ML20034D583
| ML20034D583 | |
| Person / Time | |
|---|---|
| Site: | Comanche Peak  |
| Issue date: | 11/13/1992 |
| From: | Khadijah West Office of Nuclear Reactor Regulation |
| To: | Mccracken C Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML082410566 | List: |
| References | |
| NUDOCS 9302170326 | |
| Download: ML20034D583 (17) | |
Text
-
November ~ 13, 1992 ENCLOSURE-7 MEMORANDUM FOR: Conrad E. McCracken, Chief Plant Systems Branch Office of Nuclear Reactor Regulation THRU:
Ralph E. Architzel, Chief Special Projects Section Plant Systems Branch Office of Nuclear Reactor Regulation FROM:
K. Steven West, Senior Fire Protection Engineer Special Projects Section
- Plant Systems Branch Office of Nuclear Reactor Regulation
SUBJECT:
TRIP TO OMEGA POINT LABORATORIES REGARDING TEXAS UTILITIE5 ELECTRIC COMPANY THERMO-LAG FIRE BARRIER TEST PROGRAM (TAC M83330)
On November 3 through 6,1992, I visited Omega Point Laboratories, San Antonio, Texas, to witness fire endurance tests of Thermo-Lag 330-1 fire barriers for Texas Utilities Electric Company (the licensee). The licensee declared both tests successful based on j
satisfactory post-fire inspections of the fire barriers and the cables. Subject to staff review of OPL's test reports, I concur with the licensee's conclusion. My trip report is enclosed.
.y..... + w -
K. Steven West, Senior Fire Protection Engineer-Special Projects Section Plant Systems Branch Office of Nuclear Reactor Regulation i
Enclosure:
As stated i
DISTRIBUTION See next page SPLI},g) SPLB.YV SWesP RArchitzel ll/ju92 11/2/92
[G:\\THERMOLA\\OPLTRIP2.SW]
9302170326 95bF10
~"
i PDR COMMS NRCC CORRESPONDENCE PDR
ENCLOSURE TRIP REPORT Facility:
Omega Point l2boratories, San Antonio, Texas Licensee:
Texas Utilities Electric Company Plant:
Comanche Peak Steam Electric Station, Unit 2 Docket No.: 50-446 Trip dates:
November 3 through 6,1992 Reviewer:
Steven West, NRR INTRODUCTION On November 3 through 6,1992, I visited Omega Peint l2boratories (OPL), San Antonio, Texas. The purpose of my visit was to witness fire endurance tests of Thermo-Lag 330-1 fire barriers performed by OPL for Texas Utilities Electric Company (TU Electric, the licensee). I met Joe Ulie, Region III Office of Investigations, there. My point of contact was Obaid Bhatty, TU Electric. In addition to Mr. Bhatty, I discussed the fire tests and the test program with I2nce Terry and Randy Hooten, TU Electric; Bob Braddy, Bechtel; Cal Banning and Rick Dible, ABB Impell; Chester Pructt, Fluor Daniel; Frank Collins, Stone and Webster, and Deggory Priest, OPL.
EXECUTIVE
SUMMARY
OPL performed two 1-hour fire endurance tests of Thermo-Lag 330-1 conduit fire barrier test assemblies (Scheme 9-1 and Scheme 10-1) for TU Electric. The fire barriers, which l
included a combination of vendor-recommended assembly methods and licensee-designed upgrades, were proposed for installation in Comm'.c Fed steam Electric Station (CPSES),
Unit 2. Both test assemblies were exposed to 1-hour ASTM E-119 standard fires followed by fog nozzle hose stream tests. Each cable's insulation resistance was measured before the fire exposures and immediately after the hose stream tests. After the hose stream tests, OPL and the licensee inspected each fire barrier, then disassembled each barrier and inspected its component pieces. After the post-fire megger tests, OPL removed the cables from the 7
i r
i
! conduits and inspected them for fire damage. OPL and the licensee concluded that the Thermo-Lag barriers remained intact during the fire exposures and hose stream tests without burning through or developing any openings. They also concluded that the cables were not damaged during the tests. OPL and the licensee declared both tests successful based on satisfactory post-fire inspections of the fire barriers and the cables. Due to erratic readings, OPL declared the temperatures taken from the conduit surfaces indeterminate.
I observed the megger te ts, the fire tests, and the hose stream tests and inspected the barriers and the cables after the tests. In my opinion, the barriers remained intact during the fire exposures and hose stream tests without burning through or developing any openings 1
through which either the raceways or the first layer of stress skin were visible. I did not see any visual cable damage. Subject to staff review of OPL's test reports, I believe both tests were satisfactory and qualified the subject Thermo-12g 330-1 fire barriers for installation at CPSES, Unit 2.
