Free Engineering Tool · NFPA 72 / NFPA 92
t-Squared Fire Growth Calculator
Plot the standard t-squared design fires or a custom growth rate, then cap them, model a sprinkler-controlled fire, or build a full design fire that burns at the fuel or ventilation limit and decays as the fire load runs out. Download the heat release rate table as CSV for fire models and smoke control calculations.
- Based on NFPA 72 / NFPA 92
- Free, runs in your browser; your inputs stay on your device
- Screening tool, not reviewed or sealed by a licensed engineer
- Last updated October 2026
Screening tool, not engineering advice. This calculator estimates results with simplified published methods. It has not been reviewed or sealed by a licensed engineer for your project, and DGA does not warrant that its results are accurate, complete or suitable for any use. Check every result independently and have the licensed professional responsible for your project review it before you rely on it. Use is subject to the Terms of Use.
Advanced method. The correlations behind this tool come from research and test data with narrow ranges of validity. Outside those ranges results can be wrong by a wide margin, and they are no substitute for a fire model or an analysis by a qualified fire protection engineer.
Design Fire
Peak and Sprinkler Control
Fire Load and Ventilation
Get the fire load density from the fire load density calculator. Use m = 0.8 for mainly cellulosic contents, or 1.0 if the density already includes it. For several openings, enter the total area and the area-weighted average height; leave the area blank for a fire that is not ventilation limited.
Chart and Table
Result
Design fire phases
| Curve | To 1 MW (s) | Peak | At tmax |
|---|
Heat Release Rate Against Time
About This Tool
A t-squared fire grows as Q = α · t2. NFPA 72 and NFPA 92 define the standard growth rates by the time tg to reach 1,055 kW (1,000 Btu/s), so α = 1,055 / tg2: slow 600 s (0.00293 kW/s2), medium 300 s (0.01172), fast 150 s (0.0469) and ultra-fast 75 s (0.1876). Pick a design curve or enter a custom α or growth time, and plot the standard curves beside it for comparison. An optional peak heat release rate holds each curve steady once it reaches the cap, as fuel or ventilation limits would.
The sprinkler option changes the curve after the activation time. Hold steady keeps the heat release rate at its value at activation, the usual assumption for a sprinkler-controlled fire in smoke control design. Decay follows the NIST suppression model, Q = Qact · exp(−(t − tact) / (3.0 · w−1.85)), where w is the water spray density in mm/s. Madrzykowski and Vettori fitted a bound with a 435 s time constant to furniture and wood crib tests at 0.07 mm/s (about 0.1 gpm/ft2), and Evans extended it to other densities; the time constant is 411 s at 0.07 mm/s and 59 s at 0.2 mm/s. The activation time comes from a detector or sprinkler response calculation for the same design fire.
The full design fire option follows the curve of EN 1991-1-2 Annex E. The t-squared growth stops at the lower of two limits: the fuel limit, the peak heat release rate per unit area times the fire area (250 kW/m2 is typical for offices and dwellings), and the ventilation limit for one effective opening, Q = 1,400 · Av · √heq kW, with the opening area Av in m2 and its height heq in m. That is the Eurocode expression 0.10 · m · Hu · Av · √heq MW with m = 0.8 and Hu = 17.5 MJ/kg for wood, and it goes back to Kawagoe’s burning rate of about 0.1 Av √h kg/s. The fire then burns steadily until 70 percent of the fire load has been consumed and decays linearly to zero, so the area under the whole curve equals the fire load times the combustion factor m. If 70 percent burns before the curve reaches a limit, the decay starts during growth. The phase table gives the end of growth, the start of decay, the end of the fire and which limit governs.
The results give the design curve’s peak heat release rate, its growth coefficient and growth time, the heat release rate at the end of the time horizon and the total heat released over that time. The summary table lists every plotted curve, and the CSV download gives the heat release rate of each curve at the table time step, ready for a fire model ramp or a spreadsheet. The fire load inputs are in SI units (MJ, m2, m); heat release rate units are kW, MW or Btu/s, on a linear or logarithmic scale.
Worked example
A medium fire in an office (α = 0.01172 kW/s2) activates the sprinklers at 250 s, when it is releasing 0.01172 × 2502 = 733 kW. Held steady, it is still 733 kW at 600 s and has released 317 MJ. With the decay model at 0.1 gpm/ft2 (0.068 mm/s, time constant 435 s) it falls to 327 kW by 600 s and releases 237 MJ. Left unsprinklered, the same fire reaches 4,220 kW at 600 s.
