# The Myth of Alternative Power and Hydroelectric Storage

Every time I get into a debate about “alternative” energy I point out it can’t be used for baseline power because it can’t provide reliable power, and it can’t provide reliable power because you can’t store the electricity that it episodically generates.

Immediately, someone will say, “We can use hydraulic storage!”

Hydraulic storage is basically a hydroelectric dam on a small or large scale, except instead of using water brought by a watershed, the water is pumped up behind the dam with pumps powered by the generator whose energy output you want to store. For example, you would have electric pumps powered by solar panels or wind turbines, the idea being that when the wind or cloud-free days produced a surplus of power (or you built in surplus capacity) the pumps would pump water from a lower reservoir uphill into a higher storage reservoir. The electricity would be stored as the potential energy in the elevated water. When you needed the power back, you would drain the water back downhill through turbines just like a hydroelectric damn.

Now, this certainly works and it has been done on a small scale. However, it will never, ever be a real-world, large-scale solution that can make alternative power work.

Why? Well, let’s just do some back-of-the-envelope calculations.

(Note: Below when I say “conservative assumption” I mean an assumption biased in favor of alternative power.)

Let’s say we want alternative power to produce just 30% of our current baseline power needs. Let’s make a very conservative estimate that we only need to have a 25% stored reserve. (This very optimistically assumes that alternative power will otherwise be able to provide sufficient power, when it’s needed, 75% of the time.) So, 25% of 30% equals 7.5 % of the total (0.30*0.25=0.075).

By happenstance, hydroelectric power today produces 8% of the nation’s electricity. This means that if we want to use hydraulic storage to make alternative power work, we will have to recreate the 93% of the generating capacity of every currently existing hydroelectric facility in this country. That’s right, Hoover Dam, the entire TVA, all the damns in the Rock Mountains, all the rest, all duplicated.

That alone raises the question: Where the hell are we going to put these dang dams anyway? All the places with the geography and the water supply to produce hydroelectric power are already in use. Worse, all the places that produce significant amounts of solar and wind power are simultaneously the worst places to build hydroelectric facilities. Solar energy is produced most abundantly in, wait for it, the desert, and wind power is produced most abundantly in very, very flat places. So any hydroelectric storage facilities will have to be a long, long way off from the point of generation.

[Update: 2010-6-19-8:43am Joseph Somsel, provides an article that says that the very best hydroelectric storage produces 75% effeciency. However, these plants are massive and have to be located in specific terrain.

The highest differential between high and low reservoirs in the U.S. is the Helms pumped storage facility. This 1,050 MW capacity installation, located in the Sierra Nevada Mountains in Northern California has a 1,630 foot elevation difference between reservoirs connected through an underground equipment hall carved out of solid granite.

We aren’t going to building a lot of those.

My figures below are for the much less efficient low-head generators, the ones that don’t require mountains to work. More importantly, it really doesn’t matter how efficient the storage is. Even if hydroelectric storage was 100% efficient we would still need to duplicate every existing hydroelectric dam in North America in addition to building all the alternative power generators. All that to avoid a building a few nuclear plants.
End Update]

It gets even sillier. Hydroelectric storage is only around 25% efficient. This mean that to get 1 watt back out of the system you have to put 4 watts in. This in turn means that in order to create a 25% energy store of 30% of total power, you have to actually generate 30% of total power just to get 8% of total power back out. All that in addition to 30% of the total power that goes straight into the grid.

So, to get 30% of total power from an alternative power system you actually have to build the generating capacity to supply 60% 40% of total power! Half of the alternative power will go into immediate consumption and half will be stored.

On a good day.

In the best-case scenario.

Even if alternative power’s cost is equivalent to nuclear’s at the point of generation, once you double the needed capacity to make it reliable and then build the equivalent of every existing hydroelectric plant in the country, you’re talking about a system 3-4 times more costly per watt at the point of consumption. Even if you assume that the point-of-generation cost of alternative power is half of that of nuclear, it still ends up costing twice as much at the point of consumption.

All the proposed energy storage schemes that are utterly necessary to give alternative power a hope of working have the same type of scale problem. Yeah, they might work for one small-scale facility, but when you start to scale them up to the size and geographic distribution that you actually need for them to provide baseline power, they become ridiculous.

