World’s Largest Solar-Powered Battery Is Now 75% Complete | Hacker News Reader

World’s Largest Solar-Powered Battery Is Now 75% Complete

newsroom.fpl.com

Hmm, so this 900 megawatt hour lithium installation will roughly use about 73 metric tons of lithium at 0.3 grams per amp hour. World lithium production is about 82 thousand metric tons with plenty of room for future expansion.

Using a single year's supply of lithium we would be able to produce 1 terawatt hour worth of battery as an upper limit. Currently the world uses about 64 terawatt hours daily on average. Assuming we would only need storage for 1/3 of that amount, we would need roughly 20 terawatt hours of storage.

What this means is we would need to dedicate 20 years worth of current lithium mining to produce the needed batteries or expand lithium mining by a factor of 5-10 to meet climate targets.

Back of the napkin, this seems practical in the long run, but difficult in the short run. With recycling we can build up a reserve of lithium over the years, however I think it's going to take half a century just to get the lithium we need to replace fossil fuels. In the meantime we'll take a small chunk out of fossil fuel demand. However demand is still increasing.

We may want to consider subsidizing nuclear at this point to get us over the finish line on renewables. Unless we come up with another carbon neutral energy source that provides consistent power 24/7.

> Assuming we would only need storage for 1/3 of that amount, we would need roughly 20 terawatt hours of storage

Out of curiosity, why 1/3?

I think current estimates are something along those lines. For a residential off-grid installation they suggest 100% kwh/day of battery storage. Grid power has backups and mitigation strategies that could reduce that significantly. I was mostly trying to be generous to grid battery installations. For the most part 1/3 is a bare minimum still requiring natural gas emergency peak plants.

There are other battery types such as sodium, liquid air, liquid metal, air-metal and so on.

Doing everything with lithium is not a great idea as we need that for cars mostly.

But I also always preferred nuclear.

For sure! For sure! I imagine if we're lucky we could get a grid scale chemical energy storage solution in the next 10 to 20 years that will make lithium seem like a waste of money. I was mostly curious as to how viable these grid scale batteries are materials wise right now.

Currently, lithium batteries are the only storage medium that is efficient enough (~90%) to allow photovoltaic solar to compete with nuclear and fossil fuels. Pumped water storage is okay at 80%. Molten sodium on those cool solar thermal generators in the desert is pretty good since they don't add a step to the generation process.

Liquid air is 50% efficient when used with heat exchangers, molten salt plants are at about 70%, and gravity is about 90%. Gravity is efficient but supremely low density requiring huge installations for even a few megawatt hours. As panels get cheaper and more efficient, these storage mediums will become cost effective, but for now are relegated to neat research projects and demonstrations. I imagine that molten salt will end up being our goto in 15 years or so if nothing cool like metal air batteries comes out before then.

Right now it seems lithium installations are only at a scale for grid stabilization when paired with natural gas peakers. However I can see that changing a lot as power prices go up from carbon credits and we start manufacturing more cells.

But yeah, I prefer nuclear too. We can do nuclear now. Modern designs are just waiting to be built and tested. Ones that can also do peaking like the natrium reactor. What's more, they will allow us to grow our industry at the same time without adding more carbon to our atmosphere.

I feel you undersold pumped hydro. This proven technology is cheap, efficient and at scale.

If any grid needs energy storage and has suitable undeveloped sites available, they'd be foolish to consider any other alternative you mentioned.

The significant issue with ph is site availability.

Your words seem measured and reasonable, yet your final paragraph makes no sense at all.

How does nuclear help avoid the need for battery storage on a carbon free grid?

So many people claim this, and yet as I say it makes no actual sense. Has anyone attempted to flesh this proposal out beyond a snarky comment every time renewables get mentioned on a news article?

I'd love to be proved wrong, but as far as I'm aware there's no way building any more nuclear today makes sense unless you believe that solar and wind renewables and battery storage and green hydrogen are some kind of elaborate hoax that doesn't really exist or work.

