ML20150F773

Insp Rept 50-424/88-24 on 880607-09.Violations & Weaknesses Noted Re Failure to Adequately Design & Install Water Tight Penetration Seals & Analysis of Licensee Failure.Major Areas inspected:880603 Operational Event Re Fire Protection

ML20150F773
Person / Time
Site:	Vogtle
Issue date:	06/27/1988
From:	Conlon T, Madden P NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION II)
To:
Shared Package
ML20150F772	List: ML20150F768 ML20150F773
References
50-424-88-24, NUDOCS 8807190045
Download: ML20150F773 (9)

See also: IR 05000424/1988024

Text

__ - . _.. . . _ . ._ . ..s

, .

,

-

UNITED STATES '

[km#E0pq' NUCLEAR REGULATORY COMMISSION ,

'W\ R EGION 11 i

N'I -

' - 7 ^S 101 MARIETTA STREET, N.W.

~ ('lg .h i ATLANT A, G EORGI A 30323

s p***.* j ,

>

Report No.: 50-424/88-24  ;

Licensee: Georgia Power Company i

P. 0. Box 4545

Atlanta, GA 30302

Docket No.: 50-424 License-No.': NPF-68

Facility Name: Vogtle 1  ;

Inspection Conducted: June 7-9, 1988

Inspec .: b4- _

G 7

P. M. dde , Reactor Engineer Date Signed.

Approved  : /

T. E. Conlon, Chief

I 7/87

Da'te 61gned

'

Division of Reactor Safety ,

'

I

SUMMARY f

Scope: This special announced inspection wts in response to an Operational  !

Event on June 3, 1988 in the area of fire protection.  ;

.

Results: As a result of the event, a major weakness was found in the  !

licensee's penetration seal design with respect to the seals

ability to provide a water T,ight barrier between redundant safe  ;

shutdown trains, paragraph 2. This condition resulted in a safety

issue which had not been reviewed.

'

'dithin the areas inspected, the following violation was identified:

.

,

Failure to adequately design and install water tight penetration

j seals and perform an analysis which evaluates -their failure as

"

required by License Condition 2.G, and 10 CFR 50, Appendix B,

Criteria III Design Control, paragraph 2.

x

!

8807190045 880629 $

PDR ADOCK 05000424 3

Q PDC

- _ _ _ _ _ _ _ _ . _ _ _ _ _ _ - _ _ _ _

'.

.

.

.

REPORT DETAILS

1. Persons Contacted

Licensee Employees

• G. Blockhold, General Manager

• W. Davis, Engineering Supervisor

• G. Fredrir e., QA-Site Manager

T.-Green, Plant Support Manager

• R. Lide, Engineering Support Superintendent

• J. Maddry, Unit 2 fire Protection Manager

• A. Mosbaugh, Assistant Plant Support Manager

• W. Nicklin, Regulatory Compliance Supervisor

• R. Pinson,-Vice President

• W. Ramsey, Unit 2 Project Engineering Manager

• P. Rice, Vice President and Project Director

• T. Richardson, Project Engineering Manager

• H. Varnadoe, Engineering Supervisor

Bechtel Employee

• R. Kies, Er.gineering Supervisor

Insulation Consultant & Management Service Employee

NRC Resident Inspectors

• J. Rogge

• R. Schepens

• Attended exit interview

2. Followup on Fire Protection / Operational Event (Modules 64704, 92701, &

92720)

On June 3,1988, at 0909 hours0.0105 days <br />0.253 hours <br />0.0015 weeks <br />3.458745e-4 months <br />, Central Standard Time, a fire alarm was

, received in the control room for fire alarm zone 120.1. This zone is for 8

the B train cable spreading room on level 2 of the control building.

Immediately following the receipt of the fire alarm for the '"B" train

cable spreading room a fire alarm was received, at 0910 hours0.0105 days <br />0.253 hours <br />0.0015 weeks <br />3.46255e-4 months <br />, in the

control room for fire alarm zone 179. This fire alarm zone provides-

protection for the electrical equipment room on level 3 of the control

building. Shortly after the receipt of the alarm condition _in fire alarm

zone 179 the licensee's control room operators noted that both the diesel

and the electric fire pumps started. At this time the control room

operators req. Jested the plant fire team, shift support operators, and

1

.

.

.

.

.