It doesn’t even help if you use several different technologies. Hydroelectric store is the best possible large-scale technology. The others are even more less efficient, more costly and more difficult to implement on a large scale.

Scale matters. The sad thing is, our political class doesn’t understand this so now we’re seriously trying to base our long-term energy generation on a hopeless technology.

Honestly, the stupid and expensive things that we try in order to avoid building nuclear plants never cease to amaze me.

### 18 thoughts on “The Myth of Alternative Power and Hydroelectric Storage”

1. Now, this certainly works and it has been done on a small scale.

I remember watching a doc (didn’t read it in any of the literature when I was curious about it) about grand coolie.

During the day at peak hours, they produce energy, then at night, they move water to an irrigation reservoir (apparently the designers intention was to turn eastern washington into a green field of biomass.

I can see this making sense, in a small way, simply to maintain constant action in the generators, which happens to be quite important for most mechanical devices. So Coolie was meant to designed to perform one task, while also serving another. That’s just good engineering.

But your 4-1 ratio is EXTREMELY NICE. Not only would it require increased capacity, it is also a peter and paul situation.

Wind and solar (I don’t know if solar has variable output issues in maintenance, but wind definitely does) So you actually increase the consistent output issues you end up with every mechanical device.

All of a sudden a hydro plant has to ramp up it’s production, when wind no longer can, and the dynamics of stuff that is made of stuff, that moves, changes when you have variables of basic values. Optimum is no longer the standard of operation, but you have to move from optimum to nominal, to maximum, totally effing over very large, very expensive machines that generate energy.

It’s a concept that can be used on a small scale as you say, but it’s also more likely that you will overload mass spillways, clear nutrient rich sediment, and destroy biomass.

Not to mention I lived next to a wind farm for 3 years while I was in The Corps. I think the wind farms are pretty, I really do, as the blades turn to face the wind, and rotate in coordination, it’s like a ballet, only in a not gay heavy duty I can fix anything guy sort of way. but I would say I’ve seen about 1/6 sitting silent, either cuz peak had been met, and the control station shut down the generators, or because they broke down ( I would say most of them broke down, based on what I was offered to work on them when I got out) And Solar, at least terrestrial solar is already a losing game.

I MUST admit, that in certain locations, I think geothermal looks interesting, but with our increasing energy demands, make me have a real “the world will end” thought.

2. You have a major error in your calculations. The pumped storage cycle for hydro actually requires that you input 25 to 28 % more energy in than you get out. Hydro turbines are the most efficient energy machines in the world, current technology converts 93 to 95 % of the potential energy at a dam site into actual electricity. Pumped storage projects are not “small scale”, there are several projects in the US that produce over 1000 MW of power (the biggest coal powered plant produces around 4000 MW of power).

The premise for economical pumped storage is that you pump with off-peak power (cheaper) and generate at on-peak power (most expensive). As long as you can pump with enough cheap energy to exceed the amount of high cost energy for generation, you can make money.

Now the disconnect between where your solar or wind energy produces against where you could store off-peak energy with hydro is a very vexing problem, because you would have to build expensive transmission lines to carry the energy. Just like solar and wind power, pumped storage hydro can fill a niche need. None of these will ever provide enough for the mass energy needs of the country (at most around 10-15 % cumulative and currently only possible with massive subsidies).

For the forseeable future the only answer is the current energy regime, coal, nuclear, natural gas, and petroleum. Hydrogen, fuel cells, and fusion may hold the long term answer or even space based solar. There is where our money and research should be focused.

Wayne Elkins
Hydro Engineer

3. The other major error in your assumptions is that recreating the generating capacity of every existing hydro facility in the US today will require dams with the same footprint and environmental impact.

Not so. Most of these facilities have a storage capacity of months. A site designed for off-peak storage needs a capacity of only a few days. MWD, operator of the Colorado Aqueduct and wholesale supplier of water to the LA basin, has engineered a number of these facilities.

NIMBY is a greater barrier to this than technology.

4. “All the places with the geography and the water supply to produce hydroelectric power are already in use.”

