Carbon Monoxide in Silicone Transformers
Carbon monoxide will be a lot higher in a silicone transformer than a mineral oil-filled one. It is difficult to try to determine what is producing the CO. Is it coming from normal aging of oil or from deterioration of paper from a fault condition? The only solution is a furan analysis. If the CO content is greater than the IEEE limit of 3,000 ppm  (table 19, shown later in this manual), and the generation rate G1 is met or exceeded, a furan analysis is recommended with the annual DGA. If a thermal fault is occurring and is producing CO and small amounts of methane and hydrogen, the fault may be masked by the normal production of CO from the silicone oil itself. If the CO generation rate has greatly increased, along with other gases, it becomes obvious that a fault has occurred. The furan analysis can only tell if the paper is involved (being heated) in the fault.