It’s fascinating that this is a paper - we did the math about 20 years ago and have been using excess electricity for desalination in Cayman ever since.
Long-story–short (based on my experience with these chuckleheads): it was killed because the current political establishment didn't own the desalination plants, so there was no way they could personally profit off of them; they own oil/gas/fracking & ERCOT.
Texas generally seems to have chosen an odd path politically, which to me has been strangely highlighted by their failure to supply their own people with electricity. Aside from the usual left-right dichtomy¹ one could start to consider to just elect people who at least pretend to care about the problem. Electricity has become kinda important.
¹: A divide that is seen more as a centrist — far-right divide from my european perspective.
Oof; yeah. So, Texas has a primary and then election system: candidates are selected for the ballot in the primary, and then voted on in the general election. The probability that a candidate will win the general election having won the primary is north of 95%. That means the "real" election is the primary. The turn out for the primary is the highest turnout in my district (Williams; US HD 25), which is ~25% in a contentious year, but usually 10–15%. The race is usually split 3-ways, with Williams just barely getting 50%. That means, the most competitive district in Texas chooses their rep with ~12% (high! or ~5-7%, norm) of the voting age population.
Most US representatives from Texas are elected with far fewer; many Democratic gerrymanders mean that as little as 1% of the population is choosing the representative.
As always, I advocate for some form of random election (sortition) — it'd be way more representative — and it'd be harder for people to mess with the election: we'd literally have more representative representatives and more robust elections. (I'd also require risk limiting audits, paper trails, and non-identifying ballot checking.)
Maybe you can enlighten me how the shape plays a role here? In the end isn't this about how close the absolute vote matches the political outcome? One could easily come up with rectangular districts that skew the result, right?
I am from somewhere where all important democratic elections are won by absolute numbers, so we don't even have the whole problem.
Sadly that's just about true for most things, whether it's city state or federal. God forbid Congress tries to tackle the health care industrial complex, it might hit their portfolio!
Should this kind of observation be especially surprising for anyone? There are some round-trip inefficiencies to any energy conversion and/or storage system, so using electricity directly if you can is always going to be better than storage - that doesn't mean storage is bad or less important, but it's a tool that we should try and reserve as much as possible for when it's really needed.
It's easy to demonstrate the same thing on a small scale - at my house with hot water heating for example, I run my electric (heat pump) storage hot water heater in the middle of the day, and it almost always runs 100% from solar power (sometimes with extremely cloudy or rainy weather it uses some grid power). I don't have a battery system yet, but I'm doing the sums - and if I relied on instantaneous electric hot water heating so was heating water on demand mostly outside solar hours, even before thinking about the lower round-trip efficiency I'd need a far bigger (and therefore much more expensive) battery system than I will now for the same hot water load.
If you're in an area that has less sunshine but more wind energy, the principle is basically the same, you just have to time it for when there's excess wind energy, but it just seems very obvious that if you can shift high energy use tasks to those times that's always going to be more effective than using batteries, and doing so makes the battery storage cheaper for the electricity use that you really need it for...
The question is whether the round-trip inefficiency of storing energy is greater than the cost of having energy-consuming equipment that sits idle when the sun isn't shining and gets stopped and started frequently.
For example, aluminum production is famously energy-intensive, but I've read it's not good for the equipment to start and stop it frequently because the reaction happens in molten aluminum salts and it hurts the equipment when you stop and it solidifies.
And while Bitcoin is energy-intensive, it's also capital intensive for the chips so it's not cost-effective to run it only part-time.
Reverse osmosis may have a good combination of high energy:capital ratio (so it doesn't waste a lot of capital on idle equipment) and not requiring complex startup/shutdown sequences.
In Germany they vary aluminium production rates in response to electricity supply (i.e. produce much less when it's less sunny and windy). They just dont turn it off completely, ever.
There's probably a lot more low hanging electricity demand shifting fruit like this.
That will make production a lot more expensive, if you have production capacity laying unused for a long time. The rent, write-off etc still had to be paid but is now offset against less profit.
I think you’ll find the kWh used by any water heater will be negligible, but that the peak kw of the on demand one would force you to buy giant batteries. We ended up getting a hybrid heat pump water heater. It is very efficient, but we have to wait for the water to warm up at the tap.
If it were easy to change, I’d probably put in a recirculating system that keeps the water hot in the pipes (not sure how
much electricity that wastes though).
Is the idea to use excess solar power to desalinate salt water instead of mine crypto currency? At a high level that seems smart.
It seems like the general idea is to build out excess capacity on the desalinization side and run the desalinization plant at less than full throughput during peak consumer demand (i.e. peak of the duck curve?).
The various policies we have are for Aluminum Smelting plants, Desal, and other heavy-electricity industries to use excess solar (or nuclear) power during peak-energy times, and to turn off during peak-usage times to help balance the grid.
On an individual level, we might be able to even do this for air-conditioning, washing machines, or car-charging (if you have a PHEV or EV).
Cryptocoins then perverted the models we created to earn money from this through mining (using energy during peak-energy and turning off during peak-usage). But the "intent" for those laws has always been for industry and/or individual grid-level issues.
--------------
For low-utilization machines (ex: washing machines and dryers), this makes perfect sense. The main issue is that for high-utilization machines (ie: all industry wishes to run at 100%, rather than 80% or 50%), you end up doubling or more the cost of CapEx.
Ex: You'll need twice as many Desal plants if they only operate during the most efficient 50% of times during the week. Which might be worthwhile, but you have to run the math. Every hour these plants are shutdown for power-saving purposes is an hour that they are sitting away / depreciating.
