Fire engineering analysis and performance-based design
Domain 4 tests the fire engineering approach: the alternative route where a design cannot follow the deemed-to-satisfy provisions of the FS Code 2011 and must be justified by analysis.
The process
An FEA submission defines the scope and the specific departures, proposes design fire scenarios, sets acceptance criteria, analyses with validated tools, and demonstrates compensatory or mitigation measures. In Hong Kong the report is agreed with the Building Authority and circulated to FSD where fire service installations or firefighting access are affected. Approval is project-specific and creates no precedent.
Design fire scenarios
Select a small set (typically three to six) that brackets credible outcomes: a growing fire with sprinklers effective; the same fire with sprinklers failing; a smouldering fire in a concealed space; a fire adjacent to the most remote egress point; and, where relevant, a vehicle or external fire. State the HRR curve, growth rate, soot yield, species yields (CO, CO2, HCN), radiative fraction and ventilation condition.
Tenability criteria
Common acceptance criteria at 1.8 m or head height: smoke layer temperature ≤ 60 °C, radiant heat flux ≤ 2.5 kW/m², visibility ≥ 10 m in large spaces (about 5 m in small rooms), and toxic exposure assessed as a fractional effective dose rather than one threshold concentration. ASET must exceed RSET with a stated margin, not merely equal it.
Egress and modelling
RSET = detection time + pre-movement time + travel time. Pre-movement often dominates and must be justified from occupancy type and alarm strategy. Egress models (STEPS, Pathfinder, Simulex) need realistic walking speeds, exit choice and door flow rates. CFD needs a grid resolved against the characteristic fire diameter D*, typically four to sixteen cells across it.
Worked example
For a 2.5 MW design fire, D* = (Q/(ρ∞cpT∞√g))^0.4 = (2500/(1.2 × 1.0 × 293 × 3.13))^0.4 ≈ 1.39 m. At ten cells across D*, the cell size is about 0.14 m. A 0.5 m grid gives under three cells across D*, too coarse to trust plume and ceiling-jet temperatures.
Traps
Comparing ASET at ceiling height while occupants breathe at 1.8 m; quoting a design fire that is implicitly sprinkler-controlled while the scenario is sprinkler failure; omitting pre-movement from RSET; submitting an FEA to reduce life safety without any compensating measure.