An Alumina Supported Teflon Membrane-Based Oil-Gas Separation System for Dissolved Gas Extraction from Transformer Oil

Dissolved gas extraction from transformer oil is an indispensable part of dissolved gas analysis (DGA) for electric transformer fault diagnosis. Among of different extraction approaches, membrane-based oil-gas separation represents a promising option due to ease of operation and maintenance, high extraction efficiency, and cost competitiveness. Presently, Teflon AF2400 is the most popular material for synthesis of oil-gas separation membranes due to its excellent oil-resistant property, thermal stability, and gas permeability. However, the extraction time reported in the literature is not short enough to satisfy the requirement for the practical application, primarily resulting from the low gas permeance in the oil phase. Besides, long-term stability of membranes for the oil-gas separation process is another concern in the practical use. To improve the extraction efficiency and long-term stability, the ultrathin Teflon AF2400/γ-Al₂O₃ composite hollow fiber membrane has been developed using the sol-gel method and the dip-coating technique. The result shows that 2 hours of the extraction time can be achieved by our hollow fiber membranes. Furthermore, with the aid of annealing at 330^oC (higher than the Tg of Teflon AF2400), the long-term stability can be improved by strengthening the adhesive bonding between the Teflon AF2400 layer and the γ-Al₂O₃ layer, enabling the membrane to be used in the oil-gas separation process for at least 1,000 hours without any oil leak. Finally, it was found that the gas transport in oil is the rate-limiting step in the oil-gas separation process. Gas diffusivity in oil can be improved by increasing the liquid hourly space velocity, effectively enhancing the gas permeance in the oil phase. Nevertheless, the use of high liquid hourly space velocity has a risk of peeling off the Teflon AF2400 layer to reduce the long-term stability. Therefore, the liquid hourly space velocity should be optimized in order to achieve desired extraction efficiency and long-term stability.