So what are you proposing exactly? How does subsidized nuclear prevent the use of storage on the grid?

This battery farm, has a capacity of 900 MWh. While a typical nuclear power plant has a capacity of ~3000MW. So this battery farm is just... tiny.

But solar power generation has a large potential for larger capacity, the shortest leg is energy storage.

So if something else can be used for base load (e.g., nuclear, natural gas, hydro), solar power works well for daytime peaks.

Please try to sketch this out. Grab some daily production and consumption numbers from somewhere and then increase the nuclear (basically shift it straight up) and/or increase the solar (stretching upward is a rough approx but we can also flatten it) to levels that cover roughly 100% of demand. Look at the bit where both are running at the same time and compare it's area to the rest. Why is this duplicated "baseload" useful? Try it again with wind, which is much cheaper and faster to build than nuclear. The conclusion should be fairly obvious.

It's not difficult to accumulate enough capacity for solar / wind farms, I agree. But the problem is that we have not yet solved the problem of energy storage (as I said, it's tiny, compared to the capacity it can generates). So you need something to cover the night or windless periods.

Did some search online and it seems that distributed wind farms may work as base load but I highly doubt it works anywhere due to seasonal winds. It definitely needs more research. Especially extreme weathers are becoming more frequent these years.

A nuclear plant is constantly providing energy at any time of day and can be scaled up/scaled down quickly with no need for storage. The power plant is also drastically more energy dense in terms of MW/km^2 (although that’s an important secondary benefit).

I would think that means your storage needs should go drastically down because you don’t need to store energy at all like you do with solar during sunny moments to handle night/clouds/inclement weather.

That’s an obvious explanation that comes to me when I think about it a bit, so I’m assuming you must have more expertise that indicates a flaw with that line of reasoning (I’m not super familiar with technical details). Interested to hear your viewpoint.

Edit: maybe you’re thinking of demand smoothing? Nuclear might need that to completely remove the need for carbon peaking plants, but my suspicion is that the batteries you need are drastically smaller than for solar because nuclear can natural load follow whereas solar can’t.

So nuclear can "naturally" load follow by overbuilding capacity and then just throwing away some of that energy when its not needed. But wind and solar can't do exactly that same thing?

Modern Nuclear plant designs can throttle and peak. Even older reactors can react pretty well to changes in demand. Solar produces 0% at night and wind can vary dramatically with calms lasting for days. If wind calms at night then you are on either stored energy, peakers, and/or black-outs. Ideally we would use stored power. But, at the current rate, even accounting for expansion of supply, we can expect that it will not nearly be enough to meet climate targets.

https://www.wind-energy-the-facts.org/understanding-variable...

That link doesnt seem to support your claims about wind lulls, stating almost the exact opposite but let's stick to nuclear anyway as thats far more interesting.

So in this alternative you propose, with the modern peaking nuclear reactors. What kind of capacity factor do the nuclear plants hit overall and what are you using to fill the seasonal peaks? Nukes that run only 1/12th or less of the year? It all sounds very costly if we're building lots of nuclear plants that run below their rated max. That's really going to put the price per kwh up.

Maybe we could build HVDC lines to carry the power long distances from the plants. If you had two cities with a plant each but seperated by a few tinezones you could send the power from the one with the night tine lull to the one with the daytime peak. That would bring costs down I suppose.

Or use the excess power for making hydrogen. Its not super efficient, but if your only alternatuve is a nuclear plant that could be generating more then it'll probably work out financially.

Or use the excess energy to generate carbon free fuels for airplanes.

This is starting to sound like an actual plan.

I could be wrong so please correct me. To my knowledge the generation capacity of renewables is dependent on the surrounding environment whereas nuclear capacity is constant 24/7.

Also, I don't believe nuclear power plants just throw away the energy vs just consume less fuel [1]. Finally, you can have a mix of nuclear reactor designs to handle load follow needs in addition to base needs, proving it more versatile.

