[Federal Register Volume 61, Number 70 (Wednesday, April 10, 1996)]
[Notices]
[Pages 16005-16016]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 96-8909]
-----------------------------------------------------------------------
NUCLEAR REGULATORY COMMISSION
All Licensees of Reactors With Installed Thermo-Lag Fire Barrier
Material; Issuance of Director's Decision Under 10 CFR 2.206
Notice is hereby given that the Director, Office of Nuclear Reactor
Regulation, has acted on Petitions for action under 10 CFR 2.206
received by a letter dated September 26, 1994, from the Citizens for
Fair Utility Regulation and the Nuclear Information and Resource
Service; by a press release dated October 6, 1994, from the Maryland
Safe Energy Coalition; by separate letters dated October 21, 1994, from
the GE Stockholders' Alliance and Dr. D. K. Cinquemani; by a letter
dated October 25, 1994, from the Toledo Coalition for Safe Energy; by a
letter dated October 26, 1994, from R. Benjan; by a letter dated
November 14, 1994, from B. DeBolt; and by a letter dated December 8,
1994, from the Nuclear Information and Resource Service and the Oyster
Creek Nuclear Watch. The Petitioners requested that the U.S. Nuclear
Regulatory Commission (NRC) take action with regard to the use of
Thermo-Lag by reactor licensees and that their letters be treated as
Petitions pursuant to Section 2.206 of Title 10 of the Code of Federal
Regulations (10 CFR 2.206).
The Citizens for Fair Utility Regulation and the Nuclear
Information and Resource Service requested (1) Texas Utilities Electric
Company, the licensee of Comanche Peak Steam Electric Station, Unit 1,
perform additional destructive analysis for Thermo-Lag configurations
in proportion to the total installed amount to determine the degree of
``dry joint'' occurrence, (2) the licensee perform fire tests on
upgraded ``dry joint'' Thermo-Lag configurations for conduit and cable
trays to rate the barrier as a tested configuration in compliance with
fire protection regulations, and (3) the NRC immediately suspend the
Comanche Peak Unit 1 license until the above listed corrective actions
are taken. The Maryland Safe Energy Coalition requested immediate
shutdown of both reactors at the Peach Bottom plant until the risk of
fire near electrical control cables due to combustible insulation is
[[Page 16006]]
corrected. Dr. Cinquemani and the Toledo Coalition for Safe Energy
requested that the NRC immediately shut down all reactors where Thermo-
Lag is used until it has been removed and replaced. The GE
Stockholders' Alliance requested shutdown of all reactors where Thermo-
Lag is used until it has been removed and replaced with fire-retardant
material meeting NRC standards. R. Benjan requested immediate shutdown
of all reactors where Thermo-Lag is used. B. DeBolt requested shutdown
of all reactors in which Thermo-Lag is used until it has been removed
and replaced. The Nuclear Information and Resource Service and the
Oyster Creek Nuclear Watch requested that NRC immediately suspend GPU
Nuclear Corporation's (GPUN's) operating license for Oyster Creek
Nuclear Generating Station (OCNGS) until GPUN removes Thermo-Lag fire
barrier material and replaces it with a competitive product that meets
current NRC fire protection regulations.
The Director of the Office of Nuclear Reactor Regulation has
determined that these requests should be denied for the reasons stated
in the ``Director's Decision Under 10 CFR 2.206'' (DD-96-03), the
complete text of which follows this notice, and which is available for
public inspection at the Commission's Public Document Room, the Gelman
Building, 2120 L Street, N.W., Washington, D.C., and at the Local
Public Document Room for the named facilities.
A copy of this Decision has been filed with the Secretary of the
Commission for the Commission's review in accordance with 10 CFR
2.206(c) of the Commission's regulations. As provided by this
regulation, this Decision will constitute the final action of the
Commission 25 days after the date of issuance unless the Commission, on
its own motion, institutes review of the Decision within that time.
Dated at Rockville, Maryland, this 3rd day of April, 1996.
For the Nuclear Regulatory Commission.
William T. Russell,
Director, Office of Nuclear Reactor Regulation.
Appendix A to This Document: Director's Decision Under 10 CFR 2.206
DD-96-03
I. Introduction
By letter dated September 26, 1994, the Citizens for Fair Utility
Regulation and the Nuclear Information and Resource Service (NIRS); by
press release dated October 6, 1994, the Maryland Safe Energy
Coalition; by separate letters dated October 21, 1994, the GE
Stockholders' Alliance and Dr. D. K. Cinquemani; by letter dated
October 25, 1994, the Toledo Coalition for Safe Energy; by letter dated
October 26, 1994, R. Benjan; by letter dated November 14, 1994, B.
DeBolt; and by letter dated December 8, 1994, NIRS and the Oyster Creek
Nuclear Watch (the Petitioners), requested that the U.S. Nuclear
Regulatory Commission (NRC) take action with regard to the use of
Thermo-Lag by reactor licensees and that their letters be treated as
Petitions pursuant to Section 2.206 of Title 10 of the Code of Federal
Regulations (10 CFR 2.206).
The Citizens for Fair Utility Regulation and NIRS requested that
(1) Texas Utilities Electric Company (TU Electric), licensee of
Comanche Peak Steam Electric Station, Unit 1, perform additional
destructive analysis for Thermo-Lag configurations in proportion to the
total installed amount of Thermo-Lag to determine the degree of ``dry
joint'' occurrence, (2) the licensee perform fire tests on upgraded
``dry joint'' Thermo-Lag configurations for conduit and cable trays to
rate the barrier as a tested configuration in compliance with fire
protection regulations, and (3) the NRC immediately suspend the
Comanche Peak Unit 1 license until the above corrective actions are
taken. The Maryland Safe Energy Coalition requested immediate shutdown
of both reactors at the Peach Bottom plant until the risk of fire near
electrical control cables due to combustible insulation is
corrected.1 Dr. Cinquemani and the Toledo Coalition for Safe
Energy requested that the NRC immediately shut down all reactors where
Thermo-Lag is used until it has been removed and replaced. The GE
Stockholders' Alliance requested shutdown of all reactors where Thermo-
Lag is used until it has been removed and replaced with fire-retardant
material meeting NRC standards. R. Benjan requested immediate shutdown
of all reactors where Thermo-Lag is used. B. DeBolt requested shutdown
of all reactors in which Thermo-Lag is used until it has been removed
and replaced. NIRS and the Oyster Creek Nuclear Watch requested that
NRC immediately suspend GPU Nuclear Corporation's (GPUN's) operating
license for Oyster Creek Nuclear Generating Station (OCNGS) until GPUN
removes Thermo-Lag fire barrier material and replaces it with a
competitive product that meets current NRC fire protection regulations.
\1\ The Petition submitted by the Maryland Safe Energy Coalition
expressed several concerns in addition to the fire hazard issue.
These other issues, that is other than the fire hazard issue, will
be the subject of a separate Director's Decision.
---------------------------------------------------------------------------
As a basis for their requests concerning Thermo-Lag 330-1 fire
barrier upgrades, the Citizens for Fair Utility Regulation and NIRS
Petitioners stated that (1) the licensee's records on the original
installation of Thermo-Lag fire barriers on conduits and cable trays
indicate that its contractor followed specifications for pre-buttering
all joints; (2) NRC Inspection Reports 50-455/93-42 and 50-446/93-42
found, based on destructive analysis documents, that a concern did
exist where Thermo-Lag conduit joints fell apart easily and did not
appear to have any residual material of a buttered surface, indicative
of a joint that had not been pre-buttered; (3) the ``dry joint''
deficiency appeared in Room 115A and other areas of the unit; (4) the
licensee directly contradicts an NRC inspector's findings that were
determined in part by destructive analysis; (5) the ``dry joint'' or
absence of pre-buttering of Thermo-Lag panels can be determined only by
destructive analysis and cannot be determined by a walkdown visual
inspection; (6) the findings reported in the Comanche Peak Unit 1
Region IV Inspection Reports 50-455/93-42 and 50-446/93-42, based on
the limited amount of destructive analysis conducted at the unit,
constitute a substantial documentation of installation deficiencies
found in Thermo-Lag fire barriers as documented in NRC Information
Notice (IN) 91-79, ``Deficiencies in the Procedures for Installing
Thermo-Lag Fire Barrier Materials,'' December 6, 1991, and IN 91-79,
Supplement 1, ``Deficiencies Found in Thermo-Lag Fire Barrier
Installation,'' August 4, 1994; (7) neither the NRC nor the industry,
by its agent Nuclear Energy Institute (NEI), nor a utility, have
conducted fire tests on dry-fitted or ``dry joint'' upgraded
configurations of Thermo-Lag 330-1; and (8) the presence of ``dry
joint'' upgraded configurations in Comanche Peak Unit 1 constitutes an
untested application of Thermo-Lag fire barriers.
