Guest Post by Willis Eschenbach
Inspired by an interesting guest post entitled “An energy model for the future, from the 12th century” over at Judith Curry’s excellent blog, I want to talk a bit about energy storage.
The author of the guest post is partially right. His thesis is that solving the problem of how to store city-sized amounts of electricity would make a very big difference, particularly for intermittent sources like wind and solar. And he’s right, it would. But he’s wrong not to point out how devilishly difficult that goal has been to achieve in the real world.
Storage of electricity is a very strange corner of scientific endeavors. Almost everything in a 2013 car is very different from what was in a 1913 car … except for the battery. Automobile batteries are still lead-acid, and the designs only differ slightly from those of a hundred years ago.
Figure 1. Elements of a lead-acid car battery. SOURCE
Now, we do have nicads and such, but the automobile storage battery is the bellwether for the inexpensive storage of electricity. Cars need a surprisingly large amount of energy to start, particularly if they are balky. If there were a cheaper way to store that big charge, it would be on every car on the planet. Given that huge market, and the obvious profits therein, people have been busting their heads against the problem since before Thomas Edison made his famous statement about automobile batteries.
And despite that century-long huge application of human ingenuity, in 2013 the lead-acid battery still rules. It’s an anomaly, like fusion energy, a puzzle that has proven incredibly hard to solve. Potential solutions have all fallen by the wayside, due to cost, or capacity, or energy density, or dangerous components, or long-term stability, or clogging, or rarity of materials, or a habit of exploding or melting down, or manufacturing difficulties, the number of pitfalls is legion.
So I’ll get excited when we have something other than lead-acid batteries in our cars. Because that will be evidence that we’ve taken the first step … but even that won’t be enough. The other problem is the huge amount of energy we’re talking about. Here’s some back-of-the-envelope figures.
New York City’s electricity consumption averaged over a 24/7/365 basis is on the order of 5 gigawatts (5E+09 watts) continuous. Let’s take a city a tenth of that size, there’s plenty of them on the planet, China alone has dozens and dozens of cities that big, and lets consider how much storage we’d need to provide three days of stored electrical energy for that city. The numbers look like this
5.0E+08 watts continuous times 72 hours equals 3.6E+10 watt-hours of storage times 3.6E+03 seconds/hour gives 1.3E+14 joules of storage needed
So that means we’d need to store 130 terajoules (130E+12 joules) of energy … the only problem is, very few people have an intuitive grasp of how much energy 130 terajoules is, and I’m definitely not one of them.
So let me use a different unit of energy, one that conveys more to me. That unit is “Hiroshima-sized atom bombs”. The first atomic bomb ever used in a war, the Hiroshima bomb released the unheard of, awesome energy of 60 terajoules, enough to flatten a city.
And we’re looking to store about twice that much energy …
I’m sure that you can see the problems with scalability and safety and energy density and resource availability and security for that huge amount of energy.
So while I do like the guest author’s story, and he’s right about the city-sized storage being key … it’s a wicked problem.
Finally, as usual, Judith has put up an interesting post on her interesting blog. I don’t subscribe to a lot of blogs, but hers is near the top of the list. My thanks for her contribution to the ongoing discussion.
w.
PS—Edison’t famous statement about automobile batteries? He was offered big money in those days, something like ten grand from memory, to design and build a better battery for electric automobiles than the lead-acid battery. He took the money and went back to his laboratory. Month after month, there was no news from him. So the businessmen who’d put up the money went to see him. He said he didn’t have the battery, and in fact he didn’t even have the battery design.
Naturally, they accused him of having taken their money and done nothing. No, he assured them, that wasn’t right at all.
He said there had actually been significant progress, because he now knew of more than fifty ways NOT to make a battery for an electric automobile …
Curiously, Edison ended up inventing a nickel-iron-peroxide battery, which was a commercial failure … so even he couldn’t get past lead-acid.
