Yes, which is why we need to (as a civilization) put more effort into grid-scale energy storage.
My leading candidate for that is some kind of flow battery. There have recently been some announcements about a better and more stable electrolyte that is also non-toxic and non-corrosive. But as always, it might be five years (or never) before that gets commercialized.
We'll need a lot more storage and overcapacity than people think.
For example yesterday germany had a renewable fraction of around 50% throughout the whole day. Yay? Nope. During january there was a whole week where it was in the 0-10% range during the night and 5-20% during daytime.
That means we don't just have to smooth out daily variability but on the span of multiple weeks.
Of course that doesn't mean we need to store all of that in batteries or that we need to keep a whole shadow infrastructure of fossil/nuclear plants around. With enough overcapacity from renewables we could also produce hydrogen and store that for example and run it through fuel cells or gas plants when needed.
But that still means that a bunch of smartmeters won't fix the issue, we'll need real infrastructure to handle that variability.
It's not as bad as you think. The amount of storage capacity required is easily computed from known long-term weather patterns, and we can (and necessarily will) work toward it slowly. "More than people think"? Who cares what people think. It's still cheaper than fossil in the long run, and energy investment is long-term investment.
I'm not convinced that it needs to be explicitly computed in a feedforward sense. If you just let the price float, then price will naturally form an error signal that the energy storage market (and customer demand) can respond to.
Cold snaps certainly do happen even in climates that normally only justify a heat pump. I grew up in rural SE Tennessee, so I totally get that.
Even in an environment of electric heating and an impoverished customer base, the price doesn't have to be totally inelastic to avoid a social justice nightmare. You can choose not to run the electric clothes dryer in the middle of the night, for example. Having regular large price swings will be accurately perceived by the energy storage market as an arbitrage opportunity. Even if you only allow the price to swing by some limit (say, by enforcing a maximum price ceiling on the utility through the state regulatory commission), the electric utility will be incentivised to pay even more than the natural swing to third parties who can provide that arbitrage.
There's definitely a vein of "smart grid" technology that is based around the utility shedding the customer's load for them, in a sort of fractional rolling brownout. I don't think we need to go there, and that most folks would be rightly pissed off at the imposition of such by their mammoth utility.
OTOH, I could see a Nest-style device which can read the day-ahead price and adjust for the customer based on their own preferences being perfectly acceptable.
Note that load shedding can include changing the temperature by a couple of degrees. So, not only can it be an opt-in or opt-out situation instead of being "imposed", it can be something that most folks would sleep right through.
There's no need to "impose" this. It provides economic value, so it can either be seen as a discount for opting in, or an extra charge for opting out. People already do this in a manual fashion with time-of-use rates.
Fossil fuels and nuclear carry their own weight, unlike, say, wind. Here in Finland, there would be no new, not any installations of wind without massive subsidies (we're talking about a guaranteed price more than 2 times the market prices) – just to disrupt the working grid as it is. (Solar isn't really even worth talking about here...)
Coal is dirty, and dirt cheap, but wind is simply too inefficient.
For fuel-based plants, you have the cost of the plant, the mine, the transport between the mine and the plant, as well as the clean-up. That's quite a lot of operating cost on top of the capital cost.
For a network of solar-panels and domestic batteries, it is almost all capital cost. There is no fuel transport, no mine, etc. And whereas one massive plant with one massive mine is a decade-long planning and construction process, the domestic panels and storage are each very short installations that can happen largely in parallel.
The domestic solar+storage grid should win, because it is a heck of a lot faster to marshall consumers into buying a consumer good, than to plan and build out nation-wide infrastructure projects.
Its not the infrastructure, its the price. If you let the price float more, then customers will naturally avoid buying electric energy when it is expensive, and will shift some of their demand to the less expensive times automatically. Energy storage naturally becomes an economically decentralized business that follows any other arbitrage pattern.
A floating price means smarter grids. Smart grids that don't exist yet in large swaths of the country.
We gotta upgrade all of those electrical meters in everybody's homes to take into account time AND price changes. We've gotta build the algorithms to distribute those price changes.
A lot of cities have that kind of infrastructure already of course. But lets not deny the monumental task ahead of us. The idea of "smart meters" that communicate with a central server (ie: requires perpetual and reliable internet access) was unfathomable just 10 years ago.
