AbstractThe Tanzania Gas Project aims to exploit reserves located offshore from Tanzania in East Africa. The project faces challenges in the management of liquid content due to deep waters, rough seabed terrain, long transport lines to shore, relatively steep inclinations, and very dry reservoir fluids. The narrow operational envelope associated with the water depth underlines the importance of accurate flow simulations for design and production. In addition, the low liquid loading conditions are expected to result in substantial liquid accumulation in the upwardly inclined sections of the pipeline for low production rates. A large-scale experimental campaign was launched to reduce the uncertainty in the field development. A novel experimental "screening technique", allowed for the sampling of an unprecedented number of flow rate combinations corresponding to the onset of liquid accumulation. Diameter scaling was addressed by conducting similar experiments in 8- and 12-in. pipes. Froude number similarity was utilized for scale-up and to assess model predictions for field conditions. The data confirmed That the flow model captures the correct physics, and allowed for fine tuning. The updated model was applied in an uncertainty analysis for the Tanzania field, based on a large number of combinations of key input and flow model parameters, sampled from estimated uncertainty distributions. Simulation results for gas production rate, minimum turndown point, etc. were determined as probability distributions. The effort to quantify and reduce uncertainty has been very successful. Engagement of the operating company with experimental researchers, model developers, and software suppliers greatly increased the understanding of the physics and diameter scaling of low liquid loading flows, significantly reducing the uncertainty for gas condensate field developments.