ML020950884
ML020950884 | |
Person / Time | |
---|---|
Site: | Saint Lucie ![]() |
Issue date: | 04/05/2002 |
From: | NRC/RGN-II |
To: | Stall J Florida Power & Light Co |
References | |
EA-02-033 IR-02-006 | |
Download: ML020950884 (22) | |
See also: IR 05000335/2002006
Text
April 5, 2002
Florida Power and Light Company
ATTN: Mr. J. A. Stall, Senior Vice President
Nuclear and Chief Nuclear Officer
P. O. Box 14000
Juno Beach, FL 33408-0420
SUBJECT: ST. LUCIE NUCLEAR POWER PLANT - NRC INSPECTION
REPORT 50-335/02-06; PRELIMINARY WHITE FINDING
Dear Mr. Stall:
On April 3, 2002, the NRC completed an in-office open item review for your St. Lucie facility.
The attached enclosure presents the results of that review, which was discussed on April 4,
2002, with Mr. D. Jernigan, St. Lucie Plant Vice President and other members of your staff.
This review was an in-office examination of an unresolved item (URI) which was identified in
NRC Inspection Report 50-335, 389/98-201 and forwarded to Florida Power and Light
Company on July 9, 1998. The URI was: URI 50-335, 389/98-201-09, Fire Mitigation System
Does Not Meet Plant Licensing Basis Requirements/Commitments or Minimum Industry Codes
and Standards for System Design and Testing. The URI included four fire protection program
issues concerning the design and testing of requirements of (1) fire protection systems, (2)
water suppression systems, (3) Halon 1301 fire suppression system, and (4) standpipe and fire
hose system. Issues 1 and 2 were resolved and documented in NRC Inspection Report 50-
335, 389/01-03. Issue 4 was resolved and documented in NRC inspection Report 50-335,
389/98-14. Issue 3 was unresolved pending further NRC review of the adequacy of the St.
Lucie Unit 1 Halon system.
Based on the results of this review, the inspector identified a finding involving failure to have an
installed fixed fire suppression system in an Alternative Shutdown Area. This finding was
assessed using the applicable significance determination process (SDP) and preliminarily
determined to be White (i.e., an issue with low to moderate safety significance, which may
require additional NRC inspections.) This issue was also determined to be an apparent
violation of Section III.G.3 of 10 CFR Part 50, Appendix R requirements. The apparent violation
is being considered for escalated enforcement action in accordance with the "General
Statement of Policy and Procedure for NRC Enforcement Actions - May 1, 2000" (Enforcement
Policy), NUREG-1600.
The NRC acknowledges that compensatory measures were implemented in response to the
URI, which included stationing a fire watch in the affected area. In addition, this matter was
entered into your corrective action program.
FPL 2
Before the NRC makes a final decision on this matter, we are providing you an opportunity to
request a regulatory conference where you would be able to provide your perspectives on the
significance of the finding, the bases for your position, and whether you agree with the apparent
violation. If you choose to request a regulatory conference, we encourage you to submit your
evaluation and any differences with the NRCs evaluation at least one week prior to the
conference in an effort to make the conference more efficient and effective. If a regulatory
conference is requested, it will be held in the NRC Headquarters office in Rockville, Maryland
and open for public observation. The NRC will also issue a press release to announce the
regulatory conference.
Please contact Mr. Randall A. Musser at (404) 562-4540 within seven days of the date of this
letter to notify the NRC of your intentions. If we have not heard from you within 10 days, we will
continue with our significance determination and associated enforcement processes on the
finding and you will be advised by separate correspondence of the results of our deliberations
on this matter.
Since the NRC has not made a final determination in this matter, no Notice of Violation is being
issued for the finding at this time. In addition, please be advised that the characterization of the
apparent violation described in the enclosure may change as a result of further NRC review.
In accordance with 10 CFR 2.790 of the NRC's "Rules of Practice," a copy of this letter and its
enclosure will be available electronically for public inspection in the NRC Public Document
Room or from the Publicly Available Records (PARS) component of NRCs Document system
(ADAMS). ADAMS is accessible from the NRC web site at http://www.nrc.gov/reading-
rm/adams.html (the Public Electronic Reading Room).
If you have any questions regarding this letter, please contact me at 404-562-4500.