Unrelated, but also important, nuclear can handle industrial needs that depend on extreme temperatures that are out of the range of renewables and is why, for example, Germany coal usage has remained constant/increased despite an increase in renewables (they phased out nuclear) [2]

[1] https://www.oecd-nea.org/ndd/reports/2011/load-following-npp... [2] https://www.ted.com/talks/michael_shellenberger_why_i_change...

Wind and solar also "just use less fuel" when you throttle them. The point is that for all three there's no real cost saving from doing so, as the marginal cost is very low and so finding a way to use it productively is a better option.

For example, many nuclear plants generate excess energy at night and various countries have used this otherwise wasted energy to refill pumped hydro or store heat energy in electric storage heaters or water heaters they turn on with ripple signals through the mains. Off peak price signals give incentives for power users to shift demand away from peaks to better match nculear's flat profile.

This is decades old tech in wide use across the globe. You'd think any serious attempt to advocate greater nuclear use would mention it.

At the current rate, it will take us at least a century to replace fossil fuels with battery and renewables. Nuclear makes sense in the short term until we can replace it with renewables. If we continue as we are, replacing nuclear with natural gas instead of the other way around, we will never make climate targets in time to stop catastrophic levels of CO2 emission.

This is despite gains in solar and wind generation. Battery, solar, and wind make sense in the long run. However we are running out of time. We want to limit change to 1.5 celsius and we will never make those targets at the current rate. We are reaching limits of solar and wind alone in a lot of countries and storage solutions aren't being created nearly fast enough.

In addition, if we are going to start capturing carbon from unavoidable emissions like concrete production we will need to increase our energy usage dramatically.

Essentially, we are VERY underwater and we need something to help us stop sinking to give us time to swim our way out. Nuclear is already playing a big part in that, but not nearly enough. As it is, we could have a chernobyl every year and still not see the level of devastation that is expected to occur with climate change.

We are not reaching the limits of solar and wind in any country, this is just not true.

And even if it was, you have not given a single reason as to why building new nuclear would help in that scenario. Talk me through it please.

At a certain point, base load has to be met during low wind nights. Renewable energy is cheap and efficient. However it doesn't meet base load requirements without substantial storage. Nuclear does meet those requirements. Currently we are using natural gas. Even if you had 10x the necessary capacity of solar, you would still need some sort of night time storage which currently doesn't exist. In addition you need something to offset power losses and increased demand in higher latitudes during the winter.

Baseload isnt a real thing. Its just a conceptual framework for combining nuclear and coal with gas peaking plants.

We need to ditch the gas and coal. I'm happy to keep the nuclear as long as someone can explain how it actually helps achieve a carbon free grid. But apparently no one can.

There are two sides to explaining something. A providing of information and comprehension. If someone is unwilling or unable to comprehend something then no one can explain to them how something works. In that way, no, I cannot explain this to you.

> Baseload isnt a real thing. Its just a conceptual framework for combining nuclear and coal with gas peaking plants.

Most people will always use electricity when the sun is not shining and the wind is not blowing. Offices are still open, air conditioner / heater are still running, and factories are still operating. This is the "baseload" they are talking about, and no it is not fictional.

Wind and solar fluctuates between 0 and MAX_OUTPUT. Power usage also fluctuates, but between BASELOAD and MAX_CONSUMPTION. These two are not in sync, and even if they are, you still need other power sources to handle the BASELOAD part of it.

Nuclear is able to generate a stable output power. If we have enough nuclear power plants to generate BASELOAD amount of power, we only need renewables to handle the remaining MAX_CONSUMPTION-BASELOAD, and we will need much less battery to handle the fluctuation.

See my other comment about graphing the overlap between "baseload" and solar. What are you pairing this nuclear baseload with? Gas?

At least make an effort. What about, we build enough nuclear to cover more than the yearly total energy demand, and run them all the time. Then when demand is low, like at night we use the energy to generate green hydrogen (apparently also called pink hydrogen if you use nuclear specifically).

We then use that hydrogen in converted gas turbines to meet peak needs above baseline and we sell the rest to farmers to make fertilizer and manufacturers as a feedstock and fuel.

