Skip to content
From Fire Safety Academy Wiki
Contents
  1. 1Why fire is different in nuclear facilities
  2. 2Defence in depth and regulation
  3. 3Fire hazard analysis and deterministic protection
  4. 4Cables and the Browns Ferry fire
  5. 5Fire risk assessment and fire modelling
  6. 6Detection, suppression and their reliability
    1. 6.1Worked example
  7. 7Research reactors, radiation machines and radioactive materials
  8. 8Limits and uncertainty
  9. 9See also
  10. 10Notes
  11. 11References

Nuclear facility fire protection is the fire protection of facilities that use or store nuclear materials, such as nuclear power reactors, research reactors, radiation machines and laboratories handling radioactive material. The general fire hazards in these facilities are the same as in similar non-nuclear buildings, but the consequences can be more serious. A fire may expose workers, responders and the public to radiation and contamination, cause an unintended nuclear chain reaction, or damage the systems that contain nuclear materials.[1] At a nuclear power plant the central concern is that a fire could defeat the ability to cool the reactor.[2]

Protection follows the principle of defence in depth: prevent fires, detect and extinguish them quickly, and limit the spread of fires that are not extinguished.[3] In the United States, the 1975 Browns Ferry cable fire led to the federal fire protection rules for nuclear power plants.[4] In 2004 the US Nuclear Regulatory Commission (NRC) accepted NFPA 805, a performance-based standard, as a voluntary alternative to those rules.[5] Performance-based protection often relies on fire risk analysis and validated fire models, which the NRC and the Electric Power Research Institute (EPRI) developed in NUREG/CR-6850 and NUREG-1824.[6]

Why fire is different in nuclear facilities

A nuclear reactor initiates and maintains a controlled chain reaction in a fissionable fuel. Large power reactors produce up to 3500 MW of heat or 1100 MW of electricity. Research reactors range from a few watts to many megawatts, and some occupy only a corner of a laboratory room.[3]

In 2005, 443 power reactors operated in 31 countries, 339 of them outside the United States. Nuclear power supplied less than 20 % of US electricity but 78 % in France.[1] The industry has a good safety record by any measure of acute incidents. Yet the costliest US fire involving the electric power industry occurred at an Alabama nuclear plant, and the Chernobyl accident shaped public perceptions of safety.[3]

Reactor safety rests on three objectives: shutting the reactor down safely and keeping it shut down, removing residual heat from the core, and limiting radioactive releases. Reactors meet these objectives with redundant safety systems, so that an initiating event such as a fire does not prevent the safety functions. Earlier designs have less redundancy, and then each system needs more protection from fire through passive barriers, separation, detection and suppression.[3] Control systems are vital, so the control room, cableways, emergency power supply and operated equipment need protection of prime importance.[2]

Fire risk studies therefore focus on fire damage to electrical equipment, especially cables. The large pumps, motors and valves of safety systems are relatively resistant to the effects of a distant fire. In some areas of some plants, however, the cables of redundant equipment can be vulnerable to a single fire.[7] Cable insulation and jacketing are the largest single fuel source in a nuclear power plant.[8] Beyond nuclear safety, these plants have the hazards common to all generating plants, such as large volumes of lubricating oil and oil-insulated transformers. Those hazards are treated in power station fire protection.[5]

Defence in depth and regulation

Defence in depth means that fire safety does not rely on a single fire protection subsystem, but on a combination in case one fails. The concept is widely used in building fire safety and particularly in the nuclear power industry.[9] Its first objective, preventing fires, requires a plant design and operation that minimise the probability of fire. The second relies on active protection to detect and extinguish fires early. The third relies on passive fire barriers and physical separation as the last line of defence. All three must be implemented at the same time.[3]

Three nested layers labelled prevent fires, detect and extinguish, and limit spread, around a protected safe shutdown function
Defence in depth for fire protection at a nuclear power plant: three levels that must all be implemented at the same time.