Similarly, we now know hundreds and hundreds of ways not to make a battery for a city. So I suppose that’s progress in Edison’s terms, but after a century the wait’s getting long. I suspect we’ll solve the puzzle eventually, perhaps with something like a vanadium flow battery or whatever, but dang … it’s a slow one.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
@Willis, Re Rasey at 12:47 am
Sorry, I saw the New York City and 5 Gigawatts, but missed the,
“Let’s take a city a tenth of that size,”
So your 130 TeraJoule is correct for 1.5 GW-day.
$90 Billion dollars of Li-ion batteries, or
$0.00045 Billion dollars of delivered coal.
Hydro pumped storage is the cheapest storage method we have, followed by the lead-acid battery. Of course, you can’t have a pumped storage scheme in your car! For New York, you could. But to store 1.3*10^15 Joules you’d need two lakes quite close together, each 100 metres deep and with a diameter of 41 km, with a 1,000 metre fall between the two. Those aren’t two a penny! And that hasn’t allowed for the 75 % efficiency of typical pumped storage schemes.
Storage is only part of the problem if the intention is make wind and solar dispatchable power and replace fossil fuels entirely. Even if you are lucky and manage to make it through a few poor generating days using storage, you have now exhausted that reserve and your storage is depleted. Without a follow on period of considerable length when wind and solar are operating at high production the storage takes a long time to recharge. At least it does unless you have so overbuilt the capacity of the generating system that you end up throwing away huge quantities of energy when times are good.
Surely it’s not about a big system storing heaps but lots of small systems. Take the VBR system by Prudent energy. It’ s the Vanadium system Willis mentioned.
http://www.pdenergy.com/
500kW – 5MW systems are pretty big units with up to 8 hours storage.
Hoser says:
June 30, 2013 at 1:03 am
Look, guys, waving your hands is nice but you need to run the numbers. Hoser, if you think that electrolyzing water into H2 and O2 is a feasible plan, you need to give us the numbers for:
1. The energy required to electrolyze the water.
2. The energy required to compress the H2.
3. The energy required to refrigerate the H2 (won’t liquify otherwise).
4. The efficiency of the reconversion to electricity.
5. The losses (because the atoms are so tiny, hydrogen is a bitch to contain, or to compress for that matter).
5. The cost, the cost, the cost of all of the above.
I’d be shocked, shocked if you could pull 50% efficiency out of that system.
Finally, hydrogen is one of the most dangerous of fuels, because it burns at such a wide mixing ratio of fuel to air …
As I said above, the battery problem is a wicked one. Waving your hands at hydrogen is not even a beginning of a solution … and no, Hoser, the answer is NOT “pretty simple”. If it were, Edison would have solved it.
w.
Tom says:
> Using the lead-acid battery as a comparison is not terribly relevant – as others have pointed out, it has quite different performance requirements to large-scale storage.
There are different formulations of lead-acid batteries. You are right about the car battery, which is good for cranking at 1000A but not optimal for storage. But have a look at the UPS batteries, which are also lead-acid batteries, except they are optimised for endurance. There is nothing out there that is even close to the lead-acid battery type in terms of weight, cost, reliability, and safety.
For traction, the second best has been the iron-nickel alkaline battery (don’t stomp on Edison too hard). Nothing else can sustain a similar number of charge cycles and its power density is just a little worse than that of a lead-acid battery.
For grid-scale storage, the flow batteries seemed to be promising. A humongous bromine polysulfate battery was built in Little Barford a while ago and reportedly wasn’t a total failure, but it has never been commissioned:
“DBERR part- sponsored the Regenesys’ scheme with IVTL in 2001 to build a pilot flow-cell battery as a 12MW storage system at Little Barford power station. The project encountered severe technical difficulties and in 2003, after ITVL was acquired by RWE energy, it withdrew funding. The project was subsequently discontinued.”
http://www.parliament.uk/documents/post/postpn306.pdf
Unfortunately, we are not told what kind of difficulties those were, so it could as well have been a total failure. It is there, but it does not work.
So, being the third largest energy storage (after combustible fuels and water) the lead-acid battery remains very relevant to this discussion.
but should your entire battery decide to short you’ll have those “Hiroshimas” of energy released all at once.
Remember that the “two-Hiroshimas” is just the energy stored within the battery bank.
