NFPA 92 Atrium Smoke Control: Evacuation Engineering in Large Volumes

In large atriums — malls, hotel lobbies, transit stations — a fire quickly fills the space with smoke. Before sprinklers knock down the fire, smoke can cut off egress on lower levels. NFPA 92 smoke control keeps smoke above a set height to buy evacuation time. This post summarizes smoke-layer analysis, plume math, and ventilation design.

Why Smoke Kills

• Hot smoke rises, cools, descends
• CO, HCN, HCl kill via inhalation
• Visibility drops under 10 m; egress slows
• Panic → wrong direction

NFPA 92 goal: keep smoke layer above head height (min 1.8 m + 10% safety) during evacuation.

Two Main Strategies

• Natural (smoke vents): Roof dampers open; hot smoke rises.
• Mechanical exhaust: Large fans pull smoke out. More controlled but power-dependent.

Smoke Plume Calculation

Heat release rate (HRR) × plume geometry gives mass flow. Typical design fires:

• Retail (mall shop): 2-5 MW
• Hotel lobby: 1-2 MW
• Transit station: 5-10 MW
• Theater stage: 5-15 MW

Heskestad plume: m = 0.071 × Q^(1/3) × (z - z₀)^(5/3) (m kg/s, Q kW, z m).

Smoke Layer Interface

Smoke accumulates overhead; its lower boundary is the smoke layer interface. NFPA 92 targets a minimum 1.8 m above floor.

Exhaust flow must exceed plume flow so the layer rises. If not, layer descends and egress fills with smoke.

Make-Up Air

While smoke exhausts, make-up air must enter. Otherwise atrium goes negative and doors jam. Rules:

• Make-up velocity below smoke-layer level, max 1 m/s
• Make-up dampers auto-open
• 85-95% of exhaust flow (balance is leakage)

Smoke Control Commissioning

• Cold smoke test: Low-density aerosol or theater smoke simulation
• Hot smoke test: Controlled small fire with real heat release
• Smoke layer height measured via laser/thermal camera
• NFPA 92 max 15% deviation

Turkey Perspective

BYKHY 2007 mandates smoke control for atriums. Execution varies — Istanbul's new malls run hot smoke tests, elsewhere many only verify damper operation. Systems without real performance data pass on paper, fail in fire.

Common Mistakes

1. No make-up air: Negative pressure traps doors.
2. Sprinkler plume effect ignored: Sprinklers cool and lower smoke; must be modeled.
3. Undersized exhaust: 2 MW assumed where 5 MW design fire needed.
4. No hot smoke test: Real performance unknown.

Conclusion

Atrium smoke control meets architecture with engineering. A grand glass dome is beautiful — but deadly in fire. NFPA 92 + hot smoke test + make-up air + sprinkler integration done right lets 1,000 people evacuate safely.

Atrium smoke calc in MEP Calc

Plume flow, exhaust fan sizing, make-up air calc, NFPA 92 commissioning schedule.

View on App Store