In the United States these objectives are set out in Section II of Appendix R to 10 CFR Part 50. The fire protection programme should prevent fires from starting; rapidly detect, control and extinguish fires; and protect structures, systems and components important to safety. A fire that is not promptly extinguished must not prevent safe shutdown of the plant.[10] The regulatory framework also includes General Design Criterion 3, 10 CFR 50.48, Regulatory Guide 1.189 and the NRC's standard review plan.[11] After the Browns Ferry fire, the NRC inspected operating plants and issued 10 CFR 50.48 and Appendix R, which set minimum requirements for fire protection programmes and post-fire safe shutdown.[4]

NFPA has issued nuclear fire protection standards since 1955, when NFPA 801 for facilities handling radioactive materials appeared. NFPA 802 for research and production reactors followed in 1958, and NFPA 803 for light water power plants in 1978; both were no longer current in 2008. NFPA 805, the performance-based standard that replaced NFPA 803, was the first NFPA standard to address nuclear safety. Earlier standards covered only property protection, business interruption and life safety.[2]

NFPA 805 sets performance goals for keeping the fuel in a safe and stable condition, preventing radioactive releases, life safety and limiting plant damage. Minimum requirements, such as fire protection programmes and compliance with NFPA standards for credited systems, remain prescriptive. In June 2004 the NRC revised 10 CFR 50.48 and endorsed NFPA 805 (2001 edition) as a voluntary licensing alternative.[5]

Fire hazard analysis and deterministic protection

A fire hazard analysis (FHA) assesses whether a nuclear power plant can still shut down safely and limit radioactive releases when a fire occurs. Under Regulatory Guide 1.189 it considers fixed and transient fire hazards and the consequences of a fire at any location. It then specifies measures for fire prevention, detection, suppression and containment, plus alternative shutdown capability where needed.[11] A fire scenario is the chain of events from ignition of a specific fuel package to either successful shutdown or core damage. Along the way the fire grows, may be detected and suppressed, and may damage plant equipment, usually electrical cables.[12]

NFPA 805 offers a deterministic and a performance-based route to show that nuclear safety goals are met. In the deterministic route, if the success paths needed for safe shutdown lie in the same fire area outside containment, one of three schemes is required. These are a 3-hour encapsulation of one success path; a 1-hour encapsulation with suppression and detection; or 6.1 m (20 ft) of separation without intervening combustibles, with suppression and detection throughout the area. Inside a containment that is not inerted, the options are a radiant energy shield, 6.1 m of separation without intervening combustibles, or suppression and detection throughout the area.[13] The performance-based route identifies the equipment to be protected, sets damage thresholds and fire scenarios, and checks whether the required success paths survive. This can be done by fire modelling or by risk evaluation.[5][13]

Flowchart: success paths in one fire area, inside or outside containment, inerted or not, leading to encapsulation, separation or suppression
Deterministic route of NFPA 805 (2006) for protecting redundant safe shutdown success paths located in the same fire area.

Cables and the Browns Ferry fire

Cable fires in grouped cables were recognised early as a hazard. A 1966 NFPA study described 24 cable fires at US nuclear plants and pointed out that grouped cables spread flame much faster than single cables.[4] On 22 March 1975, workers at the Browns Ferry plant in northern Alabama used a candle to check for air leaks at cable penetrations. The flame ignited polyurethane foam used to seal the openings, and fire spread along the cables, shorting out backup safety systems as the insulation burned off.[14] More than 600 of the burned cables carried circuits for the safe shutdown of one or both reactors.[15] The handbook puts the damage at $302.93 million in 2007 dollars, with a 1½-year outage of two 1065 MW units and a worldwide reappraisal of nuclear plant fire protection.[16]

The event also showed failures of detection and suppression. About 15 minutes passed before an alarm was given, partly because of confusion over the telephone number. In the reactor building the fire was fought unsuccessfully for several hours with portable carbon dioxide and dry chemical extinguishers, but water put it out within minutes. The carbon dioxide system in the cable spreading room drove smoke and toxic fumes into the control room through unsealed cable openings. The smoke detectors in the cable spreading room did not alarm, probably because the airflow drew the smoke away from them.[15]

The lessons changed practice. NRC research programmes on firestops for nuclear plants led to new firestop products.[17] Electrical cables that are difficult to ignite are available and used in nuclear plants.[18] Cables qualified to IEEE Standard 383 are more fire-resistant, but they are also more difficult to extinguish once they ignite.[19]

In NUREG/CR-6850, thermoplastic cables are assumed to fail at a radiant heat flux of 6 kW/m² or a temperature of 205 °C, and thermoset cables at 11 kW/m² or 330 °C.[20] Fire in a stack of horizontal trays is assumed to spread upwards at 35 degrees to either side. The second tray ignites 4 minutes after the first, the third 3 minutes later, the fourth 2 minutes later and the fifth 1 minute later.[21] A 2006 errata sheet limits this timing model to cable trays separated in line with NRC Regulatory Guide 1.75.[22]

Side view of five stacked cable trays above a burning cabinet, with fire spreading upwards at 35 degrees and ignition times of 4, 3, 2 and 1 minutes
Fire spread model for a stack of horizontal cable trays in NUREG/CR-6850: spread at 35 degrees to either side and successive tray ignition times.

