iiinova — where your digital self keeps evolving

Dan Schroeder @dvschroeder.bsky.social · 5h

Point taken, though isn't HEV really a subset of ICEV, and shouldn't PHEV be split out for clarity?

Dan Schroeder @dvschroeder.bsky.social · 5h

Rather than merely complaining, let me highlight this "cheat sheet" as a positive contribution to the discussion. Obviously not the final word, but if others disagree with it I think they should put out something similar. bsky.app/profile/zeke...

120

Dan Schroeder @dvschroeder.bsky.social · 5h

Solar works great! We need more of it! But we shouldn't pretend that it can do everything. There's a reason why nobody, anywhere, is running a grid of any size on 100% solar.

100

Dan Schroeder @dvschroeder.bsky.social · 5h

Nah, the storage part comes in when we get solar's reliability above 30%.

100

Dan Schroeder @dvschroeder.bsky.social · 5h

The worst thing about permitting reform is that the talking points on both sides are so drearily predictable and so obviously intended merely as talking points, not good-faith efforts to inform.

120

Dan Schroeder @dvschroeder.bsky.social · 7h

90% reliability sounds good for a single coal-fired generating unit, but it would be awful for a whole fleet of coal plants. A fleet of solar farms in a single region, though, all shuts down at pretty much the same time. So this stress tests doesn't seem so stringent to me.

200

Dan Schroeder @dvschroeder.bsky.social · 7h

My hot take: bsky.app/profile/dvsc...

000

Dan Schroeder @dvschroeder.bsky.social · 9h

I certainly agree that it can cut both ways. Even if emissions are increasing in some places, that can be good if people in those places are living better lives. But this isn't a reason to be dismissive of the effect. I'd like to see your elasticity evidence presented more fully, for one thing.

010

Dan Schroeder @dvschroeder.bsky.social · 10h

Ah, but again, what's the counterfactual? Maybe without clean energy, electricity prices would have gone up. I think it's pretty intuitive to most people that the Jevons effect is more than zero. And that's often a good thing! Though I worry about cheap wind enabling crypto mining in Texas.

100

Dan Schroeder @dvschroeder.bsky.social · 10h

RIP Ken, you had a good run. As for Glen Canyon, it seems to be tryin' to come back whether we like it or not.

000

Dan Schroeder @dvschroeder.bsky.social · 11h

All I wanna know is whether you or anyone else disagree with "If you want to electrify everything, you need a lot more electricity".

100

Dan Schroeder @dvschroeder.bsky.social · 11h

Sure but that's not the same thing as denying that we'll need a lot more electricity.

000

Dan Schroeder @dvschroeder.bsky.social · 12h

Huh. When were these previous booms of AI? I'm 64 and I don't remember any. Anyhow you seem to be talking about capability more than merit. It would be interesting, I think, to come up with a good list of past technologies that have plateaued when people thought they would become more capable.

100

Dan Schroeder @dvschroeder.bsky.social · 14h

Are you saying better isn't the same as more capable, or that it's not the norm for technologies to become more capable?

100

Dan Schroeder @dvschroeder.bsky.social · 05/10/2026

Are there no negative contributions from regions with lower than average cancer rates?

000

Dan Schroeder @dvschroeder.bsky.social · 04/10/2026

Gotta wonder how much space the book devotes to the governor's own culture-warrior Twitter habit.

010

Dan Schroeder @dvschroeder.bsky.social · 04/10/2026

Is there a simple explanation for why the method always gives positive results?

100

Reposted by Dan Schroeder

Dr. Adam Stein @astein.bsky.social · 04/10/2026

Check out our new article!

1285

Dan Schroeder @dvschroeder.bsky.social · 02/10/2026

The cynic in me wonders whether the overture for BLM land is merely setting them up to blame their eventual failure on the federal government.

020

Dan Schroeder @dvschroeder.bsky.social · 02/10/2026

As I understand it, they have a small (<1MW) test reactor. They sure don't have a commercial product, let alone one that's been proved reliable, let alone the ability to do any mass production.

120

Dan Schroeder @dvschroeder.bsky.social · 02/10/2026

Sheesh. Valar Atomics wants to use 9000 acres of Utah BLM land and they're talking about more than 400 microreactors, 25 MW each, totaling 9.6 GW. All by 2032! 🙄 🔌💡 www.npr.org/2026/09/30/n...

npr.org

A startup wants to build a massive nuclear-powered data center on public land in Utah

Valar Atomics wants to place over 450 reactors on more than 9,500 acres in Utah to power a huge data center project.

351

Dan Schroeder @dvschroeder.bsky.social · 02/10/2026

Nice! But does anyone (else) use the word "agent"?

