Detailed Fire Modeling (Task 11)

Task Overview

Background

This task describes the method to examine the consequences of a fire. This includes consideration of scenarios involving single compartments, multiple fire compartments, and the main control room. Factors considered include initial fire characteristics, fire growth in a fire compartment or across fire compartments, detection and suppression, electrical raceway fire barrier systems, and damage from heat and smoke. Special consideration is given to turbine generator (T/G) fires, hydrogen fires, high-energy arcing faults, cable fires, and main control board (MCB) fires. There are considerable improvements in the method for this task over the EPRI FIVE and EPRI's Fire PRA Implementation Guide (TR‑105928, no longer available on epri.com) in nearly all technical areas.

Purpose

In the preceding tasks, the analyses were organized around compartments, assuming that a fire would have widespread impact within the compartment. In Task 11, for those compartments found to be potentially risk-significant (i.e., unscreened compartments), a detailed analysis approach is provided. As part of the detailed analysis, fire growth and propagation is modeled and possibility of fire suppression before damage to a specific target set is analyzed.

The detailed fire modeling process generally follows a common step structure, but the details of the analyses often vary depending on the specifics of the postulated fire scenario. This task provides separate procedures for three general categories of fire scenarios: fires affecting target sets located inside one compartment (discussed in Section 11.5.1); fires affecting the main control room (MCR; Section 11.5.2); and fires affecting target sets located in more than one fire compartment (multicompartment fire analysis; Section 11.5.3).

Task 11 provides final estimates for the frequency of occurrence of fire scenarios involving a specific fire ignition source failing a predefined target set before fire protection succeeds in protecting the target set. This result is combined in the final quantification steps that follow this task, with the CCDP/CLERP given failure of the target set to estimate the CDF/LERF contribution for each fire scenario. The CCDP/CLERP may include modified human error probabilities based on fire scenario specifics.

Scope

Detailed fire modeling encompasses an analysis of the physical fire behavior (i.e., fire growth and propagation analysis), equipment damage, fire detection, and fire suppression. The fire scenarios to analyze as part of this detailed analysis task are divided into three categories:

• General single compartment fire scenarios. This general category covers fire scenarios damaging target sets located within the same compartment, exclusive of those scenarios within or impacting the MCR. In general, in this category, the fire ignition source is in the same compartment as the target set. The majority of fire scenarios analyzed generally falls into this category. The procedures applicable to the analysis of these fire scenarios are presented in Section 11.5.1.
• MCR fire scenarios. This general category covers all fires that occur within the MCR. This category also covers scenarios involving fires in compartments other than the MCR that may force MCR abandonment. The MCR analysis procedures are presented in Section 11.5.2.
• Multicompartment fire scenarios: This general category covers all fire scenarios where it is postulated that a fire may spread from one compartment to another and damage target elements in multiple compartments. In this category of scenarios, damaging effects of a fire (e.g., heat) are assumed to spread beyond the compartment of fire origin. The multicompartment fire analysis procedures are presented in Section 11.5.3.

A detailed fire modeling analysis is performed for each fire scenario in each unscreened fire compartment. For many compartments, it may be appropriate to develop several fire scenarios to appropriately represent the range of unscreened fire ignition sources (i.e., scenarios that would not screen out in Task 8) that might contribute to the fire risk. Detailed fire modeling may utilize a range of tools to assess fire growth and damage behavior, and the fire detection and suppression response, for specific fire scenarios.

The ultimate output of Task 11 is a set of fire scenarios, frequency of occurrence of those scenarios, and a list of target sets (in terms of fire PRA components) associated with the scenarios. For scenarios involving the MCR, the possibility of forced abandonment is also noted. Note that a fire scenario represents a specific chain of events starting with ignition of a fire ignition source, propagation of the fire effects to other items, and possibility of damaging a set of items identified as a target set before successful fire suppression.

Related Element of ASME/ANS PRA Standard

Fire Scenario Selection (FSS)

Related EPRI 1011989 NUREG/CR‑6850 Appendices

Appendix E, Appendix for Chapters 8 and 11, Severity Factors

Appendix F, Appendix for Chapter 8, Walkdown Forms

Appendix G, Appendix for Chapters 8 and 11, Heat Release Rates

Appendix H, Appendix for Chapters 8 and 11, Damage Criteria

Appendix L, Appendix for Chapter 11, Main Control Board Fires

Appendix M, Appendix for Chapter 11, High Energy Arcing Faults

Appendix N, Appendix for Chapter 11, Hydrogen Fires

Appendix O, Appendix for Chapter 11, Turbine Generator Fires

Appendix P, Appendix for Chapter 11, Detection and Suppression Analysis

Appendix Q, Appendix for Chapter 11, Passive Fire Protection Features

Appendix R, Appendix for Chapter 11, Cable Fires

Appendix S, Appendix for Chapter 11, Fire Propagation to Adjacent Cabinets

Appendix T, Appendix for Chapter 11, Smoke Damage

Fire Modeling Tools

Fire modeling tools include a range of complexity, from Excel-based tools which rely on physics-based algebraic relationships such as EPRI FIVE and the NRC FDT^s, to moderately complex tools such as CFAST's two-zone computational model, up to the most complex (and computationally-demanding) finite element analysis tools such as FDS.

