A novel, cost effective capacitive sensor for estimating dissolved moisture in transformer oil

The life of a transformer depends upon the condition of the insulating oil which is used inside it. One of the key factors that deteriorate its insulation strength is the presence of dissolved moisture, which eventually leads to the failure of transformer. In the past two decades more than 4% of the transformers failed due to this. Hence it will be very effective, if a sensor is used to predict the concentration of this moisture, real-time. In this work, the design of such a sensor is presented, which happens to be much cost-effective than all other similar sensors available in the market. The sensing mode used here is capacitive sensing in which the capacitance of an oil-cell changes with respect to the concentration of dissolved moisture present in the oil. In order to design this sensor, first the oil-cell was simulated in a FEA software. Then its associated signal conditioning circuitry was laid on board and the outputs obtained were verified using suitable simulations to ascertain its reliability. The conceptual prototype then developed demonstrated adequate sensitivity and hence found promising enough for replacing the existing costly moisture sensors available in the market.