Hydroelectricity times N

August 16, 2007 9:06 AM

Power generation: Let's say we have a hydroelectric dam. My understanding is that water turns a dynamo, which provides electricty. It seems like there would be a way to use the same water flow more than once. Could we make raceways of some indeterminate length to get the water back up to speed with sufficient force to turn a second dynamo? Could this be repeated so that a single water flow powers some n dynamos? What kind of constraints keep dams all over the world from doing this sort of thing?
posted by boo_radley to science & nature (22 comments total)

You have the pressure of all the dam water in the entire dam lake turning the dam dynamos. It's not speed at all. Notice the water level difference between the dam lake and the dam outlet.

If the dam river coming out were high enough in elevation, you could theoretically dam it again. I don't know if this is done anywhere or if appropriate geography exists.
posted by TheOnlyCoolTim at 9:14 AM on August 16, 2007

It is my understanding that the higher the water falls, the more power can be generated at the bottom. If you kept interrupting the flow of water it wouldn't gain the necessary momentum to turn those really massive magnets needed for power. In theory you could drop the water from the top of Everest to sea level and interrupt it many times, but how would you get the water to the top of Everest in the first place? So hydroelectric production is generally confined to places where water makes a natural drop, like Niagara Falls, which is only 200 ft or so but plenty high for generation.
posted by kuujjuarapik at 9:15 AM on August 16, 2007

Oh, wait, I see the gist of your question. The answer is probably that it would be uneconomic to build smaller and smaller generators down the raceway. Diminishing returns, higher investment costs, etc.
posted by kuujjuarapik at 9:17 AM on August 16, 2007

Given sufficient height, you could do exactly that. Many rivers have multiple dams on them, capturing electrical power as the river descends, sometimes every 100-200 feet in altitude.

See here for a cross-section diagram of a hydroelectric dam. Check explanation of potential energy there as well. Notice that the turbine is located near the base of the dam. This way, it benefits from the water pressure created by the full depth of the water backed up by the dam. A turbine placed halfway up would get only half that benefit and be able to generate half the power. Once the water passes through the turbine, the potential energy created by the head, or depth, is transferred into the generated electricity. The deeper the water, the greater the pressure, the faster the turbine will be driven. Certainly water exits the base of the dam with some force, which might turn a water wheel but would not generate nearly the amount of electricity as the turbine/dynamo does. The most efficient way to extract electricity from a river is to build a few large (tall) dams, rather than a lot of smaller ones.
posted by beagle at 9:19 AM on August 16, 2007

I suspect the turbines are high enough efficiency that there's not enough energy left in the water after one pass for it to be worth the bother.
posted by edd at 9:20 AM on August 16, 2007

I don't know if this is done anywhere or if appropriate geography exists.

The Savannah river (where I live) does exactly this; there are large hydroelectric dams for lakes Hartwell, Russell, and Clarks Hill, the Stevens Creek hydroelectric project, the diversion dam for the Augusta canal, and the lock and dam (for barge traffic that travelled the river in the past) in that order from the headwaters to just below Augusta. It doesn't answer your question exactly, but the Russell dam is equipped with reversible turbines which pump water back into lake Russell when demand is low and generate when demand is high.
posted by TedW at 9:20 AM on August 16, 2007

It's also done in Canada. Essentially the raceways you'd need to generate the power in the second set of turbines would look like the mechanism for getting the water into the first set of hydro turbines -- i.e. it would be a second dam plus big body of water downstream from the first.

I guess the space required to have multiple dams limits where this can be done -- the Canadian project I linked to above is supposedly the size of New York state.
posted by patricio at 9:35 AM on August 16, 2007

In Arkansas, the Ouachita River is dammed three times - Blakely Dam created Lake Ouachita, Carpenter Dam created Lake Hamilton, and Remmel Dam created Lake Catherine. They all generate electricity.

As to the multiple turbines issue, the total amount of energy available to be converted to electricity depends on the potential energy of the water. Potential energy is proportional to the height difference between the water levels. For example, with totally fake numbers: if you have a 100 meter difference between the lake level and the turbine, you have 100 J of energy available. A turbine is not a totally efficient machine and loses say 1 J of energy. So you get 99 J of electricity.

Now suppose you build two turbines - one 50 meters below the lake and one 100 meters below the lake. The upper turbine provides 49 J of electricity. The water then continues on to the bottom turbine, which also produces 49 J of electricity for a total of 98 J. Now you have twice the equipment costs and less electricity to sell!
posted by mbd1mbd1 at 9:36 AM on August 16, 2007 [2 favorites has favorites]

The energy that you can get out of a hydroelectric plant depends on the water pressure, which is a function of how much of a "head" of water you have behind the dam.

You can generate a lot more electricity if you have a big dam with 100' of water behind it, than if you have a little one with only 2'. The dam with a higher head of water will be under greater pressure, so it can turn the turbines faster, or turn bigger ones.

Generally, power-generation is something that scales well. So it's a better idea to build one big dam than a whole bunch of little tiny ones, each with their own generators. (There are exceptions to this, and it only scales well to a certain point, of course. But you don't build 10 1-foot dams when you could build 1 10-foot one.*)

Also, a big dam allows you to have more control over the water -- you won't "run out" of water as easily during a time of little upstream rainfall. You can use the big lake behind the dam as a battery or reserve.