2

building and grounds personnel to report to level 2 and 3 of the control

building to investigate the actuation of the fire detection / protection

system. Upon arrival of the licensee's personnel in corridor 230 on level

2 of the control building outside the "B" train cable spreading room, it

was identified that the preaction sprinkler system (Fire Protection

System) valves for systems 078 and 081 had tripped, thus charging the

headers associated with the sprinkler system protecting the "B" train

cable spreading room. None of the sprinkler heads installed in the cable

spreading room sprinkler system fused by the pending condition and

discharged water. However, the solenoid pilot valve associated with the

preaction sprinkler valve trip mechanism actuated as a result of a signal

from the cable spreading room smoke detection system. The pilot valve

upon actuation remains open and discharges water until the system is

reset. Thus, upon actuation of the sprinkler system 078 and 081 pilot

valves opened and water was discharged onto the floor of room 230 through

3/8 inch diameter tubing downstream of the pilot valve. During the event

the plant floor drain system associated with Room 230 and the "B" train

cable spreading room removed the majority of the inadvertent water

discharge. However, the water discharge from these valves seeped under

the "B" train cable spreading room door and about a 1/4 to 1/2 inch of

water accumulated over approximately 1/3 of the cable spreading floor

area. Upon entering the "B" train cable spreading room licensee personnel

identified a faint odor of smoke. The licensee's staff performed an

investigation of the cable spreading room and found no evidence of fire.

The licensee also sent personnel to level 3 of the control building to

conduct further investigations. On level 3 an odor of smoke was present

in electrical equipment room 322 and upon further investigation the

licensee's staff determined that there was no evidence of fire. Upon

completion of the licensee's investigation on levels 2 and 3 and the fact

that no fire was present, the licensee reset the fire protection systems

to stop the flow of water at 0934 hours0.0108 days <br />0.259 hours <br />0.00154 weeks <br />3.55387e-4 months <br />,

At 0954 hours0.011 days <br />0.265 hours <br />0.00158 weeks <br />3.62997e-4 months <br />, during the clean up operations, the control room received a

pressurizer high level deviation alarm and the pressurizer backup heaters

began to cycle on and off and pressurizer power operated relief valve

(PORV) PV ',56A opened. The pressurizer PORV opened for approximately 5

seconds, and system pressure in the reactor coolant system dropped to

approximately 2200 PSIG. The licensee's operations staff took prompt

action and manually shut the PORV and block valve. In addition, the

operations staff placed the backup pressurizer heaters in manual control.

At 1010 hours0.0117 days <br />0.281 hours <br />0.00167 weeks <br />3.84305e-4 months <br />, licensee personnel reported water leaking into the control

room from the cable spreading room above. Based on the licensee's

investigation it was determined that small amounts of water infiltrated

the process monitoring panels directly underneath the cable spreading

room.

1

The licensee's post event investigation revealed that Unit 2 personnel

where performing a flow balance test on the control building cable

spreading room level 2 and 3 HVAC system at the time of the event. In

performing the test the licensee's personnel opened the outside air irtake

damper (AHV-12910) to make an adjustment and the inrush of cooler outside

air activated the thermostat on electric duct heater (A-1539-H7-001).

This energized the heater coils and the initial burnoff of dust caused a

smoke odor which was carried through the ductwork and distributed into

. . . _ _ _ . .- _ _ - _ _ - _

.- . .

.

.

-

r

'

t

3

,

i

levels 2 and 3 of the common Unit I and 2 control -building. This  !

,

condition appears to have actuated the smoke detectors which. tripped  !

preaction sprinkler systems' 078 and 081. i

In response to - the event the licensee. took the follo' wing immediate

-

l

corrective actions

'

• The Fire Protection Systems involved were reset to stop water flow.

• The control room process panels were covered to prevent further

wetting, and the panel doors'were opened to prevent overheating.

i

• The PORV PV-456A was declared inoperable and the LCO was initiated.

'

• A fire watch was established for the level 2 and 3 fire detection

zones that alarmed.

'

• The Pressurizer backup heaters were placed in' manual control.

• Sand bags were placed inside the Unit 1 "B" cable spreading room door.

• All wetted plant areas were mopped and/or toweled dry.

In addition to the above corrective actions, the licensee took the

following actions to ensure that the safety systems affected by the event l

were functional and to further prevent water from entering the cable

spreading / control room:

'

,

• The licensee initiated a work request to install RTV-732 silicone caulk

'

i between the damming angle and the concrete floor around all penetration l

seal assemblies in the cable spreading room.

