And to address the other half of your question, the lower costs for solar and wind take into account the lifetime of the generation, and it's usually measured by the levelized cost of energy (LCOE) which is roughly the lifetime cost divided by the lifetime energy output. This does not take into account the value of the time that energy is delivered, so LCOE is not directly comparable for renewables and dispatchable generation like natural gas (though a huge percentage of the new solar and wind deploys for the next 5 years are adding storage to convert them into dispatchable resources.)
As for the disposal costs, that's a fairy trivial problem. Solar panel recycling has not really started because there's not many dead panels to recycle; lifetimes are tending to be longer than expected. However as there is a larger source of retired solar panels, those attempting to recycle them will get a better feel for the costs of construction of new panels from raw materials versus recycled magerials.
The issue with renewables isn't recycling, it's storage. Solar generates electricity in a sine wave. Half the day it's generating no electricity. Wind doesn't follow a sine wave, but has the potential to generate lower levels of power for long stretches of time. Wind is consistent on a yearly basis, but variable on a weekly basis.
That's a pretty big shift of the initial question, but it's an important shift!
If you look in the "most free" markets for energy in the US, where there is the most possibility for independent companies to come in and provide generation and other energy resources, we are looking at ERCOT in Texas and PJM in the north.
In these free markets, we are seeing huuuge amounts of storage resources in the interconnection pipeline over the next few years [1].
So where indecent investors are allowed to participate in energy, we are seeing massive influx of capital to build storage, from people that are profit-seeking. Meaning storage, at least this first chunk of it, is already likely to be economically efficient.
That said, when we talk about renewables, wind and solar are largely uncorrelated, and where we have good wind resources we tend to get good wind at night.
And when you look at load on grids, such as Texas' independent ERCOT grid, you see a sinusoidal pattern too [2]. And for ERCOT, with a sinusoidal amplitude of about 20GW, by sheer coincidence there are about 17GW of storage projects in the pipeline. And I'm pretty sure that doesn't include the solar projects that have storage directly on them, which is happening with a huge chunk of solar projects. And though Texas has been good at building wind resources, they have left their solar mostly untapped up until now. So as they start to deploy solar, they are going to get a ton of storage deployed with that too.
The challenge is not longer the technology or the costs, the future challenges will be entirely about regulation, and whether entrenched interests will use regulatory capture to prevent new entrants that can supply cheaper low carbon energy to consumers and industry.
What do you mean by 17GW of storage in the pipeline? I'm not seeing this in your sources. Watts isn't a measurement of storage, storage is measured in watt-hours.
To put the mismatch between available storage and required storage in perspective, the US consumes 11.5TWh of electricity daily on average. We have ~25GWh of storage currently. Most of that is in hydroelectric storage, which is geographically limited. Less than a gigawatt is in battery storage. This works out to a few minutes of storage - and a matter of seconds of battery storage.
Storage can be measured in watts as well as watt-hours. Doesn't help to have a 1TWh of storage if the output is limited to 1W. You need both. Not sure what the parent post was actually referring to though.
But watts IS actually important for balancing renewables. The storage is not so much about storing energy across days, as it is about frequency regulation. An interesting thing is - you'd think hydropower would be perfect for balancing renewables, but it's actually not on its own. You really don't want to ramp power up and down every second. It wears out the hydro power plant faster. Faster changes in pressure results in more wear and tear. So counter-intuitively, you're seeing some considering putting battery-storage in hydro power plants.
> Less than a gigawatt is in battery storage
It's funny how often numbers like this are presented with the intention of making renewables/storage look bad, when they're often incredibly impressive when you consider the context. The fact that battery storage has come this far, and most of it in only the last few years, is very impressive to me. The growth is exponential, so the fact that it's almost the same order of magnitude is promising.
> This works out to a few minutes of storage - and a matter of seconds of battery storage.
I don't think the whole country needs more than a day of total short-term storage. Especially if you combine it with more HVDC power lines. So doesn't seem that bad to me. Still a looong way from reaching the goal, but no solution is going to be complete in less then two decades.
For long term storage there needs to be other solutions. For those rare days/weeks where there's somehow significantly less solar and wind for a longer period over a larger area.. it might be a good idea to just keep some gas power plants around. If you only use them a few times a year, I think we can make enough CO2-neutral fuels for them.
For winter months in colder climates, you can also look to countries like Sweden. Make trash burning power plants that use its heat to heat nearby homes. Very efficient, and you can "store" power by storing trash. Southern areas can "transfer" power to the north by sending its trash there. Modern plants are incredibly clean burning. In Oslo they also have a demonstration that combines it with CO2 capture and storage.
ERCOT puts their panning pipeline out in the open; I'm relying on an ERCOT expert that said this as a summary of the August 2020 XLS raw data [1]. About 950 MW is as far along to have an interconnection agreement.
Grids think of energy sources primarily in terms of power, not in terms of energy, so that's the number that gets reported to ERCOT. Grid ops dispatch W, not Wh. We may never know the exact GWh for these planned projects, but since it's likely nearly all lithium ion tech, we can expect 1-4 GWh per GW for each project, and probably an average between 2-3GWh/GW overall.