As a basis for the requests concerning Thermo-Lag 330-1 fire
barrier upgrades, the Maryland Safe Energy Coalition stated that the
manufacturer of the flame retardant (Thermo-Lag insulation) was
indicted on criminal charges (of falsifying tests of the effectiveness
of the insulation as a fire barrier), and fire near the electrical
control cables, due to combustible Thermo-Lag insulation, could cause a
catastrophic meltdown.
As the bases for their requests, Dr. Cinquemani, the Toledo
Coalition for
[[Page 16007]]
Safe Energy, the GE Stockholders' Alliance, and R. Benjan stated either
individually or collectively that (1) the widespread use of Thermo-Lag
in more than 70 reactors presents a safety crisis; (2) the NRC has
known since 1982 that Thermo-Lag fails NRC performance standards for
material that protects vital electrical cables for ampacity rating and
fire resistance; (3) Thermo-Lag has failed not only NRC tests, but
almost all other independent tests; (4) Thermo-Lag is combustible,
contrary to NRC regulations, and is an ineffective fire barrier; (5)
the use of Thermo-Lag could lead to shorts, to failure of the cables in
an emergency, and to fire; (6) Thermo-Lag is faulty in that fraudulent
ampacity ratings allowed utilities to use smaller cable than permitted
by design requirements, causing the cable to overheat and its
insulation to deteriorate; (7) the NRC has stated that fire at some
nuclear power plants can contribute as much as 50 percent of the risk
to a core meltdown, and a typical reactor will have three to four
significant fires during its licensed lifetime; (8) Thermal Science,
Inc. (TSI), the manufacturer of Thermo-Lag, and its President were
indicted by a Federal grand jury on seven criminal charges related to
conspiracy to defraud the U.S. Government in regard to the
effectiveness of Thermo-Lag; and (9) the hourly fire watches at the
Davis-Besse Nuclear Power Plant operated by Toledo Edison do not
replace fire barrier material and do not prevent fires.
As the bases for his request, B. DeBolt stated that Thermo-Lag
fails to meet NRC regulations concerning combustibility and that the
manufacturer of Thermo-Lag was indicted for defrauding the Government
and the utilities. Among the many bases for their request, NIRS and the
Oyster Creek Nuclear Watch stated that (1) Southwest Research Institute
(SwRI) conducted fire tests on Thermo-Lag 330-1 specimens for GPUN and
reported that all specimens ignited approximately 2 seconds after it
was inserted into the furnace and failed specified criteria because of
flaming after the first 30 seconds of testing, an outside temperature
rise higher than 30 deg.C, and a weight loss of 50 percent; (2) GPUN's
operation of OCNGS with knowledge of the SwRI report is an example of
GPUN's reckless disregard for fire protection and public safety; (3) in
the event of fire, Thermo-Lag is likely to fail its intended function
of protecting vital electrical cables running from the control room to
plant safety systems used to shut down the reactor; (4) current
installations of Thermo-Lag are likely to fail in less time than 1 hour
(when smoke detectors and automatic sprinkler systems are present) or 3
hours (when there are no fire detection and suppression systems) that
NRC regulations require for fire barriers to withstand fire; (5) the
NRC Inspector General issued a report in August 1992 condemning NRC's
handling of the Thermo-Lag issue and documenting the NRC staff's
failure to understand the scope of the problem; (6) in April 1994,
Industrial Testing Laboratories and its President pleaded guilty to
five felony counts of aiding and abetting the distribution of falsified
test data; (7) on September 29, 1994, the U.S. Department of Justice
issued a seven-count indictment against the manufacturer of Thermo-Lag
and its Chief Executive Officer for willful violations of the Atomic
Energy Act, conspiracy to conceal material facts, and making false
statements to defraud the United States in connection with $58 million
in fire barrier material; (8) GPUN has known since at least August 11,
1992, that Thermo-Lag 330-1 as a structural base material is
combustible and that GPUN was in violation of Appendices A and R to 10
CFR Part 50 and the NRC Standard Review Plan, NUREG-0800; (9) GPUN
failed to report the SwRI test results in response to a request for
additional information regarding Generic Letter (GL) 92-08 (``Thermo-
Lag 330-1 Fire Barriers'') of February 10, 1994, when asked to describe
the Thermo-Lag 330-1 fire barriers installed as required to meet 10 CFR
Part 50, Appendix R; and (10) continued reliance on fire watches at
OCNGS is an unreasonable and unnecessary hazard to the public health
and safety because of an inoperable fire protection system for safe
shutdown of the reactor and installed combustible material on the
shutdown systems.
On November 7, 1994, I informed the Citizens for Fair Utility
Regulation and NIRS that the request for an immediate suspension of the
Comanche Peak Unit 1 operating license was denied. On December 2, 1994,
I informed the Maryland Safe Energy Coalition that the request for an
immediate shutdown of the Peach Bottom plant and for an immediate
suspension of the Peach Bottom license was denied. On December 15,
1994, I informed the GE Stockholders Alliance, Dr. D. K. Cinquemani,
the Toledo Coalition for Safe Energy, and R. Benjan that the immediate
suspension of the operating licenses of all reactors where Thermo-Lag
is used was denied. On January 3, 1995, I informed NIRS and the Oyster
Creek Nuclear Watch that the immediate suspension of the OCNGS
operating license was denied. On January 19, 1995, I informed B. DeBolt
that the request for immediate suspension of the operating licenses of
all reactors in which Thermo-Lag is used was denied. The decisions were
based on the following: (1) the staff is addressing deficiencies in
fire barriers constructed with Thermo-Lag material as part of a
Commission-approved action plan and has issued several bulletins and a
generic letter to the nuclear industry to provide information and
guidance, (2) fire barrier systems constructed with Thermo-Lag have
been identified and declared inoperable, and (3) compensatory measures
(fire watches) approved by the NRC have been instituted. Additionally
in the above correspondence, all Petitioners were informed that the
Petitions were being treated pursuant to 10 CFR 2.206 and had been
referred to this office for action pursuant to 10 CFR 2.206 of the
Commission's regulations and that appropriate action would be taken
within a reasonable time.
For the reasons stated below, the Petitions have been denied.
II. Background
The picture painted by the Petitioners of inaction by the NRC staff
in responding to the issues presented by the use of Thermo-Lag is at
odds with the facts. A review of the chronological development of the
issues shows that the NRC staff has been working diligently to resolve
the issues and has consistently sought to ensure that there is adequate
protection of the public health and safety. It is also inaccurate to
contend that Thermo-Lag generic deficiencies have been known since
1982. As can be seen from the following information, the development of
the Thermo-Lag issue has been evolutionary. Reports of problems
regarding Thermo-Lag began to surface in the late 1980s when Gulf
States Utilities, the licensee for River Bend Station, discovered some
cracks and wear damage due to installation deficiencies (Licensee Event
Report 87-005, March 25, 1987) and declared the material inoperable as
a fire barrier. The licensee further discovered that stress skin was
missing on all 3-hour Thermo-Lag fire barriers in the turbine building
as a result of an installation error. In a series of plant-specific
tests performed by Gulf States Utilities in 1989, Thermo-Lag barriers
failed to meet the fire endurance test acceptance criteria. Gulf States
Utilities categorized all 1-hour and 3-hour barriers as indeterminate
and implemented compensatory measures in the form of fire watches.
[[Page 16008]]
Other isolated plant-specific fire protection problems had been found
during NRC inspections at various utilities as early as 1982 and had
been acted on by the NRC staff. These problems were treated as plant-
specific issues and were not considered as indications of generic
problems.
In February 1991, the NRC received allegations that Thermo-Lag did
not provide fire protection for electrical cables as claimed by the
vendor. In response, in May 1991, the NRC visited River Bend Station to
review the installation procedures and the failed fire endurance tests
and concluded that a generic concern existed with 30-inch-wide cable
trays. The NRC alerted the industry of the results of the test failures
in IN 91-47, ``Failure of Thermo-Lag Fire Barrier Material To Pass Fire
Endurance Test,'' August 6, 1991.
In June 1991, the Office of Nuclear Reactor Regulation (NRR)
established a special review team to investigate the safety
significance and generic applicability of technical issues regarding
allegations and operating experience concerning Thermo-Lag fire
barriers. In its final report, which was issued with IN 92-46,
``Thermo-Lag Fire Barrier Material Special Review Team Final Report
Findings, Current Fire Endurance Testing, and Ampacity Calculation
Errors,'' June 23, 1992, the special review team reached the following
conclusions:
The fire-resistive ratings and the ampacity derating
factors for the Thermo-Lag fire barrier system were indeterminate.
Some licensees had not reviewed and evaluated the fire
endurance test results and the ampacity derating test results used as
the licensing basis for their Thermo-Lag barriers to determine the
validity of the tests and the applicability of the test results to
their plant designs.
Some licensees had not reviewed the Thermo-Lag fire
barriers installed in their plants to ensure that they met NRC
requirements and guidance, such as that provided in GL 86-10,
``Implementation of Fire Protection Requirements,'' April 24, 1986.
