6 Comments

“No country for coal gen” made my day. Thanks for these insights!!

Curious how nuclear will continue to impact overall power mix given gov incentives, DOE’s focus on it, and early days of SMR dev.

Expand full comment

Would love it see it, but even very optimistically would not expect to see anything noticeable in new nuclear on the scale of national generation this decade. For now, name of the game of is keeping legacy units running!

Expand full comment

Hi Brian. These are some very beautifully designed charts. However, I can't find any corresponding data produced by the EIA that would express generation using a capacity measure, GW. All your charts use the word "generation." But your chosen unit of account is "GW."

Your first chart, for example, portrays total US power generation as traveling along between 450 and 500 GW. There is no data from EIA that confirms this measure. All generation data from EIA is expressed in hours, like MWh, which of course can be easily translated to GWh or TWh. The US power system has been traveling along for years now, producing around 4000 TWh per year. Again, as someone who has used EIA data for 20 years, I am not aware of expressing generation using a capacity unit (whether MW, GW, or TW).

The same observation applies to all your other charts. The coal chart for example, again, expressing generation at a high of 231 GW and the current low at 95 GW. Doesn't correspond to any EIA data.

Can you explain?

Expand full comment

Hey Gregor, thanks for the thoughtful question. The units here are watts, but as labeled in the charts, they are average power production, not nameplate capacity. To be clear, the data is the same: the 480GW top line US power production in the area chart above is the same as the 4,200TWh you're familiar with; it's just an average rate (watts, joules/second) over 8760 hours in a year rather than the sum (watt-hours, joules).

Watts can express nameplate capacity but also actual power output (for example, ERCOT dashboards here: https://www.ercot.com/gridmktinfo/dashboards, or the current output or draw of any generator or load, all in watts). The value for interpretability is that power production in watts is comparable over different timescales (ERCOT wind produced ~15GW of power on average last year; it is producing ~7GW now; it will produce ~7GW in the next five minute interval... vs the same expressed in watt-hours: 21TWh in the last year, 1.9MWh in the last second, 584MWh in the next five minute interval). Average power output is also easily comparable to nameplate capacity (~7GW wind of output now vs ~30GW of ERCOT wind in commercial operation). Expressing in watt-hours IMO is more useful when comparing to something like MMBtus that is also in terms of energy (which to be fair, it typically is on an annual basis). I also know people are used to seeing annual data in terms of watt-hours instead of watts; perhaps I should have left it in energy rather than power on that basis alone. Sorry for the confusion, but hope that makes sense!

Expand full comment

Greetings Brian. Thanks so much for this explanation. I was at one time, iirc, more familiar with the practice of expressing power in these terms, but had forgotten about it perhaps because I am indeed very oriented towards expressing output over time, and habitually on an annual basis. But this is very useful, and you are of course, right--expressing power does indeed allow one to wriggle free of the time unit. My only suggestion would be to beware of using "generation" as I think something like "power" alone would suffice. But what do you think? Is "generation" really the best word to use for these units? What about "power-output" ? Many thanks. G

Expand full comment

I think you're right, power / power output would probably be the most precise way to describe it and avoid confusion

Expand full comment