If you are dead set on "never letting a drop pass your lips" then I might suggest you spray or drip a sample through a blue flame such as your galley.
If you get significant yellow sputters I'd say there was salt in it.
Best try it with a control also to see what colour you get.
(Hint petrol is NOT the best control /forums/images/graemlins/wink.gif )
Mike
Isn't this forum marvellous? You now have choices based on density, conductivity, evaporation, precipitation. There must be some optical quality that could exploited, and what about freezing point depression etc. For my part I too would taste it - perhaps boiling it first if the heads might have leaked.
It's going to be a cocktail, and will have rain-washed spray (salt) in it anyway, so chemical analysis will show salt. What's going to happen then? Paranoia? Empty it, and note the content (volume) after various activities : idle + rain, idle + dry, shaft turning etc.(log drips, saltwater cooling leaks) Where does your anchor locker drain to for instance? then fix the problem, but only if there is one. If you must, a beaker each of fresh, sea, and bilgewaters. your multimeter set to resistance, probes rubber-banded to a matchbox (to keep distance constant) dip in to compare resistance, you will see where your bilgewater falls between the fresh and salt. You will also be able to savour a delicate wine.....