The test specimens included licensee-designed upgrades. Therefore, the results of these two tests do not affect NRC Bulletin 92-01 or its supplement. It is notable, however, that some sections of the fire barriers were assembled in accordance with the vendor's recommended installation procedures.
f t
OPL will conduct additional fire endurance tests for the licensee throughout November and early December. Some of these tests will be for CPSES, Unit 1.
l t
SCHEME 9-1 OPL tested Scheme 9-1 on November 3,1992. This test assembly (attachment 1) consisted of three conduits filled with a variety of power, control, and instrumentation cables. The specimens, which were protected by a combination of vendor-recommended assembly methods and licensee-designed upgrades, were:
i I
1 i A M-inch diameter condd with a radial bend and a lateral bend (LBD). This assembly was protected with %-inch thick preshaped Thermo-Lag 330-1 conduit
[
sections upgraded with an overlay of %-inch thick preshaped Thermo-Lag 330-1 conduit sections. The Thermo-Lag thickness was increased on the radial bend with stress skin and Thermo-Lag 330-1 trowel-grade material as specified in the licensee's site specifications. The LBD was enclosed in a box constructed of
%-inch thick prefabricated Thermo-Lag 330-1 panels. The LBD box joints were reinforced with stress skin and Thermo-Lag 330-1 trowel-grade material. This is.
l a licensee-designed upgrade.
A 3-inch diameter conduit with a radial bend and a LBD. The straight conduit runs were protected with %-inch thick preshaped Thermo-Lag 330-1 conduit sections installed in accordance with the vendor's recommended procedures. The Thermo-bg thickness on the radial bend was increased with stress skin and j
i Thermo-bg 330-1 trowel-grade material as specified in the licensee's site l
specifications. The LBD was enclosed in a box constructed of %-inch thick t
prefabricated Thermo-Lag 330-1 panels. The LBD box was upgraded by reinforcing the joints with stress skin and Thermo-Lag 330-1 trowel-grade f
r material.
A 5-inch diameter conduit with a radial bend and a LBD. The straight conduit runs were protected with %-inch thick preshaped Thermo-bg 330-1 conduit sections installed in accordance with the vendor's recommended procedures. The Thermo-Lag thickness on the radial bend was increased with stress skin and Thermo-bg 330-1 trowc:-7tade material as specified in the licensee's site i
i i
specifications. The LBD was enclosed in a box constructed of prefabricated Thermo-Lag 330-1 panels. The LBD box was upgraded by reinforcing the joints with stress skin and Thermo-bg 330-1 trowel-grade material.
i r
I s
! The test specimens were coated with the vendor's topcoat material and cured for at least 30 4
days before the fire test.
Thermocouples were attached to the outside surfaces of the conduits at 12-inch intervals and on the insulation of selected cables at 6-inch intervals. A total of 200 thermocouples were installed.
Before the fire exposure, Alexander Utilities Engineers (AUE), San Antonio, Texas, an OPL subcontractor, performed conductor-to-conductor and conductor-to-ground cable insulation resistance tests with a mega-ohmmeter. After the megger tests, OPL connected the cables into a single series circuit. This circuit was energized with low voltage (12 volts DC) to monitor circuit integrity during the fire exposure and hose stream tests.
The test assembly was exposed to the standard time-temperature fire found in ASTM El19 l
for one hour. The thermocouple temperatures were recorded every 60 seconds. The Thermo-Lag material began to burn abot t 2 minutes into the fire exposure and continued to j
burn throughout the fire test. After about 14 minutes, some of the temperatures on the i
conduit surfaces began to rise faster and higher than expected. The temperatures were irregular and were inconsistent with visual observations. For example after 31 minutes, a thermocouple on the 3-inch diameter conduit read 1480 *F.
This would indicate burnthrough or an opening in the barrier. However, by the end of the one hour fire exposure, this thermocouple reading had dropped to 468 *F. The post-fire inspection revealed that the barrier was intact and that virgin Thermo-Lag material remained under the char layer in this l
area. Thermocouples on the other two conduits experienced similar behavior.12ter (see the section on conduit surface temperature measurements, below), OPL declared the conduit i
surface temperature readings indeterminate.
After trying to find the cause of the thermocouple problems discussed above, OPL rebooted l
the data acquisition computers. This caused the loss of all thermocouple data for about two minutes. ASTM E119 specifies that temperature readings need not be recorded at intervals i
t less than five minutes. Therefore, the loss of the temperature data for about 2 minutes is not a problem.
Unlike the conduit temperatures, the cable insulation temperatures rose slowly and fairly uniformly during the one hour fire exposure. The cable insulation thermocouples did not experience the erratic behavior of the conduit thermocouples. The maximum cable insulation temperatures are shown in Table 1.