But there are many rivers that have multiple dams on them. I think the Columbia River in the Pacific NW is like this; as it comes down out of the mountains it has a series of power dams. They're usually built in places where they can build up a decent head of water behind the dam, without flooding a really giant area of land.

* If your goal is to generate power. There are lots of other reasons why people build dams, and if you don't care about making electricity, small ones may be advantageous because they raise the water level less and thus flood less land.
posted by Kadin2048 at 9:56 AM on August 16, 2007

TedW, why do they bother pumping water back into Lake Russell, instead of just closing the gates and leaving it in there in periods when don't need the electricity?
posted by beagle at 9:59 AM on August 16, 2007

There's a river near me that has multiple dams. This tends to only work on smaller rivers with smaller turbines (the technical term for "turbine," because when you're generating massive amounts of power with massive dams the issues described upthread occur.
posted by croutonsupafreak at 10:09 AM on August 16, 2007

If the dam river coming out were high enough in elevation, you could theoretically dam it again. I don't know if this is done anywhere or if appropriate geography exists.

In addition to the others listed, the Colorado. If you wanted, you could view the Grand Canyon as a very long raceway connecting the Glen Canyon and Hoover dams.
posted by ROU_Xenophobe at 10:09 AM on August 16, 2007

So it's down to water pressure, then. What kind of volume does a hydroelectric (HE) dam need to operate? Or, more precisely, I guess the question is: what pressure does a certain volume of water need to generate some amount of power, say, 1KWH?
posted by boo_radley at 10:12 AM on August 16, 2007

beagle: it's acting as a capacitor, taking energy from other power stations when not all of it is needed, and storing it for later.
posted by edd at 10:29 AM on August 16, 2007

In hydro systesm, Head (in feet) times Flow (gallons per minute) times .10 (coefficient/efficiency) approximately equals Watts generated.
posted by glibhamdreck at 10:52 AM on August 16, 2007 [1 favorite has favorites]

There are numerous streams on the western slope of the Sierra Nevada range in California that have a series of hydro-power dams in them. Pacific Gas and Electric (PG&E) owns and operates the majority of them. Elevation's are sufficient for the energy of the same water to generate electricity at several points as it flows downslope.
posted by X4ster at 10:56 AM on August 16, 2007

I think you're suggesting that the water that exits the dynamo should be collected in a raceway, routed back around and sent back through the same dynamo again.

But to get the water around AND UP to the top of the dynamo takes as much energy as the dynamo puts out (plus more for friction, etc.).

This is simply a form of perpetual motion, forbidden by the Laws of Thermodynamics. Sorry about that.
posted by KRS at 10:56 AM on August 16, 2007

what pressure does a certain volume of water need to generate some amount of power, say, 1KWH?

Hydroelectric dams have basically two variables in terms of their generating capacity: flow (how much water is falling) and head (from how high is it falling). To generate 1 KW* of power you need 1 KJ of energy per second. Assuming 100% efficiency, we can assume all the potential energy is converted into electrical energy, so Energy = Mass*Height*g. 1 KJ is ~102Kg of water, falling 1m or 1kg of water falling ~102m (or whatever other combination you'd like). So 1 KJ per second (or 1 KW) is ~102L of water falling 1m per second. If that water falls for an hour, you have 1KWh.

A good setup will be around 0.9 efficiency, so make that ~113L of water per second. At least in Canada, cubic meters per second (cumecs) is the usual unit of measurement, so you can say ~0.113 cumecs.

The actual pressure at the turbine is secondary.

* KiloWatts because we are talking about power, not energy (KWh).
posted by ssg at 11:00 AM on August 16, 2007 [3 favorites has favorites]

why do they bother pumping water back into Lake Russell, instead of just closing the gates and leaving it in there in periods when don't need the electricity?

It's the same idea, it just puts even more water in Lake Russell for later use. The water is taken from the upper reaches of Clarks Hill and supposedly has a negligible impact on that lake's levels. They do kill a lot of fish and have generated some controversy for that reason. It must be pretty inefficient but it is more about timing generation rather than increasing efficiency.
posted by TedW at 11:27 AM on August 16, 2007

KRS: yes, I know that's impossible, and it's not quite what I asked. Thanks, though.
posted by boo_radley at 12:22 PM on August 16, 2007

Beagle: why do they bother pumping water back into Lake Russell, instead of just closing the gates and leaving it in there in periods when don't need the electricity?

The Savannah has a series of dams as the OP suggested. Lake Russell is between Lake Harwell to the north and Lake Thurman to the south. What they do is use electricity to pump water back up from the lower lake to the upper lake during the night to be used for peak loads required during the day. Since the generators use more water than naturally flows in from the north, the lake would eventually be drained if they didn't reuse the water.

The key to making this work is that they use electricity from other sources to run the pumps, i.e. coal and nuclear power plants. Those types of plants work most efficiently when running at full capacity 24 hours a day. But at night there is less use for that electricity so the most convenient way to store the excess energy is by pumping water uphill at night. The Lake Russell generators only run during the day to help supply power for peak loads. This scheme is only around 60% to 70% efficient overall, but that is still cheaper than building a larger coal plant to handle the daytime peaks.
posted by JackFlash at 12:45 PM on August 16, 2007

Yet another example of multiple dams is the Columbia River, particularly the sequence of the John Day dam, the Dalles dam, and Bonneville dam.
posted by Steven C. Den Beste at 1:04 PM on August 16, 2007