Fire risk assessment and fire modelling

In the nuclear power industry, probabilistic risk assessment has been the basis for safety regulation. It examines the risk of radioactive release, from a leak of contaminated water to a core meltdown.[23] One application of fire risk analysis is to determine the probability of reactor core melt due to fire.[24] A fire PRA of a nuclear plant identifies important fire-initiated accident scenarios, determines their consequences and assesses their likelihood.[25] This nuclear form of fire risk assessment uses the core damage frequency (CDF), the expected frequency of severe damage to the reactor fuel, as its main measure.[7] The NUREG/CR-6850 method covers core damage and large early release at power, but not shutdown modes, spent fuel pool accidents or sabotage.[26]

The results show that fire can matter a great deal. A quick 1975 assessment prompted by the Browns Ferry fire estimated fire-induced CDF at about 10⁻⁵ per year, about 20 % of the CDF from the causes covered in WASH-1400. Later studies gave about 5 × 10⁻⁶ per year at Zion, about 10 % of its total, and about 2 × 10⁻⁴ per year at Indian Point 2, about 40 %. Many studies found fire-induced CDFs of 10⁻⁴ per year or more, or a fire share of 20 % or more.[27] In the Individual Plant Examinations of External Events that the NRC required from 1991, more than half of the submittals identified cost-effective improvements.[28]

Bar chart on a log scale of fire-induced core damage frequency: Zion 5e-6, Browns Ferry estimate 1e-5, Indian Point 2 2e-4 per year
Fire-induced core damage frequency from early US nuclear plant fire risk studies, with the share of the total core damage frequency.

Generic fire frequencies in NUREG/CR-6850 include 4.5 × 10⁻² per reactor-year for electrical cabinets, 2.1 × 10⁻² for diesel generators and 2.5 × 10⁻³ for the main control board.[29] Weighting factors share these plant-wide values among the individual locations and ignition sources of a plant.[30] For a fire in a cabinet, the method multiplies the generic frequency by the factors that turn an ignition into core damage:[30] CDFi=λgWLWisSFPnsCCDPi Here λg is the generic fire frequency, WL and Wis are weighting factors for the location and the ignition source, and SF is the severity factor. Pns is the probability that the fire is not suppressed before the targets are damaged. CCDPi is the conditional core damage probability, given the damage in scenario i.[30]

Fire models supply times to damage and detection. The NRC's Fire Dynamics Tools are spreadsheets for first-order evaluations of fire hazards in commercial nuclear power plants.[31] NFPA 805 requires that fire models be verified and validated and acceptable to the authority having jurisdiction. NUREG-1824 therefore tested five models: the Fire Dynamics Tools, FIVE-Rev1, CFAST, MAGIC and the Fire Dynamics Simulator. The predictions showed both under-prediction and over-prediction, and the results were incomplete because of model limits and a lack of suitable experiments.[6] Most earlier validation studies used compartments of only about 9.3 m² with well-defined fires.[32]

Detection, suppression and their reliability

Automatic sprinklers or specially designed water spray systems are the first choice wherever fires may occur in reactor plants and facilities handling radioactive materials. Sprinklers keep working under radiation or contamination levels that would keep fire fighters out.[33] Water brings its own problems: it can spread contamination, so contaminated fire water must be confined and the radioactive waste systems sized to process it. Sprinklers help because they discharge limited water, and only where a fire burns.[34]

Where fissile material is present, criticality specialists may object to water because it moderates neutrons. Yet in many, if not most, such cases water spray or sprinklers can and should be provided, because criticality analyses can include their effect.[34] Liquid metal coolants react with water, which may justify excluding water from an area.[34]

The reliability of fire protection systems is the ability of a system to work under the stated conditions, including fire-induced stresses, for the stated time.[35] For US nuclear plants, NUREG/CR-6850 cites unreliability values from an industry study (NSAC-179L) of 0.04 for carbon dioxide systems, 0.05 for halon, 0.02 for wet pipe sprinklers and 0.05 for deluge or preaction systems. Smoke detection is assumed to fail with a probability of no more than 0.05.[36] Among more than 1400 US nuclear-plant fires documented by the 2005 NUREG method, none caused the widespread physical damage defined in that study, which the authors attributed to defence in depth. Fires at non-nuclear sites and at nuclear plants outside the United States had caused such damage.[37]