I can’t imagine that’s how industries could actually run. I think they can benefit from cheaper costs to absorb excess electricity or reduce capacity slightly when it gets too expensive, but shutting down an aluminum smelting plant seems like a non-starter as they need to run constantly 24/7 and require long startup/shutdown times. Same goes for desal - you’re not going to economically produce sufficient amounts of water just trying to time the energy grid.
> “We have reinvented the electrolysis process for the production of aluminium. For the first time, we will be able to vary the energy supply during operation significantly. This will allow us to react to changes in the electricity supply, which will benefit the power supply to households in Essen,” says Philipp Schlüter, CEO of TRIMET. “As an aluminium producer, we are naturally an energy-intensive company. As such, however, we are also a valuable partner for the energy revolution.”
> The €36 million trial installation converted a total of 120 furnaces in hall one of its Essen plant, which will be able to consume either 25% more or 25% less energy for up to 48 hours. The energy requirement can also be reduced to zero for up to an hour, if necessary. This means up to 2,000 megawatt hours of electricity can be stored for use in the energy revolution.
-----------
It requires new smelters, new software, new engineers. Its all very expensive. But it might be worthwhile. Just don't go like... crazy... with the idea right? There's downsides, but the upsides are immense. Keep an eye on the costs and practicality, and give it a test. It might be worthwhile.
That's my outlook anyway. Maybe Aluminum specifically is run too close to 100% utilization to be useful (we do have a very predictable amount of Aluminum...), but maybe other industries have a more start/stop and feast/famine kind of setup that would benefit from off-hours automation for energy price optimizations and/or even grid-tie stabilization.
But there are _real_ Aluminum plants giving this idea a test. So this isn't "vaporware" or even "theoretical". This is a real test occurring today.
-------------
We have to ask ourselves: is a 2GW-hr battery the best thing to build? What if Aluminum-smelters could instead vary their load by +/- 25% with this new software/manufacturing mechanism? What costs more, the 2GW-hr battery or this software + manufacturing change to Aluminum plants?
In both cases, we get 2GW-hrs of "energy storage". EDIT: IE, varying your load each day is a form of energy storage that should be considered by members of society. It may very well be cheaper to achieve 2GW-hrs of virtual energy storage by upgrading industry, rather than trying to build impossibly huge amounts of Li-ion batteries.
Varying load by +/- 25% is neat but that doesn’t solve the “I need energy at night” problem for solar. You either need batteries or wind (but even wind is risky because you’re screwed during windless nights).
All these other ideas are worth exploring and I’m 100% in favor of them, but the only path to decarbonizing the grid is nuclear & I’m so frustrated by all the solar absolutists that think renewables can power 100% of the grid by 2050 just because solar panel costs are dropping. We’d get much closer to net 0 if we were building fission plants aggressively (cue all the solar proponents claiming it’s too expensive or unsafe even when it really isn’t and solar costs are intentionally calculated in a misleading fashion to ignore the buildout needed for energy storage or altering unrelated industries to make solar more attractive).
When the grid transition began in the 2000s the hope was that solar prices, storage and nuclear would all see massive cost declines that made a fast transition possible. This has occurred for solar PV. It’s reasonably plausible that it will also occur for storage, given the huge incentives and rapid increase in manufacturing efficiencies. It has not happened at all for nuclear and there’s currently nothing on the horizon that gives much hope this will change by the 2050/60 time frame. Even China, which is blasting away at nuclear, predicts it will only be 18% of their generation by 2060 (and that’s likely optimistic.)
Yeah, nuclear is hard because it was strangled in the crib for 50 years and it’s hard to make up that ground on time scales for net 0. It still remains the better bet for decarbonizing the grid and all the other neat ideas seem mostly like party tricks and nothing more. I think China in particular is interesting because there’s a good chance that they’re building enough capacity that their costs start to decrease & they become the world leader in fission plants. Specifically interested in what they’re doing with MSR designs because they should be much much cheaper than traditional HWR designs. But yes, it’s going to take us a minute to rebuild and restart our fission know-how as a species. It’s also telling that China is building a heck of a lot of fossil plants (natural gas & coal) in addition to lots of solar. That’s because they have a lot of electrification to do (they have a massive rural population & they also have much more transportation that needs to electrify to use grid power). Solar is a fine companion to handle daytime peak growth (which is why they’re reaching TW scale with solar) but that baseload growth remains a problem which is why they’re building so many fossil fuel and nuclear plants.
I don’t want to re-litigate the “fission didn’t fail economically, it was killed” perspective (which I find unconvincing: people have proven they’ll put up with much more dangerous things if they’re profitable.) But we need to replace the entire global fossil fuel industry in 20-30 years, and the only possible way to do that is with energy sources that can be mass-produced in a factory with all the scale and efficiency improvements and virtuous cost and production flywheel that entails. Solar and wind are on this trajectory. Storage is plausibly on an earlier stage of this trajectory. Nuclear might get onto this trajectory through modular reactors, but those don’t exist and are at the “maybe” research stage with huge questions still unanswered. By contrast, commercial battery deployment and production is economical and projected to hit 6.5TWh by 2030.
As a related datapoint on “party tricks”, China built more solar and wind in the first nine months of 2023 than all 26 nuclear reactors it has under construction. They’re likely to build even more next year. The need for baseload is a valid point, but the trajectories of renewables vs. non-modular nuclear are so different that it’s hard to see nuclear catching up enough to make a difference.
