One of the possible modes of failure of a low pressure water-sealed gas holder is for the amount of water in a seal to reduce to such an extent that a release of natural gas to atmosphere is possible. A description of how the release of gas occurs during such a failure was presented in the accompanying part 1. A mathematical model to calculate the rate of gas release was also described in part 1. The purpose of part 2 of the paper is firstly to present the results of a series of wind tunnel experiments and field scale experiments that demonstrate how such releases would disperse if they were unignited. It is then shown how the results from these experiments can be used to guide the development of a mathematical model for the dispersion process. Two limiting case are identified, in which the gas is either trapped into the wake downwind of the holder or escapes completely from the wake to disperse as a buoyant, forced plume. Models for these cases are presented and it is shown how a composite model can be defined to produce predictions for the general case. The comparison of the model with the experimental data shows satisfactory agreement and the application of the model to an illustrative case is given.