We compare three atmospheric dispersion modeling approaches for a chlorine release scenario in an urban environment based on Colorado Springs geometry: (1) ALOHA 5.4.7, a widely used emergency screening tool that invokes dense-gas modeling for chlorine; (2) an independent Gaussian plume baseline using Briggs–McElroy–Pooler dispersion coefficients under Pasquill–Gifford stability class D; and (3) OpenFOAM 12, a three-dimensional computational fluid dynamics solver applied to a 5.3-million-cell urban mesh. A continuous 1 kg/s chlorine release at 2 m height is simulated under 5 m/s westerly wind for 600 s, and concentrations are sampled at 21 receptor locations across three crosswind planes at 100, 300, and 1000 yards downwind. We report Spearman rank correlations for the spatial concentration pattern across methods: ρ = 0.877 (ALOHA vs. Gaussian, n = 12, p = 0.00018), ρ = 0.913 (ALOHA vs. OpenFOAM, n = 8, p = 0.0015), and ρ = 0.905 (Gaussian vs. OpenFOAM, n = 14, p < 0.00001). Because the OpenFOAM source term is confirmed to inject at 1 kg/s in calibrated units of kg/m3, a direct magnitude comparison is also possible: at 300 m downwind on the centerline, OpenFOAM predicts 38.8 mg/m3 versus the Gaussian baseline of 34.8 mg/m3 (ratio 1.12), while ALOHA predicts 209.6 mg/m3 (ratio 6.0× over Gaussian), consistent with dense-gas dispersion enhancement. At 100 m, the OpenFOAM concentration is negligible because the release ceased 60 s before the sampling time and the plume advected past, an effect that steady-state screening tools cannot capture. All comparison data, provenance artifacts, and the Gaussian implementation code are provided as a SHA-256-verified, audit-ready artifact bundle for full reproducibility.
Devin Peters (Sat,) studied this question.