Fire Model Verification and Validation

NUREG‑1824 EPRI 1011999 documents the verification and validation (V&V) of five fire models that are commonly used in NPP applications. The models in the V&V report include:

• NRC's NUREG‑1805 Revision 1
• EPRI's Fire-Induced Vulnerability Evaluation Revision 1 (FIVE-REV 1)
• National Institute of Standards and Technology's (NIST) Consolidated Model of Fire Growth and Smoke Transport (CFAST) Version 5
• NIST's Fire Dynamics Simulator (FDS) Version 4
• Electricite de France's (EdF) MAGIC Version 4.1.1

NUREG‑1824 Supplement 1 EPRI 3002002182 updates the original NUREG‑1824 / EPRI 1011999 report with additional experiments and uses the latest versions of the fire modeling software available at the time of publication. The models in the V&V report include:

• NRC's Fire Dynamics Tools (FDT^s Version 1805.1)
• EPRI's Fire-Induced Vulnerability Evaluation (FIVE Revision 2)
• NIST's CFAST Version 7.0.0
• EdF's MAGIC Version 4.1.3
• NIST's FDS Version 6.2.0

Fire Models Included in V&V Guidance

EPRI FIVE

NRC Fire Dynamics Tools - NUREG‑1805

NIST CFAST

NIST FDS and Smokeview

EdF's MAGIC is available through EPRI for EPRI members

Fire Model User's Guide

NUREG‑1934 EPRI 1023259 provides guidance on the proper application of fire models to nuclear power plant fire scenarios. Eight (8) different example fire scenarios are developed and discussed in this report.

