Hacker News Reader

Comment by jellicle | Hacker News Reader

This is a poor article full of unrealistic assumptions. It's basically as if you took a worst case for CPU cache assumptions and then decided that, therefore, it would never be useful to put a cache on a CPU.

Not actually true.

In the real world, the wind is always blowing somewhere and the amount of sunlight falling is quite predictable. Fossil fuel plants aren't going to disappear entirely for a long time, but there is absolutely no obstacle to producing MOST of our energy needs from renewable sources. Real world storage installations are going in, in suitable places, all the time:

http://www.theglobeandmail.com/report-on-business/industry-n...

The cumulative effect of all these batteries will be quite large, and no, it isn't necessary to store 7 days * the entire energy needs of the United States in order to be useful.

View on HN

I think you're reading the article in a quite biased manner. I didn't pick up on any of the negativity you seem to have interpreted.

I don't think anyone here is against renewable energy, but certainly the storage of it represents a gigantic challenge. The article just tries to table some reasonable numbers for discussion. You've said nothing that even slightly contradicts the author's findings, and your vague insinuations that the author has an anti-renewable energy agenda is not helpful at all.

The author's calculations are for replacing the entire energy diet of the United States. He uses gross inputs to everything, rather than outputs and usable energy. He neglects any other storage capacity present in the country - no vehicle has any battery, no electrical device has any battery - and tries to calculate how much energy we need to store IF THE ENTIRE ENERGY PRODUCTION CAPACITY OF THE UNITED STATES DISAPPEARS ENTIRELY FOR A WEEK.

It's a nonsense calculation. The U.S. certainly isn't prepared today to go for a week without any electricity generation or even a single drop of fossil fuel burned, and there's no reason to demand that it be prepared to do so in the future.

The "findings" are akin to someone who says "The U.S. is borrowing a dollar a year. If we extend that to infinity, we find that the U.S. debt is infinite. PANIC!". They're nonsense. Have no meaning in the real world.

In the real world, one significant benefit of a widely distributed power generation infrastructure is that it won't all go down at the same time. The solar cells on your roof will keep generating electricity even if it's cloudy the next town over. And that means that the real-world requirements for a several-nines-reliability mostly-renewable energy infrastructure are several orders of magnitude less than what is being calculated here.

The headline on that article is terrible, the generation capacity of the planned installation is ~1/6 that of Niagara Falls (the 5x comes from the greater height difference of the planned facility). And that doesn't address the storage capacity.

The point of the HN-linked article isn't to slam storage technology, it is to establish some scale. The author even suggests dividing his results by 14 if you don't like his assumptions. You still need something like several thousand plants similar to the 15 we have to provide 1/2 day of backup capacity.

I'm sure the author would happily agree that pumped storage plants are immediately useful (this is basically obvious, the existing ones were all built because they were economical to connect to the grid).

Does anyone have good numbers for how much storage we'll need? We have data on wind and sun across the US going back a while, though most of it isn't from an energy perspective. I know that so far Europe's done well with the "it's always windy somewhere" strategy, but they haven't been doing it very long.

He is not arguing that they are not being built. He is arguing they are not going scale well going forward. The vast majority of stored megawatts is pumped hydro, solving the energy storage problem would greatly speed our transition from fossil fuels.

Reply on news.ycombinator.com