Some licensees used inadequate or incomplete installation
procedures during the construction of their Thermo-Lag barriers.
After the special review team completed its charter, the NRC staff
prepared an action plan that provided a process to resolve technical
issues identified with Thermo-Lag fire barrier systems. The NEI,
formerly the Nuclear Management and Resources Council (NUMARC), agreed
to coordinate industry efforts to resolve the issues.
In regard to the Petitioners' allegations of NRC's inaction in
responding to the issues presented by the use of Thermo-Lag, the
significant progress made by the NRC staff and the nuclear reactor
licensees in resolving Thermo-Lag issues speaks to the contrary. The
NRC staff has issued a number of generic communications related to
Thermo-Lag, which include the following: (1) two bulletins: BUL 92-01,
``Failure of Thermo-Lag 330 Fire Barrier System To Maintain Cabling in
Wide Cable Trays and Small Conduits Free From Fire Damage,'' June 24,
1992, and BUL 92-01, Supplement 1, ``Failure of Thermo-Lag 330 Fire
Barrier System To Perform Its Specified Fire Endurance Function,''
August 28, 1992; (2) two generic letters: GL 92-08, ``Thermo-Lag 330-1
Fire Barriers,'' December 17, 1992, and GL 86-10, Supplement 1, ``Fire
Endurance Test Acceptance Criteria for Fire Barrier Systems Used To
Separate Redundant Safe Shutdown Trains Within the Same Fire Area,''
March 25, 1994; and (3) 12 information notices: IN 91-47; IN 91-79; IN
91-79, Supplement 1; IN 92-46; IN 92-55, ``Current Fire Endurance Test
Results for Thermo-Lag Fire Barrier Material,'' July 27, 1992; IN 92-
82, ``Results of Thermo-Lag 330-1 Combustibility Testing,'' December
15, 1992; IN 94-22, ``Fire Endurance and Ampacity Derating Test Results
for 3-Hour Fire-Rated Thermo-Lag 330-1 Fire Barriers,'' March 16, 1994;
IN 94-86, ``Legal Actions Against Thermal Science, Inc., Manufacturer
of Thermo-Lag,'' December 22, 1994; IN 95-27, ``NRC Review of Nuclear
Energy Institute, Thermo-Lag 330-1 Combustibility Evaluation
Methodology Plant Screening Guide,'' May 31, 1995; IN 95-32, ``Thermo-
Lag 330-1 Flame Spread Test Results,'' August 10, 1995; IN 95-49,
``Seismic Adequacy of Thermo-Lag Panels,'' October 27, 1995, and IN 94-
86, Supplement 1, ``Legal Actions Against Thermal Science, Inc.,
Manufacturer of Thermo-Lag,'' November 15, 1995.
The NRC staff, the nuclear industry, and others have expended much
time and many resources to address and resolve the Thermo-Lag issues.
The NRC staff developed comprehensive fire test guidance and acceptance
criteria and worked with industry to improve existing ampacity test
procedures. The NRC staff and industry performed about 100 fire
endurance and ampacity derating tests of Thermo-Lag fire barrier
materials and full-scale test assemblies. The fire endurance tests
established the limitations and the true fire-resistive capabilities of
certain Thermo-Lag fire barrier configurations, without relying on the
fire endurance test data supplied by TSI, the manufacturer of Thermo-
Lag. On the basis of some of these tests, the NRC staff concluded that
existing Thermo-Lag barriers could be upgraded with some additional
Thermo-Lag material to satisfy NRC regulations. Precluding all use of
Thermo-Lag materials for current and future fire barrier installations
would remove a realistic option for resolving safety issues. Therefore,
the NRC staff does not object to the use of Thermo-Lag in specific
applications, where, through upgrades, NRC requirements are satisfied.
The NRC staff issued three requests for additional information (RAIs)
regarding GL 92-08 to each licensee using Thermo-Lag to obtain
information on the specific Thermo-Lag material installed at each
plant. The NRC staff reviewed and approved comprehensive Thermo-Lag
fire barrier programs proposed by TU Electric for Comanche Peak Steam
Electric Station, Unit 2, and by Tennessee Valley Authority (TVA) for
Watts Bar Nuclear Power Plant, Unit 1, which attests to the fact that
Thermo-Lag barriers can meet NRC fire protection guidelines and
requirements. The NRC staff completed toxicity tests of Thermo-Lag
material. The NRC staff and the industry completed chemical
composition, combustibility, and flame spread tests of Thermo-Lag
materials. Finally, the NRC staff reassessed previous technical
conclusions to determine the extent to which the NRC staff and industry
relied on information supplied by TSI to reach these conclusions. The
staff had concerns about the reliability of information and data
supplied by TSI that have been or could be used to make judgments
regarding Thermo-Lag materials. The NRC staff identified and
categorized the issues and previous conclusions and used the results of
the industry-wide testing program regarding the chemical composition of
Thermo-Lag, as discussed below, to determine if the in-plant Thermo-Lag
materials were consistent. The results of this reassessment indicated
that previous technical conclusions were valid independent of the
information provided by TSI. The staff therefore concluded that
additional action to reassess the issues or reverify the previous
conclusions was not needed.
The NEI testing program on the chemical composition of Thermo-Lag
analyzed samples from 18 utilities representing 25 nuclear power
plants. The samples represented Thermo-Lag material manufactured
between 1984 and 1995. NEI performed pyrolysis gas
[[Page 16009]]
chromatography evaluation of 169 samples to assess organic chemical
composition and performed energy-dispersive X-ray spectroscopy of 33
samples to assess inorganic chemical composition. On the basis of the
tests, NEI concluded that (1) all of the samples contained the
constituents identified by TSI as essential to fire barrier
performance; (2) the composition of the samples was consistent; and (3)
the test results provided a basis on which to close NRC questions about
chemical composition and product consistency and for utility use of
generic test data relative to fire endurance ratings, flame spread,
heat release, ampacity derating, and other material properties.
The NRC staff test program on the chemical composition of Thermo-
Lag was conducted by the National Institute of Standards and Technology
(NIST) during 1992 and 1995. NIST analyzed 21 samples that were either
collected by the staff during site visits to plants and test
laboratories or provided by TVA, Gulf States Utilities, Commonwealth
Edison Company, and NEI. The analysis included elemental and ammonia
analysis, pyrolysis, gas chromatography, mass spectrometry, and X-ray
fluorescence. These analytical techniques indicated that all of the
samples were similar in their bulk chemical composition. These results
were consistent with the results of the NEI chemical testing program
pertaining to the chemical composition and uniformity of Thermo-Lag.
Industry-wide progress has generally been commensurate with the
complexity of the plant-specific issues and the amounts of Thermo-Lag
installed at the individual plants. Several licensees have initiated
programs to replace Thermo-Lag and are performing plant-specific tests
of other fire barrier materials such as Mecatiss (Florida Power & Light
for Crystal River Unit 3) and Darmatt KM-1 (Carolina Power & Light for
Brunswick, IES Utilities, Inc., for Duane Arnold Energy Center,
Commonwealth Edison Company for LaSalle County Station, and Northern
States Power Company for Prairie Island Nuclear Generating Plant). The
NRC staff is reviewing the plant-specific fire endurance test programs
and has recently approved the plant-specific application of Darmatt KM-
1 fire barrier at the LaSalle plant. The remaining licensees have
submitted to the NRC staff detailed plans and schedules for resolving
the issues at their plants. Most licensees are pursuing a combination
of such options as upgrading existing Thermo-Lag fire barriers to meet
NRC fire barrier requirements, replacing Thermo-Lag fire barriers with
another type of fire barrier, reducing or eliminating reliance on
Thermo-lag fire barriers by relocating equipment and cables and by
post-fire safe-shutdown reanalysis, installing additional fire
protection features such as automatic sprinkler systems, and requesting
configuration-specific exemptions when such exemptions are allowed by
NRC regulations and are technically justified to provide a level of
safety equivalent to that prescribed by the regulations. The NRC staff
has completed its review of the plans for resolving fire protection
issues that were proposed by most of the licensees. As with any issues
as technically complex, challenging, and resource intensive as those
presented by Thermo-Lag barriers, some plant-specific questions remain.
However, the number of issues has steadily declined. The NRC staff and
the licensees will continue to address the residual questions on a
case-by-case basis as they arise, and the NRC staff will continue to
follow up with individual licensees on their corrective actions, as
appropriate. Every licensee with Thermo-Lag fire barriers will continue
to maintain NRC-approved compensatory measures, such as fire watches,
until its permanent corrective actions are implemented. Therefore, the
public health and safety are protected.
The NRC's ``defense-in-depth'' fire protection concept relies on
protecting safe shutdown functions by achieving a balance among three
echelons or levels of protection, which are (1) fire prevention
activities; (2) the ability to rapidly detect, control, and suppress a
fire; and (3) physical separation of redundant safe shutdown functions.
