Will this alternative energy idea work?

Nothing to see here. Storing water in reservoirs and letting it flow through turbines to get energy back out is super well known. It's just a big battery, though, and not a particularly efficient one. As a technique it's only used at large scales because 1) the efficiency of turbines is better at larger scales, 2) we don't have great ways to store super large amounts of energy. At small scales we have lots of options for storage, so we prefer solutions like batteries where the efficiencies of the chemical reaction are more favorable.

http://en.wikipedia.org/wiki/Micro_hydro. I saw pictures somewhere- here or reddit or somewhere else- where a guy documented how he built a little hydro installation for a house that was in the middle of the wilderness and thus grid electricity was not an option.

When amateurs look at alternative power, they ask "Can this be made to work?" And usually when the answer is "yes" they then ask, "Well, why aren't we doing it?" Engineers don't usually ask "Can this be made to work?" because usually the answer is obvious. Engineers ask "How much will it cost?" and "How big can it scale?" Usually the answer to "Why aren't we doing it?" is "It would cost too much" and/or "It can't scale up enough to make any difference." http://en.wikipedia.org/wiki/Electric_energy_consumption. That turns out to be an average of about 500 gigawatts. (It varies up and down by as much as 30%, but leave that for another day.) I generally don't take any "alternative" energy system seriously unless it can be scaled up enough to produce 10% of that, so our number to shoot for is 50 gigawatts. http://en.wikipedia.org/wiki/Ludington_Pumped_Storage_Power_Plant he referred to generates about 1.8 gigawatts when it's running full out. To reach 50 GW would require 28 such systems. If his proposed systems generated 2 MW, the target 50 GW would require 25,000 of them. Which is not practical. And they're not actually generating any power. They're storing power produced some other way, and wasting a big percentage of it due to unavoidable losses of various kinds. A better (i.e. cheaper) way to produce 50 GW is to build 25 2-GW coal burning generator plants.

The error lies in the units used in this set of calcs: (10[ft])(1200n[gpm])(.18)(.4) = (12000n)(.072) = 864n For 864n = 15000 Wh / day set n = 15000 / 864 = 17.361 g[al]. the 'n' is screwing up the units. The units on the calc looks like this: (W) = (ft)(gpm)(.18)(efficiency) to get the 15kWh in a day (15,000 W/ day) you need to continuously generate (15,000Wh/24h = 625 W). To generate 625W, plug it into the formula provided: 625W = 10ft (gpm) (.18)(1) - assume 100% efficiency for now. gpm = (625/10/.18) = 347gpm. So, you need 347gpm at 10ft of head to generate 625W, at 100% efficiency. So, for a day, you will generate 15,000Wh (625W * 24h = 15,000Wh). You will need (347gpm * 60min * 24h = 499,680 gal. at 10ft head. Note this is at 100% efficiency. (Quick check in metric units) P [kW]=Q[m3/hr] x p[kg/m3] x g[m/s2] x h [m] / 3.6x10^6 =78m3/hr * 990kg/m3 * 9.81m/s2*3.05m / 3.6*10^6 =0.64kW Pump power calc from http://www.engineeringtoolbox.com/pumps-power-d_505.html To answer the original question, sure it will work. Get a big pond.

What's the difference between this and conventional hydroelectricity? How doe the water get pumped from the lower portion of the system to the upper portion? How do you deal with evaporation? There doesn't seem to be anything unusually interesting or groundbreaking in this concept.

Another thing he doesn't pay much attention to is the "head". To make a water turbine work well, the pressure behind the water needs to be very great. When the pressure is low, the turbine efficiency suffers. When it comes to hydro, that's referred to as "head" and it means the vertical distance between the turbine and the top of water in the lake behind the dam. The Lundington system has 100 meters of head, and as such is pretty efficient. (60-100 meters is pretty typical for modern hydro generation.) For the kinds of small distributed systems he proposes, 100 meters of head would be very difficult to manage most of the time. If you reduce the head substantially, all the other calculations change for the worse.

Like dfriedman, I was also wondering what was "alternative" about this. The size?

These things work, but they DO have an environmental cost, and take up a huge area of land for the power provided (far more than wind or solar). Think of the environmental impact of one dam? well for the same power you now have twice the environmental impact and you STILL need the base load generating capacity. You do gain a little on how much base load capacity you need because you are shaving the peak and valley off the load with the storage, but you are just better off environmentally and cost wise to just build another gas power station or expand your grid to pick up that last wind farm or add tertiary capture off the existing coal or nuclear powered steam plant. Just think about it-if one reservoir is empty (and once a day if the system is designed right it will be) you have a huge stinky mud flat and the other one is full. Than vice-versa. This kind of cycling makes it impossible to establish any kind of ecosystem in either reservoir except for nasty algae that is probably toxic to boot. For a conventional hydroelectric dam you at least get some kind of ecology to replace whatever you flooded, and some mitigation for dry years as well to the downstream ecosystem (dams don't so much destroy ecosystems as radically change them-and not often in totally desirable ways). For residential systems you are best off with solar/grid tie and some kind of energy storage on site (flywheels would be way better than a big stinking algae pond out back). You can also pick up about a 20-30% gain from energy efficient measures (one of the few areas really, really ripe for conservation measures still in the US). Those roofs aren't being used for anything anyway and the average American house can fully supply its energy needs with 1/2 its roof area in solar panels (except at night-hence the flywheel/grid tie/windmill). Most peak loads happen during the day anyway so you can get the peak shaving benefit from this alone.

There are micro-hydro systems that, for suitable locations (not a trivial matter!) can provide power output of a level that can be useful to a household. I worked on doing a cost estimate for such a system as my senior design project in undergrad -- IIRC it ended up being something of a dubious proposition, payback-wise, and the place we were doing it at was basically a bed-and-breakfast on the side of a mountain (definitely not your typical suburban lot). It's been a long time since I was involved in this stuff, but as I recall the magazine "Home Power" has a lot of good information on alternative energy systems of this size. My guess, offhand, is that a system like this including the feature of pumping water back up for storage (storage of what energy? from where? where is "up"? how big does "up" have to be? how is the flow of water controlled? etc.) would be quite expensive, would probably run into some heinous water rights issues in Western states, and -- even more tentative guess -- would be hard to do efficiently. But a start would probably be to find a commercial micro-hydro system and figure out what volume of water has to go through it to store X-amount of energy, for a set of installation scenarios, and then start thinking about how to get a tank that big that high up.