For perspective: if the U.S. loaded up all 1.5 million of its rail freight cars to 30-ton capacity and sent them 4km up to the top of Mt. Rainier, that would store 494 GWh of energy.
That's just over an hour's worth of our average electricity usage.
We just need to store enough to make load-shifting environmentally (and commercially) available. The 30 Gigawatt Hour Bath County Station gets there (https://en.wikipedia.org/wiki/Bath_County_Pumped_Storage_Sta...), but even the "smaller" Tesla 50MWhr projects do too.
Compressed Air is hitting 300MW-hr designs (https://www.greentechmedia.com/articles/read/texas-calls-for...). That's enough for a metropolitan solar community to loadshift the 12:00 noon sun to the 5:00 dusk peak-energy period. (Average home: 30kwhr per day. 300MW-hr design would serve over 10,000 homes easily, and doesn't need to actually hold the electric usage for the whole day. It'd realistically only need to hold 1/10th the energy or so, to load-shift a few hours here and there).
ARES seems to be in the 12MW-hr to 100MW-hr design size. This is smaller than Pumped Hydro but still useful. Smaller, cheaper projects that partially solve the problem is still a good thing.
It seems like a perfectly good solution to the problem. Buying up cheap energy from night (or noon-power solar arrays), and selling it during the 3pm to 7pm peak-energy time. As long as there's at least a 20% differential in peak pricing somewhere in the day, "small" 1MW-hr to 300MW-hr plants will be profitable.
----------
In any case: Wind continues to work at night. Nuclear works throughout the day. Hydro works throughout the day. We don't need to store EVERYTHING, we just need to store the excess from Solar from noon, and then load-shift it to 5pm to 8pm, while the sun is setting and solar arrays generate less electricity.
To finance this, a bank will want to see they've managed to hedge their business and are in the business of capturing those spreads rather than speculating on the widening of that spread (only to have it contract). Interestingly, this may be exceedingly hard to do on the open market and may require setting up an expensive structured product with the trading group of some bank, where they pay you a fixed price each day and take on the daily/monthly (and particularly in power markets, term) risk. That won't be cheap, especially if these aren't located to one of the major grid nodes where there are liquid markets available.
Basically, I'm just wondering how hedging costs affect your numbers.
> To finance this, a bank will want to see they've managed to hedge their business and are in the business of capturing those spreads rather than speculating on the widening of that spread (only to have it contract)
The spread is grossly positive. Pumped-Hydro provides a location for baseline power plants (nuclear, coal, and even Wind / Solar) to continue to pump energy into the grid without getting shut down. Coal, Wind, and Nuclear plants, in particular, are typically very difficult to start back up again.
So it is profitable for these plants to continue to generate power, even when the plants pay money for the privilege for someone to "waste" it somewhere.
Then of course, later in the day, the "batteries" (ARES, Pumped Hydro, CAES, and what-not) sell the energy back to the grid when peak-energy occurs (usually around 5:00, when businesses still have the factories running but people have also begun to return home and turned on their air conditioning again).
The other way to look at this is to consider how much chemical energy there is in the diesel fuel tank of a train. Or that a single AAA battery has enough energy to raise a garage door (but slowly).
There was this Swedish train transporting minerals from a mine up in the mountains down to the sea harbor, it was going downhill with a load and uphill empty, and producing some energy for the neighboring town.
There is a mine in chile that does something similar with conveyor belts. The mine is in the Andes mountains, 3200 meters above sea level; the belts go downhill for 10kms to 1800 masl. They have most of their VFDs in regenerative mode, ultimately generating 25mw wich they end up selling to the interconnected grid.
That suggests a whole new type of power generation - moving rocks from mountains down to sea level. It would also have the advantage of creating more level land that could be more suitable for farming and construction. Not sustainable at all though so don't tell environmentalists :P
The patents address traction a little bit. The mass would be highly concentrated near the drive unit. So it's more of a system for stacking weights on locomotives than it is a system for pushing trains up a hill.
The problem with steep railway lines is generally BRAKING, not climbing. Norfolk Southern ran freight up and down the Saluda grade in NC for over 100 years, and that's at nearly 5%.
Generally speaking in the UK this was addressed by adding banking engines (the steepest sustained mainline gradient in the UK is 1:37 and this isn't enough to cause problems for modern trains). This suggests the problem is an economic one with a mix of banks and flat running. If you are only running a gradient you can provide more traction power and have done
The problem is adhesion (lack of). Modern traction control systems (anti-slip) help quite a bit, but eventually you'll exceed what the steel-on-steel mating surface can provide in terms of grip. Traditionally, you overcome this by using a heavier locomotive - which is one reason why the Union Pacific Big Boy was 1.2 million pounds (544,000 kg). It sounds like they're addressing this by making pretty much every car it's own locomotive.