Ignition Source Specific Fire Modeling Guidance

Bin	Plant Location	Ignition Source	Fire Modeling Guidance	Fire Modeling Reference
1	Battery Room	Batteries	Use HRR distribution for Motors (Distribution 7 of Table G-1)	EPRI 1011989 / NUREG/CR‑6850
2	Containment (PWR)	Reactor Coolant Pumps	Reactor coolant pump fires are classified as either electrical (motor) or oil. The split fraction between electrical and oil fires is provided in NUREG/CR‑6850 (0.14 electrical / 0.86 oil). Electrical (motor) fires: HRR distributions and fire durations are provided in Chapter 5 of NUREG‑2178, Volume 2 / EPRI 3002016052. The pump HRR in NUREG/CR‑6850 is bounding compared with the updated values, and is therefore still valid. Oil fire split fractions: The oil spill size fractions recommended in NUREG/CR‑6850 Appendix E.3 should be applied. Oil fire HRR: See Section G.4 of NUREG/CR‑6850 for HRR for flammable liquid fires. EPRI 3002005303, although not formally reviewed by the NRC, provides a method to more realistically characterize the HRR profile and duration for liquid spill fires.	NUREG‑2178 Volume 2 / EPRI 3002016052 EPRI 1011989 / NUREG/CR‑6850 EPRI 3002005303
3	Containment (PWR)	Transients and Hotwork	NUREG‑2233 / EPRI 3002018231 provides updated HRR distributions and zones of influence for generic transient fires and also transient combustible control locations (TCCLs). These HRRs are based upon the laboratory testing conducted by EPRI and the NRC on relevant transient ignition sources expected in nuclear power plants (see EPRI 3002015997 / NUREG‑2232). The HRR distribution (Distribution 8 of Table G‑1 in NUREG/CR‑6850) is bounding compared with the updated generic HRR distribution, and is therefore still valid. NUREG‑2233 / EPRI 3002018231 also recommends fire modeling parameters including fire growth and decay parameters, yields of minor products of combustion, heat of combustion, and the physical size and effective elevation of the fire.	NUREG‑2233 / EPRI 3002018231 NUREG‑2232 / EPRI 3002015997 NUREG/CR‑6850 / EPRI 1011989
4	Control Room	Main Control Board	Target damage: Appendix L of NUREG/CR‑6850 provides a statistical model for estimating the conditional probability of damage to a set of target items inside the main control board. § Target damage: NUREG‑2178 Volume 2 / EPRI 3002016052 Section 7 provides an alternative to the method described in Appendix L of NUREG/CR‑6850 for evaluating the risk of fire events originating in the MCB, whereby MCB fire scenarios are modeled as a progression of damage states using an event tree model. § In this formulation, each damage state requires the definition of a target set, which consists of one or more MCB functions that can be damaged by fire. The functions within the scope of this analysis are those that are represented with basic events in the plant response model and supported with cables routed within the MCB. The alternative model described in this guidance explicitly incorporates two characteristics of MCB fires observed in operating experience—relatively small fires in low-voltage panels and the ability for prompt detection and suppression by control room operators. Operating experience suggests that the majority of fires in the MCB are limited to a single subcomponent or group of subcomponents near the point of ignition. In addition, these fires are promptly detected and suppressed by control room operators. Therefore, the event tree model explicitly accounts for the operator’s ability to quickly detect and suppress the fire before growth and/or propagation. § The original NUREG/CR‑6850 Appendix L method and NUREG‑2178 Volume 2 event tree method BOTH remain viable as methods for assessing MCB fires. HRR distributions: NUREG‑2178 Volume 1 / EPRI 3002005578 provides updated HRR distributions for the main control board based on control cabinet size (either Function Group 4a (Large Enclosures) or Group 4b (Medium Enclosures)). Propagation to adjacent cabinet: NUREG‑2178 Volume 2 / EPRI 3002016052 Section 4 provides a method for refining the postulated spread of fires from one cabinet to an adjacent cabinet. This report provides screening guidance, a conditional probability (split fraction), a limitation of spread to a single adjacent cabinet only, and timing for the spread.	EPRI 1011989 / NUREG/CR‑6850 NUREG‑2178 Volume 1 / EPRI 3002005578 NUREG‑2178 Volume 2 / EPRI 3002016052
5	Control/Aux/Reactor Building	Cable fires caused by welding and cutting	FAQ 13‑0005 provides additional guidance for detailed fire modeling on both self-ignited cable fires and cable fires caused by welding and cutting. This FAQ outlines a more realistic approach for addressing these types of fires in cable trays and suggests replacement text for Section R.1 of NUREG/CR‑6850. However, the current method of evaluating cable fire risk in NUREG/CR‑6850 remains an acceptable approach.	EPRI 1011989 / NUREG/CR‑6850 FAQ 13‑0005
6	Control/Aux/Reactor Building	Transient fires caused by welding and cutting	See Bin 3 for treatment of transient fires.	See Bin 3
7	Control/Aux/Reactor Building	Transients	See Bin 3 for treatment of transient fires.	See Bin 3
8	Diesel Generator Room	Diesel Generators	There is limited guidance on modeling diesel generator fires in NUREG/CR-6850: Diesel generator fires have an electrical (motor) component and an oil component. The split fraction between electrical and oil fires is provided in NUREG/CR‑6850 (0.16 electrical / 0.84 oil). Section G.4 of NUREG/CR‑6850 provides guidance on flammable liquid (oil) fires. EPRI 3002005303, although not formally reviewed by the NRC, provides a method to more realistically characterize the HRR profile and duration for liquid spill fires.	EPRI 1011989 / NUREG/CR‑6850 EPRI 3002005303
9	Plant-Wide Components	Air Compressors	Air compressor fires are classified as either electrical (motor) or oil. The split fraction between electrical and oil fires is provided in NUREG/CR‑6850 (0.83 electrical / 0.17 oil). Electrical (motor) fires: HRR distributions and fire durations are provided in Chapter 5 of NUREG‑2178, Volume 2 / EPRI 3002016052. Oil fire split fractions: The oil spill size fractions recommended in NUREG/CR‑6850 Appendix E.3 should be applied. Oil fire HRR: See Section G.4 of NUREG/CR‑6850 for HRR for flammable liquid fires. EPRI 3002005303, although not formally reviewed by the NRC, provides a method to more realistically characterize the HRR profile and duration for liquid spill fires.	NUREG‑2178 Volume 2 / EPRI 3002016052 EPRI 1011989 / NUREG/CR‑6850 EPRI 3002005303
10	Plant-Wide Components	Battery Chargers	Table 7‑1 of NUREG‑2178 Volume 1 provides HRR distributions for Group 2 electrical enclosures, including battery chargers.	NUREG‑2178 Volume 1 / EPRI 3002005578
11	Plant-Wide Components	Cable fires caused by welding and cutting	FAQ 13‑0005 provides additional guidance for detailed fire modeling on both self-ignited cable fires and cable fires caused by welding and cutting. This FAQ outlines a more realistic approach for addressing these types of fires in cable trays and suggests replacement text for Section R.1 of NUREG/CR‑6850. However, the current method of evaluating cable fire risk in NUREG/CR‑6850 remains an acceptable approach.	EPRI 1011989 / NUREG/CR‑6850 FAQ 13‑0005