Given that China is building out more coal plants than all other countries combined, I think that’s telling about the efficacy of renewables to achieve net 0. They’re targeting 2030 for peak CO2 emissions because they’re trying to cheaply electrify as much of their population as possible as quickly as possible because the CO2 reduction efforts that will dominate 2050-2100 will impose significant economic growth challenges & China wants to make sure they’re in a more stable position to compete with the West.
Yes, China is building a huge amount of solar capacity, but AFAICT it’s all about absorbing peak growth as they add more consumers of electricity. Baseline growth remains a huge problem and China is solving that with a mix of nuclear and fossil fuels. I fully expect China will continue to ramp up nuclear as they get better at it.
And EVs are a huge iceberg problem for the electric grid as they currently don’t run off the grid and represents a massive amount of energy consumption that will be added to the grid. On top of that they consume a huge amount of battery production which means there’s not that much left over for grid scale batteries. We also have no existence proof of economical grid scale batteries at any meaningful scale. You also have to overbuild your solar by quite a bit so that you can charge the battery up when the sun is shining to time-shift that excess capacity into the night.
Solar is fine and it’s not a knock against it, but there’s simply no trajectory to reach net 0 CO2 emissions and you will end up with a hybrid grid. 18% nuclear would represent a huge CO2 reduction because all that capacity would otherwise be fossil fuels (because the alternative is not building batteries because that tech doesn’t exist yet). MSRs are a neat party trick but honestly I think China’s approach of SMRs is a far safer bet in terms of being a massive cost reduction, safer, and use waaay less water & are far less of a research project. My hunch is that they’re also standardizing their nuclear plant designs to keep costs in check. Keep in mind that France is 90% fossil fuel free in their grid even though nuclear only represents ~60% of energy produced.
I haven’t read anywhere that fission failed due to economics - it has always been price competitive with fossil fuels. It’s the regulatory burdens & concerns about safety (some valid, most not) that strangled its growth. Do you have any links suggesting nuclear construction is uneconomical?
The latest estimates are that China’s emissions will peak next year and then enter a structural decline [1]. The 2030 number is generally viewed to be obsolete now (that’s how fast things are moving.) This is even with the new coal capacity they’re building, which they’re structuring to be “idled” (through government payments not to generate.) They’re also building large amounts of solar capacity in the west and in deserts, with continent-scale HVDC interconnects.
Right now the (global) goal isn’t 100% renewable energy, and it won’t be even by 2030 or 2040: a grid that’s 70-80% renewable with 20-30% average fossil emissions will be a massive improvement and will keep us on track for decarbonization. The interesting question is whether battery storage will decrease in price fast enough to deliver that remaining fraction, or whether we’ll have to build the remaining 20% with fission by 2050/60. Batteries are already economical enough that we have 5GW deployed in CA for infra-day time shifting. It seems like an incredibly long bet to imagjne that after three more decades of technological and manufacturing improvements their cost won’t have dropped enough to make them a viable competitor for nuclear. For my own part I wish I could take the other side of that bet, since I think it would be a good one. But we’ll have to wait and see.
> Varying load by +/- 25% is neat but that doesn’t solve the “I need energy at night” problem for solar.
*48 hours* is the part of that statement that solves the "need energy at night" problem. By allowing this industry to shift power-consumption forward, or backward, by 48-hours, you effectively build a battery. A very strange battery yes, but it means the plant can collect during peak solar (+25% power usage), and then cut power at night by -25%.
>we might be able to even do this for air-conditioning, washing machines, or car-charging (if you have a PHEV or EV).
No _might_ about it, this currently happens for A/C, and is just getting going for car charging (see EV Managed Charging as a category).
Also some functions of commercial/industrial facilities that can be ramped more easily, eg shut off half of the hallway lighting in an office building. I think I recall things like rock crushers being turned off, but I might be misremembering that.
Unfortunately, solar panels get less efficient when hot, which is when your AC demand increases.
Ideally you could offset this by having people who aren't well positioned for panels do energy storage: Buy in the morning, sell back in the afternoon. Practically I don't think the utility companies are going to cut them in on a big enough part of the profits to incentivize them to do so.
Sorry I stated my original point poorly, and it wasn't exactly what you were saying. My bad.
Demand response for A/C tends to look more like pre-cooling before the hottest hours, then reduce demand during the hottest hours. So it's not about shifting demand towards maximum production, but rather shifting demand away from the demand peak (aka, peak shaving). So it's similar but a little different.
>Practically I don't think the utility companies are going to cut them in on a big enough part of the profits to incentivize them to do so.
The commercial version of what I stated in the previous paragraph (ie, demand response/peak shaving) empirically already happens, and has non-trivial dollars attached to it. Again, maybe not exactly what you're talking about, but it rhymes.
That's pretty cool! 10% was my educated guess. It's good to have some data points.
Would you say the same holds across months? Say, would cooler May have better production than hotter July? Or is July still ahead because of more sunshine hours, offsetting the hot inefficiencies?
I'm in a hot climate, I find in general with my system that the length of the day dominates over the loss in efficiency so December (southern hemisphere) is always the peak generation for me. January we usually have storms and rain due to the humidity so that's not usually as good, but it seems to be pretty reliably due to that shading from the cloudy and rainy weather more than heat.
I don't remember the numbers but I did watch this YouTube video where a guy strapped on some fans to a solar panel and actually got a better yield fan usage included.