1

l *The licensee's engineering staff wrote a special test procedure to test '

"

the affected pressurizer pressure and level channel. -i

i

• The licensee tested the affected pressurized pressure and level channel

on June 5,1988 and the PORV was placed back in service.

• The licensee placed the fire protection sprinkler system 078, 080, and

l- 081 out of service on 6-5-88 and established a continuous fire watch in

the affected areas. ,

l

'

In response to the event, on June 7, 1988, the NRC Region II Office

dispatched an inspector to the site to evaluate the event and the

licensee's interim corrective actions. The inspector performed a walkdown

of the Unit 1 "B" train cable spreading room and the control room and

-

inspected the area which was affected by the June 3,1988 event. During

i this walkdown the inspector concentrated his efforts on inspecting the

cable spreading room fire / water penetration seals which were identified as

_ . _ - _ . _ _ _ _ _ _ - _ _ _ _ _ - _ _ _ - _ - _ _ _ _ _ . _ _ _ - _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ - - . ._--____- - . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - _ - - _ _ _ _ _ _ _ _ _ _ _ _ _ - - - _ _ _ _ _ _ - _ - -

. .

.

.

.S

.

4

,

being degraded with respect to being water tight and ensuring that the

licensee's interim corrective actions were appropriate.

During the event, cable spreading room penetrations seals 347, 355, 325,

and 357 allowed water to leak into the control room and the ceramic

j damming board on the control room side of penetration seals 311, 312, 344,

and 345 showed discoloration from water seepage through these seal. The

licensee as a result of their post event investigation determined that

water entered the- control room by seaking under the steel penetration

i damming angle, which is seal welded to the penetration embed plate, then

seeping down through the concrete voids behind the concrete and. the

penetration blockout steel embed plate. The licensee to preclude this

type event from occurring in the future applied RTV-732 silicone sealant

around the perimeter of all the floor penetrations in the Unit 1 "B" train

cable spreading room. The inspector indicated that this configuration

should be tested to demonstrate that this modification would correct the

design error.

In addition, the inspector was concerned about water

impingment to the top of the seals from the overhead Fire Protection

System in the cable spreading room and water infiltration into the

control room. The inspector indicated that according to the licensee's

design criteria DC-1003, "Flooding", Section 3.2.J, states that flooding ,

'

caused by activation of the Fire Protection System shall not result in

failure of a safety-related system to perform its function and therefore, .

the permanent seal modifications, in response to this event, should take t

this into consideration. The licensee indicated that they had test 1

documentation which demonstrated that the silicone penetration sealant

material was water tight. In addition, as a result of the inspector's -

concern the licensee performed smoke tests on seven "B" train cable

spreading room penetration seal assemblies (317, 318, 319, 320, 347, 355,

& 357). As a result of these tests, it was determined that penetration

seals 318 and 347 had leakage through the cable bundles penetrating the

silicone sealant material.

The licensee provided their test documentation, test report HT-E01-08,  ;

Hydrostrahc Test For Electrical Blockouts, which the licensee claims to  :

qualify the Vogtle penetration seal design as being water tight. The

, inspector reviewed the documentation and determined that the as tested

configuration did not represent the worst case "as-built" configuration

which is currently installed in the floor of the Unit 1 "B" train-cable

spreading room. The licensee's test report documented the test of a 24 j

inch by 24 inch penetration seal with five conductors and a 12 inch i

aluminum open back cable tray penetrating the seal. The seal was i

installed in a steel open bottom box with a removable lid, then two inches I

-

of water was applied to the top of the seal along with 3 PSIG (Nitrogen) I

hydrostatic pressure for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. This seal configuration did not allow

any leakage during the specified test duration. However, based on the

'

inspector's review of this test, it appears that the only conclusion that

can be made from the results of this test is that the silicone foam

(RTV 3-6548) material is water resistant and that the test configuration

in no manner represents the in plant penetration seal installations. In

addition, the test configuration did not address the silicone-to concrete

3

interf ace, or the cable configuration (i.e. ,100 or more cables routed

through a penetration blockout) of the in plant installations.

l

l

!

l

,_ _,_!

___ _ _ . . . _ , _ _ _ _ __ __ . , . ~ , _ _ _ _ , _ , , _ _ _ -

.