If it should short out, then that energy is released and then vaporizes the metal mass of the battery, which will then burn in the air with many times more energy released than just the electrical energy. The metal-oxides will go into the air and fallout into the environment for more trouble.
A mountain of coal can burn, but there is little else that burns with it. Furthermore, a mountain of coal cannot explode or short out.
Tom says:
June 30, 2013 at 1:03 am
Thanks, Tom. You missed the point, likely my lack of clarity.
I chose Hiroshima bombs as an equivalent precisely because if such a super-battery were short-circuited, that’s how much energy it would release, and that’s how much destruction it would cause. That’s the problem that I wanted people to have an intuitive feel for, what a huge amount of energy a big city uses.
Also, I wanted people to realize how much energy you need to pack into the battery. You need to pack two atom bombs’ worth of energy into whatever medium you are storing it in, and then get it back out again.
w.
Willis mentions (responding to Hoser):
1. The energy required to electrolyze the water.
Heh. Even if it were regular drinking water, you’d have to disassemble your pile and clean the muck from electrodes and insulators, possibly replacing some of them every 1000 hours or less. Imagine how often you’d need to do that if your electrolyte is communal wastewater.
Edison was not an inventor, he was a manager who used his understanding of things like patent laws to appropriate the work of actual intelligent individuals to make himself rich, and he electrocuted cats and elephants to screw over Tesla, an actual super-genius inventor.
https://en.wikipedia.org/wiki/War_of_Currents
**** Edison, seriously.
Pumped hydro (or cousin variable release hydro) accounts for over 99 percent of grid storage. All you need is lots of water and a mountain….which is why it does not solve much of the peak shifting problem, let alone intermittency. It is unlikely any battery or capacitor system well ever find major grid scale usage, although they are practical already for distribution level frequency regulation (in devices called statcomms). There are better places to read up on the problems and the research than Judith’s excellent blog. Try EPRI for starters.
Another way to visualise the amount of energy storage required to power New York City for 72 hours is to estimate the area of each of two storage reservoirs, with 100 m vertical separation, for pumped hydro energy storage.
Assume the active storage depth in each reservoir is 10 m, then the area required is 165 square kilometres for each reservoir to store 360 GWh of energy (5 GW x 72 h)
http://www.engineeringtoolbox.com/hydropower-d_1359.html
P.S. Correction to my comment @ur momisugly June 30, 2013 at 12:31 am
Applying Willis estimates for NYC, the energy storage required is 360 GWh, not 36 GWh (5 GW x 72 h).
Therefore, the estimated cost of energy storage is:
360 GWh @ur momisugly $500/kWh = $180 billion
A battery does not actually store electricity, but it is this misconception that I think drives so many futile efforts to force its application into uses where it is unsuitable. A battery is simply a temporary storage of energy using a reversible chemical reaction. It can certainly be convenient in many applications but using a battery to store city-sized amounts of energy (or even vehicle-sized amounts) is just silly. Far better, and more efficient, to generate electricity as needed.
In contrast to a battery,supercapitors actually do store energy in an electric field. As such, they are ideal for short-term storage for vehicle purposes. Consider that a conventional vehicle engine must be sized to meet the instantaneous peak power requirement, but the 5 min, 1 min, or even 1/2 minute average peaks are all substantially less than the instantaneous peak. A significantly smaller and more efficient gasoline or diesel engine could be employed if a bank of supercapacitors were employed to store just a small amount energy to be delivered over a just a few minutes.
Oshkosh has some large trucks which use this concept.
If you have lived in the bush and been reliant on batteries you would understand the dilemma. Cost is prohibitive. I too have heard rumours of the ‘super battery’ but yet to see the real evidence. I have also heard some naïve comments on using batteries to store energy.
Edison: “…because he now knew of more than fifty ways NOT to make a battery for an electric automobile …”
Hey, Willis, don’t laugh. From my experience in designing equipment and machines and structures, I can tell you that one of the things people pay designers good bucks for is exactly that – knowing what direction NOT to go in. I.e., what doesn’t work.
Not everybody knows what doesn’t work. I’ve seen guys come in and start out in directions I know won’t work. Some listen, some don’t. The ones that don’t. . . they don’t last long.