I'd be curious to see gravitational energy storage built individually into each wind turbine. If you think about it you already have a lot of height from the tower.
Have the turbines directly lift a weight. And then lower the weight as needed to generate electricity.
A typical wind turbine is 80m high and and generates 1.5MW [1].
Gravitational storage in Joules is (mass in kg) * (height in m) * 9.8 (gravity).
Divide by 3.6 million to convert to kWh.
Suppose we raise a 30-ton weight (i.e. a big shipping container filled to capacity with rocks) all the way up that 80m turbine.
We've stored 30,000 * 80 * 9.8 / 3600000 = 6.5 kWh of energy.
That's 16 seconds worth of energy from the turbine spinning at full speed.
Impressive way to put the power generation capability into perspective.
Also makes me wonder how a spring (instead of or in addition to those 30 tons) would scale to turbine tower size. Just don't think of the destruction when something breaks!
Catastrophic energy release doesn't care if you're a spring, a flywheel, a lithium battery, or a chemical bomb - in fact all those things start to look pretty similar at high enough energy levels.
But you must admit that an 80m (~262ft or ~24 stories) wind turbine blasting off like a giant pogo stick into the sky is by far the most comical.
Maybe gravitational doesn't make sense, but compressed air might. Big compressor directly driven by the turbine (so as not to waste energy converting to electricity to turn an electric compressor motor) being used to compress normal air into large tanks inside the wind turbine tower which can later be used to spin another turbine-based generator using the same air. Not a physics major, but this seems feasible. Not sure of the potential energy storage, but I feel like the direct conversion of rotational energy to spin the pump has to be much more efficient than converting it to electricity first, then using that [many miles away] to compress air or drive a train up a mountain.
Dispatchable wind via compressed air storage seems to be quite a thing in the scientific community. Unfortunately, big subsidy schemes like the german one fail to acknowledge the importance of dispatchability, so they utterly disincentivize any development in that direction.
A 40' shipping container-sized box filled with lead would be roughly 90 metric tons, so multiply that number by three. Now you've got ~45 seconds worth of energy.
(An actual shipping container would break before you hit this point; a standard 40' shipping container has a maximum gross weight of ~30 metric tons-- you're not likely to fill it to its volumetric capacity with rocks or lead.)
There's an old paper by Salter on the possibility of doing this through hydraulic pressure, having the turbine drive a pump, the weight being a piston within the turbine tower, and the output of the system being a hydraulic generator. This was back before modern developments in inverters meant that turbines had to have either a gearbox or a hydraulic torque conversion system like a car's automatic gearbox.
You are referring to a hydraulic accumulator[0], such systems were built in the 19th century. There is a modern effort to use the technology in a massively scaled up fashion by Heindl Energy[1]. Their concept uses pumped water to raise a rock weight 100 meters or more in diameter.
You end up with an incredible amount of mass divided by 4 or 5, which most likely works out to an incredible amount of mass.
In one of the patents, they work out an example system that can handle 30 megawatts, which might be the output of 10 or 15 turbines. The system absorbing the 30 megawatts for 30 minutes has a working mass of 1200 tons (conveniently, they use a working height of ~600 meters).
So all you need is a system capable of lifting 100 tons a distance of 600 meters (to store 1/2 hour of output). Which is probably something we are capable of building (there are cranes that lift much more significant distances), but it probably isn't something that would be economical.
Based on the very lightweight comparison with pumped hydro storage, I think they realise it's going to be hard to compete with. How are the costs lower than a pipe a pump a pond and a turbine?
Yeah we don't know what the actual costs will be apart from their claim that its 60% cheaper than pumped-hydro. However, having done some run-of-the-river hydro projects - even a small 1MW mini-hydro (which is basically roughly a pumped hydro in reverse) requires 2 years of civil engineering for the penstock and weir (the water channel). I would imagine that 1-10MW pumped hydro systems would cost and take a lot longer to engineer. Plus pumped hydro is very geography/topography specific.
I don't have hard data - but wouldn't engineering a rail system not be as hard, and need less civil engineering work?
Given the mass of the cars they're talking about, they'd need substantial engineering to get a good solid level grade that's not going to subside or slip, which isn't always going to be straightforward.
As you say, however, it's very geo/topo specific - so in places like North Wales, pumped hydro is still clearly the way to go, but in the Mojave, this rail based system could work well.
It reminds me of the defunct system we had in a house I used to live in - 8" or so diameter shaft in the basement floor with a weight, a cable, a winding, and a dynamo - you'd have the servants crank it up during the day so you could have electric light at night. Quite popular in the late nineteenth century among early adopters of electricity.