Energy costs vary not only over the day but per quarter hour in each hour, e.g. during the morning the last quarter is cheaper than the first for each hour. So you don't necessarily have to shut down the machines, maybe you can just move the most energy expensive phase into another quarter.
If you are brewing beer you have times where you have to heat it and times where it needs to sit at a given temperature. Try to optimize the process to heat during cheaper quarter hours.
In Australia we've moved all the way to five-minute (!) settlement in our electricity market. Used to be 30 minutes but they switched in 2021 to give better price signals for very fast response technology like battery banks, as well as the demand response.
Do you have a auction + continuous market as well? With same or with different granularity?
Europe has an day-ahead auction with 1h products and continuous trading with quarter hours as the smallest product (still varying from country to country).
The most egregious broken window fallacy I’ve heard is a vc claiming that cryptocurrency lets you “virtually” transfer energy from areas with low prices to areas with high prices (without power lines!).
You assume that it is all linear which it may not be. Osmosis at power P may produce X litres of water per time unit while osmosis at power 4P may produce only 2X litres per time unit. And thus be viable only when the electricity price is N times lower.
Lol. Almost like making drinking water and water that can be used to grow food is a better use of energy than boiling a small ocean of water on math designed to create artificial scarcity :)
If you electrolyze seawater, to make hydrogen when the grid is over producing...and then you burn it to produce electricity when the grid is in need....can you capture the waste water from the exhaust... and call the entire process desalination?
I worked on this a little bit. You still need to energetically pay the entropic cost of separating salt from water, and the kinetic over-potential (on the electrodes). Also these cells hate running in reverse. Ruins their lifespan. Overall it is still not viable.
Your idea to use it to iron out the economics, though, might have some merit if the price fluctuations are high enough (which is what motivated us to begin with)
The cells dont like running in the reverse direction so you either research bidirectional cells or double the capital costs making cells optimized for each direction. Also, over potentials are pretty high and the catalysts foul easily but for the cleanest of inputs.
Also, is RO the cheapest for DS? I thought modern distillation is back on top, especially if you have waste heat
In theory, sure. Hydrogen fuel cells will spit out water as a waste product, that's probably the easiest way. You could directly burn the captured hydrogen or hydrox gas and power a conventional steam turbine. You'd then have to condense the exhaust gas, which is mainly steam, but that's not terribly complicated.
The problem is that the amount of energy you get back would be fairly small in comparison to the input energy. It might not be economically practical. The amount of energy you can reasonably store per unit of input energy might not ever surpass the cost of the equipment to extract the energy and put it back into the grid.
There's a minimum energy cost in separating water from salt. Salt dissociated into water has a lower energy state than separate salt and water molecules. You have to put that much energy back into the salt ions to pull them out of solution, and they will keep that energy until dissociated again.
So you have to pay the cost of entropy, then pay for all the energy loss in the combustion, capture, and conversion stages. I'm not sure if you'd get enough energy out to ever be worth the trouble.
The thermodynamic cost of separating salt from water is the same per molecule no matter what. It's a fundamental physical limit.
RO can be more efficient than electrolysis. The major factor is that electrolysis wastes a good fraction of the input energy on heating the water. On the other hand, RO requires high pressures, which is energy intensive.
RO is generally more efficient overall, but either way you have to pay the cost of reversing entropy.
But the electrolysis technique gives you the opportunity to store the energy as hydrogen. This means you can use cheap surplus energy for the electrolysis and burn the hydrogen later when energy is scarce. RO does not have that property
If energy becomes "free" does it matter? If solar continues on its current trajectory, there needs to be some kind of sink to utilize the energy or it is all going to turn into bitcoins.
The "fully charged show" sometimes have videos on this and similar topics. This [0] video goes through the pros and cons of using hydrogen in different ways
My takeaway is that as much of the energy as possible should be used without converting it. This avoids the substantial losses converting to and from hydrogen. But storage is one of the viable uses.
This seems like a special case of the general concept that matching electricity demand to supply is important, and whereas a fossil fueled grid has enough dispatchable power sources to match supply to demand, a renewable-powered grid has to instead match demand to supply (at least to a much greater degree).
So it's interesting, although at some point it seems unnecessary to enumerate that this same concept applies to every possible source of electricity demand.
> this economic efficiency “mostly vanishes” under scenarios with 100 percent renewable grids, says Conejo, because the cost of producing electricity is similar throughout the day.
I don't understand this statement. Shouldn't the cost of consuming electricity vary as the supply of renewables changes throughout the day?
In the UK we have cheaper energy tariffs at night for charging EVs. It’s about a quarter of the price as during the day. As I understand it this is because we overproduce on renewables at night.
An acquaintance of mine was looking into the general area of using cheap electricity. The pitch is "Instead of looking at storing energy efficiency, look at things that become profitable if energy is really cheap, but only for the fraction of the day". This demand in turn will drive building more generating capacity.
If you only run the desalination plant when energy is cheap, plant cost per liter of water goes up. May or may not be a win. Probably a win for small, isolated plants, say for powering and watering a small island. Large grid-connected plants, don't know.
The Carlsbad desalination plant [1] cost $1 billion to build, and it produces 190000 tons of desalinated water per day, or 5.7 million tons per month. Assuming a 6% financing cost, and 30 years amortization, you pay $6 million per month, or about $1 per ton of fresh water. That's a cost you incur whether you desalinate the water or not. If you decide to run the plant at 50%, then that's $2 per ton.
It takes 3.6 kWh to desalinate a ton of water. The cost of electricity in California is about $0.12/kWh for industrial users, so that's about $0.44 per ton of desalinated water.
