Skip to content
From Fire Safety Academy Wiki
Revision as of 03:08, 8 November 2026 by Lisa (talk | contribs) (FSE wiki: fact-check corrections: scope equivalence and FSES, clarify risk units and matrix bins (written by Claude Opus 5.5))
Contents
  1. 1Definition and terminology
  2. 2Why zero risk is not a goal
  3. 3Value judgements and risk perception
  4. 4How acceptable levels are set
  5. 5Risk measures and acceptance criteria
  6. 6Acceptable risk in decision making
  7. 7Limitations and criticism
  8. 8See also
  9. 9Notes
  10. 10References

Acceptable risk is the level of fire risk that is judged tolerable once fire safety measures are in place. It is a value judgement, made on a chosen scale of risk, about how much probability of death, injury or loss may remain.[1] The related term tolerable risk expresses that absolute protection is not possible and that some possibility of damage, injury or loss is accepted in structures.[2] Zero risk is not an achievable goal. No universally acceptable level of fire risk exists either, because acceptability depends on the problem and on who is judging.[1]

In prescriptive codes the acceptable level is usually implicit: it is the risk that remains once the code is complied with.[3] Performance-based design turns it into an acceptance criterion, either by comparison with a code-compliant reference building or by an absolute value such as a maximum expected risk to life.[4][5] In the Fire Protection Handbook of the National Fire Protection Association (NFPA, 2008), Hall and Watts describe the acceptable risk approach as one that treats risk as a constraint. Options that satisfy it are then compared on cost alone. They also point out its weaknesses.[6]

Definition and terminology

Watts and Hall define acceptable risk in the Handbook of Fire Protection Engineering of the Society of Fire Protection Engineers (SFPE, third edition, 2002). They describe it as the result of a value judgement applied to a particular scale for the measurement of risk.[1] The definition therefore needs two prior decisions. The first is the scale and method used to estimate risk; the second is whose views on acceptability count. For a practising engineer, acceptable risk is likely to be the risk acceptable to the client. For an authority having jurisdiction (AHJ), it is meant to be acceptable to society in general or to a particular community.[1]

Risk itself is the potential for unwanted, adverse consequences to life, health, property or the environment. It is usually estimated as the probability of an event multiplied by its consequence.[7] Hall and Watts (2008) note that fire risk analysis can be separated into risk estimation, which quantifies severity, probability and their uncertainties, and risk evaluation, which decides how important a value of risk is.[8] In the fire risk assessment process, the calculated risk is compared with selected criteria to determine whether it is acceptable.[9] These criteria come from goals and objectives that are value judgements, not purely scientific constructs.[8] In an acceptable risk approach, a certain level of risk is defined as acceptable, and all alternatives that meet it are evaluated strictly on cost.[6]

Regulations use the same idea. The International Fire Engineering Guidelines (IFEG, 2005) explain that the 2005 National Building Code of Canada words its objectives as "limit the probability" and refers to "unacceptable risk".[3] In that code, acceptable risk is the risk remaining after the code is complied with.[3] In New Zealand, Building Code objective C1(a), restated in Verification Method C/VM2 (second edition, 2025), is to safeguard people from an unacceptable risk of injury or illness caused by fire.[10]

Why zero risk is not a goal

Watts and Hall (2002) call it an axiom of fire risk analysis that zero risk is not an achievable goal.[1] No technology is 100 per cent reliable or immune to misuse, and lightning, wildland fires, earthquakes and wind storms can also cause fire loss. Reducing fire risk may even increase other risks, as with the potential damage to the ozone layer from halon. The goal that comes closest to zero risk is risk as low as is technically possible. In practice this may mean accepting the residual risk after all identified risk reduction strategies have been adopted.[1]

Puchovsky and Hurley (2008) state that a risk-free environment cannot be achieved under either prescriptive or performance-based design.[11] Extra resources eventually run up against the law of diminishing returns, and stricter requirements cost more to meet. They describe building regulation as a balancing act between the level of safety that society requires and the cost of providing it.[11]