Conduit Cable Type Maximum Cable Diameter Tt.mperature I
K-inch Instrument l 290 *F t
f 3-inch Power 181 *F 1
3-inch i Control 250 *F i 3-inch Instrument 309 *F i-j 5-inch Power 134 *F
! 5-inch Control l
176 "F p
j l 5-inch
! Instrument !
189 *F Table 1. Scheme 9-1 Cable Temperatures After the one hour fire exposure, the test assembly was lifted from the furnace. Within about 5 minutes of its removal, the assembly was subjected to a hose stream test. OPL used a 30" fog nozzle with a nozzle pressure of 75 psi. The nozzle was located about 5 feet from the assembly. The water spray was applied for 5 minutes. The Thermo-Lag material bumed until it was extinguished by the hose stream. The hose stream dislodged negligible amounts of the char layer from the fire barrier surfaces.
AUE started the post-fire megger testing within about 10 minutes of the hose stream test.
The megger tests took about 1% hours to complete. While AUE was conducting the megger tests, OPL and the licensee inspected each fire barrier. After this inspection, OPL disassembled the Thermo-Lag barriers and inspected their component pieces and the raceways. OPL and the licensee concluded that the Thermo-Lag barriers remained intact during the fire exposure and hose stream tests without burning through (see the section on burnthrough, below) or developing any openings. After the post-fire megger tests, OPL removed the cables from the conduits and inspected them for fire damage. OPL and the licensee concluded that the cables were not damaged during the test (see the section on cable inspections, below). OPL and the licensee concluded that the three test specimen fire barriers met the licensee's conditions for acceptance for post-fire barrier condition and post-fire cable condition (attachment 2) and were, therefore, successful.
I observed the megger tests, the fire test and the hose stream test and inspected the barriers and the cables after the tests. I agree that the barriers remained intact during the fire exposure and hose stream test without burning through or developing any openings through j
which the raceways or the innermost layers of stress skin were visible. I did not observe any visible cable damage. Subject to staff review of OPL's test report, I believe the test was satisfactory and qualified the barriers for installation at CPSES, Unit 2.
SCHEME 10-1 OPL tested Scheme 10-1 on November 5,1992. This test assembly consisted of two 3-inch diameter conduits with a horizontal junction box, a venical junction box, and LBDs (attachment 3). The licensee designated the specimens " Front Conduit" and " Rear Conduit."
The assembly was protected by the following combination of vendor-recommended assembly methods and licensee-designed upgrades:
The conduit sections were protected with %-inch thick preshaped Thermo-Lag 330-1 conduit sections installed in accordance with the vendor's recommended procedures.
l i
i,
Each LBD was enclosed in a box constructed of prefabricated Thermo-Lag 330-1 panels. The LBD box joints were reinforced with stress skin and Thermo-Lag 330-1 trowel-grade material. This is a licensee-designed upgrade.
i Each junction box was enclosed within two layers of %-inch thick prefabricated l
Thermo-Lag 330-1 panels. The inside panels did not have ribs. The outside l
panels had ribs. The stress skin of both layers faced toward the junction boxes.
The joints were reinforced with stress skin and Thermo-Lag 330-1 trowel-grade i
material.
i I
The test specimens were coated with the vendor's topcoat material and cured for at least 30 1
days before the fire test.
[
Thermocouples were attached to the outside surfaces of the conduits at 12-inch intervals, to j
each surface of each junction box (inside the junction boxes), and to the insulation of selected l
t cables at 6-inch intervals. A total of 196 thermocouples were installed.
l f
Before 'he fire exposure, AUE performed conductor-to-conductor and conductor-to-ground l
u cable insulation resistarce tests with a mega-ohmmeter. After the megger tests, OPL f,
connected the cables into a single series circuit. This circuit was energized with low voltage (12 volts DC) to monitor circuit integrity during the fire exposure and hose stream tests.
i r
i The test assembly was exposed to the standard time-temperature fire found in ASTM E119 ~
-i for one hour. The thermocouple temperatures were recorded every 60 seconds. The j
Thermo-Lag material began to burn after about 2 minutes and continued to burn throughout l
the fire exposure. Some of the conduit surface thermocouples experienced the same erratic behavior experienced by Scheme 9-1. OPL declared the temperatures recorded from these l
thermocouples indeterminate.
t k
l t
.g.
The junction box and cable insulation temperatures rose slowly and uniformly during the fire exposure. These thermocouples did not experience the erratic behavior of the conduit thermocouples. The junction box temperatures are shown in Table 2. The maximum cable insulation temperatures are shown in Table 3.
?