Let's say you switch from the current electricity providers to solar generation that is absolutely free. Then you save $0.44 for each ton, but incur an additional capital cost of $1.
All water infrastructure in mainland America is stored in large reservoirs / lakes and/or water-towers for future use.
If it takes 1GW-hr to make I dunno, a billion gallons of water or whatever, then it might be cheaper to store a billion-gallons of water + build a 2nd desal plant rather than trying to build a 1GW-hr Li-ion battery.
That's the real question that's being discussed here. What's cheaper? Additional Desal plants + water storage? Or Giant-batteries?
Presumably, pumped-hydro is off the table (which is true in some geographies). Pumped-hydro is an effective form of GW-hr sized energy storage but is somewhat frustrating at how limited the geography can be for it in practice.
Water probably isn't grid connected like electricity. Nor does the supply need to perfectly match the demand every second. There is quite a bit of water storage in the system.
There might be pumps, in addition to just gravity assisted transportation of water. When energy is cheap, we can do both desalinate + transport.
But mainland desalination plants are connected to the electricity grid.
The problem with only running them part of the time is that the fixed capital and other operational costs of the plant are now split over a smaller amount of water produced so that water is more expensive than if you ran the plant all the time paying more for electricity at other times of the day.
There is certainly a trade off there, does the cheaper electricity compensate for the fact that you need to add say 20% extra capacity and the accompanying costs to your desalination plants as they will now be idle during low renewable production/times of high electricity costs.
You're missing the point. If you need 24 m^3/day of fresh water, you can have a desalination plant producing 1 m^3/hour and a water tank to match supply to demand. But if you only want to run the plant for 3 hours/day when electricity is cheap, you now need a plant which can produce 8 m^3/hour.
There's a tradeoff -- spending more building the desalination plant makes it possible to run at a lower duty cycle so you can spend less on the electricity.
https://www.moderndescartes.com/essays/factobattery/ I looked at a variety of these ideas and most of them are bad. Desalination is another one I looked at quickly (did not add it to the blog post), but I recall it being at least 10-100x more expensive than just buying batteries.
I think the desal math changes a bit because it also produces a product beyond just batteries. For particularly arid or water poor locations where you need desal anyways, seems like this could be a win (think islands or the middle east).
That said, you'd probably still want batteries to supply power to things like a desal plant when water levels are low.
The lifetime of a Li plant is just 10 years? Batteries lose a bit of capacity at 10 years, maybe still have 80 - 90% capacity. These can run for decades.
Really impressive simulation of using fresh-water storage tanks as energy storage in regions that use desalination.
The cost of overbuilding desalination capacity is non-zero though, and energy is only about ~1/3rd of the typical cost for a cubic meter of water overall depending on the plant, so the marginal savings may not be as big as we would like here.
Also, we can charge EVs. Grid connected batteries may or may not use the energy, but EVs will use some portion of it every day, and can be charged again the next day when there is excess energy.
Power price goes negative +200M times/year, just in the US and it is increase. It would be wonderful to get paid for charging EV. Tesla Electric customers report making as much as $150 a day[1]. Car batteries are bigger than powerwall, EV owners can make quite a bit of money by participating in VPP (Virtual Power Plant).
It would take a great many discounted electric bills to justify the extra wear on the batteries that would result from the extra daily charge/discharges.
I'm not sure it would have much of an adverse impact. Driving the car is constantly charging / discharging it (regen braking).
I think it's more about how deep the cycles go -- just don't let the grid drain below like 30-40% battery and it likely would have no meaningful impact long term.
I'm not sure it a totally rational thing. This is effectively putting more miles on your battery, an item that costs tens of thousands of dollars to replace, if even possible years after purchase. There would have to be a substantial compensation for such use even if it only degraded the lifetime a few percent.
But at such a compensation level, one could probably just purchase deep cycle lead acid marine batteries. They could sit in one's basement charging/discharging at a much lower cost than the lithiums in a car.
The price of batteries has declined by 97% in the last three decades:https://ourworldindata.org/battery-price-decline. And it continues to go down! The chart ends at 2018, $181/KWH. 2024 forecast is $94/KWH (half of 2018 price), 2030 forecast is $62/KWH. Thats just Lithium Ion. Newer chemistries are even cheaper. CATL first-generation cells (Sodium Ion) cost $77 per kWh, expected to drop to $40/KWH.[1]
Current estimate for a battery replacement is 4K - 20K. With energy storage's learning curve, this will be soon under $3K.
> Driving the car is constantly charging / discharging it (regen braking).
My understanding is that the impact on battery longevity relates to the rate at which it is being charged/discharged. Accelerating and regenerative braking are small potatoes compared to charging over a typical level 2 system, I would think.
A typical L2 charger will provide in the range of 6-10kW (AC, marginally less ends up going to the battery after conversion losses). 10kW is only 13hp; forget acceleration - the car draws more than that at highway speeds.
(You can come at this result another way too - an L2 charge may take 6-10hr to refill the battery from empty. But the car would not be able to drive 6-10hr at highway speed starting at 100% charge! So the L2 must be delivering less power than the car consumes at cruising speed.)
Please don't let your imagination run away with schemes like this. The incentives might be nice, but they'll only attract people who understand them and are in a situation to take advantage of them.
I dispute that. Ontario Ultra-low-overnight rates let me charge my car overnight for $2 or during peak periods for $25. Not too many people are going to ignore savings like that.
I just tell my car to make sure it's charged & warmed by 7AM and it takes care of the rest.