Guidance documents for engineers state the same principle. The IFEG (2005) turn it into a rule for analysis: acceptance criteria must be realistic, and zero risk is not an appropriate criterion.[12] The handbook of the Australian Building Codes Board (ABCB) for the Fire Safety Verification Method (FSVM, 2020) adds that removing fire risk completely is impractical. As that target is approached, the fire safety strategy tends to conflict with the function and use of the building.[13] For communities, Granito (2008) writes that a certain level of fire loss must be accepted as tolerable because resources are limited.[14]

Value judgements and risk perception

Watts and Hall (2002) describe the identification of an acceptable level of risk as more a philosophical task than a technical one.[15] They note a contradiction in public behaviour: little attention is paid to the fire problem in general, yet every serious fire brings a clamour for change. Accepting a level of risk requires a value judgement, and people make different value judgements.[15]

Watts and Hall illustrate the role of values with four views of residential sprinklers. Experts largely agree that residential sprinklers can significantly reduce calculated fire risk. At the time of writing (2002), however, the average citizen appeared to value that reduction less than the benefit of a new car. Enforcers are often seen as guardians of societal values but must decide when it is acceptable to dictate safety choices. Managerial groups such as architects and builders are likely to be affected more by the costs than by the benefits.[15]

Perception also shapes acceptance. Studies of risk perception, as summarised by Watts and Hall, found that objectively equal risks are perceived differently when they differ in certainty, familiarity, voluntariness, detectability or how dramatic they are.[7] Hall (2008) notes that the public wants most of all to be protected from fire risks associated with strangers. Yet people are much more at risk from themselves, their families and friends. Risks are less acceptable when they are unfamiliar or involuntary.[16]

Notarianni (2002) adds that some people would accept more risk to fire fighters in return for less risk to the public, while others would not.[17] For transport vehicles, Bukowski (2002) observes that the public is more hazard averse than risk averse. Accidents with many fatalities are considered unacceptable however unlikely they are, so the concept of acceptable risk may not apply there.[18]

Watts and Hall (2002) consider the view that the present public fire risk must be acceptable, because otherwise there would be more concern, not a fully reliable generalisation.[1] People may believe the risk to be lower than it is. They may also see no feasible alternative, or have no politically effective way to show that they do not accept it. Puchovsky and Hurley (2008) add that safety goals set by society are value judgements that cannot be compromised, even by the AHJ that verifies compliance. Owners, financers and insurers may set higher goals, as long as these do not conflict with those of society.[19]

How acceptable levels are set

Hall and Watts (2008) describe several bases for an acceptable level.[6] The level is often set with an eye towards affordability and may be reset when technology changes. In their words, this makes the approach a kind of backdoor cost-benefit analysis. Other bases are historically acceptable risk, under which anything in use for a long time is accepted, and unavoidable risk, such as background radiation used as a guide for medical X-rays. In fire protection, acceptable risk has sometimes been inferred from NFPA codes and standards. The extreme case is a minimum risk approach, in which cost is ignored until all feasible safety improvements have been made.[6]

Codes often set the level implicitly. Schwartz (2008) describes codes and standards as methods, systems and uses that society deems to provide an acceptable level of safety.[20] Puchovsky and Hurley (2008) note that this level was not explicitly defined: how much safety, for whom and under what conditions.[21]

Performance-based guidance can use the code itself as the benchmark. The IFEG (2005) call this the comparative or "equivalence" approach. A trial design is compared with an otherwise identical building that follows the deemed-to-satisfy (DTS) or prescriptive provisions, using the same models and data. In the absolute approach, results are instead checked against the objectives with agreed acceptance criteria.[4]