Sincerely,
/RA/
Victor M. McCree, Deputy Director
Division of Reactor Projects
Docket No. 50-335
License No. DPR-67
Enclosure: Inspection Report No. 50-335/02-06
w/Attachment (Phase 3 SDP Analysis)
cc w/encl: (See page 3)
FPL 3
cc w/encl: J. Kammel
D. E. Jernigan Radiological Emergency
Site Vice President Planning Administrator
St. Lucie Nuclear Plant Department of Public Safety
Florida Power & Light Company Electronic Mail Distribution
Electronic Mail Distribution
Douglas Anderson
R. G. West County Administrator
Acting Plant General Manager St. Lucie County
St. Lucie Nuclear Plant 2300 Virginia Avenue
Electronic Mail Distribution Ft. Pierce, FL 34982
T. L. Patterson
Licensing Manager
St. Lucie Nuclear Plant
Electronic Mail Distribution
Don Mothena, Manager
Nuclear Plant Support Services
Florida Power & Light Company
Electronic Mail Distribution
Mark Dryden
Administrative Support & Special Projects
Florida Power & Light Company
Electronic Mail Distribution
Rajiv S. Kundalkar
Vice President - Nuclear Engineering
Florida Power & Light Company
Electronic Mail Distribution
M. S. Ross, Attorney
Florida Power & Light Company
Electronic Mail Distribution
William A. Passetti
Bureau of Radiation Control
Department of Health
Electronic Mail Distribution
Craig Fugate, Director
Division of Emergency Preparedness
Department of Community Affairs
Electronic Mail Distribution
FPL 4
Distribution w/encl:
B. Moroney, NRR
RIDSNRRDIPMLIPB
PUBLIC
OFFICE RII:DRP RII:DRP RII:DRP RII:DRP RII:DRS RII:DNMS RII:EICS
SIGNATURE RMusser HChristensen/for Rbernard for/ltr only LWert
NAME TRoss RMusser COgle VMcCree WRogers LWert SSparks
DATE 4/ /2002 3/29/02 4/5/02 4/ /2002 4/5/02 4/5/02 4/ /2002
E-MAIL COPY? YES NO YES NO YES NO YES NO YES NO YES NO YES NO
OFFICE RII:EICS NRC:NRR NRC:NRR NRC:NRR NRC:OE
SIGNATURE Cevans defer to
NAME CEvans HBerkow JHannon MJohnson DNelson
DATE 4/ /2002 4/ /2002 4/ /2002 4/ /2002 4/ /2002 4/ /2002 4/ /2002
E-MAIL COPY? YES NO YES NO YES NO YES NO YES NO YES NO YES NO
OFFICIAL RECORD COPY DOCUMENT NAME: C:\ORPCheckout\FileNET\ML020950884.wpd
U.S. NUCLEAR REGULATORY COMMISSION
REGION II
Docket No: 50-335
License No: DPR-67
Report No: 50-335/02-06
Licensee: Florida Power & Light Company (FPL)
Facility: St. Lucie Nuclear Plant, Unit 1
Location: 6351 South Ocean Drive
Jensen Beach, FL 34957
Dates: March 14 - April 3, 2002
Inspector: S. Ninh, Senior Project Engineer
Approved by: Randall A. Musser, Acting Chief
Reactor Projects Branch 3
Division of Reactor Projects
Enclosure
2
SUMMARY OF FINDINGS
IR 05000335-02-06 on 03/14-4/3/02, Florida Power & Light Company, St. Lucie Plant Unit 1.
Region-based follow up review of fire protection Unresolved Item 50-335, 389/98-201-09.
This in-office review was conducted by a regional senior project engineer. The inspector
identified one preliminary White finding with an apparent violation. The significance of an issue
is indicated by it color (green, white, yellow, red) using IMC 0609 Significance Determination
Process (SDP). Findings for which the SDP does not apply are indicated by No Color or by
the severity level of the applicable violation. The NRCs program for overseeing the safe
operation of commercial nuclear power reactors is described at its Reactor Oversight Process
website at http://www.nrc.gov/NRR/OVERSIGHT/index.html.
A. Inspector Identified Findings
Cornerstone: Mitigating Systems
To Be Determined. An apparent violation of Section III.G.3 of 10 CFR 50, Appendix R
was identified for failure to have an installed fixed fire suppression system in an
Alternative Shutdown Area. The licensees Cable Spreading Room (CSR) is designated
as an area requiring Alternative Shutdown capability. The St Lucie Unit 1 CSR Halon
1301 fire suppression system would not be able to extinguish a deep-seated fire
involving cable insulation and jacket material.
This finding appears to have a low to moderate safety significance because a
completely failed Halon system would result in a change in core damage frequency
(CDF) of 4.9E-6. (Section 1R05).
B. Licensee Identified Violations
None
3
1. REACTOR SAFETY
Cornerstone: Mitigating Systems
1R05 Fire Protection
.1 (Closed) Unresolved Item (URI) 50-335, 389/98-201-09: Fire Mitigation System Does
not Meet Plant Licensing Basis Requirements/Commitments or Minimum Industry Codes
and Standards for System Design and Testing.
a. Inspection Scope
The inspector reviewed the adequacy of the St. Lucie Unit 1 Cable Spreading Room
(CSR) Halon 1301 fire suppression system issue associated with the above URI. The
inspector also reviewed the St. Lucie Unit 1 Final Safety Analysis Report. The inspector
evaluated the licensees compliance with the requirements of 10 CFR 50.48(a)(1)(i),
Criterion 3 of Appendix A to Part 50, 10 CFR 50.48(b), and Section III.G.3 of 10 CFR
Part 50, Appendix R.
b. Findings
An apparent violation of Section III.G.3 of 10 CFR 50, Appendix R was identified for
failure to have an installed fixed fire suppression system in an Alternative Shutdown
Area (To Be Determined). The licensees CSR is designated as an area requiring
Alternative Shutdown capability. The St Lucie Unit 1 CSR Halon 1301 fire suppression
system would not be able to extinguish a deep-seated fire involving cable insulation and
jacket material.
The URI was identified during a Nuclear Regulatory (NRC) Fire Protection Functional
Inspection at the St. Lucie Plant in March and April, 1998. The URI included four fire
protection program issues concerning the design and testing of requirements of (1) fire
protection systems, (2) water suppression systems, (3) Halon 1301 fire suppression
system, and (4) standpipe and fire hose system. Issues 1 and 2 were resolved and
documented in NRC Inspection Report 50-335, 389/01-03. Issue 4 was resolved and
documented in NRC inspection Report 50-335, 389/98-14. Issue 3 was identified
concerning a design deficiency of the Halon 1301 fire suppression system installed in
the Unit 1 CSR. The inspector questioned the minimum required concentration and the
minimum soak time requirements. The licensee could not produce design-basis tests
for the concentrations and soak times of the system or demonstrate operability to the
inspectors. The inspector determined this system to be limited in its ability to mitigate a
fire since the system was not designed to extinguish the expected hazard (i.e., a deep-
seated cable fire). Issue 3 was unresolved pending further NRC review of the
adequacy of the St. Lucie Unit 1 Halon system.
Region II requested the Office of Nuclear Reactor Regulation (NRR) to evaluate this
issue in two Task Interface Agreements (TIAs), TIA 99-001, dated January 26, 1999,
and TIA 2000-04, dated May 8, 2000, to determine the design adequacy of the Halon
system for the Unit 1 CSR. NRR completed its review of the TIAs and documented their
4
conclusions by memoranda dated November 29, 1999 (TIA 99-01), August 4, 2000 (TIA 2000-04), and August 30, 2000 (TIA 2000-04). Additionally, several conference calls
with the licensee, Region II and NRR staff were conducted to discuss this issue in 2000
and 2001. The licensee provided additional information related to the CSR smoke and
thermal detection systems and the vendors performance tests of the Halon system for
NRC review.
