Hacker News Reader

>IMO nuclear is the only realistic alternative to coal to provide reliable, zero-emission "base load" power generation...

Yeah, but no one wants to pay for it. Which is why you'll see more wind and solar with hydro-storage, and even more geo-thermal energy plants in the future. Even with all the extra infrastructure, it still comes in at less than 2 or 3 cents a kwh long term. Whereas even if you take a very optimistic long term view on nuclear, it's just REALLY hard to get down below 3 cents. So for nuclear sources to get higher in the dispatch stack, government must take on more of the costs.

But then, of course, if government is taking on more of the costs, then you're still paying more than 3 cents. Probably a lot more. You'd just pay it in extra taxes instead of a utility bill. So, yeah, I guess that's why energy is a hard problem. No easy solutions really. Just hard to balance pros and cons to every known technology at this point.

(Maybe someday someone will figure out this whole fusion thing and we'll be able to consign all these other, less refined, technologies to history's dust bin?)

You're not going to be anywhere close to 3 cents in a 100% renewable system without hydro. Cheapest solar price in the US is 2.3 cents but that is in one of the sunniest places in the world, subsidised, and no storage.

Whether or not fusion is possible (it seems like it is, just incredibly difficult and meticulous, unless the MIT-ARC project works out).

There is always the eventual option to harvest solar energy in space The primary limiting factor at the moment is obviously the insane price of rocketing stuff into space.

Perhaps even before that solar power could be harvested in the upper atmosphere on massive scale by being tethered to balloons, it might even have the benefit of counter-acting some of the effects of global heating.

The primary limiting factor at the moment is obviously the insane price of rocketing stuff into space.

The people who are opposed to nuclear power because of the military proliferation risk might have something to say about the capability to transmit large amounts of power from space to the surface of this planet.

And you'd have to add to that objection the people who oppose nuclear power because of the costs. Those people would definitely be opposed to these ideas barring some new technological development that made space-based infrastructure several orders of magnitude less expensive.

Time to build also matters. How long does it take to build a storage mechanism for a solar/wind farm, vs the time to build a nuclear plant? What are the reasonable regulatory requirements? (Lest anyone be complaining about the ignorant treehuggers stopping nuclear plants, I think we can all agree that a nuclear plant is potentially more hazardous than pile of batteries, and should be regulated proportionally.) What are the up-front capital requirements of nuclear vs solar/wind + storage?

Another thing people need to keep in mind is we don't need enough storage to run the entire grid indefinitely. We just need enough to smooth the curve out to four nines or whatever number is deemed necessary. And while "the sun isn't always shining and the wind isn't always blowing", the wind is always blowing somewhere.

You wouldn't use batteries generally though.

For example, if you had hydro-storage, you'd use the power from the wind or solar to pump water into giant reservoirs. When wind or solar is unavailable, power is drawn from the hydro reservoirs by letting the stored water flow through turbines. (So a dam, basically.)

So, yeah, building reservoirs is pretty well defined. And would certainly take much less time than building a nuclear plant. (The cost would be orders of magnitude lower as well.)

Still has regulatory issues, just not nearly as many as dead batteries or nuclear power plants do.

Oh, sure. There are all sorts of interesting ways to cache energy, besides batteries - hydro, gravity, compressed air, thermal, etc. Price, scale, and local constraints should create a rich market for a wide variety of solutions.

(As an aside, a sufficiently large hydro storage is its own regulatory problem.)

>(As an aside, a sufficiently large hydro storage is its own regulatory problem.) ...

Yeah, but as I said in the comment, those regulatory issues are far less onerous than the regulatory issues with building a nuclear power plant. So the investment is less, the maintenance costs less, the kwh is less, and the time to market is less. Which technology do you think the big finance guys are gonna put their money into?

If you want money to go into nuclear plants, government has to agree to shoulder a lot of the risk. But there's drawbacks there too so...

yeah, just a hard problem.

That's why I think we'll see a lot of small-scale distributed storage, rather than giant industrial storage plants that match our current concept of the power grid. If you're running, say, a big box store, buying your own local storage for the store can make economic sense if it can pay for itself by buying dynamically priced energy. And then it doesn't need to be huge, because it's not trying to power a city, it's just trying to power your warehouse/store. A block of batteries, or a thermal system, or a water tower starts looking pretty attractive financially.

I agree. A lot of these storage solutions are starting to be so cheap that it's not hard to see different sorts of players investing in them. The benefit being, as you rightly point out, distributed storage.

Such systems are "no-brainers" for entities like, say, hospitals.

As I mentioned elsewhere, electric cars turn into pretty good storage systems, too.

And with a smart distributed grid, it's not just storage that makes sense locally - it's generation. Put a solar roof on the big box store, attach a storage system, and have a mostly self-sufficient system. And if it's generating surplus, sell it! Buy only when it's needed.

