Energy harvesting is often the only way to get power for some applications, but it is often easier in concept than implementation. Regardless, the benefits may still far outweigh the challenges, as shown by this short item in NASA Tech Briefs, "Submerged AUV Charging Station." [Note: an AUV is an Autonomous Underwater Vehicle, not to be confused with the better-known Unmanned Aerial Vehicle, or UAV.]
Obviously, replacing batteries in a sea-going data-acquisition unit is often impractical. One harvesting solution is recharging based on solar radiation, but this would be a very "iffy" proposition for many obvious reasons (although it has been done). But there is another possible energy source: the well-known differential in water temperature at different depths.
The problem is that this gradient is fairly modest - about 10°C - and extracting the potential energy from such a small ΔT is hard, but it has been done. But how do you exploit this very modest difference? How do you translate it into stored energy, with acceptable efficiency? What simple and reliable technique can you develop? All good questions, that's for sure. Obviously, whatever you do has to be both reasonably efficient and very reliable.
The approach outlined in the brief article is based on phase-change material (PCM) which melts and expands at warmer temperature, and freezes and contracts at lower temperature. This material is placed in a container filled with hydraulic oil, so as the PCM expands, it pushes the oil out into another container, eventually reaching pressures as high as 3000 psi (about 21 MPa). A valve then opens and lets the pressurized oil go out, and the expelled oil turns a generator, which in turn charges a battery. The design also has a system with valves for resetting and venting, so the process can start again as the AUV cycles up and down through the water and accompanying temperature gradient. (In some ways, the transformation from low ΔT to much-higher pressure is analogous to using a charge pump to increase the DC voltage, so you can achieve greater energy-conversion efficiencies.)
The piece in NASA Tech Briefs lacks details and images, but it does link to a pdf of the original technical paper (you have to register, but it is free). It was interesting to see block-diagram schematic details of the setup as well as some numbers on performance parameters. I also found more details at the Jet Propulsion Laboratory site, here;
This is a case where a clever yet simple design is able to exploit some energy that's freely available to harvest, but with major challenges. Think of it when you hear people (especially those know-it-all experts) who don't understand the reality of energy harvesting, but use the term easily and casually as they proclaim, "It's there, it's free, what's the big deal?"
[Incidentally, if you are not familiar with phase-change materials, they are worth learning more about; they can help solve some tough problems. Several years ago, I interviewed the team at Tufts University which designed the wet-chemistry lab for the Mars Rover (see "Mars lander's chem lab is NASA's MECA"), which landed successfully in 2008. One of the requirements was to open a small drawer to accept samples to be mixed; the drawer had to open just once. While the obvious solution is a small motor or solenoid, that wasn't the best choice in terms of weight, reliability, surviving a six-month trip through the cold and vacuum of space, and other factors. Instead, they used a piston in a cylinder filled with paraffin; when an electrical coil heated the paraffin, it expanded and pushed the piston out to open the drawer ... and it worked perfectly! ]
Have you ever faced an energy-harvesting challenge that was especially difficult? How did you resolve it?