12	Plant-Wide Components	Cable Run (self-ignited cable fires)	FAQ 13‑0005 provides additional guidance for detailed fire modeling on both self-ignited cable fires and cable fires caused by welding and cutting. This FAQ outlines a more realistic approach for addressing these types of fires in cable trays and suggests replacement text for Section R.1 of NUREG/CR‑6850. However, the current method of evaluating cable fire risk in NUREG/CR‑6850 remains an acceptable approach.	EPRI 1011989 / NUREG/CR‑6850 FAQ 13‑0005
13	Plant-Wide Components	Dryers	The transient HRR is recommended for Bin 13 dryer fires (refer to Table 11‑1 of NUREG/CR‑6850). NUREG‑2233 / EPRI 3002018231 provides updated HRR distribution and zones of influence for generic transient fires (see also Bin 3).	NUREG‑2233 / EPRI 3002018231 NUREG/CR‑6850 / EPRI 1011989
14	Plant-Wide Components	Electric Motors	Chapter 5 of NUREG‑2178 Volume 2 / EPRI 3002016052 provides updated HRR distributions for electric motors (compared with the original distribution from NUREG/CR‑6850 Table G-1). To improve realism, the HRRs in NUREG‑2178 Volume 2 are characterized by horsepower, and NUREG‑2178 Volume 2 also provides growth and decay timing.	NUREG‑2178 Volume 2 / EPRI 3002016052
15	Plant-Wide Components	Electrical Cabinets	Propagation from electrical cabinets: FAQ 08‑0042 (Section 8 of Supplement 1) clarifies the treatment of fire spread beyond the ignition source for electrical cabinets considering conditions such as the presence of ventilation, robust door construction, and seal penetration. This clarification was needed due to conflicting language in Chapters 6 and 11 and Appendix G of NUREG/CR‑6850. FAQ 08‑0042 states that the wording in Chapter 11 is correct. Propagation to adjacent cabinet: NUREG‑2178 Volume 2 / EPRI 3002016052 Section 4 provides a method for refining the postulated spread of fires from one cabinet to an adjacent cabinet. This report provides screening guidance, a conditional probability (split fraction), a limitation of spread to a single adjacent cabinet only, and timing for the spread. Propagation for Well-Sealed MCCs Greater Than 440V: FAQ 14‑0009 provides clarification for the treatment of fire propagation from well-sealed MCCs operating at greater than 440V. Heat Release Rates: NUREG‑2178 Volume 1 / EPRI 3002005578 provides updated heat release distributions for electrical enclosures. The analyst should review the equipment function or size to determine an appropriate heat release rate distribution provided in Table 7-1. Heat release rates for electrical cabinets are also found in Table G-1 of EPRI 1011989 / NUREG/CR‑6850. Fire location: FAQ 08‑0043 clarifies the treatment of fire location in electrical cabinets. Fire diameter: NUREG‑2178 Volume 1 / EPRI 3002005578 Section 4.2 provides guidance on the selection of an appropriate fire diameter. Obstructed plume model: NUREG‑2178 Volume 1 / EPRI 3002005578 Section 6 provides a method to account for the impact of the enclosure on the vertical thermal zone of influence above the enclosure during a fire. A summary of the obstructed plume methodology and the results can be found here. Obstructed radiation model: NUREG‑2178 Volume 2 / EPRI 3002016052 Section 3 provides a method to account for the impact of the enclosure on the horizontal (radial) zone of influence surrounding the enclosure during a fire. This report establishes values for the ZOI measured from the cabinet face as a function of the cabinet type, cable type, fuel loading, and fire size. Growth and suppression: NUREG‑2230 / EPRI 3002016051 includes the following updates: 1) Updated fire ignition frequency (Task 6) 2) Classification of electrical cabinet fires into one of two profiles: Interruptible fires 0.723 Growing fires 0.277 3) HRR timing for interruptible and growing fires: Interruptible fires – Option 1: Pre-growth (negligible HRR) 9 minutes Growth 7 minutes Steady state 5 minutes Decay 13 minutes Interruptible fires – Option 2: (NUREG/CR‑6850 timing profile supplemented with pre-growth period) Pre-growth (negligible HRR) 4 minutes Growth 12 minutes Steady state 8 minutes Decay 19 minutes Growing fires:  (unchanged from NUREG/CR‑6850) Growth 12 minutes Steady state 8 minutes Decay 19 minutes 4) Changes to the detection-suppression event tree to better represent the operating experience. The detection-suppression event tree was revised to better represent the manual suppression outcomes observed in operating experience (including the development of two new manual suppression curves). See EPRI 3002016051 / NUREG‑2230 for full details.	EPRI 1011989 / NUREG/CR‑6850 FAQ 08‑0042, Section 8 of Supplement 1 FAQ 14‑0009 NUREG‑2178 Volume 1 / EPRI 3002005578 FAQ 08‑0043, Section 12 of Supplement 1 NUREG‑2230 / EPRI 3002016051 NUREG‑2178 Volume 2 / EPRI 3002016052
16.a	Plant-Wide Components	High Energy Arcing Faults - Low Voltage Electrical Cabinets (480-1000 V)	Appendix M (M.4.2) provides an empirical model for determination of the ZOI from High Energy Arcing Faults (HEAFs).	EPRI 1011989 / NUREG/CR‑6850
16.b	Plant-Wide Components	High Energy Arcing Faults - Medium Voltage Electrical Cabinets (>1000 V)	Appendix M (M.4.2) provides an empirical model for determination of the ZOI from HEAFs.	EPRI 1011989 / NUREG/CR‑6850
16.1	Plant-Wide Components	HEAF for segmented bus ducts	Section 7.2.1.5 of Supplement 1 (FAQ 07-0035) provides an empirical model for estimating the ZOI for segmented bus duct fires.	FAQ 07-0035, Section 7 of Supplement 1
16.2	Plant-Wide Components	HEAF for iso-phase bus ducts	Section 7.2.1.5 of Supplement 1 (FAQ 07-0035) provides an empirical model for estimating the ZOI for iso-phase duct fires.	FAQ 07-0035, Section 7 of Supplement 1
17	Plant-Wide Components	Hydrogen Tanks	See Appendix N of NUREG/CR‑6850.	EPRI 1011989 / NUREG/CR‑6850
18	Plant-Wide Components	Junction Boxes	FAQ 13‑0006 provides a definition for junction boxes that allows the characterization and quantification of these scenarios in fire compartments that require detailed fire modeling analysis.	FAQ 13‑0006
19	Plant-Wide Components	Miscellaneous Hydrogen Fires	See Appendix N of NUREG/CR‑6850.	EPRI 1011989 / NUREG/CR‑6850
20	Plant-Wide Components	Off-gas/H2 Recombiner (BWR)	See Appendix N of NUREG/CR‑6850.	EPRI 1011989 / NUREG/CR‑6850
21	Plant-Wide Components	Pumps and large hydraulic valves	Pump fires are classified as either electrical (motor) or oil. The split fraction between pump electrical and oil fires is updated in EPRI 3002002936 / NUREG‑2169 (0.69 electrical / 0.31 oil). Electrical (motor) fires: In NUREG/CR‑6850, Bin 21 pump electrical fires were distinguished from non-pump motor fires. Research documented in NUREG‑2178 Volume 2 / EPRI 3002016052 suggests that there is little or no difference between pump motor fires and non-pump motor fires, and so electric motors and motor-driven pumps have been consolidated into a single ignition source. To improve realism, the HRRs in NUREG‑2178 Volume 2 are characterized by horsepower, and NUREG‑2178 Volume 2 also provides growth and decay timing. The pump HRR in NUREG/CR‑6850 is bounding compared with the updated values, and is therefore still valid. Oil fire split fractions: The methods panel decision letter (ML12171A583) updates the likelihood and oil spill sizes for general pump oil fires other than large hydraulic valves. Specifically: 88% of oil fires from pumps limit damage to the pump itself, 7% of oil fires from pumps produce oil pools of 10% capacity, and 5% of oil fires from pumps produce oil pools of 100% capacity. For large hydraulic valves (which are included in Bin 21), the oil spill size fractions recommended in NUREG/CR‑6850 Appendix E.3 should still be applied. Oil fire HRR: See Section G.4 of NUREG/CR‑6850 for HRR for flammable liquid fires. EPRI 3002005303, although not formally reviewed by the NRC, provides a method to more realistically characterize the HRR profile and duration for liquid spill fires.	EPRI 3002002936 / NUREG‑2169 NUREG‑2178 Volume 2 / EPRI 3002016052 Methods Panel Decision, ML12171A583 EPRI 1011989 / NUREG/CR‑6850 EPRI 3002005303
