Hacker News new | ask | show | jobs
by AlanSE 1068 days ago
I've been following Terraform Industries for a while, and it does strike me as obvious that H2 will be the preferred form of energy storage medium in the long run, not just an intermediate species.

Depending on the conditions, if you're doing day-night energy storage, then you would like to fill up the H2 tank and then burn it when demand ramps up.

The only role I see for methane is when your H2 tank is already filled up but your solar panels are still producing. This is not just possible, but deterministically required whenever solar gets deployed enough for seasonal storage to be necessary.

Right now California is something like 35% solar + wind energy, and the "duck curve" is nearly maxing out. This was predicted in extreme specificity 20 years ago, as >30% and you get stuff like zeroed-out energy prices during the day. We have not solved the day-night (daily) energy storage problem, which only requires storing energy for a matter of hours.

When you go to something like 60% renewables, then even crazier stuff happens. Overproduction during sunny times goes bananas and (assuming daily storage is solved, which it isn't) seasonal issues start to arise. I don't think H2 will ever work for seasonal storage, although I think it's fine for multi-day storage scenarios. It's not theoretically impossible, I just compare to the equivalent that we have for natural gas and realize you would need MANY TIMES that level of investment to save the same amount of energy in H2. Plus, fossil fuel gas producers are mostly at a constant level, whereas solar production outright needs summer-to-winter storage.

What I don't buy is that these units will be disconnected from the grid and be economical by their gas products. No way, no how. What you want is to run the chemicals plant with low-voltage solar power locally, and trade-off with the grid as is economical.

2 comments

>Right now California is something like 35% solar + wind energy, and the "duck curve" is nearly maxing out. This was predicted in extreme specificity 20 years ago

Legislated. Its easy to predict something if you write it into law resulting in shutdown of Nuclear power station (Diablo Canyon)

>The main driver deterring PG&E from seeking a 20-year operating licence extension is the 2015 renewable portfolio standard (RPS) of producing 50% of its electricity from qualified renewable energy sources by 2030. PG&E’s model for the future cost of operating Diablo Canyon indicated that the cost per kilowatt hour was going to almost double, since the company would be forced to lower the amount of power it could produce from the plant in order to meet the state’s requirement. Dropping the capacity factor from the current 92% to say 50% would virtually double the price per kilowatt hour since costs are largely fixed.

>The state law which effectively dictates that by 2030 Diablo Canyon should operate at lower capacity each year and buy in power from intermittent renewables has apparently sealed the fate of the plant.

California has 5 GW of grid batteries (I think that's 20 GWh of storage).

The peak demand today at CASIO is around 36 GW (interestingly, ERCOT's peak is much higher. Texas should invest in efficiency.)

The "duck curve" looks like it's already partially ameliorated by batteries, and not that much battery capacity will be needed to smooth it out entirely.