I’ve heard this several times, but it doesn’t seem likely to me. I assume this statement excludes damable rivers in or near national parks (as in the sense, the reservoir would reach into a national park). I know that in the late 1960s, Texas planned to build a series of 30-year dams (described as those whose filling would take that long) to provide water and power for the state. The nascent environmental “movement” put an end to that. IMHO, the plan will and must be resuscitated.

5. Check out Racoon Mtn. TVA, Chattanooga, TN

6. City of Denver has been using Hydro storage on a large scale for over 30 years. In California the Helms project was built to balance the proposed 5 Unit Diablo Canyon Nukes. As only 2 were built it now also balances all inputs into PG & E’s Grid.
Storage efficiency is better than batteries and everything but flywheels. You can’t compare to ideal only to real world systems to be accurate. Doc Smith Super-Capacitors would be great but they don’t yet exist. Saving part of ‘free’ power is always better than throwing it all away.

7. Wayne Elkins,

Hydro turbines are the most efficient energy machines in the world, current technology converts 93 to 95% of the potential energy at a dam site into actual electricity.

(1) The turbines recovering the energy from falling water are very efficient, the pumps that convert electricity into the potential energy of elevated water are not. Remember, half of a hydroelectric storage system is moving the water uphill. For that you essentially have a duplicate set of turbines running backwards.

(2) The efficiency of hydro turbines is directly linked to the depth and mass of water contained behind the damn. To get the efficiency of Hoover Dam, you would have to recreate Hoover Dam itself (the reservoir could be smaller but the depth would have to be the same.) Plus, you would need two reservoirs, one low and one high, so you would need two dams to ensure you always had enough water.

(3) The geography problems means that you’re looking at least 5% loss through transmission from the generators to the storage facility. Then you have another 5% lost going from the facility to point of consumption.

(4) The unreliability of alternative power will affect the efficiency of the storage pumps as well. You’ll be trying to run massive pumps off a power supply that fluctuates. To smooth the power out, you have to use the storage’s own generators to buffer the incoming fluctuating electricity.

All the sources I have seen for large scale hydroelectric storage put total recover at around 20%-30% so I think using 25% as a benchmark is fair. If you have other data you can link to, I would be happy to see it.

However, I would point out that even with 100% efficiency, you are still talking about building the equivalent of every currently exiting hydroelectric plant in North America. All that in addition to the cost of the alternative plants themselves. If you believe that CO2 is an issue, the picture becomes even grimmer.

Pumped storage projects are not “small scale”, there are several projects in the US that produce over 1000 MW of power (the biggest coal powered plant produces around 4000 MW of power).

Unfortunately, the scale we are talking about here is not the individual plants but instead the total electricity consumption of North America. If Hoover Dam was the only hydroelectric plant in North America, hydroelectric power would still be small scale on the scale of the entire continent. A handful of facilities that store half the output of a mid-sized coal plant is still small scale. Just because it is cost effective to store electricity this way in a handful of instances doesn’t mean you can scale the system to store an arbitrarily large amount of electricity.

The premise for economical pumped storage is that you pump with off-peak power (cheaper) and generate at on-peak power (most expensive). As long as you can pump with enough cheap energy to exceed the amount of high cost energy for generation, you can make money.

(1) Alternative power never really provides “cheap” power. The only reason to use it is to reduce CO2 output. To save CO2 you should be consuming all the power generated by alternative power while idling CO2 emitting systems. As such, alternative power never produces a cheap surplus it would pay to store. Instead, you would actually be caching power from large coal plant while idling the fast spin-up natural gas plants.

(2) Since the major function of the hydroelectric storage is to buffer the unreliability of alternative power, you have to store enough energy to serve as a buffer regardless of the cost of doing so. You can’t just store when it give the maximum economic benefit.

All these factors make the large scale use of hydroelectric storage a non-starter. Since alternative power cannot supply baseline power with a storage system, that makes alternative power a non-starter as well.

8. Eommy,

Saving part of ‘free’ power is always better than throwing it all away.

Buffering against changes in demand makes sense if you have large efficient (cheap) generators whose output can’t be easily adjusted. The problem with alternative power is the exact opposite. You’re buffering erratic fluctuations of very inefficient (expensive) generator against comparatively reliable consumption.

