Next Generation Energy is managed and written entirely by Seed editors and expert guest bloggers. Read more about them If the government were serious about tackling our energy problem, there's lots to do. Instead of talking about carbon taxes and compact fluorescent initiatives here (which are all essential, don't get me wrong), I'll point out a solution (that's rarely discussed) to a practical problem (that's always discussed).
The problem: we'd love to run off of solar thermal and wind and ocean currents (explained here), but these natural sources of energy aren't steady, so what do we do when the sun sets or the wind slows or the gulf stream meanders?
A solution: we coordinate our supplies to the grid, carefully investing in regions with as anti-correlated (or uncorrelated) supplies as possible and then we smooth our supply with some very large "batteries."
Our energy technology is improving rapidly with large scale projects underway. If the investment road is clear ahead, there isn't much to stop our progression to the low/no carbon economy that Gore describes. Except that the sun sets every night. It's tough for energy companies to convince their investors that they're planning for the future by building plants for renewables when their ability to supply energy depends on the weather. But, at least in the US, we have a large renewable resource base and enough variation in its supply that we should be able to solve this problem. And if the government is proactive and credibly commits to addressing this issue, investment in the sector should climb (in conjunction with carbon taxes or caps and other incentive policies, obviously).
How should the government support energy suppliers harvesting renewables? The first (and cheaper) step is to lead a coordination effort among suppliers. The wind doesn't always blow in New York. And it doesn't always blow in Pennsylvania. But its less common that its not blowing in either. By supporting efforts to coordinate investments between suppliers on the same grid, the government can reduce the risk that suppliers fall short of demand or that excess supply is wasted. In some cases, this could be as easy as funding analysis of spatial variation in renewable energy sources like wind (heck, I could probably do this if someone paid me), so that plants far from one another could be jointly planned and could "insure" one another with excess supply or demand. In other cases, this could involve supporting the establishment of new types of weather derivatives markets, not too unlike existing ones for degree-days. These are all things that private firms could, in theory, do. But the large spatial scales involved (points on the surface with low correlation in weather are probably pretty far from one another) and the cooperative aspects (maybe almost a public good) of this type of effort would benefit strongly from government support.
The second (and less cheap) step is to provide a new type of public good: national energy storage. These are the "batteries." Imagine if there was a battery large enough that it could power the entire US through the night, allowing us to charge it up during the day and thus run exclusively on solar. That would certainly quiet the critics that complain about variable renewable supply, wouldn't it? Well, this is exactly the principle behind a project on the island of Flores in the Azores archipelago where an enormous flywheel was used to smooth out the supply of energy from a small collection of hydro, wind and diesel generators [Hamsic et. al., 2007]. The flywheel is a large metal wheel that spins around a vertical axis with very little friction. When there is an excess amount of energy in the system (perhaps when the wind is strong and demand is low), the wheel is spun up faster, storing energy in its motion. Since it has so little friction, it keeps spinning long enough that it's still spinning when there is excess demand for energy on the grid, at which point its spinning motion is used to generate power and it slows back down.
Now, the grid I just described is relatively small and the flywheel can be housed in a 20-foot shipping container. It seems unlikely that such a mechanism would be appropriate for storing energy on the vast scales that we require. But a number of other technologies exist which have been used on larger scales:
Hydrogen - Unlike fossil fuels, hydrogen is relatively easy to produce by electrolysis and can be stored for the production of energy via fuel cells or internal combustion. There is some loss of energy in both processes, but with virtually no by-products, compensatory scaling-up of large hydrogen production-storage-combustion facilities would entail little additional environmental costs.
Compressed air - There are successful projects where energy is stored by compressing air into large underground caverns and then released when energy is needed again. This clearly requires special geological conditions, however there is currently work to see if a similar process can be accomplished by pumping air into large bags deep under water. The bags are inflated under high pressure, which can then force the air out again, generating power, if a valve at the surface is opened later.
Pumped water - Hydroelectric power takes advantage of water running downhill and releasing its potential energy as it does. A large number of elevated reservoirs are currently used to store energy. Water is pumped uphill into the reservoir and released into turbines later when power is needed. 70-85% of the energy used to pump the water uphill is reclaimed later.