Risk Assessment
This issue affected the mitigating systems cornerstone due to a design deficiency of a
fire defense-in-depth element, therefore, it was assessed in accordance with the NRC
Reactor Oversight Processs Significance Determination Process (SDP) as described in
NRC Inspection Manual Chapter 0609, Appendix F. The Phase III risk evaluation is
attached. This finding appears to have a low to moderate safety significance because a
completely failed Halon system would result in a change in core damage frequency
(CDF) of 4.9E-6. This finding is preliminarily characterized as a White finding in
accordance with the Fire Protection SDP. The issue was entered into the licensees
corrective action program as CR 98-0131, and the licensee has implemented
compensatory measures in the affected areas.
Apparent Violation
10 CFR 50.48(a)(1)(i) requires that each operating nuclear power plant have a fire
protection plan that satisfies Criterion 3 of Appendix A to Part 50.
Criterion 3 of Appendix A to Part 50 states that fire detection and fighting systems of
appropriate capacity and capability shall be provided and designed to minimize the
adverse effects of fires on structures, systems and components important to safety.
10 CFR 50.48(b) provides that Appendix R to Part 50 establishes fire protection features
required to satisfy Criterion 3.
Section III.G.3 of 10 CFR Part 50, Appendix R, requires that fire areas which require
Alternative Shutdown have fire detection and a fixed fire suppression system installed in
the area.
Contrary to the above, the licensee does not have an installed fixed fire suppression
system in an Alternative Shutdown Area. The licensees Cable Spreading Room (CSR)
is designated as an area requiring Alternative Shutdown capability. In 1986, the
licensee installed a Halon 1301 fire protection suppression system in the CSR. In order
to suppress a fire for special hazards such as a Halon fire suppression system, it must
be capable of extinguishing a fire. The licensees Halon 1301 system is designed to
achieve a minimum 5-7% halon concentration and minimum soak time of 10 minutes,
while a minimum 10% halon concentration and minimum soak time of 10 minutes is
necessary to extinguish a fire in a CSR. This failure to comply with Section III.G.3 of 10 CFR Part 50, Appendix R is identified as an apparent violation (AV) 50-335/02-06-01,
Failure to Have an Installed Fixed Fire Suppression System in an Alternative Shutdown
Area.
5
4OA6 Management Meeting
Exit Meeting Summary
Mr. R. Musser, Acting Chief, Reactor Projects Branch 3, Region II presented results to
Mr. D. Jernigan, Site Vice President and other members of the licensee staff via
telephone on April 4, 2002.
KEY POINTS OF CONTACT
Licensee
D. Jernigan, Site Vice President, St. Lucie
NRC
T. Ross, Senior Resident Inspector, St. Lucie
L. Wert, Acting Chief, Fuel Facilities Branch, Region II
ITEMS OPENED AND CLOSED
Opened
50-335/02-06-01 AV Failure to Have an Installed Fixed Fire Suppression
System in an Alternative Shutdown Area.
Closed
50-335, 389/98-201-09 URI Fire Mitigation System Does not Meet Plant
Licensing Basis Requirements/Commitments or
Minimum Industry Codes and Standards for
System Design and Testing.
Attachment
Phase 3 SDP Analysis: St. Lucie CSR Halon System Deficiency
1. Performance Deficiency
The St. Lucie Unit 1 Cable Spreading Room (CSR) [Fire Zone 57] Halon 1301 fire
suppression system did not meet regulatory or National Fire Code minimum
requirements for concentration and hold time to extinguish a deep-seated fire involving
cable insulation and jacket material. This condition has existed since 1986 and was
identified as Unresolved Item (URI) 50-335, 389/98-201-09 in NRC Inspection Report
50-335, 389/98-201, dated July 9, 1998, i.e. greater than 3 years.
2. Fire Scenario
The fire loading for combustibles in the St. Lucie Unit 1 CSR is over 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br />, with the
primary combustibles being electrical cables located above numerous potential ignition
sources. The potential ignition sources include the pressurizer heater transformers,
power programmer cabinets, 480V load centers, DC distribution panels, and reactor trip
switchgear. Other potential ignition sources, such as a power cable failure in a tray, or
other electrical originated failures (in distribution panels, circuit boards, electrical wiring,
internal cable fault, electrical circuit fault in switchgear cabinets, etc.) leading to ignition
of the in-situ combustibles (cables), are considered in the fire scenario. Although
creditable, outside ignition sources such as hot work (welding) or transient sources were
not factored into this analysis due to the large number of potential in-situ ignition
sources in the area.
One credible fire scenario is based on postulating that a fire develops from any one of
the potential in-situ electrical ignition sources, if undetected and unsuppressed, would
grow to a rate of heat release of 600 kW and ignite the cable insulation of electrical
cables resulting in deep-seated fires. The St. Lucie Unit 1 CSR Halon system is
actuated by a cross-zoned thermal detection system. The cross-zoned thermal
detection system requires two independent detectors, spaced some distance apart, to
reach their preset actuation temperature to initiate operation of the Halon system. After
the Halon system control panel has been activated, all the permissive circuits must be
completed to start the 30-second release delay alarm (for evacuation) followed by
system discharge. The permissive circuits include electrical confirmation that 3 HVAC
fans have stopped, 3 electrically-melted fusible links have operated and their associated
fire dampers closed, one motor-operated damper has cycled closed, and 5 fire doors
are verified closed. If any one of these permissive circuits does not confirm its operation
(e.g., electrical melting of the fusible link in fire damper FPPR-25-117 fails), the Halon
system will not automatically discharge its gaseous inventory.
