Data fusion for measurement of water holdup in horizontal pipes by conductivity rings

This paper introduces a method of fusing measurement data of a set of conductivity ring sensors to determine the average water holdup of oil-water two-phase flow within a horizontal pipeline. A six-ring conductivity sensor was built with 1kHz sampling rate, which delivers the generalized dimensionless conductivity and predicts the average water holdup based on Maxwell equation. An adaptive weighting algorithm is then brought forward to fuse the dimensionless conductivity and Maxwell equation to optimize the overall prediction upon average water holdup. The results suggest that the data fusion method reduces average relative error with 5.9%.