For the full design fire, take a 40 m2 store room with 600 MJ/m2 of mainly cellulosic contents (m = 0.8, so 19,200 MJ will burn) and one 0.9 m by 2.1 m door. The ventilation limit is 1,400 × 1.89 × √2.1 = 3,834 kW, below the fuel limit of 250 × 40 = 10,000 kW, so the opening governs. A medium fire reaches it at 572 s, burns steadily until 3,886 s when 13,440 MJ is gone, and decays to zero at 6,891 s, about 115 minutes. Doubling the door width raises the plateau to 7,669 kW and shortens the fire to 3,794 s.
Assumptions and limits
Without the full design fire option the t-squared curve describes the growth phase only. The full design fire treats the compartment as one zone with one effective opening and assumes the fire spreads over the fire area; the jump from a localized fire to flashover is not checked, and the ventilation limit applies to fully developed, mainly cellulosic fires. The decay model is a bounding estimate for unshielded furnishings in light hazard spaces with at least about 0.1 gpm/ft2; it does not apply to shielded fuel, storage, or fires that keep growing after the sprinklers open. One activation time applies to every plotted curve, while in practice faster fires activate sprinklers sooner, so compare curves with their own activation times.
When to bring in an engineer
The design fire is often the most debated input in a smoke control or performance-based design, and the authority having jurisdiction has to accept it. DGA designs smoke control systems and builds fire models for performance-based approvals.
References
- NFPA 72, National Fire Alarm and Signaling Code, National Fire Protection Association, 2025 edition, Annex B: t-squared fires and growth rates.
- NFPA 92, Standard for Smoke Control Systems, National Fire Protection Association, 2024 edition: design fires and sprinkler-controlled fires.
- EN 1991-1-2:2002, Eurocode 1: Actions on structures, Part 1-2: General actions, Actions on structures exposed to fire, Annex E (fire load densities and the design fire: growth, fuel and ventilation controlled plateau, decay after 70 percent of the fire load).
- K. Kawagoe, Fire Behaviour in Rooms, Report No. 27, Building Research Institute, Japan, 1958.
- D. Madrzykowski and R.L. Vettori, A Sprinkler Fire Suppression Algorithm for the GSA Engineering Fire Assessment System, NISTIR 4833, National Institute of Standards and Technology, 1992.
- D.D. Evans, Sprinkler Fire Suppression Algorithm for HAZARD, NISTIR 5254, National Institute of Standards and Technology, 1993.
- G. Heskestad and M.A. Delichatsios, Environments of Fire Detectors, Phase I: Effect of Fire Size, Ceiling Height and Material, NBS-GCR-77-86, National Bureau of Standards, 1977.
Frequently Asked Questions
What are the standard t-squared growth rates?
Slow (alpha about 0.00293 kW/s2), medium (0.01172), fast (0.0469) and ultra-fast (0.1876). They reach 1,055 kW (1,000 Btu/s) in 600, 300, 150 and 75 seconds.
How do I choose a growth rate?
From the fuel: its materials, arrangement and any test data. Guidance in NFPA 72 and NFPA 92 gives typical rates for common items, and the authority having jurisdiction normally has to accept the design fire.
Should a sprinkler-controlled fire hold steady or decay?
Holding the heat release rate steady at activation is the common, conservative assumption for smoke control design. The NIST decay model is a bounding estimate for unshielded furnishings in light hazard spaces; use it only where the fuel is not shielded from the spray and the density is at least about 0.1 gpm/ft2.
How long does a design fire burn?
With the full design fire option the fire grows to the fuel or ventilation limit, burns steadily until 70 percent of the fire load is consumed, then decays linearly to zero. The total heat released equals the fire load times the combustion factor, so a larger fire load or a smaller opening gives a longer fire. The phase table gives the times.
How do I use the CSV in a fire model?
Each row gives the time in seconds and the heat release rate of each plotted curve in the units selected. Divide by the peak value to build a ramp of fractions for FDS, or paste the times and values into a CFAST fire definition.
Related Tools and Services
More fire dynamics and detection calculators →
Engineering aid only, provided as is and without warranty. Results depend on the inputs and on the simplifying assumptions of the method summarized on each tool page, and DGA has not verified them for your application. They do not replace code-required design, testing or review by the licensed engineer responsible for the project. Use of the tools is subject to the Terms of Use.
Related Reading
Articles from DGA Insights on the code and the method behind this calculator.