After a while, you get so’s people think you are making snap decisions, or hip shooting. That isn’t the case at all. It’s just that you can run down the list PDQ of what you know from past experience doesn’t work. It becomes your modus operandi – to look at a new concept for those trouble spots. If you run into one of those, you have to toss the concept out the window as being unworkable. Experienced designers who talk that lingo – we communicate pretty well among ourselves. At some point, even with different backgrounds, we kind of converge.
So, don’t laugh at Edison on that one. It’s very real world.
What about compressed air storage, which can be as widely distributed as needed, or viable, using a number of smaller generating stations, rather than a few monolithic power stations.
A. Scott says:
June 30, 2013 at 1:22 am
Thanks, A.
There are a number of pumped storage systems. The most famous is Niagara Falls. Not many people realize that they turn Niagara Falls off at night.
Oh, not all the way off, but at night the water is diverted to turbines and pumps that pump water into a reservoir. Then during the day, when demand in the area is high, it’s released to provide power when it’s needed.
However, it’s not a magical solution. First off, you need height, and the more of it, the better. This is for a couple of reasons. First one is that it increases the energy density of the storage—a litre of water up 1000 metres contains much more potential energy than a litre of water up one metre. So the higher you go, the smaller your reservoir needs to be.
Ok, so are you sitting down and ready for the bad news?
One kilowatt hour is equal to 2,655,000 foot pounds of energy. (Source is UnitJuggler, my constant advisor in these matters)
Since a gallon is about 8 pounds, that’s about 320,000 foot-gallons … per kilowatt-hour.
So to store a measly kilowatt-hour of energy (assuming perfect conversion), enough energy to run a hundred watt light bulb for ten hours, you have to move 32,000 gallons up ten feet … for one kilowatt hour. Is it worth reinforcing your house and building a swimming pool on top of it, in order to keep one light bulb burning all night?
Now, imagine how much water you’d have to move just to power one house. So yes, if you’re generating power in a mountainous region, where you can pump it up say 200 metres to a natural reservoir, it might be worth doing.
In Kansas? Fugeddaboutit …
w.
PS—There’s another problem, which is that generating energy from a small head like that tends to be less efficient than from a high head …
The use of carbon nanotubes looks promising for hydrogen storage.
Stephen Rasey says:
June 30, 2013 at 1:27 am
I loved that calculation, and wanted to highlight it once you got it right … I’d go for Choice B myself, but I was born yesterday …
w.
Modern nuclear powered submarines still use lead-acid cell batteries sized to provide a restart of the power plant. We used to impress ourselves by calculating the effects of releasing its energy in various accident scenarios.
Submarines also electrolyze water for breathing oxygen in a device vulgarly known as “the bomb” (in my experience and generation of technology). My ship carried a numerically large supply of oxygen candles.
An old ship’s navigation gyroscope’s rotor weighed, if I recall correctly from forty years ago, 55 lbm and turned at 35,000 rpm. I slept with my head on the DC to AC converters so that I could manually hold the gyro erect on a loss of power – I was very impressed by photographs of the damage caused by a gyro escaping its bearings.
Reversible pumped storage (hydro) has been doing the job at roughly 85% efficiency for about 50 years. I helped design and build one just south of Chattanooga, TN in the 1970’s. The Raccoon Mountain Pumped Storage Plant is a four-unit facility on the Tennessee River with a total generating capacity of 1,740 MW. The machines include 540,000 hp synchronous motor/generators and reversible pump/generate impellers (Francis type). Net head between the river and the upper reservoir is about 1,000 feet. The plant typically pumps water up the hill during off-peak hours when the system has excess capacity and brings it back down in the generate mode during peak hours when electric power is dear. Since the market value of electric power is high during on-peak hours and relatively low during the off-peak, the plant is a “money machine”. The 85Kwh generated for every 100Kwh consumed is often worth several times that 100Kwh consumed.
The plant is completely automated (my assistant and I designed the controls) and a single operator can swing a total of over 3,400Mh of power from full pump to full generate in a matter of minutes. Although hydro pumped storage is a near perfect solution to the problem of bulk energy storage and retrieval and there are a number of such facilities around the world, getting a hydro project of any stripe permitted anywhere in the U.S. now is well neigh impossible due to “environmental intervention”.