Flywheels and hydrogen have not yet been proven for large scales (its unlikely that flywheels will ever be used for large systems, currently the maximum is about 1 MW). They can, however, be installed almost anywhere. Underground compressed air has limited capacity and requires specific conditions, but if underwater systems prove successful, then this significantly broadens the potential capacity and number of feasible locations. In my opinion, though, it is the pumped water systems that will make the biggest difference; and are the area that the government might consider using in a system of national "batteries."
The use of pumped-storage hydroelectricity is already widespread and used at large scale (a list of current plants is here). In 2000, there were 19.5 GW of capacity in the US (2.5% of the national generating capacity) and 32 GW in Europe (5.5% of their capacity). Historically, they have been used to smooth out variations in demand rather than supply, since it is easier to start and stop a hydroelectric turbine than a nuclear reactor. But current installations are at capacities that could conceivably smooth out variations in supply. Take the Racoon Mountain installation run by the Tennessee Valley Authority (and pictured above and below, courtesy of the TVA): it can generate 1600 MW for up to 22 hours.
If private companies are building these enormous energy storage systems already, what's the role of the government? Well, not so long ago the government felt that it was its responsibility to build a network of national highways. They did this because these roads were too large and too expensive for any private company to finance and the government could geted reimbursed slowly over time. If the government wants to support renewables on a large scale, they'll need to build energy storage systems on a scale larger than anything we've ever seen. The government could spearhead the construction of several enormous pumped-water storage systems around the country, systems too large to be financed by individual power firms alone. The government could buy and sell renewably-produced power itself or could charge companies a fee for access to storage capacity. Construction of a project like this wouldn't be trivial, but neither was building highways all over the country. It would take some time, some careful design and innovation, but once it was built, it would open the doors to rapid expansion of renewable power and it would be an asset for future generations.
Conceptually, this seems an appealing solution, but is it just pie in the sky? Would it be impossible to provide the power we need with this kind of system? Asking myself this, I hit the back of the envelope again.
The US consumed about 2,280 Mtoe (mega tons oil equivalent) in 2003 [IEA, 2005] or 26,563,824 GWh. One cubic meter of water flowing down 100 meters produces about 0.272 kWh, which is reduced to 0.2 kWh after our 75% efficiency loss. This suggests that it would require 1.3 x 10^14 cubic meters of water flowing down 100 meters each year to power our nation; or 1.3 x 10^13 cubic meters of water flowing down 1000 meters, the approximate elevation drop from Lake Tahoe to San Francisco. Would Lake Tahoe be able to handle this job? On a daily basis, this translates into 3.6 x 10^10 cubic meters flowing from Lake Tahoe to San Francisco (and back up again). Luckily, Tahoe holds about 1.5 x 10^11 cubic meters of water. In addition, its unlikely that we would really require the full daily production of power to be stored (eg. we usually discuss needing to store solar for just the night). For a sense of scale, if we were storing half of the daily energy consumption (nighttime demand) in a pumped-water system running from Tahoe to San Francisco, we would be varying the water in the lake by about 13%. (If we actually implemented a national system, it would probably not rely on a single system like this, but on a number spread out across the nation).
I'm not suggesting that implementing a national network of government-owned massive energy-storage systems would be a simple engineering feat (or of negligible ecological impact), obviously nothing like this has ever been attempted before. What I am suggesting is that the foundations of the technology needed exists, at large (albeit, not yet "massive") scale and that constructing such a system would make the rapid expansion of large-scale renewable energy production both economically appealing (under a regime in which carbon has a price) and sufficient to finally eliminate fossil fuels from our unhealthy energy diet.
References
Hamsic, N., Schmelter, A., Mohd, A. Ortjohann, Schultze, E. Tuckey, A. and Zimmerman, J, Increasing Renewable Energy Penetration in Isolated Grids Using a Flywheel Energy Storage System, POWERENG, 2007.
International Energy Agency, "Energy Balances of OECD Countries," 2005
[Note: in response to some interest in the potential of geothermal power, I've added some material on the global distribution of geothermal power to the end of a previous post.]
Comments
Energy storage using superconductors would be totally cool, although sadly impractical.
Posted by: Yoo | July 25, 2008 10:14 AM
Do you have any grasp of the physics involved here? Because it is clear that much of this scheme is so full of losses that its utility would be questionable.
Simple resistance, reconversion losses, massive heat losses from the compressed air storage and hydrogen systems just to name a few, would waste vast amounts of energy. Spin-down losses from flywheels, make these impractical for anything except very short-term storage, and on and on.