Australia's Fire Safety Verification Method uses the comparative route. The FSVM Standard 2022 notes that the National Construction Code has no specific safety targets, and it requires a design to be at least equivalent to the DTS Provisions.[22] These provisions are accepted as reflecting societal expectations for individual risk, societal risk and defence in depth.[22] According to the ABCB handbook, they evolved in response to, amongst other things, loss of life, and tend to mirror community values and risk appetite.[23] The DTS reference building thus indirectly defines acceptable individual and societal risk levels. The stakeholders must record their consensus that it is a reasonable benchmark.[24]

In the Netherlands, a risk-based alternative to the building regulations (Bbl) is usually referred to as "equivalence". Article 4.7 of the Omgevingswet allows an equivalent measure if it achieves at least the result intended by the prescribed measure.[25] Van Rossum (2027) notes that the Bbl aims to prevent casualties and fire spread to another plot. It does not say how many deaths or injuries may occur in a fire.[26] The residual risk can be expressed as an acceptable failure probability of a so-called risk subsystem, and one way to set it is comparison with a reference situation based on the regulations.[26] For quantitative fire safety engineering, Van Rossum (2027) notes that the Netherlands has no general risk-based limit values.[27]

Explicit criteria state a target. Puchovsky and Hurley (2008) give an objective under which the risk of death due to fire is not higher than the current level for similar properties. An alternative is a limit of a stated number of deaths per 100,000 people exposed.[28] Barry (2002) lists hazard severity limits, event likelihood limits and management's risk tolerance among the issues to agree before a quantitative risk assessment.[29]

Lane (2008) notes that the accepted level of risk varies between countries and in many cases is not explicitly documented. In Europe, it is implicitly captured in the partial safety factors of each National Annex.[30] In a risk matrix, the stakeholders decide which combinations of consequence and frequency are acceptable.[31] Performance matrices instead group buildings by importance, each group with its own tolerable level of impact.[2]

Risk measures and acceptance criteria

Barry (2002) writes fire and explosion risk as the frequency of an undesired event times its expected consequences. Risk can therefore be reduced by lowering the frequency, the consequences or both.[32] Bukowski (2002) applies the same logic to ranking: a scenario with 10 injuries that is ten times more likely than one with 50 injuries is the higher risk.[18]

R=F×C

Two measures of harm to people are common. The ABCB handbook interprets individual risk as the frequency at which a person may be expected to sustain a given level of harm from a fire in the building.[23] Societal risk relates frequency to the number of people harmed in a given population. It matters for hazards that can affect many people at once.[23] Barry (2002) presents societal risk as F–N curves, which plot the cumulative frequency of multiple fatalities against the number of fatalities. Logarithmic axes are usual because the frequencies span several orders of magnitude.[33]

For buildings, the IFEG (2005) combine several fire scenarios. Each scenario has a probability of developing after ignition and a number of occupants exposed to untenable conditions, taken as the number of deaths. The expected number of deaths is the sum of probability times deaths over all scenarios.[34] The expected risk to life (ERL) divides the expected deaths over the design life by the building population and the design life.[34]

END=∑jPjNjERL=ELLBOP×tD

In the IFEG event-tree example, the acceptance criterion is that the estimated ERL does not exceed the accepted value. That value may be a specified number, in an absolute evaluation, or the ERL of a code-compliant design, in a comparative one.[5] In a comparative evaluation, the same criteria apply to both designs, and the trial design should perform no worse than the benchmark.[35] In schematic form, the two tests are:

ERLest≤ERLaccRproposed≤Rreference

Symbol Description Unit
R Risk of an undesired event expected consequences per year
F Frequency of the undesired event 1/year
C Expected consequences of the event deaths, injuries or money
END Expected number of deaths deaths (per fire)
Pj Probability of fire scenario j developing after ignition –
Nj Occupants exposed to untenable conditions in scenario j persons
ELLB Expected number of deaths over the design life of the building deaths
OP Number of occupants in the building at the start of a fire persons
tD Design life of the building years
ERL Expected risk to life deaths per person per year