Worked example

NUREG/CR-6850 calculates the non-suppression probability for a fire in an emergency switchgear room with smoke detection and a manually activated gaseous system. A fire in a motor control centre threatens a cable tray, and fire modelling gives detection after 1 minute and target damage after 15 minutes. The fire brigade needs 7 minutes, and operators need 10 minutes to activate the gaseous system.[38] The time available for manual suppression is the time to damage minus the brigade response time and the detection time:[39] tms=tdam−tfb−tdet=15−7−1=7 min With the suppression rate of 0.12 per minute for electrical fires, the fire brigade fails within this time with a probability of e−0.12 × 7 = 0.43. Automatic detection fails with a probability of 0.05. Failure to activate the gaseous system in time has a probability of 0.15: 0.10 for human error plus 0.05 for system unreliability.[38] The two failure sequences are combined: Pns=0.95×0.15×0.43+0.05×0.15×1.0≈0.061+0.008=0.069 In the second term automatic detection has failed, and delayed detection after 15 minutes leaves the brigade no time, so its failure probability is 1.0. Under these example assumptions, about 7 % of the modelled fires would not be suppressed before the cable tray is damaged; the credited protection features lower the modelled scenario frequency by a factor of about 14.[38] A 2010 supplement revised the electrical suppression rate to 0.102 per minute. It also found that subtracting the full brigade response time can be conservative, because fire suppression often starts before the brigade arrives.[40]

Event tree with automatic detection, manual gaseous system and fire brigade branches, giving non-suppression probabilities 0.061 and 0.008, total 0.069
Detection and suppression event tree for the worked example of a motor control centre fire in an emergency switchgear room.
Symbols used in the equations
Symbol Description Unit
CDFi core damage frequency of fire scenario i 1/year
λg generic fire ignition frequency 1/year
WL, Wis location and ignition source weighting factors –
SF severity factor –
Pns non-suppression probability before target damage –
CCDPi conditional core damage probability of scenario i –
tms time available for manual suppression min
tdam time to target damage min
tfb fire brigade response time min
tdet time to detection min

Research reactors, radiation machines and radioactive materials

Research and production reactors often stand in existing buildings, sometimes of combustible construction. They may contain much combustible material such as paraffin shielding, boronated polyethylene, wall finishes and instrumentation. Combustibles should be eliminated where possible, for example by using water instead of paraffin, and automatic sprinklers are effective where they remain.[5] These reactors often lack a containment structure, so ventilation and confinement systems must be protected from fire and from airflow reversals caused by fire pressures.[13]

Particle accelerators use boronated polyethylene shielding, which has a very high heat release rate in fires, even in fire-retardant grades.[41] A 1967–1969 Swiss survey reproduced in the International Fire Engineering Guidelines (2005) lists 2100 MJ/m² for nuclear research occupancies, more than three times the 600 MJ/m² listed for a power station. These historical survey values are guides, not current design values for a specific facility.[42]

NFPA 801 applies to locations other than power plants that store, handle or use radioactive materials in quantities that require a government licence.[41] Such materials are handled on benches, in hoods, in glove boxes and in heavily shielded hot cells. A fire in a glove box can overpressurise the enclosure, and the extinguishing agent must neither overpressurise it nor increase the risk of criticality.[43]

Self-contained breathing apparatus should be used whenever exposure to airborne radioactive material is possible. The 2008 handbook cites US National Council on Radiation Protection and Measurements recommendations for a single emergency. The planned whole-body dose is at most 100 rem (1 Sv) for life saving and 25 rem (0.25 Sv) for protecting facilities or controlling fires. These are historical recommendations, not current operational limits.[44]

Emergency planning cannot rely on simply calling the public fire department. An in-plant fire protection force is recommended, and nuclear reactors need 24-hour routines for fires and emergencies. Fire fighting must be planned in advance with the local fire department, including measures to stop the spread of contamination. A reactor site needs a generous water supply and prearranged means to store or dispose of contaminated water.[33]

Limits and uncertainty

In 2002, fire PRA was considered less mature than the analysis of many other accident initiators. Its results are therefore used to support risk-informed rather than purely risk-based decisions.[28] The fire models validated in NUREG-1824 cover only some of the scenarios found in nuclear plants.[6] Most data and rules described here are American: US reactor experience, NRC regulations and NFPA standards in the editions summarised by the 2008 Fire Protection Handbook. At that time NFPA 806 for advanced reactors was still a proposal.[5] Other countries apply their own requirements, and later editions of these documents may differ.

