I must admit i do not limit myself to a single test method, it depends upon the size of the system, is it windy, how bigs the leak (repeat calls), I have found the the fluro dye and UV light has been succesful on problem jobs