356 X-204S

{ 36360 X-204S A03-1301-17, A03-1301-17, A03-1301-20 A03-1301-20 91

L F

TABLE 6 (Cont'd)

Cable Penetration No. No. Safe Shutdown Equipment L

36370 X-204M A03-1301-17, A03-1301-20 f 36371 X-204M A03-1301-17, A03-1301-20 L 36384 X-204M A03-1301-17, A03-1301-20 36385 X-204M A03-1301-17, A03-1301-20 g 36345 X-204S A03-1301-17, A03-1301-20 L 36346 X-204S A03-1301-17, A03-1301-20 36359 X-204S A03-1301-17, A03-1301-20 36357 X-204S A03-1301-17, A03-1301-20

{ 36373 36374 X-204S X-204M A03-1301-17, A03-1301-17, A03-1301-20 A03-1301-20 36387 X-204M A03-1301-17, A03-1301-20 36388 X-204M A03-1301-17, A03-1301-20 32486 X-204S MO3-1001-1A 32484 X-204S MO3-1001-1A 32485 X-204S MO3-1001-1A 32483 X-204S MO3-1001-1A 32846 X-204M MO3-1001-1B 32844 X-204M MO3-1001-1B 32843 X-204M MO3-1001-1B 32845 X-204M MO3-1001-1B

[

m i

I I

I 5

I I e2

I TABLE 7 FIRE ZONES CONTAI'!ING POLYURETHANE AND/OR POLYETHYLENE FILLER AT THE TOP OF BLOCK WALLS Percent I Fire Material Quantity (ft3)

Increase in Fire Loading 1.1.1.2 Polyurethane 3.4 0.03 I 1.1.2.2 Polyure thane 3.4 0.03 1.1.1.3 Polyurethane 10.8 0.1 I 1.1.2.3 1.1.1.4 Polyurethane Polyethylene Polyurethane 9.8 4.9 8.6 0.1 0.07 0.3 1.1.2.4 Polyurethane 9.3 0.4 Polye thelene 6.7 0.4 1.1.1.5.D Polyethylene 3.8 3.4 1.1.2.5.C Polyurethane 2.8 1.6

.E 6.2 Polyethylene 4.9 0.04 5 7.0.A Polyethylene 2.3 0.04 8.1 Polyethylene 1.8 0.001 8.2.6.A Polyethylene 13.7 0.07 I 8.2.7 Polyethylene 6.5 0.1 9.1 Polyurethane 8.2 0.1 Polyethylene 1.5 0.03 9.2 Polyethylene 0.9 0.02 9.3 Polyethylene 1.5 0.03 14.1 Polyethylene 7.0 0.3 14.6 Polyurethane 18.7 0.8 I

I I

I I

I 93

(

TABLE 8 POSSIBLE EFFECT OF A FIRE ON UNIT 2 SAFE SHUTDOWN EQUIPMENT Cable Equipment Effect 22555 MO2-1301-1 No effect. This valve is normally open 22556 and must remain open for hot shutdown.

{ 22557 Cables 22555 and 22556 are power cables, 22558 and cables 22557 and 22558 are control cables connected to a limit switch. A C- fault in these cables or a loss of these cables will not change the valve position.

22572 M02-1301-4 No effect. This valve is normally open 22573 and must remain open for hot shutdown.

22574 Cables 22572 and 22573 are power cables, 22575 and cables 22574 and 22575 are control cables connected to a limit switch. A fault in these cables will not change the valve position.

23935 2-203-3A A fire that effects these cables could 23926 disable the target rock valve. However, 25061 the mechanical function of the valve and 25060 the safety valves will be available for RPV pressure control.

23938 2-203-3B A fire that affects these cables could 22573 2-203-3C disable these electromatic valves.

22574 2-203-3D However, the mechanical function of the 22575 2-203-3E target rock valve and the safety valves 23941 will be available for RPV pressure 23905 control.

[- 25068 23944

- 23945 23906 25051 23947 23948 23940 25070

[

s 94

k r

I TABLE 8 (Cont'd) l 22655 LI2-263-106A&B A fire in the gap that effected these 22656 LI2-263-116 cables could disable these indicators.

LI2-263-ll7 However, other indicators are available I to monitor reactor levels that have l

their essential and associated circuits routed independent of the gap.

20687 M02-1301-1 In the event that a fault in these l 20683 M02-1301-2 cables affects valves A02-1301-17 and i

l M02-13Gl-3 M02-1301-4 A02-1301-17 A02-1301-20, valve A02-1301-16 can be manually closed.

could close valves M02-1301-1, A fault due to a fire A02-1301-20 M02-1301-2, M02-1301-3, and M02-1301-1 4. This will not prevent achieving a I

safe shutdown, since a f ault in the HPCI cables routed in the same penetration will not change the HPCI valve position.