I like this concept. I would think its all going to be about efficiency of the return power. Not a hydro engineer, but I would think that a turbine is not as efficient as a wheel tied to an electric motor. X Amount of wheel turns to store and the same X amount to restore the power. It all comes down to the efficiency of the motor vs efficiency of the generator side.
Its California / Nevada. They probably just don't have as many lakes out there that the local population is willing to convert into Pumped Hydro solutions.
I suspect "we're going to tie up a bunch of water for energy storage" doesn't go over that well in California and the rest of the western US these days.
It's also probably easier to add a parallel track and mass units than it is to add a penstock and double the size of your reservoir. So there's no need to develop a site to the optimum size up front.
They claim their ramp is faster than pumped storage, ramp is an ancillary service here. Also being smaller than pumped storage means they can be optimized to respond to smaller ancillary service requirements.
In general there is no silver bullet for the electricity system, rather, a better metaphor is a stack. This tech might fit one space in the stack.
If they can buy power at $20 and sell it for $50 / MWh, then their 12.5 MWh storage capacity generates $375 / cycle. They'd have to cycle it many times per day to make significant amounts of money. Does the grid require more than 1 or 2 cycles per day? (Solar cycles once / day, domestic demand has 2 peaks).
> Kelly also said that the economics of the Nevada project will be aided by recent pricing trends at CAISO, which has been seeing negative pricing from over generation from wind power at night and from solar power in the morning. Under those circumstances, the ARES project could be paid to take power off the grid, Kelly said.
Its cheaper to keep running a Wind Turbine and PAY to put energy onto the grid, than to actually power-off a Wind Turbine. If those Turbines stop spinning, its very, very difficult to get them spinning again.
But someone needs to "take" the energy. This ARES project is going to be paid in both directions. They profit as Wind / Solar farms pay for the privilege for their excess energy to be stored somewhere, and then they profit later in the day as they sell the energy back to the grid.
Use water as the working mass. Lift it by converting it to steam and sending it up the mountain through insulated pipes. Condense it at the top, recovering the heat as possible, and let it run down to a turbine.
You have to counter losses, of course, and if you could get the heat energy back down to the bottom to reuse it that would be great. Otherwise, just use Solar and waste energy to heat the water to steam.
I think you might be able to get more energy out by "mining" gravity this way than you had to put in to run the steam cycle. You're taking advantage of the fact that water falls while steam rises.
Like I said, I have no idea if this would actually be feasible.
2000 feet are 609.6 m, thus according to supahfly_remix max. 12000 kWh can be stored per train, 5.5 miles are going to take about 4 minutes to climb and another 4 minutes to descend, let's assume (from thin air) the usage is 3 times below its max, then a train will buy and sell (0.05 also taken mostly from thin air) 2.5/h * 12000 kWh * 0.05 USD/kWh = 1500 USD, which means that salary for the driver is going to be perhaps 2-4% (depending on whether fully driving, or just observing for safety).
Density helps. But environmental issues help more. There are only so many viable locations for large-scale hydro storage, and it always means drowning some valley. Just building a set of rails up a hill is much less environmentally invasive.
One of the biggest wins of this model, really, is the low environmental impact, relative to hydro or batteries. Fewer regulatory hurdles, more potential locations. But in the end, cost effectiveness is what will win, and to some degree, that will be location-dependent. Is there geography we can exploit? Will it be in a populated area? What's the proximity to wind or solar farms?
Solving the storage problem in a cost-effective way is the key to transitioning from poison fuels to clean power. Once it's cheaper than coal/LNG/nuclear, the world will switch over quickly.
Rocks have about 4 times the density of water, thus you'll need 1/4 of the volume of a reservoir to hold as much energy. Since pilling trains is much more expensive than holding water, I can assure you this project will use much more land area than an equivalent lake.
Also, this thing requires a straight line high inclination slope and plain areas both up and down. I don't know how that compares with high walled valleys, but it is not a very common formation in nature. All said, it may be useful because it uses a different kind of landscape from water storage, so it can increase the total capacity.
Anyway, I agree, costs alone will say how much of it is created.
>When an induction motor that powers a train or car is reversed, it produces electricity.
This is only true if the motor is connected to the grid (ie powerered). You need a magnetic field to create the current. If the induction motor wasn't powered when you spun it, it would not create any current as an induction motor does not need permanent magnets to create the field.
...If this works - one of the apparent benefits is scale. Water-tower, and other water-pressure based systems are hard to engineer above 1-10MW (I think).... The benefit of this system is it seems to be able to scale to 100MW+ sizes. Which is what one needs to provide effective frequency regulation, smoothing and backup to a utility-scale solar/wind park. Of course, we don't know what the costs are yet...
Seems a little weird they're using something that resembles more or less standard guage rail.
Imagine if they scaled it up where instead of a few rail cars you scaled up to something along the lines of NASA's https://en.wikipedia.org/wiki/Crawler-transporter but on rails instead of tracks.