The ABCB handbook gives a worked example of a comparison between two fire safety systems.[36] Deterministic modelling shows that nobody is exposed to untenable conditions if either system works, and 2 people are exposed if it fails. System A, the reference solution, has a reliability of 90 per cent, so it fails with a probability of 0.1. System B, the proposed solution, has a reliability of 70 per cent and fails with a probability of 0.3. The risk measures are 0.1 × 2 = 0.2 for System A and 0.3 × 2 = 0.6 for System B. System B thus exposes occupants to three times the risk, so the systems are not equivalent, although the deterministic analysis alone would rate them equal.[36]

A second example concerns the frequency scale of a risk matrix. In Meacham’s (2002) example, anticipated events exceed 10−2 per year. Unlikely events exceed 10−4 but do not exceed 10−2 per year. Extremely unlikely events exceed 10−6 but do not exceed 10−4 per year. Events at or below 10−6 per year are beyond extremely unlikely.[31]

Meacham warns that frequencies should refer to exceeding a given loss, not to one scenario. One scenario at 10−7 per year may suggest that fire is no concern, but 30 such scenarios add up to 3 × 10−6 fires per year.[31] As an illustration, each scenario alone would sit in the lowest bin, while the total falls in the extremely unlikely bin, one step higher.

Acceptable risk in decision making

Hall and Watts (2008) name cost-benefit analysis as the most common approach to risk evaluation.[8] They warn that the acceptable risk approach can produce unsatisfactory results. If risk exceeds the acceptable level by even a small fraction, no cost is too great to reach it. If risk is already acceptably low, no more money should be spent, however cheaply more safety could be bought.[6] Because the outcome can be very sensitive to the level chosen, they consider a sensitivity analysis necessary.[6]

In industry, Snyder (2008) describes risk management as dealing with the value judgements involved in setting acceptable levels of risk.[37] The decision rests on organisational goals such as profit, protection of assets, continued operation, employee and public safety, legal and insurance requirements and the environment. An acceptable risk may need no immediate action, but it must be monitored for changes. An unacceptable risk may be avoided, reduced, transferred through insurance, financed through self-insurance or controlled by loss control improvements.[37] In Australia, the ABCB handbook notes that work health and safety legislation requires safe design principles. The Safe Work Australia code of practice it cites defines safe design as integrating control measures early to eliminate risks or, if this is not reasonably practicable, to minimise them.[38]

The NFPA handbook also describes such choices by fire departments. Johnson and Price (2008) write that a community that selects a longer response travel time must be willing to accept a larger level of risk.[39] R. Wright (2008) describes how an acceptable level of risk for specialised rescue may be determined from an analysis of community risk.[40] In community risk reduction, a planning process selects the highest risks and the mitigation strategies most likely to succeed with the available resources.[41]

On the fireground, the incident commander weighs the risk to fire fighters against the possible outcomes. Once savable occupants have left the building, the level of acceptable risk falls.[42] Klaene and Sanders quote the US standard NFPA 1500, under which no risk to members is acceptable when there is no possibility to save lives or property.[42]

Limitations and criticism

An acceptable level is only as sound as the analysis behind it. Meacham (2002) counts the uncertainty in the criteria chosen for assessing acceptability among the uncertainties that often go unrecognised.[43] Notarianni (2002) notes that performance criteria had not been established or agreed. A different choice of criteria could make the same design acceptable in one jurisdiction and unacceptable in another.[44] Hall and Watts (2008) add that the acceptable risk approach runs counter to most approaches to decision making in business.[6]

Implicit levels are not consistent either. Van Rossum (2027) explains that Dutch regulations grew partly from consensus and partly from history, without a quantified goal. Per hour of stay, the chance of dying in a fire is much greater in a residential building than in an assembly building. Yet the fire detection requirements are stricter for assembly buildings.[26] In Watts’s 2002 account, the US Fire Safety Evaluation System was an index method for demonstrating equivalence to the 2000 edition of NFPA 101 for certain institutional occupancies, including health care facilities.[45] Its hardship adjustment reduced the calculated occupancy-risk factor for an existing building to 60 per cent of that for an equivalent new building.[46]

