Validation of an improved model for CO2 pipeline leakage and diffusion and prediction of hazard distance

In view of the shortcomings of the existing CO 2 pipeline leakage and diffusion models, such as large fluctuations in prediction errors, insufficient depiction of heavy gas effects, and poor adaptability to phase states, this paper constructs a phase-state adaptive leakage rate model, an SLAB model adapted to CO 2 properties, and an improved Gaussian plume model, achieving more accurate predictions of leakage rate and hazardous distance. Experimental validation showed that the average relative error of the leakage rate model in predicting density and leakage rate is respectively ≤ 5.20% and 6.90%. The prediction error under all verification conditions is ≤ 15%, which is higher than that of the existing models. The SLAB model based on CO 2 physical properties exhibited an average relative error of ≤ 14.90% in concentration-distance predictions, while the improved Gaussian plume model showed average relative errors of 15.20% and 12.40% for 4% medium-risk and 1% low-risk concentration-distance predictions, respectively. Under most operating conditions, the model prediction errors were ≤ 20%, outperforming traditional Gaussian plume models and the Phast model. Engineering case study results indicate that the dense-phase CO 2 leakage rate is the highest for the same leakage orifice diameter. For a 100 mm leakage orifice, the leakage rates for the dense-phase, supercritical, and gas-phase states are 500.39 kg/s, 370.47 kg/s, and 96.19 kg/s, respectively; For a 100 mm leak aperture, the maximum low-risk, medium-risk, and high-risk distances predictedby the diffusion model for dense-phase CO 2 reached 952.50 m, 342.10 m, and 138.70 m, respectively. These research findings provide reliable model data support for CO 2 pipeline safety design, risk assessment, and emergency response.