Weaknesses found in one area may be dealt with by enhancing the
protection capabilities of the remaining areas.2 The NRC foresaw
cases in which fire protection features would be inoperable and
required licensees, through technical specifications or approved fire
protection plans controlled by license conditions, to provide
compensation for the deficient condition. The concept of allowing
alternative actions to compensate for an inoperable condition or
component is used in various programs associated with the operation of
nuclear power plants and has long been an integral part of NRC
regulatory requirements.3
\2\ The ``defense-in-depth'' concept is detailed in the ``NRC
Standard Review Plan,'' NUREG-0800, Section 9.5.1, ``Fire Protection
Program,'' page 9.5.1-10.
\3\ NRC GL 91-18, ``Information to Licensees Regarding Two NRC
Manual Sections on Resolution of Degraded and Nonconforming
Conditions and Operability,'' issued November 7, 1991, and NRC
Inspection Manual, Part 9900, ``Resolution of Degraded and
Nonconforming Conditions,'' issued October 31, 1991.
---------------------------------------------------------------------------
The fire endurance test results contained in NRC BUL 92-01 and NRC
BUL 92-01, Supplement 1, confirmed that certain Thermo-Lag fire barrier
configurations compromise one facet of the fire protection defense-in-
depth concept. In response to NRC BUL 92-01 and its supplement, the
licensees for plants using Thermo-Lag fire barriers established fire
watches in accordance with their technical specifications or license
conditions as a compensatory measure. Fire watches are personnel
trained by the licensees to inspect for the control of ignition
sources, fire hazards, and combustible materials; to look for signs of
incipient fires; to provide prompt notification of fire hazards and
fires; and to take appropriate actions to begin fire suppression
activities. Generally, therefore, by providing additional fire
prevention activities through enhanced detection capabilities to find
fire hazards and in the case of a fire, augmented suppression
activities before a barrier's ability to endure a fire is challenged,
fire watches compensate for degraded fire barriers.
The NRC staff has carefully evaluated the issues associated with
continued use of Thermo-Lag material, including the use of fire watches
to compensate for any degradation in the effectiveness of required fire
barriers. Such compensatory actions provide an adequate level of fire
protection without an undue risk to the health and safety of the
public. Licensees have established fire watches to compensate for
degraded and possibly inoperable fire barriers. Also, licensees rely on
a defense-in-depth concept that incorporates multiple safety measures.
Automatic fire detection and suppression systems are provided in most
areas that have safe shutdown equipment. Trained fire brigades are
required 24 hours a day at all plants. All areas that have safe
shutdown equipment have manual fire suppression features. Fuels that
can feed a fire and ignition sources to start a fire are controlled.
The combination of fire watches and the defense-in-depth fire
protection features provides an adequate level of fire protection until
licensees implement permanent corrective actions.
Taken together, these factors represent an adequate means of fire
protection at the plants using Thermo-Lag to ensure, with margin,4
that operation can be
[[Page 16010]]
conducted without an undue risk to the health and safety of the public.
Nevertheless, with these considerations in mind, the NRC staff
addressed below the Petitioners' specific concerns to demonstrate that
no substantial health and safety issue has been raised.
\4\ The fact that Thermo-Lag barriers, as installed, will
provide protection for some period of time is supported by, among
others, the fire endurance test results documented in IN 92-55.
---------------------------------------------------------------------------
III. Response to Specific Concerns
The Petitioners alleged that (1) the NRC has been slow to enforce
its own regulations, (2) fire watches do not replace fire barriers and
continued reliance on fire watches is an unreasonable and unnecessary
hazard to the public health and safety because of an inoperable fire
protection system for safe shutdown of the reactor and installed
combustible material on the shutdown systems, (3) utilities are in
violation of NRC requirements because Thermo-Lag is combustible and
could contribute to a fire instead of protecting from it, and, in spite
of the danger, the NRC allows continued use of Thermo-Lag, (4) faulty
ampacity ratings could result in the use of inappropriate cables,
which, if undersized, could overheat and cause its insulation to
deteriorate, (5) the licensee for Oyster Creek did not report to the
NRC its findings regarding the combustibility of Thermo-Lag and, (6)
the Thermo-Lag barriers have been improperly installed at Comanche Peak
Unit 1, which contributes further to the poor performance of Thermo-
Lag.
The NRC staff acknowledged and has stated that certain Thermo-Lag
fire barrier configurations have failed to demonstrate the ability to
perform their fire resistance functions. In this regard, the NRC staff,
in BUL 92-01, Supplement 1, has stated that Thermo-Lag fire barriers
should be treated as inoperable until licensees can declare the fire
barriers operable on the basis of successful, applicable tests. Given
the foregoing deficiencies identified for Thermo-Lag, the NRC staff
concluded that compensatory measures are necessary until a licensee can
declare fire barriers operable on the basis of applicable tests that
demonstrate successful barrier performance.
The Petitioners also asserted that (1) the NRC should have
protected the public and not Rubin Feldman, the President of the
company manufacturing Thermo-Lag, and (2) public safety has been
compromised by NRC's seeming complicity with utilities.5
\5\ These statements could be interpreted as the appearance of
unwarranted favoritism toward the manufacturer of Thermo-Lag and
complicity with utilities. Therefore, the Petitions were referred to
the NRC Office of the Inspector General.
---------------------------------------------------------------------------
A. Regulatory Compliance
The NRC staff acknowledges that certain fire endurance tests have
demonstrated that Thermo-Lag barriers may not meet the fire endurance
rating criteria set forth in Section III.G. of Appendix R to 10 CFR
Part 50. This acknowledgment does not mean, however, that there no
longer is reasonable assurance of protection of the public health and
safety or that such actions as the shutdown of all reactors using
Thermo-Lag and the suspension of Comanche Peak, Peach Bottom, and
Oyster Creek operating licenses are warranted.
It should first be noted that Appendix R, which sets forth criteria
for specific fire protection features to protect safe shutdown systems,
is applicable only to facilities that commenced operation prior to
1979. Facilities commencing operation on or after January 1, 1979,
although not bound by Appendix R, generally are bound by licensing
commitments to follow the criteria set forth in Appendix R through
license conditions.6
\6\ In addition, there are a very limited number of plants which
commenced operation on or after January 1, 1979, that are not
subject to specific license conditions but whose licensees have made
commitments to comply with NRC fire protection requirements,
including Section III.G. of Appendix R. The NRC is elevating these
commitments to license conditions.
---------------------------------------------------------------------------
Even assuming that all of the plants in which Thermo-Lag is
installed and that commenced operation prior to 1979 are not in
compliance with Appendix R, it does not follow that the failure to
comply with a regulation indicates the absence of adequate protection.
The Commission has explained that--
[W]hile it is true that compliance with all NRC regulations
provides reasonable assurance of adequate protection of the public
health and safety, the converse is not correct, that failure to
comply with one regulation or another is an indication of the
absence of adequate protection, at least in a situation where the
Commission has reviewed the noncompliance and found that it does not
pose an ``undue risk'' to the public health and safety.
(Ohio Citizens for Responsible Energy, DPRM 88-4, 28 NRC 411
(1988).)
All the plants using Thermo-Lag have instituted fire watches as
required by their action statements regarding inoperable barriers
contained in their technical specifications or fire protection programs
subject to license conditions. Generally, action statements provide
alternative remedial actions to shutting down a plant when limiting
conditions for operation are not met. Compliance with the required
remedial actions provides reasonable assurance that the public health
and safety is protected notwithstanding the plant's continued operation
and its failure to meet the respective limiting condition for
operation. Here, since all of the plants using Thermo-Lag have
implemented the required fire watches in accordance with plant-specific
requirements, their continued operation does not pose an undue risk to
the public health and safety.
The Petitioners assert that fire watches do not replace fire
barriers and continued reliance on fire watches is a hazard to public
safety. The NRC staff acknowledges that fire watches do not replace
fire barriers. However, as will be discussed in greater detail later in
this Decision, fire watches are judged by the NRC to be acceptable
compensatory measures and are legally sanctioned remedial actions based
on 10 CFR 50.36(c)(2).7
\7\ In instances in which fire protection programs have been
moved from technical specifications and are now subject to license
conditions, the NRC's approval of the fire protection programs
subject to license conditions provides the legal basis for the
implementation of fire watches as a remedial measure.
---------------------------------------------------------------------------
In sum, notwithstanding the failure to have operable fire barriers
meeting the fire endurance rating criteria specified by Section III.G.
of Appendix R, a plant is not necessarily unsafe to continue operation.
To the contrary, fire watches are judged by the NRC to be adequate
remedial measures that provide reasonable assurance that the public
health and safety is protected. By reason of compliance by all
facilities using Thermo-Lag with their technical specifications or fire
protection program action statements requiring the implementation of
fire watches, protection of the public health and safety is still
reasonably ensured for such plants. Because the Commission has
discretion regarding enforcement of its regulations, and given the
circumstances here in which no significant health and safety issues
have been raised, enforcement action of the nature requested by the
Petitioners is not warranted.