Data also limit quantitative criteria. The 2013 commentary on the first edition of New Zealand's C/VM2 states that a probabilistic risk assessment would ideally be used. The tools and data were, however, not suitable, so the method uses a deterministic egress time comparison with additional checks.[47] The same commentary describes the risk of fire spread to adjacent property as small but not insignificant. It quantifies this risk as approximately 3 per cent of all structure fires, based on New Zealand Fire Service statistics for 2005–2010.[48]

Two misconceptions follow from these sources. Acceptable risk is never zero risk, even when people perceive it so.[1] Code compliance does not mean that the level of safety is known, because the implicit level was not explicitly defined.[21]

See also

Notes

  1. ↑ 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 Watts and Hall (2002), p. 5-4.
  2. ↑ 2.0 2.1 Meacham and Galioto (2008), p. 1-116.
  3. ↑ 3.0 3.1 3.2 National Research Council of Canada et al. (2005), Part 0 – Canada, p. 0.2-2.
  4. ↑ 4.0 4.1 National Research Council of Canada et al. (2005), p. 1.2-21.
  5. ↑ 5.0 5.1 National Research Council of Canada et al. (2005), p. 1.2-23.
  6. ↑ 6.0 6.1 6.2 6.3 6.4 6.5 6.6 Hall and Watts (2008), p. 3-138.
  7. ↑ 7.0 7.1 Watts and Hall (2002), p. 5-3.
  8. ↑ 8.0 8.1 8.2 Hall and Watts (2008), p. 3-137.
  9. ↑ Hall and Watts (2008), p. 3-136.
  10. ↑ Ministry of Business, Innovation and Employment (2025b), p. 37.
  11. ↑ 11.0 11.1 Puchovsky and Hurley (2008), p. 3-168.
  12. ↑ National Research Council of Canada et al. (2005), p. 1.2-25.
  13. ↑ Australian Building Codes Board (2020b), p. 28.
  14. ↑ Granito (2008), p. 12-19.
  15. ↑ 15.0 15.1 15.2 Watts and Hall (2002), pp. 5-1–5-2.
  16. ↑ Hall (2008), p. 5-107.
  17. ↑ Notarianni (2002), p. 5-46.
  18. ↑ 18.0 18.1 Bukowski (2002), p. 5-232.
  19. ↑ Puchovsky and Hurley (2008), p. 3-175.
  20. ↑ Schwartz (2008), p. 1-73.
  21. ↑ 21.0 21.1 Puchovsky and Hurley (2008), p. 3-167.
  22. ↑ 22.0 22.1 Australian Building Codes Board (2022), p. 7.
  23. ↑ 23.0 23.1 23.2 Australian Building Codes Board (2020b), p. 22.
  24. ↑ Australian Building Codes Board (2020b), p. 46.
  25. ↑ Van Rossum (2027), §1.
  26. ↑ 26.0 26.1 26.2 Van Rossum (2027), §1.2.
  27. ↑ Van Rossum (2027), §1.4.
  28. ↑ Puchovsky and Hurley (2008), p. 3-171.
  29. ↑ Barry (2002), p. 5-177.
  30. ↑ Lane (2008), p. 19-86.
  31. ↑ 31.0 31.1 31.2 Meacham (2002), pp. 5-166–5-167.
  32. ↑ Barry (2002), p. 5-176.
  33. ↑ Barry (2002), pp. 5-210–5-211.
  34. ↑ 34.0 34.1 National Research Council of Canada et al. (2005), pp. 2.3-4–2.3-5.
  35. ↑ National Research Council of Canada et al. (2005), p. 1.2-26.
  36. ↑ 36.0 36.1 Australian Building Codes Board (2020b), pp. 118–119.
  37. ↑ 37.0 37.1 Snyder (2008), p. 12-179.
  38. ↑ Australian Building Codes Board (2020b), p. 25.
  39. ↑ Johnson and Price (2008), p. 12-219.
  40. ↑ R. Wright (2008), p. 13-15.
  41. ↑ Kirtley (2008a), p. 12-310.
  42. ↑ 42.0 42.1 Klaene and Sanders (2008), pp. 13-6–13-7.
  43. ↑ Meacham (2002), p. 5-159.
  44. ↑ Notarianni (2002), pp. 5-46–5-47.
  45. ↑ Watts (2002), pp. 5-129–5-132.
  46. ↑ Watts (2002), p. 5-132.
  47. ↑ Ministry of Business, Innovation and Employment (2013), p. 158.
  48. ↑ Ministry of Business, Innovation and Employment (2013), p. 23.

