Flammable Vapor Storage: 2D Concentration Field & Hot-Spot Risk (2D)

Storage Room Conditions

Drum leak / evaporation severity 18 g/min

Ambient temperature 22 °C

Extraction ventilation 6 ACH

Vapor Concentration Field

— SAFE —

Peak (worst pocket)

0 ppm

Peak %LEL

Room-average

0 ppm

Room-avg %LEL

Near-vent conc.

0 ppm

Hot-spot ratio

1.0×

3D lumped model's steady-state estimate (G/Q, well-mixed)

0 ppm

How it works (2D model)

The 3D tunnel version treats the whole room as one instantly well-mixed box: dC/dt = G/V − (Q/V)·C. That is a good approximation once a fan has had time to stir the whole volume, but real leaks form a plume that is far more concentrated right next to the drum than the room average — and a spark only needs one connected pocket inside the flammable range to ignite. This 2D companion drops the well-mixed assumption and instead solves the real 2D diffusion–advection–reaction equation across the room's floor plan on a finite-difference grid:

∂C/∂t = D·∇²C − k_vent(x,y)·C + s(x,y)
s(x,y)      = G / (A_src·H)   inside the drum's footprint, else 0
k_vent(x,y) = Q / (A_vent·H)  inside the extractor's capture zone, else 0
D  = turbulent eddy diffusivity of the room's air (≈0.2–0.7 m²/min)

G(T) and Q(ACH) use exactly the same evaporation and ventilation formulas as the 3D model (G scales with vapor pressure via exp(0.045·(T−20)), Q = ACH·V/60), so both simulators agree on how much vapor enters and leaves the room overall. The 2D grid just resolves where it sits at any instant. A conservation check: at true steady state the total sink flux integrated over the vent's capture zone must equal the total source G exactly — no matter how the diffusion coefficient is tuned — so the near-vent concentration always converges toward the same G/Q value the 3D lumped model predicts, even though the field in between is far from uniform.

Peak (worst pocket) — the hottest grid cell, right by the leaking drum. This is what a spark near the source actually sees, and it is routinely 2–8× the room average, especially at low ventilation.
Room-average — the same quantity the 3D lumped model reports as its single number.
Near-vent concentration — averaged over the extractor's capture zone; converges to the 3D model's analytic G/Q estimate as the field settles, confirming the two models agree on the whole-room mass balance.
Ignition test samples the peak pocket, not the room average — a flash fire only needs one small region inside the LEL–UEL window, and burns off just that connected pocket, leaving the rest of the field intact.

Drum / leak source Extraction vent Vapor field (color = %LEL) Peak pocket marker