} 26342 A02-1301-17 In the event that a fault in one of 26343 A02-1301-20 these cables effects theca valves, valve I 26353 26356 A02-1031-16 can be manually closed.

l 26360 26370 1 26371 I 26385 g 26384 y 26352 l 26346 26345 1 26357 26359 26374 26373 22627 M02-2301-4 No effect. This valve is normally open 22628 and must remain open for hot shutdown.

I 22629 22630 Cables 22627 and 22628 are power cables, and cables 22629 and 22630 are control cables' connected to a limit switch. A fault in these cables will not change the valve position.

I 22486 22484 22485 M02-1001-1A The power feeds to this valve are routed through the penetration. In order to get to cold shutdown, the valve must be g 22483 opened. Af ter the drywell is made g accessible, these valves can be manually opened.

I I e5

k

(

TABLE 8 (Cont'd) 22846 M02-1001-1B The power feeds to this valve are routed 22844 through the penetration. In order to 22E45 get to cold shutdown, the valve must be 22843 opened. Af ter the drywell is made

{ accessible, these valves can be manually opened.

22498 M02-0202-4A The power feeds to this valve are routed

[ 22499 through the penetration. In order to 22500 get to cold shutdown, the valve must be 22497 closed. Af ter the drywell is made accessible, these valves can be manually opened.

1 22858 M02-0202-4B The power feeds to this valve are routed l C 22859 through the penetration. In order to 22860 get to cold shutdown, the valve must be 22857 closed. After the drywell is made accessible, these valves can be manually opened.

( ,

C E

[

[

C 96

TABLE 9 POSSIBLE EFFECTS OF A FIRE ON UNIT 3 SAFE SHUTDOWN EQUIPMENT Cable Equipment Effect 32555 MO3-1301-1 No effect. This valve is normally open 32556 and must remain open for hot shutdown.

32557 Cables 32555 and 32556 are power cables, i 32558 and cables 32557 and 32558 are control  !

cables connected to a limit switch.

I fault in these cables or a loss of the A

l cables will not change the valve pos i tlon.

i 32572 M03-1301-4 No effect. This valve is normally open 1

32573 and must remain open for hot shutdown.

32574 I 32575 Cables 32572 and 32573 are power cables, and cables 32574 and 32575 are control l cables connected to a limit switch. A fault in these cables or a loss of the I

j cables will not change the valve positlon.

32627 M03-2301-4 No effect. This valve is normally open 32628 and must remain open for hot shutdown.

l 32629 Cables 32627 and 32628 are power cables, I

l 32630 and cables 32628 and 32630 are control cables connected to a limit switch. A fault in these cables or a loss of the I

l cables will not change the valve position.

33668 A fire that affects these cables could I

I 33935 33936 35060 disable the target rock valve. However, the mechanical function of this valve and the safety valves will be available for RPV pressure control.

1 97

(

TABLE 9 (Con t 'd )

f 33939 3-203-3B A fire that affects these cables could L 33938 3-203-3C disable these electromatic valves.

33669 3-203-3D However, the mechanical function of the 33946 3-203-3E target rock valve and the safety valves 33670

( 33941 will be available for RPV pressure control.

33942 35068 C 33945 35051 33671

( 33944 35070 33948 33947

33672

( 36342 A03-1301-17 In the event that a fault in one of L 36343 A03-1301-20 these cable's affects these valves, valve 36356 A03-1301-16 can be manually closed.

36360 36370 36371 36384 f 36385 h '

36345 36346 36359

[ 36357 36373 36374 36387 C 36388 (L 32486 32484 MO3-1001-1A The power feeds to this valve are routed through the penetration. In order to 32485 get to cold shutdown, the valve must be 32483 opened . Af ter the drywell is made

( accessible, theso valves can be manually opened .

32846 Mo3-1001-1B The power feeds to this valve are routed

{ 32844 through the penetration. In order to 32845 get to cold shutdown, the valve must be 32843 opened . After the drywell is made accessible, these valves can be manually opened .

[  !

98

L

~

L l

TABLE 9 (Cont'd) 1 32498 MO3-0202-4A The power feeds to this valve are routed k ' 32499 through_the penetration. In order to 32500 get to. cold shutdown, the valve must be 1 32497 closed.- After the drywell is made accessible, these valves can be manually opened.

32858 MO3-0202-4B The proper feeds to this valve are 32859 routed through the penetration. In 32860 order to get to cold shutdown, this

32857 valve must be~ closed. Af ter the drywell

( ' is made accessible, these valves can be manually opened.