The economics are completely different. As I noted above, alternative power only makes sense to save CO2 generation. As such you always consume all the power it generates. This means that it never generates a surplus. The function of storage is to create a buffer to prevent the unreliability of the alternative power from brining down the grid, not to save money.

Alternative power is never “free”.

9. I did a semi-technical piece on this same topic here:

http://www.energypulse.net/centers/article/article_display.cfm?a_id=1808

Bottom line – due to the inefficiencies and high fixed cost of pumped storage for electricity, that technology works better when mated with nuclear and coal and disadvantages wind and solar (makes their cost higher!)

A quick survey of US pumped storage facilities shows that they are always part of a system that includes a large coal or nuclear generator.

All this is to agree with Mr. Love.

10. Very interesting.

My assumption (based on … well, more engineering knowledge than your average journo) has always been that coupling wind/solar with fuel cells and rail-lines always made the best sense.

It makes tons of economic sense to transport coal (very heavy relative to stored power) with trains to centers of consumption, so transporting fuel cells (very light relative to stored power) to, say, LA would work. Again, transporting gas (heavy w.r.t. to stored power) to individual gas stations versus fuel cells.

So there are some infrastructure costs, but I think a lot less than stored water power.

This solar/wind->fuel cell model also seems to solve the grid problem you get when 1M liberals put solar cells on their roof and try to sell irregular, expensive, and off-peak power back into the grid. If I could put NC summer sun energy into a fuel cell that I could use to power my heater in December, that might work.

-XC

11. There are some good storage sites left. Near the Hetch Hetchy reservoir is another large valley, which might be usable.

And I am sure Speaker Pelosi and her constituents, who already use Hetch Hetchy, would be delighted by the suggestion that we convert this other valley to some practical use.

12. Numbers aside, speaking qualitatively:

All this does is solve the problem of all the power plants you’re not allowed to build by combining the power plants you are allowed to build with yet another type that you also aren’t allowed to build.

13. Shannon Love:

You still have several errors in logic.

Pumped storage uses the same machines for generating and pumping, that is why they are called pump/turbines. You just reverse the the current with switch gear. In the pumping mode the machines are less efficient (25-28% less efficient but you do not duplicate the equipment, in the 1920’s that was the way it was done, with separate machines but not now. In today’s market it is hard to make pump storage economical due to the low differential value between on-peak and off-peak energy. Pumped storage does not need large drainage basins because you are effectively using and re-using the same water. You only have to make up for evaporation and leakage. Oglethorpe Power’s Rocky Mountain Project in Georgia uses less than a thousand total acres of reservoir (upper and lower) on a small stream with about 14 cfs average flow (total power about 850 MW).

The expense of transmission works against most renewables because of the isolation of the best locations and the transmission losses incurred. Chicago has a pumped storage plant close to the city basically recyclying water between an upper reservoir (small) and Lake Michigan with very small transmissiont costs.

The purpose of any power system is to provide electrical power at the cheapest cost possible. Any power system has consequences whether it’s CO2, bird kills, or other impacts. Renewable power requires equipment replacement ever 10-15 years while hydro can go 50-75 years or more between equipment replacements.

Hydro’s efficiency does not depend on head or flow, it is inherent in the machine. A turbine with a 1000 ft. of head can be as efficient as one with 10 feet of head. Now the power produced is a function of head and flow, P=headXflowXefficiency. 1000 ft. X 10 flow equals the same as 10 ft. X 1000 flow.

Not knocking renewables but you need to balance the facts. For the near term we need to utilize all energy sources until technology provides practical, economic solutions with lower impacts.

I trust our technology to get us there. But realize it takes many years between inventing the technology and practical implementation of it. It took over 40 years for nuclear to progress from concept to the first plant and another 30 years for it to reach 15% of our electric energy mix.

Wayne Elkins

14. Shannon Love:

The cycle efficiency of pumped storage is 1.25. That means for 4 W of electricity generation you have to put in 5 W. Not 4 for 1.

Wayne Elkins

15. Joseph Somsel – that is an excellent article.

16. Wayne Elkins,

You still have several errors in logic.