.

.

5

The licensee, on June 8, 1988 and June 9, 1988, in an effort to determine

the degree of the penetration seal leakage problem and to qualify the

RTV-732 caulk modification to the Unit I upper cable spreading room

penetration seals as being water tight, ran a series of water leakage

tests on Unit 2 upper cable spreading room seals. It should be noted that

the seal design for Unit 2 is the same as those seals which were installed

on Unit 1. The inspector witnessed these tests and the following is a

summary of the results:

Test No. 1

The purpose of this test was to demonstrate the licensee's theory on how

the water leaked into the control room the cable spreading room.

The licensee constructed a water dam assembly around the penetrations 1060

(33" by 11 1/2" penetration opening) and 1061 (17" by 11 1/2" penetration

opening) in the Unit 2 upper cable spreading room. The penetration design

of penetrations 1060 and 1061 were in conformance with the typical design

details as noted on the licensee's des'gn detail AXJG11-65-2 except

penetration seal 1060 had not been filled with the silicone taam seal

material. With the seal material not yet installed ia penetrat on 1060

direct observations during the test of the water leak path could be made.

The water dam area around penetrations 1060 and 1061 was filled with

approximately 1/2 inch of water on June 8, 1988 and approximately 5

minutes into the test, seepage was observed in penetration 1060 and the

water appeared to be seeping through the concrete voids which were present

behind the blockout embed plate. At 10 minutes into the test both

penetration assemblies showed visible signs of leakage in the control room

below.

The licensee's theory on how the water 9 into the control roen appears

to be correct.

Test No. 2

The purpose of this test was to qualify the licensee's Unit 1 siliceno

caulk modification.

On June 9, 1988, penetration 1060 and 1061 were tested to demonstrate that

the licensee's modification which installed silicone RTV-732 caulk around )

the penetration steel damming angle where it interfaces with the concrete i

floor is water tight. The water dam around the penetration assemblies was i

then filled with four inches of water. This condition was to be held for '

I hour. Approximately 4 minutes into the test leakage from penetration

1061 into the Unit 2 control roc: was observed.

The leakage appeared to be concentrated in the corners of the penetration

assembly. The water leakage through seal 1061 was at the rate of about 1

drop of water every 10 seconds. The four inch water level was held around

penetration seal assembly 1060 for approximately 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> without any signs

of leakage. The licensee's analysis of the leakage through seal 1061

- . .- . - .- . - .

.. .

-

.

.

.

I

6

indicated that the silicone RTV-732-did not cure properly due to moisture

under the steel damming angle. The moisture under the angle was directly -

attributed to the test conducted on June 8, 1988.

'

Test Nos. 3 & 4

The purpose of this test was to demonstrate that the licensee's ' silicone

caulk modification was water tight and to demonstrate that their overall

penetration seal design / assemblies were water tight.

On June 9,- 1988, penetration assembly 1117 (11" by 23" penetration

opening), which was a completed assembly with greater than 100 conductors

penetrating the silicone penetration sealant material, was tested.with

respect to it's water. tight properties. The water dam which, was r

constructed around ' the penetration assembly, was initially filled with

four inches of water and held for 30 minutes to verify that the caulk

modification was water tight. This portion of the test (Test No. 3) was

successful and no water leakage was identified in the control room during

the duration of the test. At the end of the 30 minutes the water level in

the water dam was raised to a level of two inches above the seal.

Approximately, two minutes into this test (Test No. 4) significant leakage

was identified in the control room below. The major portion of the

'

leakage into the control room appeared to be down through the cable

bundles. To avoid any damage to the control room components the licensee

terminated this portion of the test promptly once the leakage path was.

established and identified.

Test No. 5 .

.

The purpose of this test was to demonstrate that the licensee's

penetration seal design with limited conductors penetrating the sealant

,

material was water tight.

'

On June 9,1988, penetration seal 1107 (11" by 15" penetration opening),

'

which was a completed seal with only two conductors penetrating and had  ;

the silicone caulk material installed, was tested to demonstrate that a  :

penetration seal with limited conductors penetrating the silicone seal ,

material was water tight. The water dam around the penetration seal '

, assembly 1107 was filled with water until a level of two inches of water

above the seal was achieved. Approximately 20 minutes into the 30 minute

test leakage of water on the control room side of the seal wat identified.

