(32d) : Characterization of Confidence in Multiphase Flow Predictions

Confidence in multiphase flow predictions can influence key decisions on design, location, and operations of pipelines and separation systems.  These decisions in turn can profoundly affect system reliability, operability, safety, and overall project economics.   There may be no application where lack of confidence in predictions is felt more strongly than for long distance tiebacks of pipelines to subsea processing in deep water.   The uncertainties and resultant confidence in simulator outputs of flow regime, pressure drop, entrainment, and insitu velocities dictate the estimated confidence levels for opposing and complimentary systems and are key contributors in a Decision Analysis exercise.

Uncertainties in benchmarking data, input data to simulator and of internal models for various phenomena propagate to the simulator outputs of flow regime, pressure drop, entrainment, and insitu velocities.   How these discrepancies are treated will affect the scaling of the simulator to different conditions where benchmarking data may not exist.

This paper aims to raise the awareness of the importance and value of uncertainty analysis in multiphase flow systems, provide a methodology with examples, and identify some of key gaps in the uncertainty analysis.