If I had the opportunity to join VPP I'd just tell it to make sure that it was at 80% at 7AM and never below 50% any other time.
I think every EV I've driven has a charging timer - so I've been able to say "only charge overnight when prices are cheaper". Some have "Be charged to 100% by this time" so they can pick and choose when to start charging.
Please don't let your lack of experience run away concocting scenarios which don't exist.
I've had electric cars for almost a decade, and I currently have two Teslas.
> concocting scenarios which don't exist
I've had plenty of situations where my car didn't charge. (Don't buy an electric Chrysler, btw.)
The most recent situations have to do with Tesla's new feature where the car will primarily charge on solar.
> "Be charged to 100% by this time"
What I hit was with my dual car charger. (Grizzle) It gets screwed up by things like timers when two cars are plugged in.
It divides the amperage up evenly until a car finishes charging, and then gives the car that is still charging most of the amperage. If the 2nd car is on a timer, the charger will never give it full amperage until you unplug and replug it.
This is a problem with Tesla's solar charging feature, because only one car will charge at a time. When car 1 finishes and car 2 starts charging, the charger doesn't adjust, and the car 2 never gets full amperage.
You'll hit the same thing if you have two cars plugged in and they are both on timers.
Granted I could probably push on Grizzle to give me a firmware update, but my experience is that they ignore support questions.
Your assumption is there is no electricity at home or work, only at charging stations far away.
But, we have electricity anywhere we have a building. In fact, thats the first utility thats hooked up before anything is even built. Now, we can think of charging an EV anytime it is parked. Residential garages have a 240V dryer outlet (or in the laundry room next to garage). That outlet can be used or an additional outlet can be added.
Cars are parked 22 - 23 hours a day. They can charge anytime they are parked.
> Your assumption is there is no electricity at home or work
I've driven electric cars for a decade and have two.
That is not my assumption at all. My assumption is that most people will consider the car defective if they plug it in, and then it doesn't have a full charge when they come back to it. (Either overnight or at work.)
> Residential garages have a 240V dryer outlet
Where I live that was a recent code change. I had to install an outlet for my first electric car. When I built my house in 2017, I had to really push on the builder to include an outlet.
If you've chosen to go to a fast charger, you're always going to get a charge and you'll just pay what it costs. That's not going to change.
But for a destination charger (e.g. shopping center), or a charger at home or work, it doesn't have to be like that. I would have a system where there's three options (can be three big light up buttons) - 1. I need the battery charged right now, 2. I'd like the battery charged over the next 6 hours, and 3. I have enough charge for now, but do charge me up if it's free or prices are negative.
Option 2 would probably be the default, because it would be able to modulate the demand and charge very cheaply, because it'd mostly only have to avoid the two-hour or so peak periods (like the solar duck curve). Option 1 is there when you need it, but you might pay a few dollars more for a charge. During the day in summer, in areas with lots of solar (like my state in Australia) you might be able to always select Option 3, and usually get a full charge (well, I'd set the 80% limit but fully up to there) every time without ever paying anything.
It doesn't have to be hard - you just tell it you want charge right now, over the next little while, or you don't care, and you pay (or don't pay) a differently depending on what you select. It can be a very small number of super simple and extremely intuitive options, not complex settings you need to think about. People scare-monger about "Governments taking away our ability to use power when we want", but it's actually going to be "Utilities will offer incentives like free electricity if you choose to shift your usage" - just offering more options, not taking away freedoms.
> New research shows that hybrid energy grids that rely on a mix of renewable and fossil fuel energy are up to the task, so long as the timing of freshwater production is right. In some situations, it may even be more financially and energy efficient to put that generated energy toward desalination rather than store it in batteries for use later.
And this is why I remain skeptical of renewables as base load energy. The only carbon free base load energy tech remains nuclear which is also perfectly capable of performing desal during off-peak times.
Unfortunately nuclear is now horribly expensive to build so we’re just not going to build that much of it. By contrast renewables (and batteries) are getting cheaper and cheaper so we are building lots of them.
In 2022 Denmark generated 62% (+12% from 2021) of its electricity from renewables (excluding biomass which seems suspect to me) while France generated 63% (-5% from 2021) from Nuclear.
I think you can make a case that nuclear is competitive for 100% carbon free grids (though by the time it’s built batteries will have gotten far cheaper.) But it seems clear that renewables are by far the cheapest and quickest way to get to ~80% carbon free grids.
Yes yes. Renewable proponents really love trotting out costs while ignoring the only reason costs have gone up is because we significantly ramped down our nuclear production efforts. Same would happen if we did that for solar. But that’s a choice we get to make. Also nuclear construction costs are particularly outsized in the US where regulations are intentionally constructed to strangle it. Japan, China, and Europe build it much more cheaply.
Existing heavy water reactors could be built more cheaply than they are now. And MSR/thorium designs are a step function cheaper beyond that because they generate even less waste, don’t need expensive fail-safe mechanisms because it’s inheritently physically impossible to melt down, and don’t need water to cool down which is both a cost savings & avoids the issue of warming waters due to global warming impacting the ability for reactors to run at times.
China has already started building MSR designs and it sucks that the DOE is dragging their feet in approving these designs & focusing instead on SMR designs. The regulatory capture of the US beuracracy by fossil fuel companies is a huge problem.