When you combine the technical difficulties with the problem of operating the forward dispatch part of the electric power market, these ideas are simply unworkable.
Posted by: DV82XL | July 25, 2008 12:26 PM
DV82XL:
This is why I advocate the use of pumped water storage. As I mentioned, currently it achieves 70-85% efficiency in its recovery and has been installed with enormous global capacity >50 GW. This suggests its utility is quite real.
The other storage mechanisms may hold promise in the future (in regions without the correct geomorphology for pumped-water storage), however there seems to be little reason why we cannot move forward now with pumped-water storage.
Posted by: Solomon Hsiang | July 25, 2008 2:45 PM
While pumped storage is the best of them, it unfortunately depends on geography, and it is not clear just how much of this can be economically exploited. The places that are suitable are not necessarily convenient meaning more transmission lines need be built.
Now it may come as surprise to you that there is still a significant amount of undeveloped traditional hydro potential left in North America. The National Hydropower Association (U.S.) river basin studies show a potential of 73,200 MW of additional U.S. hydroelectric capacity in 5,677 undeveloped sites. The situation is the same for Canada, including the Far North where eight major rivers draining into the Arctic Ocean are considered ideal for major projects
Considering this, it would probably be more cost effective to exploit these than construct pumped storage.
But unfortunately despite the widespread belief that hydro is the ideal clean source of renewable energy the bald fact is that it is hugely destructive to local environments and can and does create disruptions to the hydrology of an area several orders of magnitude greater.
For these reasons I feel that the best option is still nuclear energy, which along with being easier to build, has a smaller environmental footprint and doesn't require a full overhaul of the grid.
Posted by: DV82XL | July 25, 2008 3:20 PM
The world's largest battery is...the World, isn't it? Maybe not a battery, but certainly a generator. Isn't there some way we can harness the Earth's magnetism to produce current?
Is it practical? Is anyone working on it?
Posted by: Ian | July 25, 2008 3:26 PM
Actually, there's a more efficient alternative to pumped storage: stop water flow through the dam when there's plenty of power from elsewhere, and restart it when the power is needed. A dam produces a constant quantity of energy per rain season, not a constant amount of power.
Also, the efficiency of pumped storage really depends on how efficient your plumbing is, which is to a substantial degree a function of length. Pumping water from SF to Tahoe would be highly inefficient.
On a second point, while I'm not opposed to government action in principle, in this case I really don't see the benefit.
Posted by: Anthony | July 25, 2008 4:38 PM
I agree on using hydro as both an energy source and for energy storage... and am a bit surprised by how much power your back-of-the-envelope calc indicates (not disagreeing). Yeah, hydro has downsides, but a number of existing sites can be utilized.
The real punchline, IMO, is that a single hydro site can not only be used to produce power (from rainfall), but also store power produced by other means.
--
On quibble though...
Why only focus on massive scale? We need high efficiency and low cost per unit of capacity. If we can achieve this with a lot of small or mid scale facilities, that would be even better.
Often, the larger the scale the more efficient cost wise. However, that doesn't always hold. Even when it does, it doesn't necessarily offset the benefits of a more distributed system.
--
Also, don't ignore thermal storage.
Both universities I've worked at have installed big chillers to store power (bought at lower overnight rates) which is used for cooling during the day. Such systems are becoming quite common, and work even at quite small scales.
Solar Tres being built in Spain will store up molten salt and be able to run overnight. I don't know the efficiency of such things, but they are pretty easy to do.
--
BTW: Gore is all over promoting updating the power grid.
Posted by: travc | July 26, 2008 6:20 AM
Chillers are becoming more common on small gas peaking units, too, but the reason is the huge difference in the cost of electricity at different times. Natural load curves ensure that power is cheapest at night and when the weather is nice, and most expensive - far, far more expensive - in hot summer days (or cold winter ones, for winter peakers). The wholesale price varies a LOT.
That's why these storage plans are used today. They store power when it's cheaper, and release it when it's expensive. But they're inefficient. They only make sense at peak usage (and prices). Thus, my local utility uses them on hot summer afternoons. But they're not used at all in the winter, spring, and fall, and not all the time in the summer, either. Expanding them to smooth out normal electrical usage (and the variable generation of solar and wind) would add hugely to our electric bills. Yeah, we may have to do it, but that's going to make such generation even MORE expensive.