1 s
Horizontal Junction Box Vertical Junction Box j
l Time Average
!. Maximum Average Maximum O min 63 *F l
63 *F 64 *F 65 *F i
i 60 min 172 *F 186 *F 146 "F 197 *F i
Table 2. Scheme 10-1 Junction Box Temperatures l
Specimen
! Cable Type Maximum Cable Designation j
Temperature I
Front Conduit Power 158 *F t
I Front Conduit Control i 233 *F i
Front Conduit
- Instrument 232 'F Rear Conduit i Power 2% *F i-Rear Conduit
! Control 174 'F i
Rear Conduit
! Instrument 232 *F Table 3. Scheme 10-1 Cable Temperatures I
After the one hour fire exposure, OPL removed the test assembly from the furnace and, within about 5 minutes of its removal, subjected it to a hose stream test. OPL used a 30*
fog nozzle with a nozzle pressure of 75 psi. The nozzle was located about 5 feet from the assembly. The water spray was applied for 5 minutes. The Thermo-Lag material burned '
untilit was extinguished by the hose stream. The hose stream dislodged negligible amounts of the char layer from the fire barrier surfaces.
3
(
F n
-e--
1 9
AUE started the post-fire megger testing about 10 minutes after the hose stream test. The megger tests took about 1% hours to complete. While AUE conducted the megger tests, f
OPL and the licensee inspected the fire barriers. After these inspections, OPL disassembled f
the Thermo-Lag barriers and inspected their component pieces and the receways. OPL and the licensee concluded that the barriers did not burn through and there were no open seams or joints. After the post-fire megger tests, OPL removed the cables from the conduits and inspected them for visual fire damage. OPL and the licensee concluded that the Thermo-Lag barriers remained intact during the fire exposure and hose stream test without burning f
through or developing any openings. They also concluded that the cables were not damaged.
OPL and the licensee concluded that the two test specimen fire barriers met the licensee's conditions for acceptance for post-fire barrier condition and post-fire cable condition and were, therefore, that the test was successful.
I observed the megger tests, the fire test, and the hose stream test and inspected the barriers and the cables after the tests. In my opinion, the barriers remained intact during the fire exposure and hose stream test without burning through or developing any openings through which the raceways or innermost layers of stress skin were visible. I did not observe any j
cable damage. Subject to staff review of OPL's test report, I believe this test was satisfactory and qualified the barriers for installation at CPSES, Unit 2.
i
_ CONDUIT SURFACE TEMPERATURE MEASUREMENTS i
During both fire tests, some conduit surface temperatures were inconsistent with visual observations. As discussed under the section on Scheme 9-1, above, the temperatures were i
irregular and higher than expected. The post-fire inspections revealed that the insulation of -
l some of the conduit thermocouple wires were saturated with a dark brown gummy substance i
that appeared.o be a mixture of water and decomposed Thermo-Lag material th'at had l
nigrated into the enclosures and condensed on the cool conduit surfaces. Mr. Priest informed me that the thermocouple wire used for the fire tests (fiberglass insulated) was not i
designed to operate in wet environments. He theorized that the saturation of the i
i i
i thermocouple insulation with the Thermo-Lag residue affected the temperature measurements and declared the conduit surface temperatures indeterminate. The thermocouples installed inside the conduits, i.e., on the cable insulation and inside the junction boxes, did not experience this problem. OPL will document this anomaly in its final fire test repons.
t The licensee's conditions of acceptance specify that the fire test is successful if the j
transmission of heat through the fire barrier does not raise the average temperature of the raceway surface 250 *F above its initial temperature and if the transmission of heat through f
the barrier does not raise the temperature of any single thermocouple 30 percent in excess of the specified average temperature. If the temperature rise exceeds either of these limitations, the licensee's criteria specifies that the post-fire condition of the fire barrier and the cables be assessed. If the barrier is ictact and the cables do not have any visual fire damage, the test is considered successful even if the temperature criteria are exceeded. The Scheme 9-1 and Scheme 10-1 fire tests met the licensee's conditions of acceptance for post-fire barrier I
condition and post-fire cable condition. 'Iherefore, the conduit surface temperatures are not needed. Declaring these temperatures indeterminate does not affect the fire test results.
Mr. Priest informed me that as a lesson learned, OPL will use either teflon insulated or Inconel shielded thermocouple wires between the Thermo-Lag material and the raceway for test specimens assembled in the future.
l BURNTHROUGH I
Burnthrough is defined as an opening in the fire barrier through which the raceway or a cable is visible. By design, Thermo-Lag barriers typically have at least one layer of stress skin between the raceway and the Thermo-Lag 330-1 material Cenain design upgrades and f
enhancements result in two or more layers of stress skin within the cross section of the fire barrier. Stress skin is not likely to burn away during the fire exposure or to be washed away j
by a fog nozzle hose stream test. The stress skin would, however, obstruct the view of the raceway. Therefore, burnthrough has also occurred if an opening develops in the Thermo-
i I -
Lag 330-1 material through which the innermost layer of stress skin is visible. That is, if the Thermo-12g 330-1 material was consumed during the fire and the resulting char layer is dislodged from the stress skin.
