Updated and re-edited video showing installation of a pico-hydropower generator to power the instruments measuring sediment transport in the Porter Creek Project.
The video shows the tail-end of installation of a pico-hydropower generator to supply power for the Porter Creek Project by Jon Sanfilippo (MS student in Water Resources Engineering) and Stephen Lancaster (Assoc. Prof., College of Earth, Ocean, and Atmospheric Sciences), both of Oregon State University. Supporters of the Porter Creek Project include the Siuslaw National Forest, the National Science Foundation, and OSU's Institute for Water and Watersheds.
In the Porter Creek Project, we are studying hydraulics and sediment transport in a gravel-bed stream in the Oregon Coast Range. Like any field study site, the location is off the grid. Some of the instrumentation can run just fine for a year or more on its own internal batteries, but other instruments draw a relatively large amount of power. In particular, we have an antenna array for detecting the movement of gravel embedded with passive integrated transponder (PIT) tags, and these antennas draw about 1/2 amp at 12 volts. Okay, that's less than 10 watts, but it adds up. We have been powering the array with 12-volt deep-cycle batteries and visiting the site every two weeks to recharge them with a gasoline-powered generator. This system has two big problems. First, our not-so-smart charger has to be continually reset about every 10 minutes, and the constant babysitting of the recharging process makes getting any other work done nearly impossible. Second, even deep-cycle batteries don't hold up well to repeated draw-down, and the batteries have stopped holding a charge.
Many remote data collection stations use solar panels. While we're not absolutely sure that wouldn't work at our site, the site is at the bottom of a deep, steep-walled valley in the Oregon Coast Range, in a seasonal rain forest. The most important times for data collection are when it's raining the hardest, and rainy conditions can last for weeks at a time. These do not seem like ideal conditions for solar power. However, there is a steep perennial stream next to the site, and this stream provides abundant energy. We discovered a "recipe" online (http://www.five-gallon-bucket-hydroelectric.org/) for a generator specifically designed to be relatively cheap and well suited for relatively small power demands (less than 1 kilowatt) and available power in the form of streams or canals. In our case, the stream is so steep that we required relatively little piping to get enough head to power the generator, roughly 120 meters of piping to get 40 m of head. The most expensive single item is the permanent magnet alternator, a little over $200.
Update: We went back to the site to install fuses and bypass switches (to be sure the batteries won't explode) and also to take some measurements with a real ammeter (one that encircles the wire without making conductive contact). It's producing 0.7 to 0.8 amps. Another problem: we need another generator to power the two logging multiplexers, because they interfere with one another when connected to the same power source. The plan is to stick a Y- or T-joint onto the 2-inch pipe and have the two generators side by side. If we can get around the same amount of current from each of two generators as we're getting from one now, then that should be ample. Otherwise, there are things we could do to improve the efficiency. For example, rough calculations indicate the ratio of turbine velocity and jet velocity is 0.6, and optimum is 0.46. To get there with the current setup, we'd need to reduce the diameter of the nozzles to increase the jet velocity, but that might actually result in a loss in total power, because smaller-diameter nozzles would produce less flow, and ultimately the power produced is proportional to the flow rate. Alternatively, if we could somehow increase the load, that would increase the resistance felt by the rotor and decrease the turbine velocity. How to do that? Short the circuit to ground!
See the old version: http://youtu.be/6rkqo06V9Ck
New video about "system 2.0": https://youtu.be/O-TAt6ekFck