And I just have to note (it was mentioned already) the huge environmental damage from hydro power. Environmentalists in my state (myself included) fight and fight to keep more dams from being built. Yes, there are problems with everything, but I still think that the expansion of nuclear power would make far more sense (and be MUCH less damaging to the environment). But, of course, that's probably not feasible politically, is it?
Posted by: WCG | July 26, 2008 8:41 AM
"Often, the larger the scale the more efficient cost wise. However, that doesn't always hold. Even when it does, it doesn't necessarily offset the benefits of a more distributed system."
This is a mantra that is sung over and over with very little hard proof to back it up. The fact is that scale is not just cost efficient on initial build, but on operating expenses, and maintenance costs. At any rate it is six of one and half a dozen of the other because it is the large scale of having many small generators that makes this concept posible. The difference being that they are not in one place where they can be looked after efficiently.
Also many of the proposals for distributed generation involve retasking the grid, and in many instances this critical issue is breezed over in discussion. This is a mistake. This system is huge and complex, with a bewildering number of control nodes and operates under protocols that been less designed then they have accumulated. It has not been built for two-way traffic end-to-end, and even in cases where bi-directional flow is physically possible it is often achieved only by overriding system fail-safes, and potentially compromising product integrity. Refitting to allow for this, while certainly doable from the engineering standpoint, would be horrendously expensive, and in some cases would require that large chunks of the network go offline or isolate for extended periods of time and in most cases this factor alone makes conversion unfeasible.
The grid has one other important property that cannot be ignored. It is there. Vast sums of money, time and material have been expended over the better part of the last century building it and polishing the procedures to make it run. The only way we were able to afford this huge system in the first place is that it grew slowly and the product that it moved was so inexpensive to produce that the consumer could absorb the cost of construction almost without noticing.
This issue has not been given enough attention, or examined by supporters of distributed generation in detail. This is a huge mistake.
Posted by: DV82XL | July 26, 2008 1:13 PM
WCG writes:
"Expanding them to smooth out normal electrical usage (and the variable generation of solar and wind) would add hugely to our electric bills."
Indeed.
In fact if they became commonplace to the point that night time demand was similar to daytime demand, especially with people re-charging their electric cars overnight, the price differential would rapidly disappear.
In terms of generation efficiency measurements this may not be a bad thing but the economics of the devices (chillers, cars and anything else that currently benefits from off-peak low rates) would be shot to pieces.
Meanwhile the greatly reduced demand for 'fossil' fuels would destroy the options (in the case of oil perhaps) or reduce the cost of more accessible fuels, like coal, to a level impossible to resist, in economic terms, for discrete manufacturing processes. Those operation would move to wherever coal was available and the use of coal was not banned. There will always be such places.
Posted by: Grant | July 26, 2008 3:56 PM
All of the methods for storing electrical energy are expensive, and have efficiency much less than desired. For this reason I don't expect us to build a great deal of storage capacity. As mentioned a few comments up thread using dispatchable sources to cover peak demands, and lulls in renewable availability is more practical. Aside from the obvious hydro, we also have gas turbines, which are already used for peaking. Natural gas is storable, so diverting the use of NG for power generation from baseline use to cover for intermittences would seem to be a natural choice. I also note, that biogas generation could be a significant renewable source of methane, and could be used to extend the supply of natural gas for such purposes. Adding nuclear baseline capacity would also significantly reduce the threat (or cost to mitigate) low renewable power periods. And of course we can add demand management, whereby certain uses of power are curtailed or reduced during periods of high demand or low supply.
Now things like flow batteries, are being build to handle short term load balancing, buying time to bring up peaking plants. But the cost per megajoule is likely to mean we don't try to store energy for periods of days or weeks.
Posted by: bigTom | July 26, 2008 7:23 PM
Great discussion.
My thoughts on a developing a few installations of massive scale, instead of a large number of smaller installations, have to do with the topographical requirements of pumped water storage. As others mention, pumped water storage requires two large basins near each other at different elevations. My casual guess would be that its more likely that we could find the right geographic conditions to for a few very large installations (that would meet the needed capacity constraints) rather than a very large number of smaller units. I imagine that this could translate into some economies of scale and also limit the ecological impact to a small number of systems.