Post-fire inspection of the fire barriers revealed that most areas had virgin Thermo-Lag 330-1 material remaining. In some areas, the virgin Thermo-lag material was consumed during the fire, but a char layer remained attached to the underlying stress skin. The depth of the char layer ranged from about %-inch thick to more than 1 inch thick. I did not observe any areas that burned through.
h CABLE INSPECTIONS 1
After the fire, *~ 'se stream, and post-fire megger tests, OPL removed the cables from the l
raceways. The Scheme 9-1 cables were removed from the conduits about 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> after the i
fire exposure. The Scheme 10-1 cables were removed from the conduits about 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> after i
the fire exposure. OPL and the licensee visually inspected the cabies for jacket swelling, splitting, blistering, cracking, hardening, and discoloration; exposed shielding, conductor
- i insulation, and bare copper conductor; and conductor insulation degradation and discoloration. OPL and the licensee did not identify any of these attributes for any of the cables.
i I agreed with the results of these visual inspections. However, I noted that when I flexed the f
t cable segments that were exposed to the fire environment (heated cable) by hand, they felt stiffer than those that were not heated by the fire (virgin cable). Several cycles of flexing -
appeared to restore the flexibility of the cable segments that had been heated during the fire exposure. At my request, the licensee dissected two sections of the instrumentation cable l
that was installed in the M-inch conduit (Scheme 9-1). One section was heated. The other l
section extended out of the conduit and, therefore, was not heated during the fire test. My inspection of the heated cable section revealed that the cable jacket had not hardened, but the shielding material was constricted around the conductors, not unlike shrinkwrap. The r
_. shielding in the virgin cable was wrapped around the conductors but was not constricted.
The licensee informed me that the shrinkage of the shielding that I observed would not affect cable functionality.
CABLE INSULATION RESISTANCE MEASUREMENTS (MEGGER TESTING)
In a letter of October 29,1992, the staff documented the results of its review of the licensee's fire test methodology and acceptance criteria. In this letter the staff stated that megger tests (cable insulation resistance measurements) should be performed for instrumentation cables before the fire exposure, at least once during the fire exposure, and immediately after the hose stream test. AUE performed conductor-to-conductor and conductor-to-ground megger tests of the instrumentation cables before the fire exposure and after the hose stream test but did not megger them during the fire exposure. The licensee informed me that mcgger testing of the instrumentation cables was not performed during the fire exposure because: (1) the megger tests would require that the circuit integrity series circuit be disconnected, which would interrupt the circuit integrity monitoring, (2) OPL would not allow megger testing of cables while its data acquisition computers were connected 1
to the cable thermocouples, and (3) the cables will be at or near their maximum temperatures immediately following the hose stream test. The licensee believes that the worst case insulation resistance values would exist at the end of the test, when the cables are at their maximum temperatures, rather than midway through the test.
3 i
The licensee informed me that the cable insulation resistance values obtained during the I
megger tests were within their specified acceptable limits. I did not review the megger test
{
data during my trip.
l The licensee's acceptance criteria requires cable functionality tests only when the cables show
)
signs of fire damage or when the fire barrier burns through or opens up. These conditions were not observed during the subject fire tests. Therefore, the licensee's failure to conduct megger tests of the instrument cables during the subject tests is not a concern.
)
i
i
, CIRCUIT INTEGRITY MONITORING t
The licensee's test methodology calls for low voltage circuit integrity monitoring. Wh n I questioned the value of monitoring circuit integrity at low voltage, the licen.see stated that it was their understanding that the NRC staff required circuit integrity monitoring. I informed the licensee that staff guidance did not include circuit integrity monitoring and that the staff did not believe it would be beneficial unless it was done at rated voltage and current. I also stated that cable functionality could oe demonstrated without circuit integrity monitoring if F
the tests identified in the staff's letter of October 29,1992, were performed.
AMPACITY DERATING TESTS The licensee is constructing the following test assemblies for its upcoming ampacity derating tests: M-inch diameter conduit,1-inch diameter conduit,1%-inch diameter conduit,2-inch diameter conduit,5-inch diameter conduit, and 24-inch by 4-inch cable tray. Mr. Bhatty informed me that the ampacity derating tests will be conducted during December 1992. He informed me that the licensee has not provided any information on the test program for staff review, but plans to discuss it with the staff bcfore conducting the tests.