22	Plant-Wide Components	RPS MG Sets	The motor HRR is recommended for Bin 22 RPS MG Sets (refer to Table 11‑1 of NUREG/CR‑6850). See Bin 14.	EPRI 1011989 / NUREG/CR‑6850 See Bin 14
23a	Plant-Wide Components	Transformers (oil filled)	See Section G.4 of NUREG/CR‑6850 for HRR for flammable liquid fires. EPRI 3002005303, although not formally reviewed by the NRC, provides a method to more realistically characterize the HRR profile and duration for liquid spill fires.	EPRI 1011989 / NUREG/CR‑6850 EPRI 3002005303
23b	Plant-Wide Components	Transformers (dry)	Chapter 5 of NUREG‑2178 Volume 2 / EPRI 3002016052 provides updated HRR distributions for dry transformers (compared with the original distribution from NUREG‑6850) based on power rating, as well as growth and decay timing.	NUREG‑2178 Volume 2 / EPRI 3002016052
24	Plant-Wide Components	Transient fires caused by welding and cutting	See Bin 3 for treatment of transient fires.	See Bin 3
25	Plant-Wide Components	Transients	See Bin 3 for treatment of transient fires.	See Bin 3
26	Plant-Wide Components	Ventilation Subsystems	Ventilation subsystem fires are classified as either electrical (motor) or oil. The split fraction between electrical and oil fires is provided in NUREG/CR‑6850 (0.95 electrical / 0.05 oil). Electrical (motor) fires: HRR distributions and fire durations are provided in Chapter 5 of NUREG‑2178, Volume 2 / EPRI 3002016052. Oil fire split fractions: The oil spill size fractions recommended in NUREG/CR‑6850 Appendix E.3 should be applied. Oil fire HRR: See Section G.4 of NUREG/CR‑6850 for HRR for flammable liquid fires. EPRI 3002005303, although not formally reviewed by the NRC, provides a method to more realistically characterize the HRR profile and duration for liquid spill fires.	NUREG‑2178 Volume 2 / EPRI 3002016052 EPRI 1011989 / NUREG/CR‑6850 EPRI 3002005303
27	Transformer Yard	Transformer - Catastrophic	The catastrophic failure of a large transformer is defined as an energetic failure of the transformer that includes a rupture of the transformer tank, oil spill, and burning oil splattered a distance from the transformer. The analyst should use the frequency and 1.) determine availability of offsite power based on the function of the transformer(s) and 2.) consider propagation to adjacent (not nearby) buildings or components. A propagation path may be considered at the location of open or sealed penetrations, e.g., where a bus-duct enters from the Yard into the Turbine Building. Structural damage need only be considered only where appropriate shields are not present to protected structures and components against blast or debris.	EPRI 1011989 / NUREG/CR‑6850
28	Transformer Yard	Transformer - Non Catastrophic	In this failure, oil does not spill outside the transformer tank and the fire does not necessarily propagate beyond the fire source transformer. Analyst can use all the frequency and assume total loss of the "Transformer/ Switch Yard" or may split this frequency equally among the large transformers of the area and assume loss of each transformer separately. Loss of offsite power should be determined based on the function of the affected transformer(s).	EPRI 1011989 / NUREG/CR‑6850
29	Transformer Yard	Yard Transformers (Others)	In the screening phase of the project, the analyst may conservatively assign the same frequency to all of the items in this group. If the scenario would not screen out, the frequency may then be divided among the various items in this group. A relative ranking scheme may be used for this purpose. The ranking may be based on the relative characteristics of the item and the analysts' judgment.	EPRI 1011989 / NUREG/CR‑6850
30	Turbine Building	Boiler	See Section G.4 of NUREG/CR‑6850 for HRR for flammable liquid fires. EPRI 3002005303, although not formally reviewed by the NRC, provides a method to more realistically characterize the HRR profile and duration for liquid spill fires.	EPRI 1011989 / NUREG/CR‑6850 EPRI 3002005303
31	Turbine Building	Cable fires caused by welding and cutting	FAQ 13‑0005 provides additional guidance for detailed fire modeling on both self-ignited cable fires and cable fires caused by welding and cutting. This FAQ outlines a more realistic approach for addressing these types of fires in cable trays and suggests replacement text for Section R.1 of NUREG/CR‑6850. However, the current method of evaluating cable fire risk in NUREG/CR‑6850 remains an acceptable approach.	EPRI 1011989 / NUREG/CR‑6850 FAQ 13‑0005
32	Turbine Building	Main Feedwater Pumps	Main feedwater pumps are classified as either electrical (motor) or oil. The split fraction between electrical and oil fires is provided in NUREG/CR‑6850 (0.11 electrical / 0.89 oil). Electrical (motor) fires: HRR distributions and fire durations are provided in Chapter 5 of NUREG‑2178, Volume 2 / EPRI 3002016052. The pump HRR in NUREG/CR‑6850 is bounding compared with the updated values, and is therefore still valid. Oil fire split fractions: FAQ 08‑0044 (Section 9 of NUREG/CR‑6850 Supplement 1) clarifies the severity factors for small fires (0.966 for a leak that impacts the pump), large fires (0.0306 for 10% inventory spill), and very large fires (0.0034 for 100% inventory spill). Oil fire HRR: See Section G.4 of NUREG/CR‑6850 for HRR for flammable liquid fires. EPRI 3002005303, although not formally reviewed by the NRC, provides a method to more realistically characterize the HRR profile and duration for liquid spill fires.	NUREG‑2178 Volume 2 / EPRI 3002016052 EPRI 1011989 / NUREG/CR‑6850 FAQ 08‑0044, Section 9 of Supplement 1 EPRI 3002005303
33	Turbine Building	Turbine Generator Excitor	Appendix O (Section O.2.1 & Table O-2) recommends assuming the excitor fire is limited to the excitor itself.	EPRI 1011989 / NUREG/CR‑6850
34	Turbine Building	Turbine Generator Hydrogen	Appendix O (Section O.2.2 & Table O-2) provides guidance for both limited and severe T/G Hydrogen fires. Table O-2 also provides a conditional probability for a catastrophic T/G fire involving the hydrogen, oil and blade ejection.	EPRI 1011989 / NUREG/CR‑6850
35	Turbine Building	Turbine Generator Oil	Appendix O (Section O.2.3 & Table O-2) provides guidance for both limited and severe T/G oil fires. Table O-2 also provides a conditional probability for a catastrophic T/G fire involving the hydrogen, oil and blade ejection.	EPRI 1011989 / NUREG/CR‑6850
36	Turbine Building	Transient fires caused by welding and cutting	See Bin 3 for treatment of transient fires.	See Bin 3
37	Turbine Building	Transients	See Bin 3 for treatment of transient fires.	See Bin 3