On the cost of generation and investments in transmission infrastructure, this is why I argue that there is a role for the government to step in (also in response to Anthony). I agree that the costs will be large, massive actually. But the government can finance these investments, spreading out the financial burden over a number of generations and preventing modern consumers from experiencing an unfairly heavy load. After all, these investments are designed to improve the standard of living for the future, so it doesn't seem unreasonable for them to pay for some of it.
In general, I'm open to other solutions to the problem. My overarching point to the post is that we often talk about economic mechanisms (like cap and trade) that would fix the whole problem if we had the guts to just implement them, but that there are other major obstacles that we'll probably have to overcome if we seriously want to rework our energy system. And if the government is feeling serious about cap and trade (or an analog) they might as well feel serious about these other issues too.
As for nuclear, a response may have to wait for a future post.
Posted by: Solomon Hsiang | July 28, 2008 12:48 AM
Solomon,
In parallel you would need to be building the generating capacity to make use of the storage. Does this double the required investment? Are there enough skilled resources around? Can quality and longevity be guaranteed?
If you have mapped out a risk assessment, does it include re-purposed water usage (and positioning?) considerations and a study of the potential for, say , evaporation adversely affecting other aspects of the solution?
In terms of economics and mindful of the warnings from some that there are a mere handful of years to put a solution into place before the situation becomes irretrievable AND that the solution needs to be global to have any effect, how do the economics work out?
Monetary economics with government funding is all very well but where does the capital come from it everyone is doing the same thing at the same time? I assume the oil/coal producing nations would be the only obvious source since the engineering work required would provide them with large cash flow increases (or equivalent) in the short term.
There must be few previous examples of investment on such a scale in modern times. Perhaps China currently and maybe Germany between WW1 and WW2 and then in their reconstruction post WW2 might be relevant examples? But I am not sure they would be suitable models for what is proposed here.
One danger of massive infrastructure investment with long term payback plans is that the infrastructure technology becomes obsolete or the economic model changes well before it has been paid for. Think IBM mainframes and telecomms networks. One also loses the option to adapt to progress or constrains progress through being committed to past investment decisions.
In addition infrastructure development typically comes with a high carbon output price tag. Not quite the best way to challenge an approaching watershed for CO2 levels. This is for energy fuels and construction materials. Do you have a way of mitigating for that? Do you have any trade off numbers related to such a large and focused investment versus opportunity costs in other areas?
In WW2, to follow the 'this is a war' theme, the USA was in a position to be able to provide both a safe manufacturing powerhouse and financial input and took the opportunity very well. In the aftermath it took the UK several decades to pay off the outstanding debt, though the rebuilding of Germany seemed less financially troubling to them after just 20 years or so.
In a WWCC scenario I could see China and Russia being the bankers equivalent to the US role in WW2. Perhaps also India although I am less certain that it would have the same attitude.
What does your fiscal model produce as a likely result?
Posted by: Grant | July 28, 2008 1:45 PM
When we're all driving electric cars, there will a huge capacity for storage in their batteries. Presumably the infrastructure will already be in place. All the cars not on the road could be hooked up to the grid, and you could agree to sell back some of that power from your Blackberry.
Posted by: Johnboy | July 28, 2008 2:48 PM
Pumped storage can work but it also can be a bit of a problem:
http://en.wikipedia.org/wiki/Taum_Sauk_pumped_storage_plant
Posted by: Jim | July 28, 2008 5:21 PM
This was just released by the MIT news office. Its a discussion of a new process that hopefully will allow us to store solar power.
Posted by: Solomon Hsiang | August 1, 2008 9:59 PM
Perhaps the most efficient of all possible energy storage media is ammonia. It's needed everywhere, as fertilizer, so conversion loss is effectively zero, and transport costs are minimal. It can be produced whenever there is power available, and then not when there isn't.
Whoever develops a self-contained fertilizer production module that may be plugged into any variable-output generator will do more for the planet than anybody developing clean power generation.
Such a module would be immediately usable attached to wind generators in places with wind but no transmission grid, and next to oil wells that presently flare off natural gas as too expensive to condense and transport. It could be attached to solar power towers in deserts far from population centers, and operate directly off the high temperature in the tower without first converting to electricity.
Posted by: Nathan Myers | August 7, 2008 8:08 AM