I b
b R
e b
b i
6 l
l UPChaDC LSD 903 UFs%
f 3.Dia LSD
.-e
\\ */
S k
- O I
I l
p h sia tao '
b g
Y
_ _ _ -.1_
_E_.
__ _/
7
- Die LSD 5
g g
g
,g ll
- - ff-4
r--------,----
I I
-6 3.,y PLAN VIEW
$* Des Ccf gesmt CSCf F tAf SEAL ST TU ELECTRtC -
- 50 bO i
E I E
Fant$ top wati
%*Tes ( Cov[A (fYPitt Las i
l l
l l
I r
l l
/,. r
,c,auari.
i l
lt BY LAS tivPs l
1 I
,,st.or..cos 41, A I,
b 5
[
l
[
l '* """'M pf en e sw-,
Pw-s=>ata oc l
t s
g4 - _ _I_ _
L_
ssst
't l
i I
I i
i t,,.
ELEVATION 1-1
+
TU ELECTRIC CPSES GLEN ROSE. TEXAS FIRE TEST ASSEMBLY
-}
- _ j...
3 m--
f
?
ATTACHMENT 1
~
Test Sequenct Build Sample Megger Temperature Fire Test Acceptance Sat Measure Conduit / Cable Tray Continuity
< Amb + 250 F avg Temperature
< Amb + 325 F peak
. Fog NozzelTest llorMegger Potential Deviation Fire BarncrInspection y
Functionality Tests
- Visual Cable Inspection
- Post Fire Sat Sat Fire Barrier Inspection Visual Cable inspection r
o if Required No Burn Through o
Potential Deviation u
Test Sat Potential Deviation Cable Functionality Verification
~
n Sat b
ATTACllMENT 2
__-__.u_
____._m._s
____._-__._.m__--_
-_.__-_am_-__.a_-.__m__h_.___---m_____a__._..Amm.
... - ~
~~.
~.
s 3'Osa cact M -
eartms.at l
csci aestr
- 1AL SY Tu CLECTRIC i
r bi E3 F 880 h
StV>n, TOP matt
%* fas ( Covin
?
1 LAR 2
s i
h P
l
=
l' !
I i \\ weir =r e Lk i
.,T
,ws,r,.,a.r o,,
t j
j i
t,,
>.. ir..
g Titweaso tac 330-s l
- M-SaAPG Cic
~'$ 3a3r%
SLC1sDN l-I l
i
\\
\\,
i 3 %>
.r....
! *s 3'
- [.D.ia. Lac,
sc5 I
..i t-s y t
a u
I m
s t
j
_m.-
___ f 4
v.o-t e
h ELEVATION 6
I I
i L
9 OI'
.4 '88 a
g330.i seig.e 8 90ast0 asaft
\\ _ _y tL a
i --
o m
-yg E T" 3'"'"
_ _ _j
! O - @. c_c d Z
t i
wi
_i
,,,r, A _
,;- - c t
4 s
,e..
m.
.i SECTION 1-1
'k 1
I
(
f I
i a
i 6
t TU ELET' ev
{
GLEN V FIRF
~
=&M1 liffe
}
t
.uk esh i
I i
I
ENCLO5URE 8
[
o,j UNITED STATES I
NUCLEAR REGULATORY COMMISSION e
- r, r-wAswiecToN D.C.20555 k....+ p October 23, 1992 6
MEMORANDUM FOR:
Conrad E. McCracken, Chief Plant Systems Branch Division of Systems Safety and Analysis THRU:
Ralph Architzel,. Chief Special Projects Section Plant Systems Branch Division of Systems Safety and Analysis' FROM:
Patrick M.. Madden Senior Fire Protection Engineer Special Projects Section Plant Systems Branch j
Division of Systems Safety and Analysis
SUBJECT:
TRIP REPORT CONCERNING TU ELECTRIC'S QUALIFICATION FIRE f
ENDURANCE TESTING OF THREM0-LAG 330 FIRE BARRIER SYSTEMS, l
COMANCHE PEAK STEAM ELECTRIC STATIONS, DOCKET NOS. 50-445 1
AND 446 On June 17-20, and August 19-21, 1992, I visited Omega Point Laboratories,.in
'i San Antonio, Texas, to witness fire endurance testing of 1-hour Thermo-Lag 330' i
fire barrier systems. This testing was being conducted by TU Electric to-i qualify the existing Thermo-Lag configurations _ currently installed in Comanche-Peak Unit I and configurations proposed to be used on Unit 2.
The following summarizes the observations made during the June 17-20, testing.
June 17, 1992, Test Article Tested - junction box with a 3/4,1, and 5-inch ' conduit
[
~
entering and exiting through the junction. box.
