I would not commit to saying more likely to be fresh water in...

I would not commit to saying more likely to be fresh water in land based wells. However, if the SP method is applied to estimate effective water resistivity, Rwe, then it will require a conductive medium (The SP value is taken from the shale body nearest the sand of interest I believe). Rather than play safe and decide to run a carbon oxygen log (generally employed in fresh water reservoirs), an NMR log should provide useful info such as which phases are present (oil, gas or water) and their relative amounts in the pore space, as well as how much water exists as clay bound water.