Meanwhile, poor developing nations or rural areas that don't have the capital or the chops to build giant nuclear plants can easily invest in very small scale projects. Go to some remote village, put up a wind turbine and a compressed air tank you can just bring in on a truck. Now that village has power! It may not be as financially efficient long-term, but by increasing the potential productivity of remote areas, it can pay for itself easily at the local level.

It's an exciting thing to see.

There's not many geologically good locations for pumped storage, they tend to be a bit environmentally sensitive, and they're not that quick to build - the one nearest me, Cruachan, took 6 years to build in the 1950s at the reduced safety standards of the time.

That's only one way that you would do hydro-storage these days though.

Think, for example, if you drill a hole deep into the ground, lower a cylinder into it, poke a hole in the bottom of the cylinder, then fill that cylinder with water. If I put a platform with a giant weight on top of it, the platform will press all of the water in the cylinder down through the hole in the bottom. That water will travel through a tube to the top of the cylinder and pour back into the cylinder on top of the weighted platform. In that tube, is where you capture the stored energy with turbines.

Then, when the wind picks back up, or the sun comes out, or power is just cheap even from the coal plant, you use that power to lift the platform back up and suck all the water back into the bottom of the cylinder.

Benefits of this are, as the HN User "beat" implied below, it's so cheap and simple to build that even a single, say, shopping mall might invest in it to lower their power draw from the grid. They would build one that only serves their mall for instance. You start to build out a truly distributed power storage infrastructure.

"Solar With Batteries Cheaper Than Gas in Parts of U.S. Southwest"

https://www.bloomberg.com/news/articles/2018-09-17/solar-wit...

Tesla built the Hornsdale Power Reserve in under 90 days. Hydro storage is never going to be able to compete with that, and nuclear will never be able to compete with how quickly you can deploy wind and solar. Batteries are rapidly coming down in price, and will continue to do so as EV sales scale up (1 million EVs are sold every six months, and this cadence is increasing). There are no regulatory issues with batteries. You can ship them to a concrete pad and install them, no questions asked.

I have started to view "baseload" generation more in a negative sense (as in "can't be easily turned on/off").

Also the daytime/nighttime power consumption differs almost twice, so almost any solar added only smoothens out the difference.

Also the new offshore wind turbines (>10MW) offer some extraordinary capacity factors (~60%). Makes me wonder what these capacity factors could be if we could get to 20MW turbines.

Traditional "baseload" plants are a product of cost and engineering. Nuclear and baseload coal plants are designed to deliver maximum power for minimum cost. Due to engineering constraints, this means that it's hard to scale the output up or down to meet changes in demand (hours? days?). So short-term demand is dealt with by "peaker" plants, which are very expensive. They're less efficient, and they aren't always in use, which means tying up capital in something you don't always need.

I think of solar/wind as a sort of weird baseload, with a wide variance in output that is not controllable and not always predictable in the short term.

Where I think we have a real opportunity is the application of modern software to this unpredictable power grid, by pricing power dynamically at a rate that matches the fluctuations in both supply and demand. At this point, building storage to stabilize the grid is arbitrage, a market imperative. Independent storage systems can buy power from the grid when it's cheap (sunny, windy days), and sell it when it's expensive (cloudy days with no wind and high demand). Then it's just a numbers game, and the market itself will provide the right amount of storage.

This gets more fun as electric cars become more and more common. Every electric car is just a big battery, right? Plug it into this smart grid, and you can make a few bucks using the same arbitrage as big power companies. Sell the power in your car!

Why is nuclear capacity easier or less expensive to build than storage capacity for wind/solar?

Be sure you're not starting with a conclusion and working backwards to the data required to support it.

> so you need a huge storage infrastructure (batteries, presumably)

The approach in the UK is to build Combined Cycle Gas Turbine [0] stations rather than storage infrastructure, as CCGT can be spun up quickly if renewables aren't cooperating. When I checked this awesome dashboard [1] just now, 48% of the UK's power was CCGT compared with approx 12% wind solar, presumably partly today has been still foggy. On good days solar and wind get to 30%.

[0] https://en.wikipedia.org/wiki/Combined_cycle

[1] http://www.gridwatch.templar.co.uk/index.php

> IMO nuclear is the only realistic alternative to coal to provide reliable, zero-emission "base load" power generation. Wind and solar could make sense in some use cases but not in general.

How much heat energy does a reactor with n meters of concrete around it, located on a water supply in order to use water in an open closed loop, protected with national security resources, waste into the environment?

I'd be interested to see which power sources the authors of this study would choose as a control for these just sensational stats.

From https://news.ycombinator.com/item?id=17806589 :

> Canada (2030), France (2021), and the UK (2025) are all working to entirely phase out coal-fired power plants for very good reasons (such as neonatal health).

Would you burn a charcoal grill in an enclosed space like a garage? No.