How possible is this? Very possible. Now, replace the nukes with a similar output of solar and wind. And locate some of it in the sunniest regions of the world and ship the hydrogen to customer by boat or pipeline. Is it much cheaper. Yes it is. So why do we need nuclear?

> At the current rate, it will take us at least a century to replace fossil fuels with battery and renewables. N

While I am as much of a pessimist as you are, you also need to recognize that the growth in solar, wind, and storage capacities is _exponential_, with the exponent being _dramatically_ higher than even the most optimistic of expectations.

Keep in mind, using current technology, we would need to mine 3-5x our yearly output JUST for grid energy. This doesn't include cell phones, laptops, electric bikes, and Teslas. Even then it would be the bare minimum just required to replace baseline natural gas and coal at the current power usage. Not including replacing peaker plants for winter and heat waves.

Sure, we are increasing battery production pretty fast. However mined lithium has gone down in the last few years while demand has soared for non-grid batteries. We need either a new storage technology, a vast increase in mined lithium, or a base load generator which can compete with natural gas like lithium does.

So far we have 0.1% of the grid scale batteries we need. It took 6 years to go from 6 gigawatts to 28 gigawatts. (At a power to storage scale of 0.5 you can roughly convert to energy storage by doubling the power) Even at an exponential rate it's still going to take decades. Eventually we will hit the top of the logistic curve when we find a limiting factor to production. Current estimates are 52 gigawatts by 2025. Roughly a doubling rate of 4-5 years.

Assuming a 5 year doubling rate with no logistical limitation, an exponential expansion would reach necessary capacity in 44 years. Which would be awesome, if we can manage it. However, in 44 years, power demand will have increased as well. Currently we are not building enough renewables to even touch the increase much less batteries.

Year - Power - Percent of 10tw

2025 - 52 gw - 0.5%

2030 - 100 gw - 1.0%

2035 - 200 gw - 2%

2040 - 400 gw - 4%

2045 - 800 gw - 8%

2050 - 1600 gw - 16%

2055 - 3200 gw - 32%

2060 - 6400 gw - 64%

2065 - 12800 gw - 128%

Did you just switch from solar / wind, which is totally workable and cheapest form of energy, to arguing about batteries?

Batteries for storage are problematic, especially grid scale lithium - globally unfeasible regardless of nuclear.

We need to move a lot of consumption onto electric (or direct solar) so we will need to manage the time and location aspects of consumption - easier with nuclear (if you get it built) but not a deal-breaker as argued elsewhere. Technology already exists - least sophisticated one is called a switch.

Solar by itself is already rated at 3.7% of world electricity consumption, so your table is .. grid batteries? Yes, unworkable and besides the point.

   y    GW cap    % of 30TW
  2020     700    2.3
  2025    2200    7.3
  2030    7000   23  
  2035   21000   71
  2040   65000  210

If we keep the pace of x1.25 year-to-year increase, PV solar achieves global domination by 2040 and Li-ion is considered an archaic technology (at least for grid scale load shifting).

The challenging problem is not storing the electricity produced during the day to consume it at night, but to store electricity produced during summer to consume it during winter. At latitudes such as Paris, or London, or Seattle, or higher, the summer day is much longer than the winter day, and the rays come from an angle much closer to the vertical, they go through a thinner layer of atmosphere so the end result is that solar panels produce about 6 times more electricity. But people consume on average more electricity during the winter days. You can address this problem by building massive overcapacity, but this is expensive (solar power may be cheaper than natural gas power, but not 6 times cheaper). Or you can try to find a way to store the summer electricity for 6 months. That is not easy either.

I think there are solutions like high voltage DC transmission that could help with this. That would allow power sharing schemes over ~1000 mile distances without too much loss. But yeah, aside from HUGE overprovisioning in southern latitudes and monstrous transmission losses, we may be stuck with nuclear for northern countries and states until something new comes up. Also I don't see Russia being happy with being dependent on middle eastern, Chinese, or European power generation during the winter.