I don't think the meter needs to be that smart. First off, I'm using the existing day-ahead markets that many of the independent system operators are already using to predict demand. These markets clear once per hour, for the price of energy 24hr in advance in one hour windows. They also account for somewhat fine-grained regional variation (ie, power line usage). There are additional markets that clear only 15 min in advance (the "spot" price for power) and these can have much more volatility.
So, if I were king of the electric market, I would only expose retail customers to the day-ahead price. This way, existing web-scale publishing tech can tell people what their price will be up to 24hr in advance. Retail customers' price gets fixed at that time, and they pay a small energy premium for reducing their volatility exposure. Now, the meters only need to record a list of {kWh, hour-of-day} pairs, one for each hour between the last reading and today. So if the local electric utility reads off the consumption once a month, then you only store ~700 entries or so. That's much cheaper than needing a reliable Internet connection to the meter.
Anything smarter than that is considered a "smart" meter. A dude comes out one a month to see how many KWhr the meter has ticked up, then they send you a bill in the mail for you to pay.
All this talk about "spot" price and "day of pricing" simply doesn't work with current electricity meters that are deployed around the country.
And honestly, if we're going to be upgrading to "smart" meters... we should all have "smart bi-directional" meters to account for the ability to "sell back to the utility". (IE: Net Metering).
I haven't seen those meters in years (in the US, at least)
For as long as I can remember living in a suburban house (since ~2008), there was a smart meter. At some point it was replaced with another smart meter.
Neither have I seen a person come out and manually get readings, because that seems incredibly inefficient, time consuming, and costly for the utility, so it was probably one of the first things to get automated.
The smart meters are medium-range wireless for the most part. The utility still has someone drive around in a truck, but they just need to drive by. The driver doesn't need to get out or even stop.
What would the picture look like if you included a wider area? Germany is relatively small, and geographical diversity can substitute for storage to an extent.
I recall seeing a blog post about that issue, assembling publicly available data, and they found that even over larger geographical regions there were still fairly significant lulls lasting longer than a day.
I look at maps like this http://www.electricitymap.org/ which show the solar and wind potential across Europe along with imports and exports between countries.
It makes me wonder how much more of a baseload renewables could provide with a more integrated energy market and continent wide capacity planning. At least in the US, electrical grids are highly fragmented and poorly connected between states and regions.
When Germany was in the 10% range, was there a region that could provide high amount of wind generation to pick up the slack? Or on sunny, windy summer days when electric prices go negative, is there somewhere to put that energy?
https://energytransition.org/2014/05/german-power-prices-neg...
I noticed in Germany, there are solar panels on houses and barn roofs everywhere, even in not so sunny regions - maybe that investment would be better put into grid scale wind, while Spain and Portugal build out grid scale solar.
If investment in renewables took place on a continental scale, with high capacity connections between solar installations in Spain, wind turbines in the North Sea, nuclear plants in France, renewable and lower carbon utilization could go up and some of the gas and coal plants could be spun down.
> If investment in renewables took place on a continental scale, with high capacity connections between solar installations in Spain, wind turbines in the North Sea, nuclear plants in France, renewable and lower carbon utilization could go up and some of the gas and coal plants could be spun down.
Yes, but countries don't typically like being dependent on each other for their energy needs. We are even in an era where people are openly questioning the future of the EU due to the rise of Euro sceptic parties in France and the Netherlands.
The power grid operates at a snail's pace compared to the rest of the high tech industry. Getting the permission to build something, passing all the environmental impact assessments, financing, etc, can take years before the first shovel of dirt is moved to build the project.
> I noticed in Germany, there are solar panels on houses and barn roofs everywhere, even in not so sunny regions - maybe that investment would be better put into grid scale wind, while Spain and Portugal build out grid scale solar.
This I do agree with. Financially, I can't see how it makes sense for someone in Germany to install solar, when places like Spain or Greece get far more sun in the year. It simply a question of location, and northern localities such as Germany receive much less annual irradiation than more equatorial localities do.
However I think most people in Germany would install solar not for the electricity, but because they are conscious of the environment and believe that by installing solar, they are reducing their greenhouse gas emissions.
Ökostrom (ecological electricity) is very big in Germany.
Ökostrom is indeed big here, but the solar thing seems logical enough for me if you look at the policy picture: The German government needs to reduce the country's overall CO2 emissions by X percent over the next Y years to keep political promises. I mean promises in both the sense of treaty obligations, and also social expectations from its own citizens as you alluded to.