The CSR area also has a fire detection system equipped with ionization smoke
detectors that only alarms in the Main Control Room (MCR). The Halon system was
designed to maintain a concentration of 5 to 7 percent for a minimum of 10 minutes for
extinguishing fires. However, the NFPA Standard 12A states that "deep-seated fires
usually require much higher concentrations than 10 percent and much longer soaking
times than 10 minutes." Thus, a fully developed deep-seated fire in the CSR fire zone
57 would overpower the suppressive capability of the Halon fire suppression system
resulting in damage to electrical cables performing safe shutdown functions and loss of
plant control from the MCR. In the event of an uncontrollable fire in the cable spreading
room, alternate shutdown of the plant outside the MCR would be required from the hot
shutdown control panel room in the Electrical Equipment Room1B which is located
adjacent to the cable spreading room.
In this fire scenario, it is assumed that a fire starts from a specific ignition source (e.g.,
pressurizer heater transformers, power programmer cabinets, etc.) and has sufficient
flame spread (i.e., flame height and radius) to ignite a cable tray closest to the ignition
source. Deep-seated fires (which are the concern of the finding related to the Halon
system design concentration and soak time at St. Lucie Unit 1 CSR) can be either
severe or non-severe (i.e. depending on whether fire damage is limited to the
component of fire origin). Severe fires are defined as fires that have grown beyond the
incipient stage and have the potential to damage structures, systems or components
(SSCs) beyond the SSC in which the fire originates. However, all severe fires, as
defined above, involving cables routed in cable trays are expected to result in a deep-
seated fire due to the combustible characteristics of cable insulation (i.e., the fire burns
internally as opposed to a surface fire). Based on the SPLB fire hazards analysis (see
attachment), it is postulated that a credible fire propagation pathway exists in the power
programmer cabinets, 480 V reactor auxiliary common MCC 1AB, vital AC supply
panels, and DC bus 1AB-1. The failure of 480V reactor auxiliary common MCC 1AB,
vital AC supply panels, and DC bus 1AB-1 would cause preheating of cables leading to
cable failure, thus initiating a secondary fire in the cable trays. Three different cases of
cable tray fire were postulated to bound the fire scenario: (1) a cable tray fire with
mechanical ventilation on (i.e., HVAC system was not shut down during the fire), (2) a
cable tray fire with exhaust fan on only (i.e., smoke purge during suppression), and (3) a
cable tray fire with mechanical ventilation off. In modeling the fire growth and damage
potential, Case 2 was considered to be the better representation of actual conditions in
the CSR. In the event of a fire, supply fan(s) in the CSR should shut down automatically
and the ventilation is limited only by the exhaust flow rate from the CSR.
The CFAST (Consolidated Model of Fire Growth and Smoke Transport) computer code
was used to model the fire growth of cable tray fire that is, by definition, considered to
be a deep-seated fire. The cable trays were assumed to ignite at the bottom of the
lowest tray which means that the entire cable tray will burn to produce a larger peak
heat release rate (HRR). For this analysis, the HRR for cable tray fires will be
approximated as slow fire growth rate. Results from the CFAST simulation of the
three ventilation conditions in the CSR show that there is sufficient oxygen available in
the CSR, even without mechanical ventilation and the door closed, such that a
significant fire can be sustained for some period of time. The upper gas layer
temperature with mechanical ventilation on (supply and exhaust) and with exhaust fan
on only, can lead to a flashover. Flashover is a phenomenon, which defines the point in
a compartment fire where all combustibles in the compartment are involved and flames
appear to fill the entire volume. During a fire, the exhaust fan will remove smoke from
the hot gas layer and raise the elevation of the layer interface. This exhaust will allow
the fire to burn at a higher intensity since more fresh air will be entrained into the fire
from the lower layer. Due to high temperatures of the upper gas layer, it is possible that
the exhaust fan could fail during the fire. In the case of cable tray fire with exhaust fan
on only (Case 2), the CFAST results show that the time to flashover ranges from 23
minutes to 29 minutes depending on the tray surface area.
During normal operations, the doors from CSR to other areas are closed. As such,
there are no large openings available (according to the licensee) to allow air into the
CSR to feed the fire. This will result in the fire becoming ventilation limited such that it
will burn at a lower intensity with a greater production of toxic smoke. The peak HRR is
ultimately determined by the amount of fresh air available to the fire. However, the CSR
is not perfectly sealed, gaps around the door and small cracks around the CSR will allow
the passage of air from the outside. Flashover occurs when all of the combustible
materials located in the compartment are involved and the fire is ventilation limited.
Backdraft is a phenomenon that results when a relatively well-sealed compartment is
ventilated near the floor level after the fire has been burning for an extended period of
time (i.e., hours) and almost all of the oxygen has been consumed, the compartment
accumulates a large quantity of non-combusted fuel-rich gases and the compartment
temperatures have increased significantly due to a smoldering fire in a semi-quiescent
state. The sudden addition of air (containing oxygen) results in the ignition of the fuel-
rich gases in a detonation or deflagration (i.e., backdraft occurs). The CFAST modeling
appears to indicate that flashover can occur without additional ventilation. If this is the
case, a backdraft cannot occur regardless of the fire brigade actions and the likelihood
that the fire would be allowed to develop for hours is small. Since there is sufficient
ventilation in the compartment for flashover, the fire brigades actions are not relevant to
the issue.
The effects of CSR fire were modeled using the HRR from the PE/PVC cable jacket
insulation with various cable tray exposed surface areas. Temperatures and products of
combustion in the CSR could result in damage to the entire CSR and all in-situ
combustibles therein. Without prompt automatic/manual fire suppression, hazardous
conditions are expected to occur in a relatively short period of time, i.e., about 23
minutes, in the CSR.
Non-IEEE-383 qualified cables were installed in the St. Lucie 1 CSR during its original
construction. These cables installed in the CSR cable trays were coated with Flamastic
fire-retardant coatings in order to offset the rapid flame spread of the thermoplastic
cables in the event of fire. However, Flamastic cable-coating is combustible, i.e., it does
not fire-proof the cables as initially thought. Table III of the Sandia National
Laboratories (SNL) report, A Preliminary Report on Fire Protection Research Program
Fire Retardant Coatings Tests, SAND 78-0518, 1978 provides the following data points:
(a) Test 13, non-IEEE-383 cable without cable coating had a burning duration of 36
minutes and an affected length of 70 inches.