TLDNR Excess renewable energy is fed to electric rail cars which move boxes of rocks up a slope. When renewables are not producing power the rail cars roll down the slope and regenerative braking produces power.
A very creative combination of existing technologies. Zero emissions, nearly zero waste, and can probably switch back and forth between consumption to production on a minute-by-minute basis. Also, looks like a crazy Factorio project.
Vastly more costly than pumped water, high upkeep costs.
Does have relatively high efficiency though.
PS: "Cost comparisons with pumped hydro are difficult, he said, because there are so few projects and not many data points." Is pure BS, here is someones coursework on the subject: http://large.stanford.edu/courses/2014/ph240/galvan-lopez2/
Or stop farming Almonds which use 1/2 as much water as all non farm use combined. Agriculture alone is 80% of all water use in the state. So, running out is less of an issue than trying to farm land that's just shy of a desert.
The real problem is all the alfalfa and the cows. Per pound of protein/carbs/whatever, cows consume way more water than pretty much any plant. And much of our alfalfa is just being shipped to China.
Cutting Alfalfa production is fairly cheap in the long term which fits well with California's regular droughts. However, using prehistoric water to farm Alfalfa is crazy IMO.
The problem with Almonds is they are trees, so letting them die is far more costly.
But almonds are a high-value crop. If the price of water went up substantially, the almonds would get a little more expensive, but demand wouldn't collapse. I don't think it's right to say we shouldn't grow almonds in California, though they should pay for their water.
> using prehistoric water to farm Alfalfa is crazy IMO
Agreed -- completely insane. And we wouldn't, if water cost what it should cost.
Basically, if the water table is dropping then people are pumping out water faster than its being replaced. If you look into when that water was put there you quickly get into geologic timeframes. Aquifer's often do flow meaning that specific drop may only be decades old, but replenishment rates in many places are vastly smaller than extraction rates.
EX: Some places in China have seen 10+ feet drops in the water table per year.
> Rail storage also does not have life cycle limits that batteries do. “There is zero degradation,” he said.
They must be using rails, wheels, and bearings made of unobtainum. I know little about conventional rail maintenance, but a quick search suggested annual grinding of rails and 10-year replacement intervals.
California already uses a lot of pumped storage built into their long water distribution network.
As to land use these tracks take up space at both end plus the space for the tracks between them unlike a pipe which can cheaply be placed under ground. So, per kWh stored these are likely far worse in terms of land used.
This totally highlight the problem with Imperial Units:
"ARES wants to lay a nearly 5.5 mile track up an 8 degree slope, gaining about 2,000 feet top to bottom. ARES would then put up to seven 8,600-ton trains on the track"
If the units were in SI[1], 1000 * 5.5[km] * 8/100 = 55 * 8 = 440m (no calculator/conversion needed)
And for the energy stored, just do 440m * ~10m/s2 (g) * 8,600,000kg
So 200m long cars, given that there are four in a train of that mass from TFA? I suppose it's not inconceivable, but it seems weird to be re-inventing the railway carriage format.
Also, having a long train is surely undesirable for this application, unless you have a plateau at the top and bottom at least as long as the train.
If you look at the video, they have a neat system where the mass is stored perpendicular to the rails in the storage yard, then cars come in, pick up the masses, rotate them parallel to the rail for transport, then re-rotate them back to perpendicular for storage at the other end.
Totally true, however if "8,600-ton trains" just means 8.6/8,6-ton trains then that would be certainly too little (unless it means per car and not per train).
An example in one of the patents I linked discusses having 240 regenerative units in a single train. So it sounds like the plan is to make them quite large.
But they say in that same sentence that it'll be two locomotives and four cars, weighing that in total... but then again perhaps if you don't have to deal with any curves, you can have reaaaaaaaaaallly long cars.
I would imagine an advantage of the rail design is that you can build them in existing factories already hooked up to the rail system and just drive them to the location, rather than having to truck-transport the parts and assemble locally.
Oh I was totally wrong, in Spain slopes for roads are usually in % of m(up)/100m(length) [1] so I misread it as 8m up for every 100m of length, still not sure why. Thanks for the correction :)
Edit: Also, they are claiming 93% mechanical and electrical efficiency. I'm not sure if that is each or total, but either way it sounds impressive for storage and retrieval in a wide demand range.
Edit2: Is this company publicly traded? Or is it owned by a publicly traded parent company?
Edit3: It is a startup (and they are looking for funding).
For perspective: if the U.S. loaded up all 1.5 million of its rail freight cars to 30-ton capacity and sent them 4km up to the top of Mt. Rainier, that would store 494 GWh of energy.
That's just over an hour's worth of our average electricity usage.
Gravitational storage is hard.