On the other hand, Canada seems to do pretty well with wind power despite the northern latitudes. But it is still only 6% of their winter power output.

Leaving out wind and solar during the winter, you are kinda left with geothermal and nuclear as your only low carbon options.

> monstrous transmission losses

When people claim monstrous transmission losses, I'm not sure what to think of. Is it 80%? Maybe 50%? Is 30% too much of a hit for reliable clean power? Surely you're not hung up about 15% for a good PV spot?

Transporting HVDC across the globe raises questions about efficiency, but for 1000-2000 km I suggest you compare it to the efficiency of batteries and report which one has better efficiency.

> HUGE overprovisioning in southern latitudes

If you design it well, it's not overprovisioning, it's displacing production from consumption. Which is mostly problematic for socio-economical and political, not technical reasons.

I think HVDC losses over 1000 km are about 10%. This is in addition to power storage losses for responding to grid variability. Grid sharing makes a lot of sense but it will only get you so far. Not even considering the political situation of northern countries being dependent on southern countries. In addition it would turn countries like Turkey, Greece, and Ukraine into strategic assets leading to wars of acquisition to secure energy for Russia and northern Europe.

However you are right about it being a socio-economical and political problem rather than a technical problem. We could totally just build a ring of solar around the tropics and transmit it north at 50% losses over 6000km. If Solar with battery reaches half the price of natural gas I could totally see it happening as energy demand increases.

From https://en.wikipedia.org/wiki/High-voltage_direct_current :

> HVDC transmission losses are quoted at [around] 3.5% per 1,000 km, about 50% less than AC (6.5%) lines at the same voltage

1000 km of HVDC has lower losses than a Tesla powerwall or almost any battery system I know of; likely less than the last 10km of transport to your device; also less than most AC->DC converters.

Transport over 6000km would be about 20% losses, which is notable, but not too bad and could offset the need for quite a bit of storage. Though note that grid scale batteries are not workable. We need other solutions - create hydrogen, methane, ammonia/fertilizer, smelt aluminum, iron, cook concrete etc.

The cheapest source of electrical energy has been onshore wind farms for like a decade and they've been improving since then while most alternatives have become more costly. Why dont you just use that in the winter?

Yes. In some regions, this is viable.

One issue is that we'll soon use up all the low-hanging fruit of windy NIMBY-free onshore sites.

Offshore wind is progressing rapidly. It should be ready to scale up just at the point the issues above get critical.

The largest solar battery capacity in the world is filled by less than one hour of production from the smallest of the 52 French nuclear power plants. Guys, that's going to be a lot of batteries. And by a lot I mean a SHIT-TON of batteries.

Nuclear is not what they're looking to replace with batteries, so comparing to that doesn't make sense. How many hours of smallest French coal power plant?

By far the best metric would be "how many minutes or seconds of French (or possibly regional) electricity consumption can this battery cover".

This gives you a rough idea of the total bridging capacity for the case of "no sun, no wind" situations.

That said, I must admit I am a bit afraid of the huge batteries, at least if they are going to be located close to human dwellings. The videos of fires of vehicle batteries that I have seen so far were pretty bad, and here we are talking about orders of magnitude bigger energy store.

I think a highly useful metric would be how many homes this battery installation could power alone for a one day period.

Homes are just one thing that consumes energy, and they're no more important than street lights, factories, night clubs, gas stations, electric charging stations, and hotels. there are no bonus points for renewable suburbs

In an excessive heat wave, blizzard/ice storm, other severe inclement weather events being able to power homes directly is frankly all of the points.

Out of all the users of energy that you listed perhaps the industrial sector is worth mentioning, otherwise street lights are essentially in the same magnitude of lights-per-capita so that can be a constant tacked onto the # of homes metric. Not to mention the fact that quite a substantial portion of them now use LEDs. The others are a magnitude less significant in comparison to homes, whether those homes are rural through urban.