See also

Notes

  1. ↑ 1.0 1.1 Holmes (2008a), p. 9-191.
  2. ↑ 2.0 2.1 2.2 Holmes (2008a), p. 9-193.
  3. ↑ 3.0 3.1 3.2 3.3 3.4 Holmes (2008a), p. 9-192.
  4. ↑ 4.0 4.1 4.2 Iqbal and Salley (2004), p. 7-1.
  5. ↑ 5.0 5.1 5.2 5.3 5.4 5.5 Holmes (2008a), p. 9-194.
  6. ↑ 6.0 6.1 6.2 Najafi et al. (2007), pp. xv–xvi.
  7. ↑ 7.0 7.1 Siu (2002), p. 5-215.
  8. ↑ Iqbal and Salley (2004), p. 1-7.
  9. ↑ Mowrer and Rosenbaum (2008), p. 3-190.
  10. ↑ Iqbal and Salley (2004), pp. 1-1–1-2.
  11. ↑ 11.0 11.1 Iqbal and Salley (2004), p. 1-3.
  12. ↑ Iqbal and Salley (2004), p. 1-4.
  13. ↑ 13.0 13.1 13.2 Holmes (2008a), p. 9-195.
  14. ↑ Kampmeyer (2008), p. 18-73.
  15. ↑ 15.0 15.1 Iqbal and Salley (2004), pp. H-8–H-9.
  16. ↑ Sheridan (2008), p. 9-187.
  17. ↑ Davoodi (2008), p. 18-19.
  18. ↑ Davoodi (2008), p. 18-11.
  19. ↑ Iqbal and Salley (2004), p. 1-2.
  20. ↑ Nowlen et al. (2005b), p. H-1.
  21. ↑ Nowlen et al. (2005b), pp. R-9–R-10.
  22. ↑ NUREG/CR-6850, EPRI TR 1011989 Errata Sheet (2006), p. 2.
  23. ↑ Hurley and Bukowski (2008), p. 3-121.
  24. ↑ Hall and Watts (2008), p. 3-140.
  25. ↑ Siu (2002), p. 5-214.
  26. ↑ Nowlen et al. (2005a), p. 2-1.
  27. ↑ Siu (2002), p. 5-222.
  28. ↑ 28.0 28.1 Siu (2002), p. 5-223.
  29. ↑ Nowlen et al. (2005b), pp. 6-3–6-4.
  30. ↑ 30.0 30.1 30.2 Nowlen et al. (2005b), pp. M-16–M-17.
  31. ↑ Beyler et al. (2008), p. 3-101.
  32. ↑ Beyler et al. (2008), pp. 3-106–3-107.
  33. ↑ 33.0 33.1 Holmes (2008a), p. 9-199.
  34. ↑ 34.0 34.1 34.2 Holmes (2008a), p. 9-200.
  35. ↑ Modarres and Joglar-Billoch (2002), p. 5-24.
  36. ↑ Nowlen et al. (2005b), p. P-6.
  37. ↑ Nowlen et al. (2005b), p. P-1.
  38. ↑ 38.0 38.1 38.2 Nowlen et al. (2005b), pp. P-14–P-15.
  39. ↑ Nowlen et al. (2005b), p. P-9.
  40. ↑ Hyslop et al. (2010), pp. 14-8, 14-11.
  41. ↑ 41.0 41.1 Holmes (2008a), p. 9-196.
  42. ↑ National Research Council of Canada et al. (2005), pp. 3.4-2, 3.4-6.
  43. ↑ Holmes (2008a), p. 9-197.
  44. ↑ Holmes (2008a), p. 9-198.

