Skip to content
Upcoming courses

Acceptance criteria

Tools
From Fire Safety Academy Wiki
Revision as of 23:50, 1 November 2026 by Peter (talk | contribs) (New FSE wiki article: Acceptance criteria (written by GPT-6))
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Contents
  1. 1From objectives to assessable outcomes
  2. 2Absolute and comparative criteria
  3. 3Tenability thresholds and evacuation margins
  4. 4Risk measures and the population being protected
  5. 5Uncertainty, safety factors and sensitivity cases
  6. 6Agreement, documentation and review
  7. 7See also
  8. 8Notes
  9. 9References

Acceptance criteria are the conditions used to judge whether a fire safety analysis demonstrates an acceptable outcome. They translate objectives into limits on damage, failure, exposure or risk that can be compared with the results. Meacham connects their selection with consequence analysis: identifying what is valued, how fire could harm it and which impacts can be tolerated.[1]

Criteria are part of the design basis, alongside the scenarios and methods used to evaluate them. The International Fire Engineering Guidelines require criteria to suit the objectives and analysis methods, to be numerical unless the assessment is qualitative, and to be realistic. An assertion of zero risk does not provide an appropriate general criterion.[2]

A criterion may specify an absolute limit or compare a proposed design with a defined benchmark. Either approach requires clarity about the measured quantity, relevant location or population, and conditions being assessed. Passing one numerical check demonstrates only the performance addressed by that check. It does not automatically establish that every fire safety objective has been satisfied.[3][1]

From objectives to assessable outcomes

Fire safety objectives describe what the design is intended to protect or achieve. Acceptance criteria state how the assessment will determine whether that outcome is adequately achieved. Meacham identifies life safety, property, business continuity, heritage and environmental protection as different value areas. The relevant criteria therefore depend on the objectives of the particular analysis.[4]

A general objective such as protecting occupants does not identify a unique calculation. It could lead to criteria concerning evacuation time, smoke exposure or expected harm. Protecting adjacent property could instead involve radiant exposure or flame impingement. The IFEG examples demonstrate that different objectives require different output quantities, even when they arise from the same underlying fire scenario.[5]

The distinction matters in performance-based design because broad performance language does not necessarily contain a numerical threshold. Van Rossum discusses the difficulty of translating prescriptive limits directly into quantified safety goals. A measurable rule about a building feature does not necessarily state an acceptable probability of harm. The assessment must identify the performance meaning of the criterion it adopts.[6]

Criteria also need to distinguish regulatory objectives from additional client objectives. A design that satisfies an occupant protection requirement may still permit property damage or business interruption that the owner considers unacceptable. Meacham highlights the difference between code compliance and protection of assets or operations. Agreement about the objective prevents these distinct outcomes from being treated as interchangeable.[7]

Consequence analysis helps establish that agreement. It considers how injury, damage or interruption would affect the people and assets concerned. Some impacts are readily represented by physical thresholds, while others require additional decisions about their significance. The criterion is therefore informed by technical evidence and the values of the interested and affected parties, rather than by modelling capability alone.[1]

Absolute and comparative criteria

An absolute assessment compares the proposed design with specified performance objectives through agreed criteria. It does not require a prescriptive reference design as the numerical benchmark. The selected threshold still needs a defensible connection to the objective, an appropriate analysis method and treatment of uncertainty. Calling the assessment absolute does not mean its result is free from assumptions.[8][9]

A comparative assessment supports equivalency by comparing a trial design with a defined reference. The same criteria are applied to both, and the trial design should perform at least as well. The IFEG approach also calls for consistent assumptions, models, calculations and inputs. This consistency allows the comparison to focus on the design differences under evaluation.[8][9]

The direction of a comparison depends on the quantity. A longer available evacuation time may be favourable, while greater exposure or loss may be unfavourable. The decision rule must preserve that meaning. Describing two results as similar without defining acceptable direction or magnitude leaves the criterion incomplete and can hide a deterioration in the relevant performance.[10]

A reference value is not automatically a universal threshold. It belongs to the defined reference design, scenario and analysis conditions. The ABCB Handbook describes deriving comparative acceptance criteria together with the design fire, protection system status and occupant characteristics. Changing those conditions can change the benchmark as well as the proposed design's result.[11]

Several criteria may be needed simultaneously. The IFEG allows relevant parameters to be used individually or together. An evacuation result, structural condition and property exposure can therefore form separate parts of the assessment. Success against one does not provide a numerical offset for failure against another unless the adopted analysis explicitly establishes and justifies such a relationship.[5]

A qualitative assessment still requires an acceptance basis. The IFEG discusses qualitative evaluation where the issues are sufficiently limited and the reasoning can be documented with relevant references. In that case, the criterion may be expressed through an argued performance comparison rather than a computed threshold. That reasoning should identify the relevant design difference and the evidence supporting its claimed effect on performance. The absence of a numerical result does not remove the need to explain why the evidence answers the objective.[8][2]

