Oh, interesting. So that would preclude open slides, but doesn't rule out insulated flex, or sealed slides.
(For example, a two-lip rubber seal pinches closed around each electrical rail, insulating them from the salt water. Where the sliding contact enters, the rubber is pushed away a bit and some leakage current flows, but only around the contact, not the full length. Low frequency AC is used to avoid electrolysis, and inert materials are used besides (stainless, monel?). This might be a product that exists already, but I'll admit it would be pretty expensive to produce, from scratch, in continuous lengths!)
How conductive of a salt are we talking here? You may be limited on transformer size because the salt will make a big shorted turn.
How long is the axis?
I could certainly give you a fast and wrong answer. 12V RMS sine, 1kHz, 30 turns, Ae = 300 mm^2, 3C90 ferrite, checks out. But then, you'll need some pretty damn thick wire if the path is "really long". And higher frequencies and voltages may not help you due to induction losses along the wire pair and salt water.
How about shaking the thing back and forth to keep a turbine spinning?...
Tim