Abstract
The predictive power of a two-phase flow design method, consisting of one-dimensional flow models depending both on flow regime and pipe inclination, has been investigated by comparing predicted pressure losses and liquid hold-ups with pipeline data collected under actual conditions. For this purpose, Shell’s database of controlled field tests in large-size gas pipelines was extended with data obtained in an 8-in. high-pressure two-phase flow test facility with gas and condensate as flowing media. It was concluded that the predictive power was reasonable for wide ranges of flow conditions, but that liquid entrainment is to be incorporated in models to remove conservatism from liquid hold-up predictions.