(b) Test 9, IEEE-383 cable without cable coating had a burning duration of 13 minutes
and an affected length of 27 inches.
(c) Test 3, Same cable as Test 9 with Flamastic coating had a burning duration of 7
minutes and an affected length of 40 inches.
The Sandia cable test program did not test non-IEEE-383 cables with Flamastic
coatings (which were installed in St. Lucie 1 CSR cable trays). The cable trays in the
Sandia tests were only tested in the horizontal configuration which is not the worst case
configuration for cable tray fires. (Vertical cable tray configurations provide a more
challenging configuration, in that heat from the burning cables in the lower trays will
expose the cables above, causing the exposed cables to offgas and support more rapid
combustion. St. Lucie 1 CSR has both vertical and horizontal cable tray arrays.) Based
on this limited test data, it can be conjectured that the non-IEEE-383 qualified cables
with Flamastic coating in the St. Lucie 1 CSR, will perform better than Test #13 (Non-
IEEE 383 cables without coating) but not as well as Test#9 (IEEE-383 cables) or Test
- 3 (IEEE-383 cables with Flamastic coating). It is also assumed that any post-1980
modifications at St. Lucie 1 included the installation of IEEE-383 qualified cables, such
that the current as-installed plant configuration has both IEEE-383 qualified and non-
qualified cables with Flamastic coatings. Based on this information, even though non-
IEEE-383 qualified cables had been installed in the St. Lucie 1 CSR, no penalty has
been assessed in the analysis. Likewise, no credit for the Flamastic coatings was
assessed in the analysis.
Based on the results of the fire modeling, SPLB fire protection staff concluded that a fire
in the CSR would have a significant impact on the CSR. Depending on the ventilation
condition and exposed surface areas involved, it is possible for the room to flashover
and this could result in failure of the CSR structure.
3. Assumptions
(a) Fire Barrier - The licensee had replaced the Thermo-Lag 330 fire barrier wall
separating the Unit 1 cable spreading room and the Electrical Equipment Room 1B, with
a sheet metal and ceramic fibre barrier (See NRC Inspection Report 50-335/99-08,
January 31, 2000). As a result of the modification, the barrier can be considered to be
in the normal operating state. Therefore, in the event that both equipment trains in the
CSR are affected by fire, the remote hot shutdown panel in Electrical Equipment Room
1B could be used to achieve safe shutdown of the plant.
(b) Fire Detection and Manual Fire Suppression - The Unit 1 CSR is equipped with a
cross-zoned thermal detection system, which activates the Halon system. The cross-
zoned thermal detection system has two independent detectors per ceiling bay, spaced
some distance apart and preset to reach their actuation temperature to initiate the
discharge of the Halon system. The CSR area is also equipped with a detection system
of ionization smoke detectors that provide alarm conditions in the Main Control Room
(MCR). The 1998 Fire Protection Functional Inspection (FPFI) found a number of
problems with the Fire Alarm Control Panel (FACP) annunciator located in the MCR,
most notably the audible levels in the MCR (Section F7.4.1 of FPFI report). The
licensee had initiated a Condition Report, CR98-0453, to correct this problem. With this
condition corrected, alarms from the ionization fire detection system would result in the
dispatch of an operator to investigate the CSR, and plant-specific procedures (1-ONP-
100.01) would be implemented to notify the fire brigade for fire-fighting response. The
time taken for operator response to the fire alarm and arriving at the CSR area should
be reasonably rapid since the CSR is located directly below the MCR.
Based on previous four fire drills, all (five) members of the fire brigade arrived at the
CSR area in about 10 minutes after the MCR personnel sounds the alarm. The arrival
time of the assembled fire brigade in 10 minutes may not include the time from fire
ignition to the time that the alarm is sounded by MCR operators. This unaccounted time
begins when the fire must ignite (time = 0), grow and generate products of combustion
in sufficient quantity to initiate the local spot detector, the spot detector then sends an
alarm to the Local Fire Alarm Control Panel (LFACP) which in turn transmits an alarm to
the main FACP in the MCR. The operator must then acknowledge the alarm and
determine the location of the fire. By normal procedures, a member of the operations
staff would be dispatched to investigate the origin of the alarm. Upon confirmation of
the fire, the operations staff will then determine if the fire brigade should be activated. In
IPEEE reviews, the IPEEE Senior Review Team considered that the time from the start
of the fire until fire brigade arrival was a minimum of 30 minutes across industry-wide
experience. Consistent with the Reactor Oversight Process, a fire brigade rating of
normal operating state is attributed to the fire brigade response capability in order to
evaluate the significance of the degraded halon system.
The fire-induced core damage frequency (CDF) equation for the CSR area can be
defined as follows:
FCDF = Fi * Sf * P1 * P2 * P3
where Fi = Fire ignition frequency of ignition source
Sf = Severity factor for a challenging fire
P1 = Failure probability of Halon system
P2 = Failure probability of manual suppression by fire brigade
P3 = Conditional core damage probability, or failure probability of operator to
operate remote hot shutdown panel in Electrical Equipment Room 1B
This fire-induced CDF equation does not account for geometry factors for a large fire to
damage the critical equipment. Geometry factors are not recognized as an accepted
partitioning method for fire events. The use of geometry factors was proposed in some
early fire PRAs. However, this approach was rejected by NRC reviewers during the
IPEEE review process. Most fires that have challenged the safe shutdown capability of
a plant have not been characterized as large (Nureg/CR-6738), and also, the definition
of severe fires in the EPRI Fire PRA Implementation Guide include more than large
fires.