References

  • Beyler, C. L., DiNenno, P. J., Carpenter, D. J., & Watts, J. M., Jr. (2008). Introduction to fire modeling. In A. E. Cote, C. C. Grant, J. R. Hall Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 3-93–3-110). National Fire Protection Association.
  • Davoodi, H. (2008). Confinement of fire in buildings. In A. E. Cote, C. C. Grant, J. R. Hall Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 18-3–18-21). National Fire Protection Association.
  • Hall, J. R., Jr., & Watts, J. M., Jr. (2008). Fire risk analysis. In A. E. Cote, C. C. Grant, J. R. Hall Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 3-135–3-143). National Fire Protection Association.
  • Holmes, W. D. (2008a). Nuclear facilities. In A. E. Cote, C. C. Grant, J. R. Hall Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 9-191–9-201). National Fire Protection Association.
  • Hurley, M. J., & Bukowski, R. W. (2008). Fire hazard analysis techniques. In A. E. Cote, C. C. Grant, J. R. Hall Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 3-121–3-134). National Fire Protection Association.
  • Hyslop, J. S., Kratchman, J., Siu, N., Afshar, C., Najafi, B., Joglar, F., Li, Y., Henneke, D., & Warner, M. (2010). Fire probabilistic risk assessment methods enhancements: Supplement 1 to NUREG/CR-6850 and EPRI 1011989 (NUREG/CR-6850, Supplement 1; EPRI 1019259). Electric Power Research Institute; U.S. Nuclear Regulatory Commission, Office of Nuclear Regulatory Research.
  • Iqbal, N., & Salley, M. H. (2004). Fire Dynamics Tools (FDTs): Quantitative fire hazard analysis methods for the U.S. Nuclear Regulatory Commission Fire Protection Inspection Program (NUREG-1805). U.S. Nuclear Regulatory Commission, Office of Nuclear Reactor Regulation.
  • Kampmeyer, J. E. (2008). Penetration sealing. In A. E. Cote, C. C. Grant, J. R. Hall Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 18-73–18-78). National Fire Protection Association.
  • Modarres, M., & Joglar-Billoch, F. (2002). Reliability. In P. J. DiNenno, D. Drysdale, C. L. Beyler, W. D. Walton, R. L. P. Custer, J. R. Hall Jr., & J. M. Watts Jr. (Eds.), SFPE handbook of fire protection engineering (3rd ed., pp. 5-24–5-39). National Fire Protection Association; Society of Fire Protection Engineers.
  • Mowrer, F. W., & Rosenbaum, E. R. (2008). Overview of performance-based fire protection design. In A. E. Cote, C. C. Grant, J. R. Hall Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 3-183–3-194). National Fire Protection Association.
  • Najafi, B., Joglar, F., & Dreisbach, J. (2007). Verification and validation of selected fire models for nuclear power plant applications: Vol. 1. Main report (NUREG-1824; EPRI 1011999). U.S. Nuclear Regulatory Commission, Office of Nuclear Regulatory Research; Electric Power Research Institute.
  • National Research Council of Canada, International Code Council, Department of Building and Housing, & Australian Building Codes Board. (2005). International fire engineering guidelines (2005 ed.). Australian Building Codes Board.
  • Nowlen, S. P., Najafi, B., Wyant, F. J., Forester, J., Kazarians, M., Kolaczkowski, A., Joglar, F., Funk, D., Hannaman, G., & Anoba, R. (2005a). EPRI/NRC-RES fire PRA methodology for nuclear power facilities: Vol. 1. Summary and overview (NUREG/CR-6850; EPRI 1011989). Electric Power Research Institute; U.S. Nuclear Regulatory Commission, Office of Nuclear Regulatory Research.
  • Nowlen, S. P., Najafi, B., Wyant, F. J., Forester, J., Kazarians, M., Kolaczkowski, A., Joglar, F., Funk, D., Hannaman, G., & Anoba, R. (2005b). EPRI/NRC-RES fire PRA methodology for nuclear power facilities: Vol. 2. Detailed methodology (NUREG/CR-6850; EPRI 1011989). Electric Power Research Institute; U.S. Nuclear Regulatory Commission, Office of Nuclear Regulatory Research.
  • NUREG/CR-6850, EPRI TR 1011989 errata sheet. (2006).
  • Sheridan, D. J. (2008). Electric generating plants. In A. E. Cote, C. C. Grant, J. R. Hall Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 9-179–9-189). National Fire Protection Association.
  • Siu, N. (2002). Fire risk assessment for nuclear power plants. In P. J. DiNenno, D. Drysdale, C. L. Beyler, W. D. Walton, R. L. P. Custer, J. R. Hall Jr., & J. M. Watts Jr. (Eds.), SFPE handbook of fire protection engineering (3rd ed., pp. 5-214–5-226). National Fire Protection Association; Society of Fire Protection Engineers.