Suppression agents & fire dynamics (NFPA 2001/12)

modified2026-07-21statusfinished

NFPA 2001/12 gaseous suppression sizing and SFPE fire dynamics — clean-agent and CO₂ total flooding mass, design-fire growth, flame height, and radiant exposure, in one reference.

Index

Clean agent flooding mass
Agent specific volume
Enclosure hold time
Total flooding CO₂ mass
t² fire growth & HRR
Heskestad mean flame height
Radiant heat flux & exposure distance

Relation sets with a natural input-output curve include a Voici¹1Voici compiles a notebook to a static, serverless dashboard — mocked here as a static page, since this pipeline has no Jupyter kernel to run against. dashboard: a matplotlib-styled chart with the live input marked as a red point.

1.19 · Clean agent flooding mass

[1.19] Governing equations

$$W = \frac{V}{s}\left(\frac{C}{100-C}\right),\quad s = k_1 + k_2 T$$

$V$ enclosure volume, $C$ design concentration (%), $s$ specific volume of superheated vapor at temperature $T$. Source: NFPA 2001 §5.2.

FIG. 01 — total flooding clean-agent system: cylinder bank discharging into a sealed enclosure of volume V

Enclosure volume V (m³) Design concentration C (%) Temperature T (°C) Agent

—

1.20 · Agent specific volume

[1.20] Governing equation

$$s = k_1 + k_2 T_C$$

Temperature-dependent specific volume of superheated agent vapor. Source: NFPA 2001 §5.2.

Agent Temperature T (°C)

—

import numpy as np, matplotlib.pyplot as plt
T = np.linspace(-20, 60, 200)
s = k1 + k2*T
plt.plot(T, s); plt.scatter([T_input],[s_input], color='red', zorder=3)

FIG. 02 — specific volume s vs. temperature T, current T marked in red

1.21 · Enclosure hold time

[1.21] Governing relation

$$t \propto \frac{A_f\sqrt{H}}{\text{EAL}}$$

Simplified proxy for the NFPA 2001 Annex C door-fan retention-time model: larger floor area and enclosure height slow the agent/air interface descent; larger equivalent leakage area (EAL) speeds it. Consult NFPA 2001 Annex C for the full buoyancy-driven model. Source: NFPA 2001 Annex C.

Enclosure height H (m) Floor area Af (m²) Equivalent leakage area EAL (cm²)

—

1.22 · Total flooding CO₂ system mass

[1.22] Governing equation

$$M_{\text{CO}_2} = V \times \text{Volume Factor} + A_{\text{openings}} \times \text{Leakage Factor}$$

Volume factor depends on hazard class and design concentration; leakage factor accounts for unclosable openings. Representative values shown. Source: NFPA 12 §5.3.

Hazard class Room volume V (m³) Unclosable opening area (m²)

—

1.23 · t² fire growth rate & HRR

[1.23] Governing equation

$$\dot{Q}(t) = \alpha t^2$$

Standard growth coefficients $\alpha$: Slow 0.00293, Medium 0.01172, Fast 0.0469, Ultra-fast 0.1876 kW/s². Source: NFPA 92, SFPE Handbook.

FIG. 03 — parabolic t² design fire growth curve

Growth rate Time t (s)

—

import numpy as np, matplotlib.pyplot as plt
t = np.linspace(0, 600, 200)
Q = alpha * t**2
plt.plot(t, Q); plt.scatter([t_input],[Q_input], color='red', zorder=3)

FIG. 04 — heat release rate Q̇ vs. time, current time marked in red

1.24 · Heskestad mean flame height

[1.24] Governing equation

$$L_f = 0.235\,\dot{Q}^{2/5} - 1.02\,D$$

$\dot{Q}$ heat release rate (kW), $D$ effective pool/fuel bed diameter (m). Source: Heskestad (1983), SFPE Handbook.

Heat release rate Q̇ (kW) Fuel bed diameter D (m)

—

import numpy as np, matplotlib.pyplot as plt
Q = np.linspace(10, 10000, 200)
Lf = 0.235*Q**(2/5) - 1.02*D
plt.plot(Q, Lf); plt.scatter([Q_input],[Lf_input], color='red', zorder=3)

FIG. 05 — mean flame height Lf vs. HRR, current Q̇ marked in red

1.25 · Radiant heat flux & exposure distance

[1.25] Governing equation

$$q'' = \frac{\chi_r \dot{Q}}{4\pi R^2}$$

Point-source radiation model; $\chi_r$ radiative fraction (~0.3 typical), $R$ target distance. 12.5 kW/m² is a commonly used piloted-ignition threshold for exposed wood. Source: NFPA 80A.

FIG. 06 — radiant heat flux from an exposed burning facade decays with the square of separation distance

Heat release rate Q̇ (kW) Radiative fraction χr Target distance R (m)

—

import numpy as np, matplotlib.pyplot as plt
R = np.linspace(1, 30, 200)
qf = chi_r*Q/(4*np.pi*R**2)
plt.plot(R, qf); plt.axhline(12.5, ls='--'); plt.scatter([R_input],[qf_input], color='red', zorder=3)

FIG. 07 — incident heat flux vs. separation distance R, current R marked in red (dashed: 12.5 kW/m² threshold)