B. Ability of Fire Watches to Compensate for a Degraded Barrier
One of the Petitioners' allegations is that the measures taken by
licensees to compensate for degraded barrier conditions, specifically
fire watches, are not adequate to protect the public health and safety.
The Petitioners have questioned the continued reliance on fire watches
in the light of an inoperable fire protection system for safe plant
shutdown and the combustibility of Thermo-Lag. In addition, the
Petitioners
[[Page 16011]]
claim that a fire watch does not replace a fire barrier in that fire
watches are not preventive.
Despite the acknowledged shortcomings identified with certain
Thermo-Lag fire barriers and after fully considering the arguments
presented by the Petitioners regarding the ability of fire watches to
provide adequate compensation, the NRC staff has determined that
compensatory measures using fire watches are adequate and acceptable to
ensure public health and safety until permanent corrective measures are
implemented.
The use of fire watches in instances of degraded or inoperable
barriers is an integral part of NRC-approved fire protection programs.
In general, these NRC staff-approved compensatory measures specify the
establishment of a continuous fire watch or an hourly fire watch in
cases in which automatic detection systems protect the affected
components. Although it is true that Thermo-Lag is intended as a
barrier and fire watch personnel cannot act as physical shields, a fire
watch provides more than simply a detection function. Personnel
assigned to fire watches are trained by the licensee to inspect for the
control of ignition sources, fire hazards, and combustible materials;
to look for signs of incipient fires; to provide prompt notification of
fire hazards and fires; and to take appropriate action to begin fire
suppression activities. Fire watch personnel are capable of determining
the size, the actual location, the source, and the type of fire--
valuable information that cannot be provided by an automatic fire
detection system.
During a plant fire, compartment temperatures are likely to be less
severe at the early stages. On the basis of enhanced capabilities
provided by fire watches and notwithstanding that the level of barrier-
type protection may be reduced, the NRC staff has determined that there
is an adequate margin of safety to ensure protection in cases in which
fire watches are approved.
The goal of the NRC staff's Thermo-Lag Action Plan is directed
towards restoring the functional capability of fire barriers as soon as
practicable. There is not a time limit associated with the use of fire
watches as a compensatory measure. Given the margin of safety a fire
watch brings to a fire protection program, as discussed above, the NRC
staff has determined that continuing the use of fire watches while
barriers are inoperable is acceptable. However, the NRC believes that
notwithstanding interim reliance on compensatory measures, appropriate
actions must be taken by licensees to restore operability of Thermo-Lag
barriers. Individual licensees have provided schedules for restoring
operability and these are being tracked by the NRC staff.
The NRC staff has carefully evaluated the use of fire watches to
compensate for any degradation in the effectiveness of required fire
barriers and has concluded that fire watches continue to ensure
protection of the public health and safety. Therefore, the Petitioners'
assertion that the measures taken by licensees to compensate for
degraded fire barrier conditions, specifically fire watches, are a
hazard is without merit.
C. Combustibility
The Petitioners alleged that, contrary to NRC regulations, Thermo-
Lag is combustible.
The NRC staff recognizes that Thermo-Lag is combustible. To assess
Thermo-Lag combustibility, the NRC staff conducted a testing program at
the National Institute of Standards and Technology (NIST) based on the
American Society for Testing and Materials (ASTM) Standard E-136. Under
this testing standard, the material is considered to be ``combustible''
if three out of four samples tested exceed the following criteria: (1)
the recorded temperature of the specimen's surface and interior
thermocouples, during the test, rises 54 deg.F (30 deg.C) above the
initial furnace temperature; (2) there is flaming from the specimen
after the first 30 seconds of irradiance; and (3) the weight loss of
the specimen, due to combustion during the testing, exceeds 50 percent.
Of the four Thermo-Lag specimens tested, all experienced a weight loss
of greater than 50 percent and flaming continued in excess of 30
seconds. In 92-82, which provided licensees with the results of the E-
136 tests and confirmed the combustibility of Thermo-Lag, restated the
NRC fire protection requirements of Section III.G. of Appendix R to 10
CFR Part 50 and asked that licenses review the information for
applicability to their facilities.
The NRC's basic fire protection regulation for commercial nuclear
power plants is Section 50.48 of 10 CFR Part 50 ``Fire protection.''
Section 50.48 references General Design Criterion (GDC) 3 of Appendix A
to 10 CFR Part 50, ``Fire protection,'' Appendix R to 10 CFR Part 50''
Fire Protection Program for Nuclear Power Facilities Operating Prior to
January 1, 1979,'' and various NRC fire protection guidance documents.
Specifically, Section 50.48(a) states that each operating nuclear power
plant must have a fire protection plan that satisfies GDC 3, and
Section 50.48)b) states that Appendix R to 10 CFR Part 50 establishes
fire protection features required to satisfy GDC 3 with respect to
certain generic issues for nuclear power plants licensed to operate
prior to January 1, 1979. \8\ These issues are addressed in Section
III.G, ``Fire protection of safe shutdown capability,'' Section III.J,
``Emergency lighting,'' and Section III.O, ``Oil collection system,''
of Appendix R. Of these three sections of Appendix R, Section III.G
addresses the use of fire barriers to protect one train of systems
necessary to achieve and maintain hot shutdown conditions in the event
of a fire and, therefore, is the regulation of interest here.
\8\ While Appendix R is applicable only to facilities that
commenced operation prior to January 1, 1979, as discussed earlier
in this Director's Decision, facilities commencing operation on or
after January 1, 1979, are bound to satisfy the criteria of Appendix
R through license conditions or licensing commitments.
---------------------------------------------------------------------------
Section 50.48(a) notes that fire protection guidance for nuclear
power plants is contained in two NRC documents. These are (1) Branch
Technical Position (BTP) Auxiliary Power Conversion Systems Branch
(APCSB) 9.5-1, ``Guidelines for Fire Protection for Nuclear Power
Plants,'' for new plants docketed after July 1, 1976, and (2) Appendix
A to BTP APCSB 9.5-1, ``Guidelines for Fire Protection for Nuclear
Power Plants Docketed Prior to July 1, 1976.'' These two NRC documents
specify preferred methods for fire protection program design including
the use of fire barriers to satisfy Section III.G of Appendix R. Fire
barriers that meet the criteria of Section III.G of Appendix R to 10
CFR Part 50 and these NRC guidance documents satisfy GDC 3. NUREG-0800,
``Standard Review Plan,'' (SRP) Section 9.5-1, ``Fire Protection
Program,'' incorporates the guidance of BTP APCSB 9.5-1 and Appendix A
to BTP APCSB 9.5-1 and the criteria of Section III.G of Appendix R to
10 CFR Part 50. Therefore, fire barriers that meet the guidelines of
SRP Section 9.5-1 also satisfy 10 CFR 50.48 and GDC 3.
As stated in 10 CFR 50.48(a), the purpose of the fire protection
plan is ``to limit fire damage to structures, systems, or components
important to safety so that the capability to safely shut down the
plant is ensured. '' In general, a fire protection plan consists of
administrative controls and procedures, personnel for implementing the
plan and for fire prevention and manual fire suppression activities,
fire detection systems, automatic and manually operated fire
suppression systems and equipment, and fire barriers.
[[Page 16012]]
Section III.G of Appendix R to 10 CFR Part 50 is the only part of
the fire protection regulations that addresses the use of fire
barriers. It addresses the use of fire barriers to protect one train of
systems necessary to achieve and maintain hot shutdown conditions in
the event of a fire. Fire barriers are required to have either a 1-hour
or 3-hour rating depending on the specific requirement. However,
Section III.G does not provide acceptance criteria for fire barriers,
nor does it address the combustibility of fire barrier materials. The
criteria are set out in BTP APCSB 9.5-1, Appendix A to BTP APCSB 9.5-1,
and SRP Section 9.5-1. These NRC documents do not preclude the use of
combustible materials for construction of fire barriers required to
have a 1-hour or 3-hour rating. On March 25, 1994, the staff
consolidated and clarified in Supplement 1 to Generic Letter (GL) 86-
10, the fire barrier criteria specified in the BTPs and the SRP. This
GL supplement provides detailed staff guidelines for assessing the
combustibility of fire barrier materials, but it does not preclude the
use of combustible materials for fire barriers required to satisfy a 1-
hour or 3-hour rating. In fact, the fire barrier criteria are
appropriately focused on the performance of the fire barrier and its
ability to achieve its intended design function, that is, its ability
to limit temperature rise within the barrier enclosure and to prevent
the passage of flame or gasses hot enough to adversely affect the
functionality of the safe shutdown components (e.g., cables) enclosed
within the fire barrier.