Precisely. New renewable energy installations should replace the worst offenders not the other clean energy generation infrastructure. There was an article on HN a few days ago about how the 5% most polluting power plants pollute as much as 75% of the total.

My previous thought provoking argument was to say that if you plan to produce electricity in a developed country using mostly wind turbines and solar panels, you'll need an unreasonable amount of batteries.

In that sense, it makes totally sense. If you are Greenpeace (for instance) and argue that all nuclear plants have to be close (for reason X) you need to translate that production with something else. And that amount can't be reasonably replaced by solar + wind + batteries.

> Nuclear is not what they're looking to replace with batteries, so comparing to that doesn't make sense

Some people, including many "greens", do want to replace nuclear with renewables + storage (at best, the other day Greenpeace Belgium said that "natural gas is a necessary evil for transitioning away from nuclear and to renewables" (as in, we should retire nuclear ASAP to move to natural gas until renewables are ready), so it does make sense.

Thankfully there are places with a "SHIT-TON" of space throughout the USA and Canada.

This add to the land competition for farming, solar panels and wind turbines. Also, remember the World is not USA + Canada, and there are plenty of ppl living in the World. The battery solution might likely not be the magic wand to will solve the energy problem as some might think.

I haven't forgotten the world is not just USA + Canada.

The rest of the world will set up their own primary energy storage providers as well. Imagine each continent had a primary battery hub. Secondary storage can be in the countries that have the space or can afford the additional infrastructure.

Also, remember this is how a lot of the world works now with existing energy infrastructure.

Don't try to make the simple complex. Solar "tunnels" where existing roads are. Vertical building for batteries in space restricted areas.

I don't understand why people see this as a problem. Energy generation being a local thing feels like a good thing -- sure you'll need a lot of batteries, but every city/state/municipality already has power plants, sewage plants, water treatment plants, etc. If we get to the point where energy is just this easy thing where you buy off-the-shelf batteries and have some solar panels or wind turbines that need replacement every now and then, that sounds way better than investing billions over decades to build a centralized fleet of nuclear reactors.

> I don't understand why people see this as a problem. Energy generation being a local thing feels like a good thing -- sure you'll need a lot of batteries, but every city/state/municipality already has power plants, sewage plants, water treatment plants, etc. If we get to the point where energy is just this easy thing where you buy off-the-shelf batteries and have some solar panels or wind turbines that need replacement every now and then, that sounds way better than investing billions over decades to build a centralized fleet of nuclear reactors

Not every locality will have the place, budget, geography to enable that.

Furthermore, "a lot of batteries". Isn't there a shortage of battery production capacity currently? Isn't there a massive shift to battery electric vehicles which will need multiple times the battery production capacity available today?

Wasting enormous amounts of space, money and production capacity to be able to offset a couple of hours of no wind/sun seems wasteful and risky.

The shift to EVs drives down the cost of batteries and scales up manufacturing capacity. Nuclear is never getting built, takes too long and isn’t cost effective against storage and renewables. There is more than enough unused space worldwide for solar, wind, and batteries without impacting more productive land uses.

The only thing holding back renewables is cost -- particularly that there aren't really off-the-shelf battery storage solutions yet, the reason why this article is a headline is in itself indicative that it's still immature as of yet. However, it's obvious that this is going to be solved soon, the actual battery tech is there.

and if cost is the main thing holding back renewables, nuclear will never win, because it's an order of magnitude more costly. It takes a LONG time for something that costs a billion dollars to break even versus renewables that cost $100M.

Woah. An article with MW and MWh in the subheading.

They're running a 2:1 capacity vs power ratio, so this is likely a demand smoothing installation rather than a demand shifting one.

> demand smoothing installation

Depending on definition of smoothing vs. shifting - yes, most of them are. Demand shifting with current batteries is not really workable at scale - neither economically nor practically.

I may have missed it in the article. What chemistry does it use?

After reading the headlines I question myself: What is the eqivalent in bathtubs filled with iPhones?

But still an interesting project..

Reply on news.ycombinator.com