Actually, they are errors of fact. The logic is fine. Unfortunately, logic will manipulate false axioms as well as it will true ones.

Pumped storage uses the same machines for generating and pumping, that is why they are called pump/turbines.

You can but it is unlikely we will be able to do so with alternative power. The current systems seek to store energy from a constant and reliable source against a slowly fluctuating demand. The systems under discussion have to buffer fluctuating demand against a wildly fluctuating source. This means that at least part of the generation system will have to operate continuously. That in turn means the pumps will have to do so as well. Some or all of the generator/pumps will have to be duplicated.

Pumped storage does not need large drainage basins because you are effectively using and re-using the same water. You only have to make up for evaporation and leakage.

Well, you have to have enough drainage to hold all the water that passes through the system, especially if you only have one set of pump/generators. If you just continuously cycle the water, then you can’t use the turbines as pumps.

Hydro’s efficiency does not depend on head or flow, it is inherent in the machine. A turbine with a 1000 ft. of head can be as efficient as one with 10 feet of head. Now the power produced is a function of head and flow, P=headXflowXefficiency. 1000 ft. X 10 flow equals the same as 10 ft. X 1000 flow.

As a practical matter, you need elevation if you don’t have huge amounts of water available. Again, where are you going to put all these millions of acres of shallow water? In fact, where is all this water going to come from anyway? Most regions of the country don’t have that much excess water at all? What happens to the system in droughts or floods.

See, this is the error you keep making. You think that because we have a few dozen hydroelectric storage facilities that store reliably generated off-peak electricity that therefore we can easily scale that up massively to any arbitrary number of such facilities that can efficiently store unreliably generated electricity. It took us a century to build the hydroelectric infrastructure we have know. To reach the goal of providing 30% of our current consumption by 2030 we would need to recreate that entire infrastructure all over again in less than 20 years.

None of the nitpicking you’ve done, correct or otherwise, changes this. As I said in the parent, even with the most wildly optimistic efficiency assumptions about both alternative power and hydroelectric storage, this simply cannot be done.

Without that storage buffer, alternative energy is simply a non-starter. Without a means to convert the wildly fluctuating generation of alternative power into reliable current, alternative power cannot provide more power to the grid than the reliable systems can back up. This caps, alternative power’s contribution to the total grid at 15% at most with a near 100% backup with non-alternative means. There is no technical reason to choose wind and solar over nuclear in that case.

I trust our technology to get us there.

We can have energy certainly but we can’t have it from any randomly imagined technology. Nature puts limits on what technology is optimum at any particular stage of technological development. The problem with alternative power is that a wide swath of the electorate has simply decreed that this technology will work and that it will do so at acceptable tradeoffs. Well, it won’t. Political will power cannot dictate the laws of nature.

It took over 40 years for nuclear to progress from concept to the first plant and another 30 years for it to reach 15% of our electric energy mix.

No, it took form 1937 to 1943 when the first large scale nukes where built. The Hanford reservation reactors produced power in addition to fissionable materials. The first commercial reactor went online in 1962. Moreover, it was always obvious from the outset that nuclear reactors could work in the existing power grid. After all, they we just steam plants that used a different heat source. By contrast it is not obvious that alternative power will work at all. It is a radical departure from all existing technology that produces power on a human schedule instead of natures. No one, at any scale, has made alternative power the main source of power for any community.

The key point here is that investing in alternative power is a waste of time and resources. It won’t even work on paper. Instead, we devote all those resources into building nukes. For just the resources of the hydroelectric storage alone, we could build enough nukes to provide 15%-20% of our power.

If someone comes up with an efficient storage system for alternative power great but until it has that keystone element, it’s useless. We need to stop making policy based on flights of fancy. That is my key point and one you haven’t even touched upon.

17. If you had an efficient way to produce hydrogen from electricity and efficient (and cheap) fuel cell technology and cheap AE….. We are a long way from that combination.

Now if AE was really cheap and you could produce methane….. Except that the US has at least a 100 year supply of really cheap methane.

Should we be doing pilot programs on this stuff? Of course. It is how you learn in the real world. Should we be doing a massive roll out? Don’t be insane.