Thermo-Lag 330-1 is a sacrificial material. When it is exposed to
elevated temperatures, such as those experienced during a fully-
developed room fire, it sublimes and transitions from a solid to a
vapor. The vapors go through an endothermic decomposition process
(pyrolysis) which absorbs heat from the fire. As a result of the
pyrolysis, the unreacted Thermo-Lag material is replaced by an
insulating char layer which is composed of small interconnecting cells
having a large surface area. The char layer re-radiates energy and
limits heat transfer through the Thermo-Lag material. The low thermal
conductivity of the char layer provides additional thermal insulation.
Therefore, even though Thermo-Lag is classified as a combustible
material when testing in accordance with the guidance of Supplement 1
to GL 86-10, properly designed, qualified, and installed Thermo-Lag can
yield fire barriers with a 1-hour or 3-hour rating which will protect
safe shutdown components from the effects of the fire. Therefore, such
barriers can satisfy the requirements of 10 CFR 50.48 and GDC 3.
To provide reasonable assurance that Thermo-Lag fire barriers
installed in the nuclear power plants can meet their intended function,
representative Thermo-Lag fire barrier assemblies have been subjected
to full-scale qualification-type fire endurance tests conducted in
accordance with the guidance of Supplement 1 to GL 86-10. This guidance
provides standard and uniform test methods and acceptance criteria for
assessing the fire-resistive capabilities of these barriers. The staff
has found the use of Thermo-Lag acceptable as a fire barrier material
when it is used in accordance with existing NRC regulations and
guidance and where supported by appropriate tests and analyses.
However, there are two types of applications where the use of
Thermo-Lag material is not appropriate. These are (1) enclosing
combustible materials (e.g., insulated cables) within Thermo-Lag fire
barriers to eliminate the combustible materials as a fire hazard and
(2) using Thermo-Lag as radiant energy heat shields inside noninerted
containments.
Section III.G of Appendix R (and the equivalent SRP guidance)
specifies three options for protecting redundant trains of systems
necessary to achieve and maintain hot shutdown conditions located
within the same fire area outside of containment. Two of the three
options (Sections III.G.2.a and c) rely on the use of fire barriers
with a 1-hour or 3-hour rating, as discussed above. The third option,
Section III.G.2.b, specifies the separation of redundant safe shutdown
trains by a horizontal distance of more than 20 feet with no
intervening combustibles or fire hazards. (A typical example of
intervening combustibles is a cable tray loaded with cables, because
cable jacket materials are combustible.) Therefore, spacial separation,
and not fire barriers, are used to meet Section III.G.2.b. However, to
meet this requirement, some licensees have enclosed combustibles that
are installed between redundant shutdown trains within a fire barrier.
In theory, the fire barrier prevents an exposure fire from igniting the
intervening combustible materials and spreading along them from one
redundant train to the other. Thus the fire barrier effectively
eliminates the intervening combustible as a fire hazard. If the fire
barrier itself is noncombustible and the redundant safe shutdown trains
are separated by a horizontal distance of more than 20 feet, then the
configuration meets Section III.G.2.b of Appendix R. However, if the
fire barrier material used to enclose the intervening combustibles is
also combustible, such as Thermo-Lag, then the licensee has simply
installed one combustible material over another and has not eliminated
the intervening fire hazard. In a limited number of cases, licensees
have enclosed intervening combustibles within Thermo-Lag fire barriers
under the incorrect assumption that the Thermo-Lag fire barrier would
eliminate the intervening combustibles as a fire hazard. Corrective
actions will be required in these cases.
As an alternative to the three options discussed above, Section
III.G.2.f of Appendix R (and the equivalent SRP guidance) provides a
fourth option for noninerted containments, that is, the separation of
redundant safe shutdown components with noncombustible radiant energy
heat shields. Thermo-Lag is classified as a combustible material when
tested in accordance with the guidance of Supplement 1 to GL 86-10.
Therefore, it does not meet the criteria for radiant energy heat
shields. Licensees using Thermo-Lag in this fashion will also be
required to take corrective action.
To assure that corrective actions are taken in these cases, the NRC
staff issued IN 95-27. In that IN, the staff addressed enclosing
combustible materials within Thermo-Lag fire barriers in an attempt to
eliminate the combustible materials as a fire hazard and using Thermo-
Lag to construct radiant energy heat shields inside noninerted
containments. The staff identified such solutions for reevaluating the
use of Thermo-Lag for these applications as: (1) reanalyzing post-fire
safe shutdown circuits inside containment and their separation to
determine if the Thermo-Lag radiant energy shields are needed, (2)
replacing Thermo-Lag barriers installed inside the containment with
noncombustible barrier materials, (3) replacing Thermo-Lag barriers
used to create combustible-free zones with noncombustible barrier
materials, (4) rerouting cables or relocating other protected
components, or (5) requesting plant-specific exemptions where
technically justified.
One of the Petitioners also asserted that subsection 5a(3) of
Section 9.5-1 of the SRP states that fire barrier designs ``should
utilize only non-combustible materials.'' This section of the SRP does
not apply to fire barriers which are used to separate redundant safe
shutdown components located within a nuclear power plant fire area.
Rather, it applies to fire barrier penetration seals, which are
typically installed in fire area
[[Page 16013]]
boundaries. Thermo-Lag 330-1 is not used in such applications.
The principal consideration for 1-hour and 3-hour rated fire
barriers installed to meet NRC fire protection requirements and
guidelines is that they can achieve their intended design function.
That is, that they can limit temperature rise within the barrier
enclosure and prevent the passage of flame or gasses hot enough to
adversely affect the functionality of the safe shutdown components
enclosed within the fire barriers. The fact that Thermo-Lag material is
combustible does not preclude Thermo-Lag fire barriers from achieving
the intended function of preventing fire damage if the fire barriers
are properly designed, qualified, and installed. The Petitioners'
contention that Thermo-Lag material should not be used because it is
combustible is without basis.
D. Ampacity Derating
The Petitioners assert that Thermo-Lag could contribute to starting
a fire instead of protecting from it. They further alleged that faulty
ampacity derating factors could result in the use of inappropriate
cables that, if undersized, could overheat and cause its insulation to
deteriorate.
Ampacity derating is the lowering (derating) of the current-
carrying capacity of power cables enclosed in electrical raceways
protected with fire barrier materials because of the insulating effect
of the fire barrier material. This insulating effect may reduce the
ability of the cable insulation to dissipate heat. If not accounted for
in the plant design, the increased cable insulation temperature could
lead to premature insulation failure. Other factors also affect
ampacity derating, including the extent of cable fill in the raceway,
cable type, raceway construction, and ambient temperature. The National
Electrical Code, Insulated Cable Engineers Association (ICEA)
publications, and other industry standards provide ampacity derating
factors for open air installations. These standards do not provide
derating factors for fire barrier systems. Although a national standard
test method is in the process of being developed but has not yet been
established, ampacity derating factors for raceways enclosed with fire
barrier material are determined by testing for the specific
installation configurations.
TSI, the manufacturer of Thermo-Lag, has documented a wide range of
ampacity derating factors that were determined by testing, for raceways
enclosed within Thermo-Lag fire barrier materials. On October 2, 1986,
TSI informed its customers that, while conducting tests in September
1986 at Underwriters Laboratories, Inc. (UL), it found that the
ampacity derating factors for Thermo-Lag barriers were greater than
previous tests indicated. However, the cable fill and tray
configurations were different for each test than those tested
previously. In addition, the NRC staff learned that UL performed a
duplicate cable tray test that resulted in an even higher derating
factor. The NRC staff also learned of the determination of other
derating factors during its review of other tests conducted at
Southwest Research Institute (SwRI).9
\9\ The test procedures and test configurations differed among
the testing laboratories. Therefore, the results from the different
ampacity tests may not be directly comparable to each other.
The NRC staff is concerned that the ampacity derating factors,
as determined in UL tests for Thermo-Lag barrier designs, are
inconsistent with TSI results for similar designs because different
times were allowed for the temperature to stabilize before taking
current measurements. Inconsistent stabilization times would call
into question the validity of previous TSI results. The NRC also
noticed during the review of the Industrial Testing Laboratories
(ITL) test reports that ambient temperature and maximum cable
temperature were allowed to vary widely for some tests. Therefore,
those tests in which the ambient and maximum cable temperatures were
not maintained within specified limits may be questionable.
Additionally, a licensee discovered a mathematical error for the
ampacity derating factor published in an ITL test report. A
preliminary assessment of the use of a lower-than-actual ampacity
derating factor indicates that higher-than-rated cable temperatures
are possible for Thermo-Lag installations. Higher-than-rated cable
temperatures could accelerate the aging effects experienced by the
cable.
