In my experience monitoring water, small amounts of organics in the water, such as oils, do not affect its resistivity much. That is why we used chemical methods to remove oils before purifying the water.
In your case, since you mention crude oil, I suspect the water is not a know entity either. In other words, it does not appear you are mixing crude oil with a characterized sample of "reagent grade" water and monitoring the change. Therefore, I still think IR (or maybe Raman) is your best bet. Water does not have carbon-hydrogen and other organic bonds; oil does.
John