The leakage which was identified appeared to be originating and seeping

through the northwest corner of the assembly where the silicone sealant

material came in contact with the steel damming angle. In order to

preclude any damage to control room equipment / components this test was

terminated by the licensee once the leakage path was identified.

As a result of the June 3,1988 event, the failure of Unit 1 "B" train

cable spreading room penetration seals to pass the smoke test, and the

failures associated with the Unit 2 penetration seal -tests as documented

above, the licensee's penetration seal design and installation failed to

1

l

l

. ._ _ _. ___ .__ _ _-_._ __ _ _ . .

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

. .

,- .

.

.

7

meet Vogtle Unit i license condition 2.G. License condition 2.G., ,

'

requires that the licensee implement and maintain in effect all the

provisions of the approved fire protection program as described in the

final safety analysis report for the facility, as approved in the NRC's-

Safety Evaluation Report (SER NUREG-1137) through Supplement 5. In the  ;

Vogtle Unit 1 SER dated June 1985, Section 9.5.1.1, "Fire Protection- i

Requirements", states that 10 CFR 50, Appendix A, General Design criteria

(GDC) 3 requires that fire fighting systems be designed.to ensure that

rupture or inadvertent operation does not significantly impair the safety

capability of structure, systems, and components important to safety. The

licensee by letter dated February 22, 1985 stated the components required

for hot shutdown are designed so that rupture or inadvertent operation of

the fire suppression systems will not adversely affect the operability of

these components. In addition, the licensee stated that redundant trains

of components that are susceptible to damage from water spray are '

physically separated so that manual' fire fighting activities will not

adversely affect the operability of the components not involved in the -

postulated fire.

Contrary to the above, prior to the June 3,1988 event, a postulated fire

in the Unit 1 "B" train cable spreading room with either automatic-

actuation of the cable spreading room sprinkler system and/or manual fire

suppression activities, could have rendered the "B" train of safe shutdown

equipment inoperable. In addition, as demonstrated by the June 3, 1988

event, the condition of the cable spreading room penetration seals with

respect to their ability to preclude water from entering the control room,

"A" train safe shutdown components could have been rendered inoperable or

could have spuriously operated in an adverse manner. The auxiliary 4

shutdown panel under this postulated fire condition would not be available ,

for shutdown since this panel is associated with the "B" train. Thus, the  :

postulated fire in the "B" train cable spreading room causes a loss of the

"B" train safe shutdown functions, the auxiliary shutdown panel, and due

to the water leakage into the control room from the fire suppression

activities in the cable spreading room above, "A" train safe shutdown

process monitoring functions, since these circuits are not isolated from

the control room, could have been rendered inoperable. Failure to perform

a complete analysis of the impact of the above postulated event is

contrary to license condition 2.G., and 10 CFR 50, Appendix B,

Criteria III requirements and is identified as violation 424/88-24-01.

i

The inspector reviewed the licensee's interim corrective actions as

previously described above and found them to be acceptable. In addition, .

the licensee has established a continuous fire watch in the Unit 1 "B"  !

train cable spreading room as a compensatory measure and the licensee will  !

maintain it in effect until an acceptable water tight penetration seal I

design can be tested and installed. The inspector found the licensee's  !

compensatory measures (which exceeded the associated fire protection

action statements) and interim actions to provide adequate justification l

with respect to continued plant operation.

i

l

1

,

)

. .

, .

.

4

8

3. Exit Interview

The inspection scope and findings vere summarized on June 9,1988, with

those person identified in paragraph 1 above. The inspection findings

identified in this report were originally identified as a Unresolved Item

pending NRC management review. Based on NRC's management's review of the

issues a subsequent telephone exit interview was held with the licensee on

June 16, 1988 and the unresolved item was identified as a violation. The

inspector described the areas insp?cted and discussed in detail the

results of the inspection and the ffnding listed below. The licensee did

not identify as proprietary any of v.he material provided to or reviewed by

the inspector during this inspection. Dissenting comments were not

received from the licensee.

Item Number Description and Reference

424/88-24-01 Violation - Failure to adequately design and

install water tight penetration seals and

perform an analysis which evaluates their failure

as required by License Condition 2.G and

10 CFR 50, Appendix B, Criteria III Design

Control, paragraph 2.

l

l

l

1

- - - - - - - - _ - -

\