(

[ .

(

[ '

[

C

[ -

[

[

[

[

' 99

i m

?

q. Typical Pipe Penetration Joint Exhibit 1 r

5

?

L f

Fiberglass cover panet

. .h e@0,iT,Ni,5.%5*;9 .#

Potential path of hot slag v 6 0.i *" * **" M SP c Pipe sleeve o.0*' *

~}

'N' r b

, . . d 'i- V i$ _* e o 6- Epony caulking k 2 4nch gap 6

Penetration

[

O 2-inch gap

/

ff . . a'ffy

- Polyurethane foam Po,.n,1.,,a  !(thdIN$D h o ho,.,ag

. sn #easi ,..e..a.,

[

(

[

[

0 - 30

, Post-Fire Condition of Containment Coating Exhibit 2 I O *

>e Y

• h, $ 1 illi/li!

~

.un E l Pei a th

'I dEIE i1l-i 93I!

Ic h h'lsilgi ne i u .. .. /

j s' j h -

e

,4lt

' y i 2 c:. . c I

n

~

d.g,jf '; } _ .

.. o ......., _;

F 2

$ a k S

W O !jD t

6j)_1 ' ..

P

= v I

w

• ' D {31 Y e L

ks il Q -

~

PE .* .s a= _h y f g

g i .

, m' ~ @ &^ n ~*

I l

g "l

I

.o;e / .nr esu t,5

.h 4 .s Jg 5[2 a sa Ie c ..,-

T g e --

4 l

kj .iii~ g

• I HT7S1 _

l - -

...c....-7.,

g 3 ,

l 1

l v .{j!@}.(b

-i*. eus

+n h- =

9 ',f [

3 M flkh 8 L M f

s I f

=

I BE e5L 2 V 6800-2 03 86 - 30 I

w St:ci CcntrinmInt Tcmp;raturo Exhibit 3 Profiles

~

H f

L l

u Steel temperature (* F)

' 81 83 g .

1 7..

[ f f((j'18 "

)

v8 I

I g . ...p....... ,,

I I

I ra

L_.

I L Vertical Cross Section of the Exhibit 4 -

Steel Containment and the r Concrete Shield L

u

- l L

,.se w.. n s ,.y.scu, v. o . n. o . w .m ,c

. s p :e o.. .u. n ,s. < eo ..e . . . . . .: a . an;,8

? L o-:o..un,m.. m w ....:., sog I

k so,: D..

. ?9 P3 V:a t

1. . o5 be.,

I g.g c, ,

r j

i= -

ib. , ,

Concrete shield wall 18 ft 5 in, radius .,

s, 6 ft 6 in. 4-e. *-. = '

~ f& 'e .

g  %(.. ..

b Typical a .ee .

l '*

7. . , . 4 ,

O,? ' //o Fire location Elev. 537 ft 0 in.

2.in. gap Drywell

'oo ,,r i

4. . . around vessel no<-

ey, *e g{

gog between steel and concrete I

_ {g f g-5 y. g - Elev. 525 ft. 4 in.

l

Id w Steel d  %

containment ' .**%

'Ogf i ..

_ g g-' '

l . '.iY

6. - .'o Vent Elev. 502 ft in. Vent l g:

.w

.o Sand pocket .

.. Sand pocke

u..

8<

lRBasBNlt l =

Axisymmetric Shell Finite Element Model of Exhibit 5 the Steel Containment Vessel for Overall Analysis I

L t

[  !

i p ,

ei... .o. ,i s i..

I 66 R = 17 ft 4 in. _T l l l gso I l _

s -18 I

R = 18 ft 6 in.

l '

f' le l

1 46 3 4

-13 5

=

i j ele.. sso f t 1 1 2 in.

I I

l l i ..... s 3 o ,. .. , . .

-u l

~

o . 33,, -

, _ll l

l .

l l l +

l l

j

/ e,... soo ,, 3 . ...

=>,

~

Three-Dimensional Finite Element Model of Exhibit 6 I a Local Area of the Steel Containment Vessel l I

E I

E Ele,. $50 pg 1-1/2 in.