4: Fire Ignition Frequencies
The various ignition sources in the CSR area and their associated fire ignition frequency
estimates, as provided on the Ignition Source Data Sheet (ISDS) for the St. Lucie Unit 1
CSR, are shown below:
Ignition Source Fire Ignition Frequency
1. Transformers 1.09E-3
2. Electrical Cabinets 3.20E-3
3. Ventilation Systems 3.39E-4
4. Cable Runs 7.65E-4
5. Fire Protection Panels 1.75E-4
The ignition frequency estimate of 1.09E-3/yr for transformers represents the total
contribution from 10 transformers in the CSR. However, only two of the 10 transformers
are high voltage, dry-type transformers. The other transformers are small and low
voltage transformers, and are not located near to the combustible loads. In the case of
high voltage, dry-type transformers, vendor-specific information from ABB Power T& D
Co., indicated that only two transformer fires involving the VPE and VPI dry-type
transformers have occurred during 100,453 service years. This provides an estimate of
2.0E-5/yr for the fire ignition frequency of transformers in the CSR. This component-
specific fire ignition frequency estimate, rather than the IPEEE generic estimate of
1.09E-3/yr for transformers in the CSR, was used in the risk analysis because the
estimate of 2.0E-5/yr reflects realistic operational experience of the transformers.
The ignition frequency of 3.2E-3/yr for electrical cabinets represents the total
contribution from all electrical cabinets including the reactor trip switchgear (240V AC,
with solid bottom tray and solid cover above), 125 V DC bus, battery chargers, 480V
motor control centers, and programmer cabinets. It was conservatively estimated that
there are at least 80 electrical cabinets in the CSR. As identified in the CSR layout
drawing, the only electrical cabinet housing medium voltage electrical equipment is the
480V Motor Control Center 1AB cabinet. The licensees fire hazard analysis showed
that MCCs have concentrated combustible loading of 830,000 BTUs. The remaining
electrical cabinets located in the CSR contain low voltage electrical equipment, and
each appear to have low combustible loadings (i.e., the volume fraction of cable
insulation and non-metal combustibles is estimated to be 10 percent or less). Based on
these considerations, the evaluation of risk impacts from electrical cabinets is divided
into two scenarios, beginning with: (1) 480V MCC 1AB cabinet, and (2) remaining low
voltage electrical cabinets. Therefore, the ignition frequencies for the 480V MCC 1AB
cabinet and remaining low voltage electrical cabinets used in the risk analysis were
estimated as follows:
(1) 480V MCC 1AB cabinet: (1/80)x (3.2E-3) = 4.0E-5 events/yr
(2) Low voltage electrical cabinets: (79/80)x(3.2E-3) = 3.16E-3 events/yr
The ignition frequency estimate of 1.75E-4/yr for the fire protection panels represents
the total contribution from two supervisory electrical cabinets containing relays and
annunciator lights alarming the fire zones affected by fire.
5: Conditional Core Damage Probability (CCDP)
In the various fire scenarios considered (i.e., each scenario initiated by a different
ignition source), the conditional core damage probability was estimated to be 3.0E-4
(using licensees PRA model) if there is one equipment train available to perform
mitigating functions. In the event that both equipment trains in the CSR are affected by
fire, the CCDP would be dominated by operator actions to achieve safe shutdown at the
remote hot shutdown (HSD) panel at the Electrical Equipment Room 1B.
In the licensees IPEEE study, the human error probability (HEP) of failing to control the
plant safe shutdown at the HSD control panel was estimated to be 0.1. In the deep
seated fire scenario, performance shaping factors (PSFs) for the HEP of operator
failure to start the HVAC fan to prevent heat and smoke buildup before operating safe
shutdown equipment must be considered. Using the ASP human error worksheet, the
total PSF value for the nominal HEP of the action to start the HVAC fan was determined
to be 300 based on the multiplication of each PSF (i.e., nominal time = 1, extreme stress
= 5, moderately complex = 2, low experience, or training = 3, available, but poor
procedures = 5, nominal ergonomics = 1, nominal fitness for duty = 1, poor work
processes = 2). Since the nominal HEP for a single action of starting the HVAC fan is
1E-3, the estimated HEP is 0.3 for the case of the available procedure having poor
instructions for achieving proper ventilation and smoke control. In a hostile, fire-induced
environment, an important PSF for consideration is the opacity effects of smoke on the
operator to manipulate the HSD control panel. If the smoke effects on the operators
vision to manipulate the HSD control panel and potential ineffectiveness of the HVAC
system after its restart (based on an inspection finding discussed in NRC IR 98-201) are
taken into account, an additional probability value of 0.1 should be included in the
overall estimation of the operator failure to achieve safe shutdown. Therefore, a
reasonable value of the HEP for operator failure to manually start the HVAC and
operate the HSD control panel was estimated to be 0.4 for the stated case. This
probability value is more conservative than the probability of 0.1 assumed in the
licensees analysis.
6. Integrated Assessment of Fire-Induced Core Damage Frequency
The fire-induced CDF estimate for fire in the CSR with a failed Halon system is
calculated as shown below:
Ignition Source Fi Sf P1 P2 P3 FCDF
1. Transformers 2.00E-5 1.00 1.0 0.5 4.0E-1 4.0E-6
2. 480V MCC 1AB Cabinet 4.00E-5 0.12 1.0 0.5 4.0E-1 9.6E-7
3. Low Voltage Electrical Cabinets 3.16E-3 0.12 1.0 0.5 3.0E-4 5.7E-8
4. Ventilation Systems 3.39E-4 0.08 1.0 0.5 3.0E-4 4.1E-9
5. Cable Runs 7.65E-4 0.01 1.0 0.5 4.0E-1 1.5E-6
6. Fire Protection Panels 1.75E-4 0.12 1.0 0.5 3.0E-4 3.2E-9
Total CDF 6.5E-6
(a) Severity Factors, Sf - A fire severity factor is a fractional value (between 0 and 1) that
is used to adjust fire frequency estimates to reflect some specific mitigating pattern of
behavior of the fire event. The severity factor is applied to reflect a split in challenging
versus non-challenging fires. In the absence of plant-specific information, the severity
factors for the electrical cabinets, ventilation systems, and fire protection panels were
based on the EPRI Fire PRA Implementation Guide (FPRAIG), December 1995 (Section
D.3). The FPRAIG severity factors ranged from 0.08 to 0.2, and engineering judgment
was used to determine these severity factors. The question of double counting was
raised concerning the use of severity factors and credit for manual suppression by the
fire brigade in this risk analysis. The severity factors used for the fire risk analysis of the
electrical cabinets, ventilation systems, and fire protection panels reflect the ratio of
severe fires to the total number of fires observed in each fire source group. The severity
factor approach used in the risk analysis model is to account for the probability of fire
growth, and the severity factors are independent of the credit for subsequent fire-
fighting efforts by the fire brigade.
