Energy Evaluation of Differential Pressure Signal and its Relationship with Flow Pattern of Two-phase Flow

A new threshold probability statistical technique based on energy evaluation was proposed to explore the relationship of the flow pattern of oil-gas two-phase flow with the differential pressure signal that was a representative non-stationary signal. The differential pressure signal was firstly decomposed into a series of IMFs (intrinsic mode functions) by EMD (empirical mode decomposition) according to different frequency scales. The normalized energies of IMFs were evaluated. The energy variation in beta-band (5-30Hz) was noticeable when the flow pattern was changed. A computable way of probability statistical technique was provided to identify the flow pattern of the oil-gas two-phase flow by the scan of time-window. Application of energy evaluation to the oil-gas two-phase flow is discussed and results show the proposed technique can be a good alternative for the flow pattern identification of two-phase flow