> it seems clear that renewables are by far the cheapest and quickest way to get to 80% carbon free grids
Got any existence proof for this claim? We’re building renewables as fast as we can and countries seem to generally tap out at ~20% ammortized across the year and are growing extremely slowly (~1% each year). By comparison France runs 90% carbon free and that’s because of their nuclear grid. Batteries will decrease over time but we still don’t yet know what grid-scale renewable base load looks like & I don’t think that 80% number is correct considering that baseload is 30-40% of maximum load (i.e. best case peak without batteries is 60-70%). But all the solar proponents also seem to ignore a major headwind for solar which is that while we build out more & more capacity, our energy demands grow faster than that. We’ve electrified 2% of consumer cars, 0% of trucks, 0% of ships, 0% of airplanes. Electrifying cars is going to add an insane amount of demands on the grid & solar can’t keep up (batteries either). And yes, there’s the argument of using batteries in EVs to do the grid balancing but we don’t actually have that tech & 0 regulations requiring it meaning that in 2035 which is the target when ostensibly we stop creating new ICE cars, we still won’t have that ability.
Building out nuclear capacity is still by far the quickest & cheapest option to not only convert our existing energy to carbon free but also to keep up with ever growing demands. Remember - the more capacity you build, the cheaper it gets per MW. Also, nuclear is way more land efficient than solar which is a separate unrelated discussion but also relevant. Again, my argument is not to stop solar investments but to remove all the regulatory and bueracratic roadblocks that inhibit nuclear fission so that those companies can compete fairly.
Denmark is an absolute best case because they are geographically advantaged for wind. Wind is much more resilient and cheaper than solar but is also more geography dependent. Fission can be installed anywhere.
> Yes yes. Renewable proponents really love trotting out costs while ignoring the only reason costs have gone up is because we significantly ramped down our nuclear production efforts. Same would happen if we did that for solar. But that’s a choice we get to make.
The huge difference here is that each individual nuclear power station is a megaproject while solar panels are mass manufactured items produced in the hundreds of millions per year.
> Also nuclear construction costs are particularly outsized in the US where regulations are intentionally constructed to strangle it. Japan, China, and Europe build it much more cheaply.
Europe is not building new nuclear cheaply or quickly either. All three EPR projects in Europe are hugely over budget (at least $10B per reactor) and more than a decade behind schedule.
China has been able to build new nuclear more quickly but the factors that make building new nuclear faster in China also make building new renewables faster:
"Every year, the combination of wind and solar, and usually both individually, outstripped new nuclear generation, both in raw nameplate capacity and in additional TWh of annual generation."
> Denmark is an absolute best case because they are geographically advantaged for wind. Wind is much more resilient and cheaper than solar but is also more geography dependent. Fission can be installed anywhere.
Onshore wind and utility scale solar seem to be pretty similar cost-wise now, which is cheaper will depend on the site. Northern Europe is really bad for solar given the northerly latitude, but southern Europe and the US are far more suited to it.
Offshore wind is expensive for renewables though getting cheaper. It's now about the same cost as gas (before prices went up when Russia invaded Ukraine) in Europe and under half the cost of new nuclear.
> Got any existence proof for this claim? We’re building renewables as fast as we can and countries seem to generally tap out at ~20% ammortized across the year and are growing extremely slowly (~1% each year).
Between 2000 and 2020 renewables went from 2.8% to 43.1% of UK generation. It's being built at 2% a year. That's about 25% in the time it takes to build a new nuclear reactor.
China is increasing the percentage of both solar and wind in its energy mix by about 1% each per year (much faster than nuclear.) In 10 years wind and nuclear have gone from neck and neck to wind being almost double, while solar has gone from nothing to neck and neck with nuclear.
> By comparison France runs 90% carbon free and that’s because of their nuclear grid.
I pretty much lost hope in nuclear when it turned out even the French can't build it cost effectively anymore. 20 years ago nuclear looked like the only realistic option. Since then nuclear build costs have spiralled while renewables have gotten cheaper and cheaper.
Somebody always wants to tout nuclear, even when that's not the topic at hand. Nuclear is equally unsuitable to solar as a stand-alone energy source. Nuclear energy is produced constantly throughout the day, and would not match the daytime peaks in electricity consumption.
Anyway, this article is discussing one of many new ways that power usage can be shifted around during the day, which makes solar able to provide a LARGER share of our overall energy.
No, this article is saying that if you have a hybrid grid of fossil fuels & solar, desal is a cheaper application than storing in batteries & using solar power banked in batteries at night. If you have a full renewable grid, that advantage disappears. That’s because the price differential comes from arbitrage & that price differential makes desal cheaper than building capacity.
I bring up nuclear whenever there’s a discussion of grid energy to provide a reality check that solar is not able to get us to net 0 by 2050 in any way regardless of clever new ideas. By contrast nuclear actually does have that capability and it’s not purely theoretical nor “hey what if we productionized some new hypothetical theoretical idea at massive scale and ignore all unintended side effects” - we have plenty of existence proof that nuclear is a drop-in replacement for fossil fuels & thus actually very quickly reduces the need for them in the grid energy mix whereas no such existence proof really exists for renewables (yes - someone brought up Denmark but that’s not a repeatable situation because they are one of the windiest places on earth and wind doesn’t have the same critical downsides that solar does for grid power).
>very quickly
Vogtle 3 just came online in July, 17 years after the initial application was filed. Unit 4 hasn't come online yet. They're going to cost more than double the initial estimate.
I just don't get the fascination with a power source that is so much more expensive and slower to bring online.
