Carbon Capture and Storage (CCS) is of interest to the scientific community as a way of achieving significant global reduction of atmospheric CO2 emission in the medium term. CO2 would be transported from large emission points (e.g. coal fired power plants) to storage sites by surface/shallow high pressure pipelines. Modelling of CO2 atmospheric dispersion after leakages from transportation facilities will be required before starting large scale CCS projects. This paper deals with the evaluation of the atmospheric dispersion CFD tool Fluidyn-PANACHE against Prairie Grass and Kit Fox field experiments. A description of the models for turbulence generation and dissipation used (k–? and k–l) and a comparison with the Gaussian model ALOHA for both field experiments are also outlined. The main outcome of this work puts PANACHE among the “fit-for-purpose” models, respecting all the prerequisites stated by Hanna et al. [Hanna, S.R., Chang, J.C. and Strimaitis, D.G., 1993. Hazardous gas model evaluation with field observations. Atmospheric Environment, 27, 2265–2285] for the evaluation of atmospheric dispersion model performance. The average under-prediction has been ascribed to the usage of mean wind speed and direction, which is characteristic of all CFD models. The authors suggest a modification of performance ranges for model acceptability measures, within the field of high pressure CO2 transportation risk assessment, with the aim of accounting for the overall simplification induced by the usage of constant wind speed and direction within CFD atmospheric dispersion models.