And it has to do that at the very same time that it is constructing new coal-fired power plants to keep the base load power it will lose by shutting down its remaining nuclear plants. So there is some urgency to this. Rooftop PV is really the only option that can be quickly deployed en masse through incentives, right across the country, skipping all that red tape you mentioned.
If solar in Germany can be very effective for half the year, then it is still going to make a sizable impact on those annual CO2 figures. So obligations are being met on paper while also having a pretty visible presence on the ground in cities and towns.
> With enough overcapacity from renewables we could also produce hydrogen and store that for example and run it through fuel cells or gas plants when needed.
Aren't lithium-ion batteries more efficient than hydrogen fuel cells?
More efficient, but also more expensive and harder to scale.
To expand battery capacity you need to dig up lithium and half a dozen other rare elements, deal with the toxic byproducts (or dump them in a lake), manufacture batteries, control circuits and run ongoing maintenance.
To expand hydrogen capacity, you need more pressure vessels. These can be made of steel or aluminum, two of the cheapest materials in existence, and with practically zero pollution. Ongoing maintenance amounts to checking for rust and occasionally replacing valves, so it scales far better.
Hydrogen has serious downsides, but there's definitely a case to be made.
Lithium's not just a little bit more efficient, it's like 2x more efficient.
I think converting H2 to methane and storing that is far more realistic than storing H2. Methane slots into all of our current natural gas infrastructure. Liquid natural gas cars are more practical than H2 cars.
However, we're going to be harvesting TWh-worth of lithium for our car batteries anyway, so we're going to have a huuuuuge secondary market of used car batteries that still have a consistent 50%-80% charge capacity that could be used for grid storage.
I think it's too early to see exactly how the technology will work out, but a huge lithium-ion economy seems inevitable. I truly hope that electricity -> H2 -> methane (or a carbon chain) happens though. Because that is carbon sequestration, and I think that in 2050 we're going to want to start paying people to do that and not let the result get emitted again.
> To expand hydrogen capacity, you need more pressure vessels.
Even simpler, you can store it in salt domes which are currently used for natural gas reserves. The pipes and related infrastructure may need to be replaced due to hydrogen embrittlement, but the storage itself is already there.
At the risk of sounding like an idiot, would there be any value in capturing some of the energy as compressed air/purified oxygen to be injected in to the hydrogen burning process... thingy? Perhaps in a manner similar to how a SABRE engine works? [0]
Redox Flow Batteries can be expanded by also just adding more "pressure vessels" to store the charged electrolyte.
Which is why I'm more interested in Redox Flow than Lithium Ion, for grid-scale applications anyway. Lithium Ion's efficiency is necessary for cars or other mobile applications. But if you're building it on the ground, you just build a bunch of cheap heavy tanks all over the place.
Nah, what I already wrote already covers about everything I feel sure about, and half of it is just common sense. Shouldn't be too hard to find better sources, though.
They are, but we probably want lots of hydrogen anyway. I don't know if we can easily substitute methane with hydrogen in gas power stations or gas ovens, but it is certainly possible.
Yes, but the methane here is generated from CO2 and H2, then combusted back to CO2. With no leaks, that cycle is carbon neutral, and leaks can be minimized.
If this infrastructure is built out, it may also provide the path to pulling CO2 out of the atmosphere or ocean and permanently sequestering it. That is something that I think we'll probably be very interested in doing in coming decades. Having affordable technology for that (perhaps $20-80/ton?) would enable CO2 emissions for those cases that really require it, and possibly also some clawing back of the worst parts of anthropogenic climate change.
Last I read they're working on issues with the high-temperature seals. It'd be awesome to have a big shipping container in my neighborhood that just stores all of our energy for use during peak/off hours.
Yeah a lithium-ion battery that doesn't catch fire and may be able to double energy density from PBS Nova [1] though I don't think the clip talks about energy density.
I think it will be multiple things, that would slowly converge to one or two solutions.
Already you can see that Tesla's power wall works really well. Then you have other technologies that would allow cities to store energy. I think we are slowly getting there.
My leading candidate for that is some kind of flow battery. There have recently been some announcements about a better and more stable electrolyte that is also non-toxic and non-corrosive. But as always, it might be five years (or never) before that gets commercialized.