Salinity-independent measurement of void fraction with temperature variation consideration in homogeneous saltwater-gas flow by a sensor-ANN method

Measuring void fraction in fluids is an essential task that has attracted considerable attention from researchers over the years. Variations in salinity of water significantly affect the measurement accuracy of void fraction across different fluid types and flow regimes. This study aims to measure the void fraction of a two-phase water – gas fluid by considering temperature variation and independent of changes in salinity. To achieve this, a combination of a capacitance-based sensor and an Artificial Neural Network (ANN) is employed. The presented network takes two inputs, measured capacitance and temperature, and produces one output, volume fraction of the gas phase. The novelty of this metering system lies in its ability to estimate the void fraction of a two-phase homogeneous fluid by considering temperature changes and independent of salinity variations using a single non-invasive sensor. The system achieves a Mean Absolute Error (MAE) of 0.63 and 1.52 for the training and testing sets, respectively, demonstrating its accuracy and reliability.