The severity factor used for the pressurizer heater transformers in the risk analysis is
1.0 because it is assumed that the transformer fire would be a single-point source of
flames with height of at least 5 feet, and the distances from the top of the transformers
to the lowest cable trays were less than 5 feet. The use of the severity factor of 1.0 is a
conservative assumption because the reported transformer fires (based on vendor
historical information) were self-extinguished when the fires were not sustained after
ignition.
In the case of cable runs in the CSR, it is the analysts judgment that the severity factor
used in the risk analysis is 0.01 based on the assumption that one percent of self-ignited
cable tray fires may grow into a large fire if the original fire is undetected and
unsuppressed. Although cable self-ignition events have occurred in the past, none of
the self-ignited cable fires in the current operating U.S. plants have led to a large fire
(see NUREG/CR-6738). Furthermore, some of the cable trays in the CSR are vented
trays with solid covers, and the average fill of the cable trays near to the major ignition
sources (i.e., pressurizer heater transformers and programmer cabinets) is, at the most,
about 25 percent.
(b) Halon system failure probability, P1 - The failure probability of the Halon system to
extinguish a fully-developed fire in the CSR was assumed to be 1.0. The major in-situ
fire hazard in the CSR is of electrical origin. The major in-situ fuel loading in the CSR is
cable insulation and jacket (for both IEEE-383 and non-IEEE-383 qualified cables).
Electrical cable fires are deep-seated fires. St. Lucie Stations NFPA Code of Record
(COR) for the installation of the Halon system is NFPA 12A, Halon 1301 Fire
Extinguishing Systems, 1980 Edition. The code recommends a minimum of 10%
concentration Halon 1301, held in the enclosure for a minimum of 10 minutes (Section
A-2-4), or allows a different concentration and hold time that is approved by the
Authority Having Jurisdiction (AHJ). The NRC, i.e., the AHJ, had sponsored testing of
fire suppressant agent effectiveness, and established the minimum concentrations and
hold times for fire suppressant agents used to extinguish cable fires in nuclear power
plants. As documented in NUREG/CR -3656, Evaluation of Suppression Methods for
Electrical Cable Fires, October 1986, Table 8 shows that a minimum of 6%
concentration of Halon 1301 should be held for 15 minutes for IEEE-383 qualified cables
and 10 minutes for non-IEEE-383 qualified cables to result in no re-ignition from fully
developed, five-tray cable fires. At the lowest cable tray elevation, the CSR Halon 1301
system drops below 6% concentration after 4 minutes and continues to decrease to
approximately 4.5% at 15 minutes. Concentrations measured at the lowest levels of the
CSR (i.e., on the floor far below the cable fire hazard) are less than the minimum 6%
concentration at 15 minutes. Therefore, no credit was given to the Halon 1301
suppression system. Furthermore, research studies have shown that the ineffective
Halon 1301 agent which thermally decomposes will increase the generation of toxic
products of combustion (smoke). Most notably, halogen acid products, particularly
hydrogen fluoride, are generated. This condition will further hamper fire brigade
personnel in performing manual suppression activities, and affect operations personnel
in adjacent locations performing alternate shutdown.
(c) Manual Suppression by Fire Brigade, P2 - Historical records of fire brigade drills
performed for the CSR (on 11/14/85, 9/19/97, 3/6/98, and 3/8/98) have shown that all
required members of the fire brigade team arrived at the CSR in less than 10 minutes
after notification. Additionally, historical records of fire brigade response to fires in the
CSR due to capacitor failures in battery charger cabinets (on 12/6/83 and 11/14/85)
indicated that the two fires were successfully extinguished by operators investigating the
scene prior to the arrival of the full fire brigade in 6 minutes after notification from the
Control Room.
Based on the above considerations, together with the discussion in Section 3b, entitled:
Fire Detection and Manual Fire Suppression, and since the fire brigade response
capability was considered to be in the normal operating state (due to the Reactor
Oversight Process guidance to consider findings separately in significance evaluations
unless a common root cause exists), the failure probability of manual suppression by the
fire brigade associated with severe fires was estimated to be 0.5 in this risk analysis.
(d) Conditional Core Damage Probability, P3 - In the fire scenarios involving ignition of
the electrical cabinets, ventilation systems, and fire protection panels, an analysis was
performed to determine whether a redundant cable train would be affected by fire
damage from the specified ignition source. In this analysis, the critical radiant flux
distances were estimated for various peak fire intensities (e.g., 65 BTU/sec to 500
BTU/sec) using the single-point source equation on page 10.4-23 of the Fire-Induced
Vulnerability Evaluation (FIVE) Report. The critical radiant flux distance varies from 2.0
ft. for 65 BTU/sec to 5.6 ft. for 500 BTU/sec heat release rate and 0.5 BTU/s/ft2 for
unqualified cables. These critical heat flux distances were then compared to the
horizontal distances between the redundant trains of cable trays with the specific ignition
source directly below its nearest cable tray. Based on review of layout drawings of the
cable tray configurations and ignition source locations, and information on measured
distances provided by the licensee, the minimum horizontal distance between redundant
cable trains was estimated to be 6 ft (for the low voltage electrical cabinets such as 125
VDC Bus 1AB-1, 1AB Battery Charger, 1AB DC Switchgear, and Vital AC Supply
cabinets), while the maximum distance between A and B trains is about 19 ft (for the
pressurizer heater transformers). Based on observations of low combustible loadings of
the low voltage electrical cabinets (i.e., the volume fraction of cable insulation and non-
metal combustibles is estimated to be 10 percent or less) and the minimum distance
separation of redundant cable trains being greater than the critical radiant heat flux
distances with conservative heat release rates, it is reasonable to assume that one
equipment train would be available to perform mitigating functions if any of the low
voltage electrical cabinets were to ignite. Therefore, the CCDP estimate of 3.0E-4 was
used in the risk analysis of the fire scenarios involving the low voltage electrical
cabinets. Similar arguments can be made to use the CCDP estimate of 3.0E-4 for
analyzing the fire scenarios involving the ventilation systems and fire protection panels.
