Investigation of the exploitation of a fast-sampling single gamma densitometer and pattern recognition to resolve the superficial phase velocities and liquid phase water cut of vertically upward multiphase flows

The application of a single gamma densitometer unit to determine both the phase volume fractions and velocities to yield the individual phase flow rates of vertically upward multiphase flows was investigated. A fast-sampling (250 Hz) gamma densitometer was installed at the top of the 10.5 m high, 108.2 mm internal diameter, stainless-steel catenary riser in the Cranfield University multiphase flow test facility. Gamma radiation attenuation data was collected for the caesium-137 radioisotope-based densitometer for a range of air, water and oil flow mixtures spanning the facility’s delivery range. Investigation of the gamma densitometer signal response identified the presence of quasi-periodic waveforms in the time-varying multiphase flow densities. Classification of the multiphase mixtures’ flow regimes was investigated through the analysis of the probability distributions from the gamma densitometer data and self-organising feature maps. Two pattern recognition models were developed: a single feedforward multilayer perceptron and a multilayer hierarchical flow regime specific model. The neural network models were exploited to extract the individual phase flow rates of the multiphase mixtures from the quasi-periodic density waveforms inferred by the gamma densitometer measurements. Initial results yielded superficial phase velocity and water cut measurements to within ±10% based on flow regime specific correlations.