Underwater Power Generation?

An anonymous reader writes: "We keep seeing stories here about tidal power, and that's cool, but I don't see how it can be done without a column rising all the way to the surface. So here are the ideas I've got right now." Read on for some interesting thoughts on the subject...

"Keep in mind that the device will probably be housed in a length of 4-inch PVC or ABS pipe, and it needs about 0.5 ma at 1.5 volts:

• Surge power. Put a couple of funnels back-to-back with a CPU cooling fan-sized turbine and generator in the middle, and run the output through a rectifier and capacitor. But how reliable will those moving parts be after years underwater?
• Self-winding watch concept. Float the thing tethered to the bottom and install some sort of pendulum inside with a magnet on it, moving through a coil. The moving parts are protected, but will it be enough power?
• Yank the chain. Again, tether it, but use the varying tension on the tether to drive a dynamo of some sort. Not sure how this would work.
• Magnetohydrodynamic generator. Like the surge power thing, but using the flow of cunductive seawater through a magnetic field to generate a current. I have no idea how much power this would generate, if any, or how to deal with ion accumulation at the electrodes.
• Nukes. Anyone got a spare radioisotope thermoelectric generator? Any idea how many smoke detectors I'd need to cannibalize to get enough Americium-241?

The generator need not fit inside the 4-inch cache tube, but it shouldn't be huge, either. It needs to be practical to build, and not terribly expensive. Above all it's got to be reliable and enduring. Any ideas?"

Comments

Differentials

[Nyarly]

I wonder if the device couldn't be placed with an end on either end of a thermocline.

The idea would be to use the temperature differential (which can be significant - 10 degrees Faranheit isn't unusual) to generate current. A decent thermocouple would be able to do this without too much difficulty.

The difficulty, of course, is getting the thermocouple on either side of the thermocline. If the device can be of sufficient length (describing it as housed in a pipe suggests as much) this shouldn't be an issue. However, remaining on the thermocline could be difficult. Perhaps some sort of diaphram that would maintain position at the boundary? Tricky.

Re:6 watt hours per year.

[Rorschach1]

If the /. 'editors' hadn't mangled my submission, you'd have seen that I already considered that. Here's the full thing:

Ok, so I've gotten into this geocaching thing lately, and while working on a cache to be hidden in about 60 feet of water off the coast, it occured to me that a blinking LED might make it easier for divers to spot. No problem, whip up a blinker circuit with an LM3909 and a super-bright green LED and we're set. But what about power? Sure, four D-cells would let it run for close to a decade, but where's the fun in that? The undersea environment is quite dynamic, and there's got to be some power down there that can be harnessed. What I need are some ideas on how to do that.

We keep seeing stories here about tidal power, and that's cool, but I don't see how it can be done without a column rising all the way to the surface. So here are the ideas I've got right now. Keep in mind that the device will probably be housed in a length of 4-inch PVC or ABS pipe, and it needs about 0.5 ma at 1.5 volts:

• Surge power. Put a couple of funnels back-to-back with a CPU cooling fan-sized turbine and generator in the middle, and run the output through a rectifier and capacitor. But how reliable will those moving parts be after years underwater?
• Self-winding watch concept. Float the thing tethered to the bottom and install some sort of pendulum inside with a magnet on it, moving through a coil. The moving parts are protected, but will it be enough power?
• Yank the chain. Again, tether it, but use the varying tension on the tether to drive a dynamo of some sort.
• Nukes. Anyone got a spare radioisotope thermoelectric generator? Any idea how many smoke detectors I'd need to cannibalize to get enough Americium-241?
• Magnetohydrodynamic generator. Like the surge power thing, but using the flow of cunductive seawater through a magnetic field to generate a current. I have no idea how much power this would generate, if any, or how to deal with ion accumulation at the electrodes.

The generator need not fit inside the 4-inch cache tube, but it shouldn't be huge, either. It needs to be practical to build, and not terribly expensive. Above all it's got to be reliable and enduring. Any ideas?