Because the problems of the US regulatory systems are not endemic to the tech. The fascination is we know it should be very cost competitive (in fact cheaper than fossil fuels), can run 24/7 (not true for renewables except hydro), and can fully replace all existing fossil fuel plant use cases (not true for renewables except for hydro), and basically has as much energy as we would ever need to build (not true for hydro as there’s only so many rivers in the world and same for wind which is similarly not evenly distributed), and can be built anywhere (applies more to wind than solar but still a little true for solar). In other words a lot of proven upside and little downside. The frequently mentioned ones are safety (completely solved by next gen designs of SMRs/thorium/etc) and cost (improved by SMRs but actually largely a political self inflicted problem).
Basically it and hydro are the only green tech that have shown the ability to completely displace the need for fossil fuels. With solar/wind we’re left with a hybrid grid and hoping that battery tech improves to the point where we overbuild enough solar that we can recharge batteries with excess capacity for nighttime use (the price of which is frequently ignored when discussing solar as a grid energy). In other words - it’s an open question as to solar and wind can replace base load requirements (which are substantial). If they can’t then we’re left with fossil fuels, nuclear, and wind. As we see in China, they’re building a crapton of fossil fuel (specifically coal) power plants that they will run overnight / emergency situations during the day. That’s progress but it’s a long way away from net 0. That’s why they’re also investing in their capability to build fission. that’s why you’ll see their fission projects finish on time and get cheaper over time. Fission is “more expensive” not because of the tech but because we don’t make enough of it to recoup economies of scale. You’d see the same effect with solar but because solar doesn’t pose as much threat to fossil fuels in the near term people don’t really care (it’s a very long transition period - I fully expect us to not achieve net 0 by 2100 unless we switch to nuclear). Keep in mind that solar and wind also can’t solve major transportation use cases with ships and planes which can’t be electrified (even an open question for semis but at least plausible there over time). Planes likely are the hardest but putting reactors that can’t meltdown onto ships sounds like a no brainer and would cut a huge green house gas emitter and an insane amount of pollution going into the oceans.
> Nuclear energy is produced constantly throughout the day, and would not match the daytime peaks in electricity consumption.
Nuclear is perfectly good at load following until the U is nearly burnt: the French do it.
The reason nuclear is most suited to base load is because the capital costs are so high it makes sense to use a reactor as close to 100% utilisation as you can.
Yes, someone really should. Germany is currently one of, if not the, worse polluters in the EU, by far. Compare it to France, which produces most of its electricity from nuclear power, and thus produces almost half the CO2.
Aiming is easy. Achieving your aims is an entirely different problem.
Given their track record, it seems more plausible that they'll close down wind turbines and replace them with more coal, like their "green" government has been doing with nuclear.
> freshwater tanks are cheaper than electrical batteries, it seems appropriate to fully exploit the ability of tanks to displace energy production/consumption, and then, if needed, to use electrical batteries for the same purpose
What about the cost of additional desalination equipment needed to take advantage of energy peaks? In a situation like this, is extra desalination equipment needed when renewable energy peaks? Or, do current desalination systems have periods where the equipment is idle?
I get this is useful for a huge overload of wind or solar power, but this doesn't address base load and offpeak generation usage.
The headline reeks of oil/gas companies pushing an agenda to keep gas turbine a critical part of the baseload infrastructure and keep grid storage from developing.
I think we should build giant desalination and water distillation plants in the American Southwest where there's a lot of sun for solar power. Then we can pump the water inland and reclaim enormous amounts of currently unusable land.
This is just naturally solved by capitalism, no magic.
When electricity on the grid is cheaper, all intensive applications can run at that time.
Just make the difference of price worth the pause of the industry when electricity is scarce.
Storing it lets you feed back that energy at a different time (e.g. night) but requires a lot of expensive hardware (i.e. batteries) + you have to deal with losses involved with storing & retrieving that energy. Consuming it directly doesn’t require batteries and desal plants are cheaper than battery plants + going to be fewer losses.
It’s a stupid comparison though because the point of batteries is to get to net 0 and desal plants don’t help with that. But it’s not a great sign for renewables achieving net 0 at the grid level that a process that is so energy intensive for getting so little water out is a better use of electricity than batteries.
If you live somewhere that desal is the way you get your drinking / irrigation water, it makes sense that rather than storing energy to run desal continuously you can run desal intermittently and store water.
You already need infrastructure to store fresh water for later use, and storing water for later use is relatively low loss (depending on the scale --- outdoor reservoirs can have a lot of loss). There's always a question of capital costs, of course: does it make more sense to have a higher capacity plant with a lower duty cycle where the duty cycle that uses mostly off-peak priced energy, or a lower capacity plant with a higher duty cycle which uses on-peak priced energy much of the time.
Most energy consumers don't care about getting to net 0, they want to pay less for energy, if energy is a large part of their input costs.
Intermittent operations do help get to net 0 though --- if all energy intensive uses could handle intermittent operations, it allows for more intermittent generation. It certainly doesn't get you all the way though; there's lots of uses that won't work well without continuous energy.
If you read the article, this take actually is not true if you have a grid powered fully by renewables. It only makes sense when you have a hybrid grid where baseload is powered by fossil fuels and the arbitrage difference makes desal attractive. Note that a nuclear grid doesn’t have this problem & it has enough capacity to reliably power everything + continuously run desal. To run desal off of solar requires a much larger capacity installed because you have to make up the lack of 24/7 operation with excess capacity whereas something like nuclear can make up the production gap due to smaller excess capacity by running 24/7.
Only because you have to overbuild more solar capacity than you need to with nuclear/hydro/fossil fuel plants which have a more reliable method of operation.