Recommended HRR Values

The following tables summarize the latest research on HRR probability distributions. These distributions were developed to increase realism in modeling electrical cabinet fires and transient fires. As such, HRR probability distributions available in earlier publications (such as Appendix G of NUREG/CR-6850) are bounding. In the case of electric motors and transformers, the latest HRR probability distributions are based on equipment sizes so that the fires can also be realistically characterized.

Electrical Cabinets (NUREG‑2178 Volume 1)

NUREG‑2178 Volume 1 / EPRI 3002005578 provides HRR distributions for electrical enclosures.

Electrical Enclosures

Enclosure Class / Function Group	Enclosure Ventilation (Open or Closed Doors)	Fuel Type† (TS/QTP/SIS or TP Cables)	Gamma Distribution
			(a) Default				(b) Low Fuel Loading				(c) Very Low Fuel Loading
			α	β	P75 (kW)	P98 (kW)	α	β	P75 (kW)	P98 (kW)	α	β	P75 (kW)	P98 (kW)
1 - Switchgear and Load Centers	Closed	TS/QTP/SIS	0.32	79	30	170	Not Applicable				Not Applicable
	Closed	TP	0.99	44	60	170
2 - MCCs and Battery Chargers	Closed	TS/QTP/SIS	0.36	57	25	130	Not Applicable				Not Applicable
	Closed	TP	1.21	30	50	130
3 - Power Inverters	Closed	TS/QTP/SIS	0.23	111	25	200	Not Applicable				Not Applicable
	Closed	TP	0.52	73	50	200
4a - Large Enclosures     >1.42 m3 (>50 ft3)	Closed	TS/QTP/SIS	0.23	223	50	400	0.23	111	25	200	0.38	32	15	75
	Closed	TP	0.52	145	100	400	0.52	73	50	200	0.88	21	25	75
	Open	TS/QTP/SIS	0.26	365	100	700	0.26	182	50	350	0.38	32	15	75
	Open	TP	0.38	428	200	1000	0.38	214	100	500	0.88	21	25	75
4b - Medium Enclosures     ≤1.42 m3 (50 ft3) and     > 0.34 m3 (12 ft3)	Closed	TS/QTP/SIS	0.23	111	25	200	0.27	51	15	100	0.88	12	15	45
	Closed	TP	0.52	73	50	200	0.52	36	25	100	0.88	12	15	45
	Open	TS/QTP/SIS	0.23	182	40	325	0.19	92	15	150	0.88	12	15	45
	Open	TP	0.51	119	80	325	0.3	72	25	150	0.88	12	15	45
4c - Small Enclosures     ≤ 0.34 m3 (12 ft3)	Not Applicable	All	0.88	12	15	45	Not Applicable				Not Applicable
† Legend for Fuel Type: TS = Thermoset, TP = Thermoplastic, QTP = Qualified Thermoplastic, SIS = Synthetic Insulated Switchboard Wire or XLPE-Insulated Conductor

Motors and Dry Transformers (NUREG‑2178 Volume 2)

NUREG‑2178 Volume 2 / EPRI 3002016052 provides HRR distributions for motors and dry transformers.