Tenability thresholds and evacuation margins

Tenability criteria identify environmental conditions relevant to occupants' ability to remain or move safely. In an evacuation analysis, they help determine the available safe evacuation time, commonly abbreviated ASET. The required safe evacuation time, RSET, comes from the evacuation assessment. Comparing those times is a further decision step, distinct from choosing the environmental thresholds used to calculate ASET.[12][5]

The IFEG timeline example requires an agreed margin between available and required time for an absolute evaluation. It does not prescribe one universal margin for every building. In a comparative evaluation, it describes comparing the trial design's margin with that of the reference. The definition of the margin and the applicable acceptance rule must therefore be stated.[12]

An additive margin can be written as the difference between the two times. For the illustrative rule below, acceptance requires that difference to reach a separately specified minimum.[12]

M = ASET − RSET;   accept the stated time comparison when M ≥ Mreq
Symbol Meaning Unit
ASET Available safe evacuation time under the selected tenability criteria s
RSET Required safe evacuation time for the assessed occupants and strategy s
M Calculated additive time margin s
Mreq Required margin selected for this illustrative decision rule s

Consider hypothetical ASET of 600 s and RSET of 450 s. Their difference is 150 s. If an illustrative criterion requires at least 90 s, this case satisfies that criterion. The 90 s value is selected only for the example; it is not a general requirement from the cited guidance.[12][9]

Now suppose a slower evacuation case gives RSET of 540 s while ASET remains 600 s. The margin falls to 60 s and fails the same illustrative criterion. Although RSET is still less than ASET, the agreed margin has not been achieved. Merely reporting the order of the two times would miss the specified decision rule.[12]

The arithmetic does not validate either time estimate. Their credibility depends on the fire scenario, population, evacuation behaviour and the methods used. The IFEG identifies reduced movement speeds as one possible sensitivity study. It also requires the treatment of such studies, including any different acceptance basis, to be agreed rather than assumed after the results are known.[13]

Risk measures and the population being protected

A probabilistic acceptance criterion needs a defined risk measure as well as a numerical limit. Individual risk concerns harm to an individual under the specified exposure assumptions. Societal risk concerns events affecting multiple people and the relationship between frequency and the number harmed. The ABCB Handbook distinguishes these quantities and calls for both to be considered in its comparative fire safety strategy.[14]

Different risk measures answer different questions. A location-based contour can describe the frequency of a specified harmful exposure at a position, whether or not someone is present there. A measure involving actual occupants requires assumptions about their presence and exposure. Barry's discussion shows why the population and exposure basis must be stated when interpreting a reported risk value.[15]

An expected loss measure combines consequences and their likelihood across the scenarios included in the analysis. It does not mean that the expected consequence will occur in each fire. The IFEG probabilistic example compares an expected risk-to-life result with an agreed criterion. The scenario set, probabilities and population denominator are therefore part of the meaning of that comparison.[16]

A deterministic analysis does not automatically provide those probabilities. It predicts an outcome for specified initial and boundary conditions. Passing a deterministic scenario can establish that its criterion is met under those conditions, but it does not by itself quantify the frequency of that outcome. Treating the result as a numerical individual or societal risk would require additional information and analysis.[12]

Risk criteria also involve decisions about which consequences matter. Meacham discusses injury, lasting impairment, property damage and interruption of operations as potentially distinct impacts. Collapsing them into one value can obscure those distinctions. The assessment should preserve enough information for the selected criteria to address the objectives actually agreed with the affected parties.[4]

Uncertainty, safety factors and sensitivity cases

A predicted value close to a limit needs to be interpreted with the uncertainties in its inputs and method. The IFEG distinguishes sensitivity studies, which change inputs to examine their influence, from uncertainty studies, which examine how uncertain information affects outputs. It also distinguishes redundancy studies that examine alternative means of providing a safety function. These studies inform different parts of the acceptance judgement.[17]

A safety factor or margin should have a stated basis. The IFEG identifies system reliability, redundancy, analysis methods, assumptions, uncertainty and consequences as relevant considerations. The selection is consequently not reduced to choosing an attractive round number. Its relationship to the uncertainty and potential harm needs to remain visible in the assessment.[9]

The guidance also cautions against applying safety factors repeatedly throughout a calculation chain. Doing so can accumulate conservatism without showing how much has been introduced. A transparent assessment identifies where the allowance enters and how it affects the final comparison. This makes its relationship to the acceptance criterion easier to understand and review.[9]

Sensitivity cases can have an acceptance basis different from the principal design cases where that approach is agreed and justified. The IFEG explicitly discusses this possibility to avoid excessive conservatism. A report should therefore identify which criterion applies to each case. Quietly relaxing a limit after an unfavourable result would not preserve the agreed analysis basis.[13]