The air handling units, HVA-4 and HVA-5, located in the CSR contained less than 4 lb.
of plastic filter material and wiring insulation and a few ounces of grease associated with
the fan and motor. The air handling units are about 3 ft. below the nearest cable trays,
and the distance between redundant cable trains appear to be greater than 6 ft. The
two fire protection panels are small enclosed metal boxes with small indicator lights that
show the location of specific detectors in the alarm mode. Each fire protection panel
contains about 10 lb of plastic/insulation, and are located about 3 ft horizontally and 4 ft
vertically from the nearest cable tray.
In the case of the transformers and cable runs, it is assumed that a fire from these
sources would result in a deep-seated fire and the operators would have to gain access
to the remote HSD panel to achieve safe shutdown of the plant. Based on the ASP
Human Reliability Analysis method, the operator failure to operate the HSD control
panel was estimated to be at least 0.4, and this probability was used in the risk analysis
as the CCDP estimate. The risk impact of spurious actuations from hot shorts is
subsumed in this CCDP estimate of 0.4.
In the case of 480V Motor Control Center 1AB cabinet, the licensees fire hazard
analysis showed that MCCs have concentrated combustible loading of 830,000 BTUs.
This 830,000 BTU combustible loading can result in a fire of peak fire intensity of 332
BTU/sec (using the equation on page E-8, Fire PRA Implementation Guide). The critical
radiant flux distance for this fire peak intensity was estimated to be 4.7 ft. Although the
minimum separation distance between redundant cable trays for the nearest cable tray
above the 480V MCC 1AB cabinet is about 6 ft., it is assumed that a fire from the 480V
MCC 1AB could damage both redundant cable trains as a bounding analysis. Therefore,
the CCDP estimate of 0.4 was used in the risk analysis of a fire from the 480V MCC
1AB.
7: Incremental Fire-Induced CDF
The baseline CDF (conforming case) for the cable spreading room without a degraded
Halon system is calculated by assuming the Halon system failure probability of 0.05, and
the manual suppression failure probability of 0.1. These values were used for the non-
degraded Halon system and manual suppression capability because the values were
considered to be appropriate for the entire population of fires (including deep-seated
fires) arising from an ignition source. It is expected that failure probabilities for these fire
protection systems and features would be much higher if a severity factor was used
since the severity factor represents the case where only severe fires are considered.
Severe fires are defined as fires that have grown beyond the incipient stage and would
have the potential to damage structures, systems or components (SSCs) beyond the
SSC in which the fire originates. Severe fire events are a subset of the entire population
of fires.
Deep-seated fires (which are the concern of the finding related to the Halon system
design concentration and soak time at St. Lucie Unit 1 CSR) can be either severe or
non-severe (i.e. depending on whether fire damage is limited to the component of fire
origin). Based on data from currently available databases of fire events, it is presently
not feasible to partition the deep-seated fire events from the non-deep-seated fire
events to develop a deep-seated fire factor to adjust the fire ignition frequency similar
to that done for severe fires. However, all severe fires, as defined above, involving
cables routed in cable trays are expected to result in a deep-seated fire due to the
combustible characteristics of cable insulation (i.e., the fire burns internally as opposed
to a surface fire). Additionally, an accurate determination of the effectiveness of a code
compliant Halon system on a severe fire is beyond the current state-of-the-art methods.
The basis for using the failure probability value of 0.05 for the non-degraded Halon
system in the conforming case of this analysis is found in EPRIs Fire-Induced
Vulnerability Evaluation (FIVE) guidance (page 10.3-7). This probability value is the
accepted value used by the community of fire risk analysis practitioners for a non-
degraded Halon system. The basis for using the probability of 0.1 for failure of manual
suppression in this conforming-case analysis is, as discussed above, that the value of
0.1 is appropriate for the entire population of fires. The fire protection SDP
methodology, which uses the entire population of fires as the basis to derive an ignition
frequency, also uses the probability value of 0.1, in general, for the failure probability of
nondegraded manual suppression capability.
Based on the preceding discussions, the baseline CDF for the St. Lucie Unit 1CSR
without a degraded Halon system is calculated as shown below:
Ignition Source Fi P1 P2 P3 FCDF
1. Transformers 2.00E-5 0.05 0.1 4.0E-1 4.0E-8
2. 480V MCC 1AB Cabinet 4.00E-5 0.05 0.1 4.0E-1 8.0E-8
3. Low Voltage Electrical Cabinets 3.16E-3 0.05 0.1 3.0E-4 4.7E-9
4. Ventilation Systems 3.39E-4 0.05 0.1 3.0E-4 5.1E-10
5. Cable Runs 7.65E-4 0.05 0.1 4.0E-1 1.5E-6
6. Fire Protection Panels 1.75E-4 0.05 0.1 3.0E-4 2.6E-10
Baseline CDF 1.6E-6
The incremental CDF change due to a failed Halon system in the CSR would be:
6.5E-6 - 1.6E-6 = 4.9E-6
CONCLUSION: The change in CDF due to a completely failed Halon system is 4.9E-6.
Therefore, the significance characterization of this issue is WHITE.