Motors

Motor Classification Group	Motor Size (horsepower)	Gamma Distribution
		α	β	P75 (kW)	P98 (kW)
A	>5-30	1.34	3.26	6	15
B	>30-100	1.17	8.69	14	37
C	>100	1.10	24.19	37	100

Dry Transformers

Transformer Classification Group	Transformer Power (kVA)	Gamma Distribution
		α	β	P75 (kW)	P98 (kW)
A	>45-75	0.38	12.84	6	30
B	>75-750	0.41	28.57	15	70
C	>750	0.46	50.26	30	130

Transients (NUREG‑2233)

NUREG‑2233 / EPRI 3002018231 provides HRR distributions for both generic and "transient combustible control location" (TCCL) type transient fires. The report also provides values for total energy release (TER) and zones of influence (ZOIs), but only HRRs are included here.

Transients

Type	Gamma Distribution
	α	β	P75 (kW)	P98 (kW)
Generic	0.271	141	41.6	278
TCCL	0.314	67.3	24.6	143

Additional Fire Modeling Considerations

Time-to-Damage Models for Cables

Three approaches are documented for assessing the time-to-damage for cables.

Exposure threshold The method described in EPRI 1011989 / NUREG/CR‑6850 Appendix H consists of using the threshold exposure gas temperature or heat flux for determining cable failure. See below for damage criteria. This is the simplest of the approaches, but it can be fairly conservative because it does not account for the time it takes for cable heating to actually result in damage.

Heat soak The method described in Appendix A of NUREG‑2178 Volume 2 / EPRI 3002016052 considers exposure integrated over time based upon the time to failure data provided in Appendix H of NUREG/CR‑6850. This method is less conservative than the above "exposure threshold" method but still conservative when compared with THIEF. Time to failure data for Kerite-FR materials are provided in NUREG‑2232 / EPRI 3002015997.

Heat conduction (Thermally-Induced Electrical Failure, "THIEF") The THIEF approach presented in NUREG/CR‑6931 Volume 3 and NUREG‑1805 Supplement 1 performs a one-dimensional (1-D), cylindrical heat transfer calculation for a cable exposed to a time-varying exposure to determine when the cable jacket will fail based on the jacket’s inner temperature. Validation of the model shows that it does well at computing the temperature rise of the cable jacket; however, because it requires cable-specific data (dimensions and mass), it cannot be applied in a generic manner such as the exposure threshold or heat soak methods.

Location Factor

When the fire is located near a wall or in a corner, less air can be entrained into the fire plume. Less air entrainment into the fire plume produces higher plume temperatures. The flames from fires in contact with wall and corner surfaces tend to be longer, also resulting in higher plume temperatures. For such fires, a location factor, traditionally 2 for fires near a wall or 4 for fires near a corner, has been applied as a correction to the plume temperature calculation. NUREG‑2178 Volume 2 / EPRI 3002016052 Section 6 demonstrates that the traditional approach is overly conservative, and presents new factors based on the distance from the source to a corner or wall:

Configuration	Location Factor
	0–0.3 m [0–1 ft]	0.3–0.6 m [1–2 ft]	>0.6 m [2 ft]
Corner	4	2	1
Wall	1	1	1

EPRI 3002005303 provides the technical basis for the work in NUREG‑2178 Volume 2.

Radiation effects modeling

Chapter 2 of NUREG‑2178 Volume 2 / EPRI 3002016052 evaluates radiation emission models used to assess horizontal zone of influence. The two commonly-implemented empirical models – the point source method and the solid flame method – are compared against a computational model (Fire Dynamics Simulator). The results of this chapter recommended that the adjusted solid flame model should generally be considered a preferred method over the point source method because the adjusted flame model shows somewhat better characteristics in terms of a) NOT under-predicting and b) improved statistical error and bias. This applies to all fire types, where the flame is unobstructed. The modeling of obstructed radiation circumstances as present in electrical cabinets is discussed in the context of Bin 15 electrical cabinet fire modeling.

High Energy Arcing Fault (HEAF) Research

EPRI and the NRC are currently developing further methods and data on the risk impact of HEAF events; for example frequencies, fault duration, and zone of influence (e.g., copper versus aluminum). EPRI has issued the following white paper reports:

• EPRI 3002015992 provides an overview of nuclear power station electrical distribution systems and covers fault protection system concepts, fault isolation times, the potential impact of HEAFs on Class 1E electrical distribution systems, and typical industry practices and programs that help ensure proper operation. This report also provides some preliminary risk insights based on a review of existing data.
• EPRI 3002011922 reviews the operating experience to gain insights about equipment type, event characteristics, and the range of damage for HEAF events occurring at nuclear power plants within the United States and internationally. This paper also explores recent U.S. and international HEAF test programs for low- and medium-voltage electrical equipment and summarizes the insights gained from these test programs, including the potential role of aluminum oxidation in HEAF severity.
• EPRI 3002015459 demonstrates that an effective preventive maintenance program is important in minimizing the likelihood and/or severity of a HEAF event. Sixty‑four percent (64%) of HEAF events were determined to be preventable, and the most prevalent cause of failure was inadequate maintenance. These data demonstrate that proper maintenance can prevent most HEAF events. Effective maintenance practices and strategies are summarized in this report by equipment type, including circuit breakers, bus ducts, protective relays, and cables.

Fire Propagation and Suppression Guidance

Detection-Suppression Event Tree

For electrical cabinet fires, Section 5 of NUREG‑2230 / EPRI 3002016051 presents a revised detection-suppression event tree model for characterizing fire detection and suppression activities in response to a fire event (revised compared with the original model described in Appendix P of NUREG/CR‑6850 and Chapter 14 of NUREG/CR‑6850 Supplement 1). This modification is intended to capture the potential for plant personnel suppression during the early stages of an electrical cabinet fire. For other fire types, the original model described in Appendix P of NUREG/CR‑6850 and Chapter 14 of NUREG/CR‑6850 Supplement 1 should be used.