The number printed by a model is not, on its own, the acceptance decision. Meacham stresses that uncertainty, variability and unknowns influence both consequence assessment and selection of criteria. The result must be interpreted with those limitations and with the scope of the represented scenario. A precise numerical output does not establish equally precise knowledge of the real outcome.[18]

Protection system status is another part of the acceptance context. A result obtained with every system operating cannot silently represent a scenario in which a particular system fails. The ABCB Handbook identifies the status and impact of active and passive protection among the elements used to define reference scenarios. Keeping that information beside the criterion makes clear which assumed conditions the decision covers.[11]

Agreement, documentation and review

The IFEG process calls for objectives, trial designs, methods and acceptance criteria to be agreed before the full analysis. Preliminary calculations may help establish whether a proposed solution is plausible. If the design proves unacceptable, the process can return to the brief and develop another trial design. Revisions use the same consultative approach rather than silently altering the basis of acceptance.[19]

Agreement does not mean that every participant will necessarily share the same view. The ABCB Handbook requires dissenting views to be recorded and carried through the process where consensus is not achieved. This preserves unresolved issues for the decision and approval stage. A report that removes disagreement from the record would conceal information relevant to interpreting the analysis.[20]

The documented basis should connect each criterion with its objective, scenario and output quantity. It should also identify the method, relevant assumptions and treatment of uncertainty. The IFEG requires methods to generate outputs that can actually be compared with the agreed criteria. A detailed simulation is of limited value if its reported quantities cannot answer the acceptance question.[21]

Method selection should also address validation and suitability for the intended application. The IFEG asks for the chosen methods to have the necessary validation and to produce relevant outputs. A criterion can therefore be well defined while the proposed calculation remains inadequate to demonstrate it. Review needs to examine both the decision rule and the capability of the evidence offered in support.[21]

See also

Notes

  1. ↑ 1.0 1.1 1.2 Meacham (2002), pp. 5-154–5-155.
  2. ↑ 2.0 2.1 Australian Building Codes Board et al. (2005), p. 1.2-25.
  3. ↑ Australian Building Codes Board et al. (2005), pp. 1.2-21–1.2-26.
  4. ↑ 4.0 4.1 Meacham (2002), pp. 5-153–5-155.
  5. ↑ 5.0 5.1 5.2 Australian Building Codes Board et al. (2005), pp. 1.2-25–1.2-26.
  6. ↑ Van Rossum (2027), §1.2.
  7. ↑ Meacham (2002), p. 5-155.
  8. ↑ 8.0 8.1 8.2 Australian Building Codes Board et al. (2005), p. 1.2-21.
  9. ↑ 9.0 9.1 9.2 9.3 9.4 Australian Building Codes Board et al. (2005), p. 1.2-26.
  10. ↑ Australian Building Codes Board et al. (2005), pp. 1.2-22–1.2-26.
  11. ↑ 11.0 11.1 Australian Building Codes Board (2020), p. 59.
  12. ↑ 12.0 12.1 12.2 12.3 12.4 12.5 Australian Building Codes Board et al. (2005), p. 1.2-22.
  13. ↑ 13.0 13.1 Australian Building Codes Board et al. (2005), pp. 1.2-24–1.2-26.
  14. ↑ Australian Building Codes Board (2020), p. 22.
  15. ↑ Barry (2002), p. 5-210.
  16. ↑ Australian Building Codes Board et al. (2005), pp. 1.2-22–1.2-23.
  17. ↑ Australian Building Codes Board et al. (2005), p. 1.2-24.
  18. ↑ Meacham (2002), pp. 5-154–5-156.
  19. ↑ Australian Building Codes Board et al. (2005), p. 1.2-2.
  20. ↑ Australian Building Codes Board (2020), p. 32.
  21. ↑ 21.0 21.1 Australian Building Codes Board et al. (2005), p. 1.2-23.

References

  • Australian Building Codes Board. (2020). Fire safety verification method handbook (Version 1.1).
  • Australian Building Codes Board, National Research Council of Canada, International Code Council, & Department of Building and Housing. (2005). International fire engineering guidelines.
  • Barry, T. F. (2002). Quantitative risk assessment in chemical process industries. In P. J. DiNenno (Ed.), SFPE handbook of fire protection engineering (3rd ed., Chapter 5-13). National Fire Protection Association.
  • Meacham, B. J. (2002). Building fire risk analysis. In P. J. DiNenno (Ed.), SFPE handbook of fire protection engineering (3rd ed., Chapter 5-12). National Fire Protection Association.
  • Van Rossum, E. (2027). Introduction to Fire Safety Engineering. Fire Safety Academy.
Retrieved from "https://fire-safety.academy/w/index.php?title=Acceptance_criteria&oldid=497"