References

  • Australian Building Codes Board. (2020b). Handbook: Fire Safety Verification Method (Print version 1.1).
  • Australian Building Codes Board. (2022). Fire Safety Verification Method Standard 2022.
  • Barry, T. F. (2002). Quantitative risk assessment in chemical process industries. 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. 5-176–5-213). National Fire Protection Association; Society of Fire Protection Engineers.
  • Bukowski, R. W. (2002). Fire hazard assessment for transportation vehicles. 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. 5-227–5-233). National Fire Protection Association; Society of Fire Protection Engineers.
  • Granito, J. A. (2008). Planning for public fire-rescue protection. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 12-3–12-22). National Fire Protection Association.
  • Hall, J. R., Jr. (2008). Using data for public education planning and decision making. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 5-103–5-118). 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.
  • Johnson, R., & Price, M. (2008). GIS for fire station locations and response protocols. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 12-215–12-231). National Fire Protection Association.
  • Kirtley, E. (2008a). Community risk reduction. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 12-309–12-315). National Fire Protection Association.
  • Klaene, B. J., & Sanders, R. (2008). Fireground operations. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 13-3–13-11). National Fire Protection Association.
  • Lane, B. (2008). Approaches to calculating structural fire resistance. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 19-77–19-126). National Fire Protection Association.
  • Meacham, B. J. (2002). Building fire risk analysis. 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. 5-153–5-175). National Fire Protection Association; Society of Fire Protection Engineers.
  • Meacham, B. J., & Galioto, C. (2008). Protecting against extreme events. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 1-109–1-136). National Fire Protection Association.
  • Ministry of Business, Innovation and Employment. (2013). Commentary for Building Code clauses C1–C6 and Verification Method C/VM2.
  • Ministry of Business, Innovation and Employment. (2025b). Verification Method C/VM2: Framework for fire safety design (2nd ed.).
  • 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.
  • Notarianni, K. A. (2002). Uncertainty. 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. 5-40–5-64). National Fire Protection Association; Society of Fire Protection Engineers.
  • Puchovsky, M. T., & Hurley, M. J. (2008). Performance-Based codes and standards for fire safety. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 3-167–3-181). National Fire Protection Association.
  • Schwartz, A. E. (2008). Legal issues for the designer and enforcer. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 1-73–1-80). National Fire Protection Association.
  • Snyder, M. (2008). Industrial fire loss prevention. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 12-177–12-198). National Fire Protection Association.
  • Van Rossum, E. (2027). Introduction to Fire Safety Engineering. Fire Safety Academy.
  • Watts, J. M., Jr. (2002). Fire risk indexing. 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. 5-125–5-142). National Fire Protection Association; Society of Fire Protection Engineers.
  • Watts, J. M., Jr., & Hall, J. R., Jr. (2002). Introduction to fire risk analysis. 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. 5-1–5-7). National Fire Protection Association; Society of Fire Protection Engineers.
  • Wright, R. (2008). Organizing rescue operations. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 13-13–13-21). National Fire Protection Association.