Test Observations-- Throughout the 1-hour fire endurance test, the-cabling routed inside the conduits was monitored in accordance with the American Nuclear Insurer's criteria for low voltage circuit integrity and continuity. None of the cables experienced a failure in circuit integrity. The' maximum temperature on the inside cover of the junction.
box reached 539'F and hot spots (temperatures on the cable in excess of f
500*F) on the 3/4-inch conduit and the 1-inch conduit occurred. The fire.
[
barrier _ material ignited within 2-minutes from the start of the test and j
continued to burn throughout the. fire test.
Flaming of the material-t continued after the specimen was removed from the furnace. (Note All
- l the test specimens observed during this series and.the_ August 19-21, j
series of' tests exhibited similar combustible characteristics).
In:
(
addition, I_ observed that the fire-barrier material on the test assembly.
did not stay intact when subjected to the hose stream test. The barrier
[
assembly showed signs of structural j
Yl
.]
Conrad E. McCracken I weakness and'a projection of water penetrated the barrier assembly in j
multiple locations.
1
-Post Test Observations.- On June 18, 1992, the cables were pulled from the test article. There were no visible signs of thermal degradation on'-
the cables routed in the 5-inch conduit. The cable inside the 3/4-inch conduit was thermally damaged in two locations.and cable in-the 1-inch-
.}
conduit was damaged'in one location.
June 18, 1992, Test Article Tested inch wide tray configuration without T-section.
Structural supports for the cable tray were protected for their entire length with Thermo-Lag. fire barrier material.
i Test Observations -LThroughout the fire endurance test.cthe thermocouple temperatures on the cables inside the test article were less than'325 "F.
The observations made during the hose stream test indicated that the fire barrier system showed signs of structural weakness and allowed 1
a projection of water to penetrated the barrier assembly in multiple locations.
Post Test Observations - A visual inspection of the ' cables was performed I
on June 19, 1992..The results of this inspection did not identify any.
signs of visual thermal-damage to the cables.
June 19, 1992, Test i
Article Tested - a 30-inch wide ladder back tray configuration with a'T-section.
Structural supports for the cable tray were protected for
~!
their entire length with Thermo-Lag fire barrier material.
]
t Test Observations - At 17 minutes into the test, the Thermo-Lag 330 panel on the bottom of the test. article began'to sag..'At 18 minutes, j
the joint at the interface between the tray. support and the tray howed j
signs of weakening and separation.
Internal temperatures.within teas of the test article exceeded 325'F at 25 minutes. ~ The. joint fully l
separated in 41 minutes resulting in cable circuit integrity failure and fire damage to the cables.
-l i
Note - As a result of these tests and the observations made by Ralph Architzel-j during the' construction of the test articles, several additional concerns were addressed with the licensee. The following is a summary of these concerns: 1) i ihe raceway supports for the test specimen are protected however they are not
-(
protected in the plant. The raceway supports for the in-plant configurations lj t
a l
'i t
Conrad E. McCracken !
are not protected with fire resistive materials; 2) As a result of the tray specimens tested (e.g.,30-inch wide tray with a T-section), the tests did not test the nominal band spacing allowed by TU Electric procedures for straight tray runs; 3) The amount of steel exposed to test the 9-inch coverage requirement to prevent a thermal short was not adequate. The 1.icensee did extend the steel to test for thermal short and demonstrated that protecting 9-inches of penetrating steel with the fire barrier material as the steel exits the barrier acts as a thermal break; and 4) thermocouples were not place on the cables near the cable tray side rails.
On August 19-21, 1992, TU Electric sponsored a second series of tests at the Omega Point Laboratory to aid in qualifying its Thermo-Lag 330 electrical raceway fire barrier systems for its Comanche Peak Steam Electric Station.
Accompanying NRR representatives at this series of tests were Ashok Thadani, Director, Division of Systems Technology, and Isabel Moghissi, Plant Systems Branch, Division of Systems Technology. The following is a summary of these tests:
August 19, 1992, Test Article Tested - Conduit configuration consisting of a 3-inch, 2-inch and 1-1/2-inch conduits and their associated cable lateral bend Doxes.
.l The 3/4-inch conduits were constructed using a Thermo-Lag 330 conduit pre-shape as a base material. The two 3/4-inch conduits were divided at the middle of the test specimen, and four different enhanced barrier systems were tested. The first of these consisted of a 3/4-inch conduit run, one half of which was protected by a 3/4-inch Thermo-Lag 330 fire barrier conduit pre-shape, and the other half protected with a 1/2-inch 7
thick conduit pre-shape with stress skin applied on the exterior and 1/4-inch of trowel grade Thermo-Lag applied to the stress skin. One half of the second 3/4-inch conduit run was protected by a 1/2-inch thick conduit pre-shape with a 1/4-inch thick Thermo-Lag flexi-blanket wrap. The other half was protected by a 1/2-inch thick conduit pre-shape with a 1/4-inch thick pre-shape overlay.