---------------------------------------------------------------------------
The NRC special review team concluded that the ampacity derating
test results completed at the time of the review, including the UL test
results, were indeterminate. This conclusion was based on observed
inconsistencies in the derating test results of the various testing
laboratories. The special review team found that there was no national
consensus test standard (e.g., Institute of Electrical and Electronics
Engineers (IEEE) or American National Standards Institute (ANSI)) for
conducting these tests, and that some licensees had not adequately
reviewed ampacity derating test results to determine the validity of
the tests and the applicability of those test results to their plant
design. The special review team recognized that, in hypothetical cases,
nonconservative ampacity derating factors could have been instrumental
in the installation of inappropriate cables, which as a result, could
suffer premature cable jacket and cable insulation failures over a
period of time. However, since that time, the NRC staff has determined
that in practice the ampacity derating factor resulting from Thermo-Lag
insulating properties represents only one of many variables used in
determining the design ampacity for power cable systems and that, as
discussed below, sufficient margin exists in this area to preclude any
immediate safety concerns.
For actual installations, various derating factors are typically
applied to the ICEA ampacity values provided for each cable size. In
general, the cables typically used in actual installations have higher
current-carrying capacity than the ICEA ampacity values.10 Also,
cables are sized based on full-load current plus a 25 percent margin to
account for starting current requirements of the load. Given the short
duration of typical equipment starts, this margin is available to
compensate for any errors in ampacity derating. Further, use of a cable
size larger than normal may be required as a result of voltage drop
considerations for long circuit lengths. In typical applications this
also provides additional current-carrying capacity. Given these
conservatisms inherent in the design ampacity of cable systems and in
addition the fact that most power cables required for safe shutdown are
not normally energized, but are typically operated during surveillance
testing for short time periods, the likelihood that cables could ignite
as a result of Thermo-Lag ampacity derating errors has been judged by
the NRC staff to be unlikely. In addition, based on these conservatisms
and the currently available information on existing plants, ampacity
design, and operating history, the NRC staff believes that the ampacity
derating issue is not an immediate safety issue but rather is an aging
issue to be resolved over the long term.11
\10\ ICEA ampacity values include conservatisms to compensate
for skin and proximity effects and shield and/or sheath losses which
may or may not apply in specific situations.
\11\ Generic Letter 92-08 requires licensees to review the
ampacity derating factors used for all raceways protected by Thermo-
Lag 330-1 (for fire protection of safe shutdown capability or to
achieve physical independence of electrical systems) and to
determine whether the ampacity derating test results relied upon are
correct and applicable to the plant design. Presently, the staff is
conducting reviews of followup actions to close out ampacity
derating concerns with licensees pursuant to GL 92-08.
---------------------------------------------------------------------------
E. Oyster Creek Failed To Report Test Results on Combustibility to the
NRC
The Petitioners requested that Oyster Creek's license be suspended
based on the following: (1) SwRI conducted fire tests on Thermo-Lag
330-1 specimens for GPUN, the licensee for Oyster Creek, and reported
that all specimens ignited approximately 2 seconds after they were
inserted into the furnace and failed specified criteria because of
flaming after the first 30 seconds of testing, an
[[Page 16014]]
outside temperature rise higher than 30 deg.C, and a weight loss of 50
percent; (2) GPUN's operation of Oyster Creek with knowledge of the
SwRI report is an example of GPUN's reckless disregard for fire
protection and public safety; (3) in the event of fire, Thermo-Lag is
likely to fail its intended function of protecting vital electrical
cables running from the control room to plant safety systems used to
shut down the reactor; (4) current installations of Thermo-Lag are
likely to fail in less time than the 1 hour (when smoke detectors and
automatic sprinkler systems are present) or 3 hours (when there are no
fire detection and suppression systems) that NRC regulations require
for fire barriers to withstand fire; (5) the NRC Inspector General
issued a report in August 1992 condemning NRC's handling of the Thermo-
Lag issue and documenting the NRC staff's failure to understand the
scope of the problem; (6) in April 1994, ITL and its President pleaded
guilty to five felony counts of aiding and abetting the distribution of
falsified test data; (7) on September 29, 1994, the U.S. Department of
Justice issued a seven-count indictment against the manufacturer of
Thermo-Lag and its Chief Executive Officer for willful violations of
the Atomic Energy Act, conspiracy to conceal material facts, and making
false statements to defraud the United States, in connection with $58
million in fire barrier material; (8) GPUN has known since at least
August 11, 1992, that Thermo-Lag 330-1 as a structural base material is
combustible and that it was in violation of Appendices A and R to Part
50 of Title 10 of the Code of Federal Regulations (10 CFR) and the NRC
Standard Review Plan, NUREG-0800; (9) GPUN failed to report the SwRI
test results in response to GL 92-08 of February 10, 1994, when asked
to describe the Thermo-Lag 330-1 fire barriers installed as required to
meet 10 CFR Part 50, Appendix R; and (10) continued reliance on fire
watches at Oyster Creek is an unreasonable and unnecessary hazard to
the public health and safety because of an inoperable fire protection
system for safe shutdown of the reactor and installed combustible
material on the shutdown systems.
Several of the issues listed above have been addressed earlier in
this decision. Therefore, the NRC staff will only address below the
remaining plant-specific issues. As discussed earlier in this decision,
the NRC issued IN 92-82 to inform the industry of the results of
combustibility tests performed by NIST in early August 1992. These
tests confirmed the combustibility of Thermo-Lag. As a result of
discussions with the NRC staff on the subject of Thermo-Lag
combustibility, GPUN decided to independently verify the results of the
E-136 tests performed by NIST and contracted SwRI to perform the E-136
tests. The results of these tests, as documented by the telecopy
transmittal sheet submitted with the Petition, confirmed the
combustibility of Thermo-Lag. Contrary to the Petitioners' allegations,
the NRC staff does not require that licensees report the results of
their independent testing. It should be noted here that, prior to the
SwRI testing that confirmed combustibility, the NRC was aware of the
combustibility of Thermo-Lag and that the NRC was also well aware of
the results of the E-136 tests performed by GPUN through telephone
conversations with GPUN personnel, even though there was no requirement
for GPUN to report these test results.
The Petitioners also alleged that GPUN did not report to NRC its
findings of the SwRI test results in its ``Response to Request for
Additional Information Regarding Generic Letter 92-08, `Thermo-Lag Fire
Barriers,' '' (RAI) dated February 10, 1994.
The RAI quoted by the Petitioners did not request that GPUN report
to NRC its findings of the SwRI test results and, in addition, the NRC
staff does not require that licensees report the results of their
independent testing. Therefore the NRC staff has concluded that,
contrary to the Petitioners' allegation, GPUN did not have to report to
the NRC its findings of the SwRI test results.
For the reasons stated above, the suspension of Oyster Creek's
license, as requested by the Petitioners, is not warranted.
F. Dry-Joint Issue at Comanche Peak Unit 1
The Petitioners requested that (a) the Comanche Peak Unit 1 license
be suspended, (b) the licensee perform additional destructive analysis
for Thermo-Lag configurations, and, (c) the licensee perform fire tests
on upgraded ``dry-joint'' Thermo-Lag configurations based on the
following: (1) the licensee's records on the original installation of
Thermo-Lag fire barriers on conduits and cable trays indicate that its
contractor followed specifications for pre-buttering all joints; (2)
NRC Inspection Report Nos. 50-445/93-42; 50-446/93-42 found, based on
destructive analysis documents, that a concern did exist where Thermo-
Lag conduit joints fell apart easily and did not appear to have any
residual material of a buttered surface, indicative of a joint that had
not been pre-buttered; (3) the ``dry joint'' deficiency appeared in
Room 115A and other areas of the unit; (4) the licensee directly
contradicts an NRC inspector's findings that were determined in part by
destructive analysis; (5) the ``dry joint'' or absence of pre-buttering
of Thermo-Lag panels can be determined only by destructive analysis and
cannot be determined by a walk down visual inspection; (6) the findings
reported in the Comanche Peak Unit 1 Region IV Inspection Reports 50-
445/93-42 and 50-446/93-42, based on the limited amount of destructive
analysis conducted at the unit, constitute a substantial documentation
of installation deficiencies found in Thermo-Lag fire barriers as
documented in NRC IN 91-79 and Supplement 1; (7) neither the NRC nor
the industry, by its agent NEI, nor a utility, have conducted fire
tests on dry fitted or ``dry joint'' upgraded configurations of Thermo-
Lag 330-1; and (8) the presence of ``dry joint'' upgraded
configurations in Comanche Peak Unit 1 constitutes an untested
application of Thermo-Lag fire barriers.
These allegations were based on the Petitioners' interpretation of
NRC Inspection Report 93-42 issued on February 21, 1994. By letter of
November 29, 1994, TU Electric, the licensee for Comanche Peak Unit 1,
sent a letter to the NRC staff responding to the Petition.