Fire Damage Criteria

Cable Damage Criteria

FAQ 16‑0011 provides radiant heating and temperature criteria for bulk cable tray ignition (which was not previously provided in NUREG/CR‑6850). The bounding cable damage and ignition criteria remain the same. A summary of the results are shown below. The analyst should refer to both NUREG/CR‑6850 Appendix H and FAQ 16‑0011 for full guidance.

	Bounding Cable Damage / Ignition Criteria		Bulk Cable / Tray Ignition Criteria
Cable Type	Radiant Heating	Temperature	Radiant Heating	Temperature
Thermoplastic	6 kW/m2	205°C	25 kW/m2	500°C
Thermoset	11 kW/m2	330°C

For Kerite cables, refer to NUREG/CR‑7102 for damage criteria. Originally FAQ 08‑0053 was initiated to clarify failure thresholds for Kerite cables and the resolution can be found in the closure memo dated June 6, 2012 following the publication of NUREG/CR‑7102.

Treatment of Sensitive Electronics

FAQ 13‑0004 provides supplemental guidance for the application of the lower damage thresholds provided in NUREG/CR‑6850 Section 8.5.1.2 and H.2 for solid-state components. Fire Dynamics Simulator (FDS) modeling results support the recommendation that a generic screening heat flux damage threshold for thermoset cables, as observed on the outer surface of the cabinet, can be used as a conservative surrogate for assessing the potential for thermal damage to solid-state and sensitive electronics within an electrical panel (cabinet). Since the conclusions of the FDS analysis are based on heat flux exposure to the cabinet, the 65°C temperature damage criterion must still be assessed for other types of fire exposures to the enclosed sensitive electronics.

Cable Tray Fire Propagation

FAQ 08‑0049, Section 11 of Supplement 1 clarifies the limits of the empirical cable tray fire propagation model in EPRI 1011989, NUREG/CR‑6850. The model can lead to conservative estimates of cable fire growth rates and unrealistically short room burnout times when used outside the ZOI (i.e., outside the fire plume that extends above the ignition source).

NUREG/CR‑7010 documents the results of experiments to better understand and quantify the burning characteristics of grouped electrical cables commonly found in nuclear power plants. Volume 1 studies horizontal cable trays and Volume 2 studies vertical shafts and corridors. The experiments in Volume 1 address horizontal, ladder-back trays filled with unshielded cables in open configurations. The results of the full-scale experiments have been used to validate a simple model called FLASH‑CAT (Flame Spread over Horizontal Cable Trays). The document also provides verification and validation material for the FLASH‑CAT model. Volume 2 performed experiments on vertical cable tray configurations and enclosure effects. Volume 2 also extends the FLASH‑CAT model to address cable trays within enclosures and vertical tray configurations.

Manual Non-Suppression Probability Estimates

Various reports have documented updates to the manual non-suppression probability data. The latest updates for each event type are summarized below.

Probability Distribution for Rate of Fires Suppressed Per Unit Time, λ

Suppression Curve	Number of Events in Curve	Total Duration (minutes)	Rate of Fire Suppressed (λ)				Calculation Source Document
			Mean	P5	P50	P95
Turbine-generator fires	30	1167	0.026	0.019	0.025	0.034	NUREG‑2169
Control room	10	26	0.385	0.209	0.372	0.604	NUREG‑2178 Volume 2
Pressurized water reactor containment (at power)	3	40	0.075	0.020	0.067	0.157	NUREG‑2169
Containment (low power-shutdown)	31	299	0.104	0.075	0.103	0.136	NUREG‑2169
Outdoor transformers	24	928	0.026	0.018	0.026	0.035	NUREG‑2169
Flammable gas	8	234	0.034	0.017	0.033	0.056	NUREG‑2169
Oil fires	50	562	0.089	0.069	0.088	0.111	NUREG‑2169
Cable fires	4	29	0.138	0.047	0.127	0.267	NUREG‑2169
Electrical fires ‡	74	653	0.113	0.093	0.113	0.136	NUREG‑2230
Interruptible fires (Bin 15)	43	288	0.149	0.114	0.148	0.189	NUREG‑2230
Growing fires (Bin 15)	18	179.5	0.100	0.065	0.098	0.142	NUREG‑2230
Welding fires	52	484	0.107	0.084	0.107	0.133	NUREG‑2169
Transient fires	43	386	0.111	0.085	0.111	0.141	NUREG‑2169
HEAFs	11	385	0.029	0.016	0.028^	0.044	FAQ 17‑0013
All fires	401	5661	0.071	0.065	0.071	0.077	NUREG‑2230
‡ Electrical fires include non-cabinet electrical sources, such as electrical motors, indoor transformers, and junction boxes, among other electrical equipment. ^ FAQ 17-0013 reported the 50th percentile as 0.029. When calculated using the chi-squared distribution the calculated 50th percentile 0.028 as shown in NUREG‑2230.

Incipient Detection

NUREG‑2180 NRC guidance on crediting incipient detection systems in fire PRA is discussed in NUREG‑2180. The issuance of NUREG‑2180 retires FAQ 08‑0046 (Chapter 13 of NUREG/CR‑6850 Supplement 1) as documented in the July 1, 2016 letter to NEI.