I ei... sso ,, ,.,,, , .

pj% ,

r iy,i l , 1 ,

i, ' ,

% . ( 's

\ s 4 ji E -

\ g g j'.A, I

• - \  ;

N

-l-

.Qtl-/s h{ ,

I TTO - -

l -

LL p.,1i

l ----- - - /

W , / ,Y

' /

,: : j.

a-- - - - -

/ ,: / . /

i '

- +- ' -

,g

,jI

'3 - , ' -

, /,

- u-LIi L L l-'

- ' /> / - />

/ /

1IIIJ-IIIIlL l -

- 'Q '

c

'L /ls.d}"1}-l,oEle. 530 og g.1'8 in-llll

l f 1

'L j. '

llll i Ellll' ~

z E

///// , ,

///ll I////f p'-

ei... sso ,i s.1,. io.

g E y 6800-6 Y CJ.86-30 E

F Merldlonal Stress Distribution in the Exhibit 7 Meridional Direction at 95: Azimuth -

Local Analysis

{

F L

Elev. 542.3 ft -

r 7

\

E E )

Elev. 537 f t I

h Location of fire i

I -

I ' I I I I I I Elev. 531.2 f t

-10 -10 -30 -40 -50 20 0 20 -60 Stress (ksi)

, =>.

b

{ Maridional Stress Distribution in the Hoop Direction at Elevation 537 Feet 0 Inch -

Exhibit 8 Local Analysis I

u Stress (ksi)

L, Ik

-40 -

+30 -

+20 -

+10 -

1 I I I l d% I l 11.2' 9.6

• 1 l l I 8.0

• 6.4

• 4.8

• 3.2 * ,1.6 ' o\F 1.6

• 3.2

• 4.8

• 6.4

• 8.0

• 9.6 ' 11.2' Azimuth 10 -

I

-20 -

\

-30 h Location of fire I \

-40 g .0

-,0 -

I I

I I

I =,

I

~ Typical Electrical Penetration Exhibit 9 Assembly Canister

- y  :

g

. 2 E -

E D d

- - \ _ S 5% - -

g N OOD

! i L >

% f5'&

F 3 -

g l

s Y////////// b

- 8 s s 5

\ s w 0

' s a s e N  : s E

\ C \ -,

( -s-N s 4'.

.a \ [ s . *-

s A-(

N C

s

?

N , . .

kO b ' *..'^ '  % '  %.

W - - ' ' d 7, f gW -

c' L w$

en g . . * . . , . Ns - - -_ -

1s s

d: -. - \ \ .... E r W 5 N \ < -

0

%W

.,'*- \ \

{ 5E

.?

.- i, N N 7, c i 1.

Il t

. N N , ' ' . '

meAq N P. Ml

[ 58 # h // S ei NN q /####//' k N s?

[ 9! # h t' 5

!!NP s-a

[

E P W' E c y- , a g NE

[ W n .m @

=

{ 6 '////////////////

E-ANNNN\\\\\\\\' N'N ij '////////////////

[ 98 n

O$

3k

[ ?H ' m

s Typical Mechanical Penetration Assemblies Exhibit 10

~

(Sheet 1 of 2)

L F

~ ,

i r , /) 5

- r -fm/ ,

'A

/

x

.o

$ # 1 w

c

/, -

L e o

r

( m Zr

!=

S

% [!" =3 w

c u .

i = V  : s %

}

L s~N e

sm -

)~( (X4/

f /, I g .1 e

~"

i, .. @

s  :

a

.g v # T "' s L ]

(7 J # e -

((J g -

9  ;

x= # s m 3

• l

.$ .9 9 .-

E i., 4 ;, s A  :, , J s urs__ ..

. u wv

( g I e -

w 3 b

_tu _.

m -_

i G

eN 42 &l  ;

~,

sG~

~

c

[,C, g; j b4 8 g g , -  % ,.

[ m e g

+ 4

- / ,+

w M

< A b h A

[

e n E / # y W e 8

[ E e

E N

s. 5

[  ! O v #

[ 6850-2 04 389

1 L

I L Typical Mechanical Penetration Assemblies Exhibit 10 (Sheet 2 of 2) r L

r  ;

a L sC o e

b $

I -

S u

g 0

=

3 C '

R E s= .8- s '

REW > i h b r '

255 x r

W 7 $

l  ;/ g L d -

/ O f

Y $  %

g

[

- z

[ I

$ 4 *c*.

A m & 24 I f l $34* c*.3C\

h. .T y , c .,

c$

i '

c,4  ;  ; a.

c

~

+ . sc I

k

, y(g

- C l l( Y $

' i ,4_

'y g i c

aj I J 5!

! fl4 ftV f( i I

i 'V ca l S- i

{ =

\

.I ,

S 6

, ',.-, . .i C

e, r j u a,., d. .- c .

l t =Q 'ae  ;  ;

M%% , EN % W Y -.-

l

\

/

/

e ,, gy e  ;- a 8

I EU2 "s

S P

5 7

_\t' 68W2 I 04 389

\

_ - . . _ _ . - _ J