Going from less than a GWh storage currently in the entire US, to probably 20x that planned for Texas alone in the near future shows just how scalable battery tech is and how quickly it is advancing. This switch is unprecedented, and is going to catch a ton of energy folks by complete surprise. People aren't used to things changing so quickly when it comes to electrical grids. The next decade is going to be a wild ride and I can't wait to see the grid drop carbon-based generation faster than nearly anyone is imagining. Next, we need to decarbonize industry.
PS: I originally replied to a "manfredo" and your username is "Manfredo_", which seems like a remarkable coincidence. Given the prevalence of paid astroturfing on these subjects, it would be good to clarify if this is just a coincidence, or if you have two accounts, or what else might be going on! Given our past discussions I'm sure your an honest commentator but it would probably help people that don't recognize your username.
4 hours of 17 Gigwatts would represent 20% of the ~300 GWh global lithium ion battery production [1]. That's a massive storage plant, and such a plant would represent an order of magnitude increase in battery storage. But consider the fact that the US consumes 11.5 TWh of electricity daily. Plans for an economy powered by wind and solar call for anything from 12 hours to 3 weeks of storage depending on how much overproduction we're building.
Manfredo_1 was the account I created on mobile, responding to a comment I made on desktop (password manager, so no way to log in on mobile). Everyone so far seems to have inferred continuity from the name similarity, but if it's a cause for confusion, I've imported my password to mobile.
It's not a single plant, and it's not all going to be installed in a year; these are the interconnection plans for all sorts of much smaller projects.
We are only at the infancy of lithium ion production, I'm actually surprised that we are at 300GWh. With this very limited capacity, and typical grid storage batteries being warranties for around 15 years, our current production supports a global deployment of 4.5TWh of storage. We are at 9 hours of global storage, and we haven't even really started scaling battery production.
I know you didn't put a label on these numbers, as being big or small, but we are on track to solve this problem with battery storage, it seems!
If your takeaway is that we're on track to solve energy storage with lithium ion batteries, then you have severely misinterpreted the situation.
Remember 300 GWh is global battery production. 4.5 TWh is 9 hours of storage for just the United States. Furthermore, actually using this quantity of batteries would entail ceasing the use of lithium ion batteries for electric vehicles and electronics - not a realistic option. Furthermore batteries only last a few hundred to a thousand cycles, depending on depth of discharge. These batteries need to be replaced every 3-5 years if they're being used for diurnal use (cycled 365 times a year), not 15. If you produce 300 GWh of batter every year and hook it up to the grid you'll only reach 0.9 or 1.5TWh of battery storage - the first batteries you add will have worn out before 15 years. This can be extended if you reduce depth of discharge. But that entails building an over-capacity of storage. E.g. if you're only withdrawing or storing half the energy you can increase the lifespan, but then you need to build 2 TWh for every 1 TWh of actual storage that can be used.
Global daily energy consumption is 60TWh. If we want to provide 12 hours of storage then this amounts to 30 TWh. If this is cycled daily, a battery with a life of 2,000 cycles (which is very, very good) then we'd need to be contributing 5.5 TWh of batteries to grid storage every year. Most lithium ion battery production is expected to be dedicated to electric vehicles, with grid storage making up only ~10% of produced lithium ion batteries [1].
And 12 hours of storage is likely only enough to get us to ~80% renewables. 100% renewables is estimated to require 3 weeks of storage [2].
We're not anywhere close to solving grid storage with lithium ion batteries, even if we assume its exponential growth will be sustained.
Existing storage lithium ion batteries have a minimum of 10 year warranties [1,2], and the Tesla Hornsdale battery has a fifteen year warranty. [3] it’s not as though the batteries suddenly stop working then, they just have less capacity. So 3-5 years of equipment life is not a realistic estimate.
Putting 12 hours of battery storage right next to 3 weeks just makes the 12 hours look suspect too. 3 weeks is just crazy numbers from people that haven’t studied the problem much. Most zero carbon grid modeling selects far far less than 12 hours of battery storage.
But even if we stick with that 12 hours of world energy of batteries, 60 TWh is only 5x more than the 12 TWh of US energy. I apologize for misinterpreting your stats and being off by 5 in the comparison of numbers, but a 5x or 10x or even 100x growth is well within the range of sustained exponential growth.
As grid storage plus solar becomes cheaper than natural gas, we will hit an inflection point that will drive massive expansion of battery storage production capacity, even beyond what is currently being built for EVs.
Will there be a massive interchange of technologies and production capacities. Yes, it will be huge, so from one perspective we are a long ways away from our target, but from the perspective of our current trajectory and from the perspective of what is currently being planned just for EVs, an entire industry is being built. It will take a little while for energy guys to wake up to this new reality because they are a slow changing industry, but they will be brought along for the ride whether they know it now or not. There’s a good reason that NextEra surpassed the market cap of Exxon recently, as the market is finally waking up to the new reality too.
As for the disposal costs, that's a fairy trivial problem. Solar panel recycling has not really started because there's not many dead panels to recycle; lifetimes are tending to be longer than expected. However as there is a larger source of retired solar panels, those attempting to recycle them will get a better feel for the costs of construction of new panels from raw materials versus recycled magerials.