Pumped storage is high capital cost and subject to escalation, reference Bakun in Sarawak You also need a large volume of water and a large fall.
The splitting water, then using the resultant gases to make electricity when required looks the simplest and possibly cheapest. Mucky electrodes covered by having several. Switch out for cleaning as required, automatic with sonic cleaning. Problem fixed, hopefully.
Nice to see some cost figures on the process.
@Steve R June 30, 2013 at 2:02 am:
“In contrast to a battery, supercapitors actually do store energy in an electric field. As such, they are ideal for short-term storage for vehicle purposes. Consider that a conventional vehicle engine must be sized to meet the instantaneous peak power requirement, but the 5 min, 1 min, or even 1/2 minute average peaks are all substantially less than the instantaneous peak. A significantly smaller and more efficient gasoline or diesel engine could be employed if a bank of supercapacitors were employed to store just a small amount energy to be delivered over a just a few minutes.
Oshkosh has some large trucks which use this concept.”
Industrial electric motors almost universally use “motor starters”, which are basically capacitors to provide the instantaneous high energy requirements to get the rotor rotating – the same thing an automobile starter motor does to an engine.
This allows the size of electric motors as small as possible – no reason to size the motors according to that initial energy requirement! It saves money and saves energy to do it that way.
* * * *
On that front, I have a question to ask everyone:
Does anyone know if hybrid cars use the electrics at any time in tandem with the gasoline engines? Or is it one or the other?
The reason I ask is that back in the days of the oil embargo there happened to be a study of engine efficiency in St Louis. The study found that, for a car going only 30 miles per hour, when it stopped for a stoplight and then accelerated back up to 30 mph, a car used up 14 times as much gasoline during that decel-accel period compared to if it had kept going at 30 mph. I’ve never forgotten that study.
But my point in asking is this: If the electrics were used just to BOOST the torque during acceleration, that would serve the same function as a motor starter or a super-capacitor in the way Steve R has described here.
Whenever I have my car’s air-conditioner on, I can readily sense the added strain of acceleration, so I got in the habit of turning off my air-conditioner when I have to accelerate hard. It makes a big difference in rate of acceleration. If an electrical boost could be used instead, it seems like that would be a good use in a hybrid.
But that can only take place if the electric and gasoline can be used at the same time. (Hybrids were always out of my price range, so I never inquired enough to find out…)
So, does anybody know?
Steve Garcia says:
> Industrial electric motors almost universally use “motor starters”, which are basically capacitors to provide the instantaneous high energy requirements to get the rotor rotating – the same thing an automobile starter motor does to an engine.
A small correction: the motor starter capacitor does not “provide” the energy to get the rotor rotating; it conducts the energy from the power source to the starter coil, if this is the the type of motor you are referring to:
http://en.wikipedia.org/wiki/AC_motor#Capacitor_start_motor
The purpose of this capacitor is to excite the starter coil out of phase with the drive coil. Not quite the same thing as the automobile starter does.
Re: My post
Typo – Make that “…swing a total of over 3,400Mw of power….”
Compressed air energy storage.
From Wikepedia; “As of 1896, the Paris system had 2.2 MW of generation distributed at 550 kPa in 50 km of air pipes for motors in light and heavy industry”
~
“Adiabatic storage retains the heat produced by compression and returns it to the air when the air is expanded to generate power. This is a subject of ongoing study, with no utility scale plants as of 2010, but a German project ADELE is planned to enter development in 2013.[3] The theoretical efficiency of adiabatic storage approaches 100% with perfect insulation, but in practice round trip efficiency is expected to be 70%.”
~
“Thus if 1.0 m3 of ambient air is very slowly compressed into a 5 L bottle at 20 MPa (200 bar), the potential energy stored is 530 kJ. ~ theoretical energy densities are from roughly 70 kJ/kg at the motor shaft for a plain steel bottle to 180 kJ/kg for an advanced fiber-wound one, whereas practical achievable energy densities for the same containers would be from 40 to 100 kJ/kg.”