Contrera, I seem to recall that NASA determined that 50% mix is the threshold whereby EAN starts to behave like pure oxygen. I don't know of any compressor tests specifically. The 40% number for compressors, etc is, like all risk assessment, based on statistics. A lot of compressor simulator testing would have to be done to build a statistical model and that would be expensive. Well, US Divers and some other companies subjected their regulators to some extremely testing and claimed that they proved that a NITROX reg must be O2 cleaned. However, that was a hail Mary pass which would be difficult to duplicate, IMO, and provided no statistical probabilities. The best source of statistics are the divers themselves who gradually evolved a working knowledge and used this in their own problem management. The first brave soul who pumped 40% EAN survived and others followed. There are the Lindbergs and there are the Earharts I know a guy who insisted he was going to use a compressor to jam a welder supply tank with oxygen. On his first (and last) test run he hid behind a barracade. It was fortunate that he did so.