Test Observations - Throughout the fire endurance test, the thermocouple temperatures on the cables inside the 3/4-inch conduit protected by the overlay never reached 325'F. All other conduit configurations exceeded 325*F on the cables during the test. TU Electric did not conduct a hose stream test after the fire endurance test.
Post Test Observations - The visual inspection of the test specimen revealed that the interface joints between the vertical conduit runs and the cable pull boxes had opened and exposed conduit metal surfaces to the fire.
In addition, the ' cables exhibited visible fire damage to I
1o
- ij f
Conrad E. McCracken- *,
i cable jackets _'in all conduits, except for the 3/4-inch conduit ;
protected by the 1/2-inch thick conduit pre-shape with the-1/4-inch'
~'
' pre-shape overlay.
August 20, 1992, Test Article Tested inch wide ladder back tray with a T-tray configuration.
1 Test Observations - The thermocouples indicated that' internal.
-l temperatures in certain areas of the test article exceeded 325'F at 47
.4 minutes. The maximum monitored cable temperature during the test was.
1 381*F. _The hose stream test used for this' test consisted of applying a 30* fog spray of water onto the te'st specimen at a flow rate of 75 gpm at 75 psi form a distance of 5 feet for 5 minutes. This hose stream' 4
test had little structural impact on the barrier system.
l; Test Observations - The' visual inspection. of this specimen' revealed:that '
i five joint and seam type failures had occurred. These failures were.
'i both in horizontal and vertical; runs of the cable. tray.
Fire damage to, the cables was also identified during'the post-fire inspection.
j j
August 21, 1992, Test
}
Article Tested inch wide ladder back tray configuration.
Test Observations - Thermocouples indicated that internal temperatures -
1
~
in certain areas of the test article exceeded 325*F at 30 minutes. :The
- i maximum monitored cable temperature during the test was approximately
]
700*F.
Post Test Observation - The visual inspection of this_ specimen. five-joint and seam type failures were identified in horizontal and vertical-runs of the cable tray. The Thermo-Lag barrier also experienced areas -
of loss of its material, leaving spots of bare stress skin' covering the l
tray.
Fire damage to the cables was identified-during this inspection.
j l.
'i i
5 l
q
.l 1
i
.]
-Conrad E.'McCracken.
'i Attached is a list of individuals which were' present during this. series of j
' test-and post-test evaluation.
If you should have any questions or need any
,i additional information, please' advise.
angense W W l
1 L
Patrick M. Madden 1
Senior Fire' Protection Engineer.
.i Special' Projects Section i
Plant Systems' Branch j
Division of Systems Safety and Analysis'
{
i DISTRIBUTION l
SPLB TSI File Central File / TSI X005 NRC PDR FMiraglia 12 G 18 WRussell 12 G 18 AThadani
-j GHolahan l
CMcCracken RArchitzel-l PMadden j
MWidmann' 1
IMoghissi JFouchard (0PA). 2 G S GMulley (IG)
EW-542 EPawlik REG III SWest 9 A121 l
JLee 9A2 j
MCallahan (0CA) 17 A 3 J
'l l
.i
!l h
//
cn U PLY: SSA SPLB: SSA-S LB:DSSA-PMadden;cf RArchitzel CMcCracken 10/fg92 10/D/92 10/g/92 j
1
[G:\\ THERM 0LA\\TUTRIP.PMM]'
j
?
~
r
Attachment The following individuals were present for one or more of the fire endurance tests of the Thermo-Lag 330-1 Fire Barrier System sponsored by Texas Utilities and conducted at Omega Point Laboratories, Incorporated on June 17-22, and August 19-21, 1992.
Impell Corporation Calvin Banning Rick Dible National Institute of Standards and Technolooy Kenneth Steckler (at June 17-22, testing)
Omeca Point Laboratories Richard Beasley Kerry Hitchcock Connie Humphrey Deggery Priest Deggery Priest II Peak Seals Michael Jordan Sandia National Laboratory i
Steven Nowlen (at June 17-22, testing)
Stone and Webster Enoineerino Corporation Frank Collins Texas Utilities Electric Company Gene Beckett Obaid Bhatty Robert Braddy Randy Hooten Chester Pruett Melvin Quick Tim Wright Thermal Sciences. Incorporated Ben Evans James Rippe I
t b
4 e
i W
? 1>
U.S. Nuclear Reaulatory Commission James Gagliardo Patrick Madden Joe Ulie (at the June 17-22, testing)
Ashok Thadani (at the August 19-21, testing)
Isabel Moghissi (at the August 19-21, testing) r i
f I
I t
l I
L i
i l