The term ``joint'' refers to the interface between two adjacent
Thermo-Lag surfaces. Comanche Peak Unit 1 installation procedures for
Thermo-Lag fire barriers specify that, during the initial installation
process, the joints should be pre-buttered (or covered) with Thermo-Lag
trowel grade material before the mating surfaces are joined to ensure
adhesion of the surfaces. The term ``dry joint'' refers to the lack of
Thermo-Lag trowel grade material in a joint. The failure to pre-butter
a joint with trowel grade Thermo-Lag could result in a weakening of the
joint during a potential fire exposure and could provide an exposure
path in the fire barrier envelope. The NRC performed an inspection at
Comanche Peak Unit 1 on November 2-5, and 23-24, 1993, and January 26-
28, 1994, to compare the Thermo-Lag test specimens with the upgraded
Thermo-Lag configurations on site. The results of this inspection are
documented in NRC Inspection Report 93-42. The report stated that there
appeared to be a large number of deficiencies with the installed fire
barriers and that an example of these deficiencies involved dry joints
on conduit overlays installed on pedestal hangers. The NRC inspector
did not personally observe the dry joints in question. His statements
were based on
[[Page 16015]]
observations made by TU Electric and documented in an Operations
Notification and Evaluation (ONE) form. However, the ONE form in
question did not identify a dry joint. Instead, the ONE form identified
a condition that was conservatively reported as an apparent dry joint.
Upon further evaluation of the ONE form, TU Electric determined that
the joint in question had in fact been pre-buttered with trowel grade
Thermo-Lag. These facts are discussed in more detail below.
On November 25, 1992, a speed memo was written by a TU Electric
contractor identifying ``apparent unsatisfactorily conditions on Unit 1
commodities.'' This memorandum identified ``an apparent'' dry joint on
an oversize coupling section (on top of a pedestal hanger). The speed
memo also stated that, ``we have decided that the best vehicle to call
attention to these apparent deficiencies would be a letter to your
attention for further evaluation of the situation. . . .'' The letter
was forwarded to the appropriate TU Electric engineering section.
The cognizant TU Electric engineer performed a walkdown of the
described areas and evaluated the commodities. He conservatively
initiated a ONE form (the process used by TU Electric to report
problems and develop resolution for the identified problems). A
comprehensive evaluation of this condition determined that the joint
had been pre-buttered. Therefore, the engineering resolution for this
condition was that ``this is not a deficient condition, and there are
no generic implications.''
The originator of the speed memo initially believed that the
condition in question was a dry joint because of the appearance of the
joint. During alignment of Thermo-Lag panels, the leading edge of one
panel contacts the outer edge of a preceding panel and forces most of
the trowel grade along the initial contact edge toward the inside of
the ThermoLag envelope. Subsequent shrinkage of the trowel grade in the
joint can give the appearance of a dry joint because the trowel grade
material is not visible. Therefore, contrary to the Petitioners'
allegation, there was no ``dry joint'' deficiency on the pedestal
hanger.
The Petitioners also alleged that dry joints appear in other
Thermo-Lag installations at Comanche Peak Unit 1. In response to the
Petition, TU Electric performed an electronic search of its ONE form
data base. The search did identify additional ONE forms related to dry
joints. However, Thermo-Lag rework crews and the quality control
inspectors at Comanche Peak Unit 1 have used the term ``dry joints''
and ``no visible trowel grade material'' synonymously. Upon further
investigation of these ONE forms, it was determined that trowel grade
material had in fact been applied to the joints in question. Therefore,
these ONE forms were also dispositioned as ``not a nonconforming
condition.'' These findings support the NRC staff's conclusion that,
contrary to the Petitioners' allegations, there is no evidence of dry
joints at Comanche Peak Unit 1. The Petitioners' allegations regarding
dry joints at Comanche Peak Unit 1 are based on premises that are
faulty and contrary to the information contained in Inspection Report
93-42.
In regard to the Petitioners' request that the licensee perform
fire tests on upgraded ``dry joint'' Thermo-Lag configurations and
additional destructive analysis, the NRC staff has reviewed the
documentation provided by the licensee in response to the RAIs
regarding GL 92-08 and concluded that the licensee's quality assurance
program gave adequate confidence that the as-installed Thermo-Lag
configurations at Comanche Peak Unit 1 conform with NRC specification
requirements for both material and installation attributes.
Accordingly, suspension of the Comanche Peak Unit 1 license, as
requested by the Petitioners, is not warranted.
G. Protection of Rubin Feldman
The Petitioners assert that, rather than protecting the public, the
NRC is protecting Rubin Feldman, President of the company that
manufactures Thermo-Lag.
As discussed earlier, the NRC received allegations in 1991 that
questioned the adequacy of Thermo-Lag fire barriers. In response (1)
the Office of the Inspector General (OIG) and the Office of
Investigations (OI) formed a joint task force to investigate the
allegations and (2) the Office of Nuclear Reactor Regulation (NRR)
established a special team to review the safety issues raised by the
allegations. Throughout its review, the special team gave expert
technical advice and assistance to the OIG/OI task force. The Director
of NRR tasked the NRR staff to resolve the technical issues raised by
the special team. The NRC staff continued to cooperate fully with the
investigative task force. Further, the NRR staff carried out a full-
scale test program and developed other technical data and information
for the investigative task force. These NRC staff efforts contributed
significantly to a referral to the Department of Justice of possible
wrongdoing by TSI. The referral resulted in a seven-count criminal
indictment of TSI, the manufacturer and supplier of Thermo-Lag fire
barriers and of its President, Rubin Feldman, by a Federal Grand Jury.
The NRC staff continued to support the Department of Justice throughout
the criminal case.\12\ In addition, throughout the trial, the NRC staff
continued to pursue corrective actions consistent with its action plan
for the resolution of the Thermo-Lag issues. The above facts contradict
the Petitioners' assertion that the NRC was protecting Rubin Feldman.
\12\ The jury returned a verdict of ``not guilty'' on all counts
of the indictment against TSI and Mr. Feldman.
---------------------------------------------------------------------------
H. NRC Seeming Complicity With Utilities
The Petitioners also assert that there is seeming complicity
between the NRC and the licensees and that licensees seek to avoid
costly replacement of the Thermo-Lag.
In May 1991, the NRC Office of the Inspector General performed an
inspection of the NRC's staff performance in regard to Thermo-Lag
barriers and found indications of inadequate performance by the NRC
staff in the acceptance and review of Thermo-Lag barriers.
Subsequently, the NRC staff initiated an aggressive program of
corrective actions to rectify the deficiencies identified in the review
and response process, as summarized earlier in this decision.
In addition, the staff has expended considerable time and effort to
address and resolve Thermo-Lag issues to ensure that licensees return
to compliance with existing NRC fire protection requirements. The NRC
staff issued three requests for additional information regarding GL 92-
08 to each licensee using Thermo-Lag to obtain information on the
specific Thermo-Lag material installed at each plant, details about the
corrective actions each licensee intended to take to return to
compliance with NRC fire protection requirements, and schedules for the
implementation of these corrective actions. The response of each
licensee was evaluated by the NRC staff. As a consequence of this
substantial NRC staff effort, a number of licensees have already
returned to compliance with NRC requirements by a variety of means
which include replacing, rerouting, or upgrading existing Thermo-Lag
barriers, performing post-fire safe shutdown reanalysis, and installing
additional fire detection and suppression features. All of these
measures involve some burden on licensees. In addition, some licensees
have initiated costly programs to perform plant-specific fire endurance
[[Page 16016]]
tests of other fire barriers with the intention of replacing Thermo-Lag
with these barriers. All licensees who utilize Thermo-Lag will need to
expend resources commensurate with their reliance on Thermo-Lag to come
into compliance with NRC fire protection requirements. NRC staff
oversight will ensure that this is the case.
The Petitioners' assertion of seeming complicity with utilities on
the part of the NRC staff is unfounded in the light of the significant
NRC staff efforts to ensure that licensees expend the resources
necessary to return to compliance with NRC requirements.
IV. Conclusion
The Petitioners request that the NRC order the immediate shutdown
of all reactors using Thermo-Lag and the suspension of Oyster Creek,
Peach Bottom Units 1 and 2, and Comanche Peak Unit 1 operating
licenses.
For the reasons discussed above, I find no basis for taking such
actions. Rather, on the basis of the review efforts by the NRC staff, I
conclude that the issues raised by the Petitioners are being addressed
by licensees in a manner which assures adequate protection of the
public health and safety. Accordingly, the Petitioners' requests for
action pursuant to 10 CFR 2.206 are denied.
A copy of this Decision will be placed in the Commission's Public
Document Room, Gelman Building, 2120 L Street, N.W., Washington, D.C.,
and at the Local Public Document Room for the named facilities. A copy
of this Decision will also be filed with the Secretary for the
Commission's review as provided in 10 CFR 2.206(c) of the Commission's
regulations.
As provided by this regulation, the Decision will constitute the
final action of the Commission 25 days after issuance, unless the
Commission, on its own motion, institutes a review of the Decision
within that time.
Dated at Rockville, Maryland, this 3rd day of April 1996.
For the Nuclear Regulatory Commission.
William T. Russell,
Director, Office of Nuclear Reactor Regulation.
[FR Doc. 96-8909 Filed 4-9-96; 8:45 am]
BILLING CODE 7590-01-P