100

Dan Schroeder @dvschroeder.bsky.social · 01/10/2026

Hmm. @hossenfelder.bsky.social says the anonymous funder may have been Bill Gates. x.com/skdh/status/...

020

Dan Schroeder @dvschroeder.bsky.social · 01/10/2026

If Ds try to demand more in 2027, Rs can just walk away and then there's no deal. Which would be fine with the many Ds who don't buy your stipulation. Or put another way, these negotiations are too delicate and chaotic for the outcome to be predictable from such a high level.

010

Dan Schroeder @dvschroeder.bsky.social · 01/10/2026

I can't figure out how to relate the chart to the "almost two million" claim. What am I missing?

120

Dan Schroeder @dvschroeder.bsky.social · 30/09/2026

Yes, let's remove regulatory barriers whenever we reasonably can. But this kind of rhetoric seems like an attempt to deflect attention away from still bigger barriers: inherently high costs of some technologies, and political opposition from the fossil fuel lobby.

020

Dan Schroeder @dvschroeder.bsky.social · 30/09/2026

Yeah but their definition of useful energy is unintelligible.

Screen capture from page 29 of the IEA "Electrification" special report (September 2026): "Useful energy refers to the energy available for end-users to satisfy their need for energy services such as heating, cooling, lighting and mobility. Ultimately, it is these energy services that improve consumer welfare."

101

Dan Schroeder @dvschroeder.bsky.social · 26/09/2026

How should we calculate what the fraction of climate-focused stories in the New Yorker should ideally be? I have no clue. Here in academia every department thinks their subject is the most important and most deserving of resources. And each can make a truly compelling case!

100

Dan Schroeder @dvschroeder.bsky.social · 26/09/2026

(I think one reason I was unaware of this is because so many reports on data centers focus on what's proposed for coming years rather than what's already up and running.)

000

Dan Schroeder @dvschroeder.bsky.social · 26/09/2026

As a belated follow-up, it seems Oregon really is just an outlier: probably (a distant) third place in current data center capacity after TX and VA. Many thanks to @bistline.bsky.social for pointing me to this report which gives state-level estimates. powering-intelligence.epri.com

powering-intelligence.epri.com

Home | Powering Intelligence 2026

EPRI projects U.S. data centers to consume 9-17% of national electricity by 2030 as rapid growth driven by AI, streaming, and cryptocurrency accelerates. This report explores scenario-based forecasts,...

130

Dan Schroeder @dvschroeder.bsky.social · 25/09/2026

Many thanks to the @sltrib.com for providing this helpful resource!

121

Dan Schroeder @dvschroeder.bsky.social · 25/09/2026

Here's a direct link that should display the monthly Cape Station generation data, once there's data to display: www.eia.gov/electricity/... Or you can always go to www.eia.gov/electricity/... and download the whole spreadsheet.

120

Dan Schroeder @dvschroeder.bsky.social · 24/09/2026

(Also let's enjoy a laugh at the word "annually" tacked onto the end of that paragraph in Fervo's news release, indicating confusion over whether megawatts is a measure of energy or power. 🙄 #iTeachPhysics)

010

Dan Schroeder @dvschroeder.bsky.social · 24/09/2026

The Fervo news release adds some clarity to their capacity accounting. 33 MW is the net winter capacity of each Phase I unit, as reported to EIA. The latest EIA list still doesn't include any Phase II units, despite their being "under construction, with an expected COD in 2028".

Excerpt from the Fervo news release under the heading "Cape Station Details": "Cape Station Phase I is an approximately 100 MW installation comprising three 33-megawatt GeoBlocks, the first of which has achieved First Power and is expected to reach its contractual commercial operations date (COD) by October 1, 2026. Fervo will continue ramping up the first GeoBlock as it brings the full well system online while commissioning the remaining two GeoBlocks, which are expected to reach contractual COD by January 1, 2027. The next phase, an additional 400 MW, is already under construction, with an expected COD in 2028. Fully contracted offtake at Cape Station sits at approximately 900 MW, enough to power the equivalent of nearly 1 million U.S. homes annually."

Screen capture from EIA Form 860M spreadsheet data for August 2026, downloaded from https://www.eia.gov/electricity/data/eia860m/ and sorted to place the three planned Cape Generating Station geothermal units in consecutive rows. The table shows that each unit will in Beaver County, Utah, and under the LDWP balancing authority, with nameplate capacity 43 MW, net summer capacity 30 MW, and net winter capacity 33 MW. The planned operation dates are October 2026 for the first unit and January 2027 for the other two. The status is listed as "construction complete" for the first two units and "under construction" for the third.

120

Dan Schroeder @dvschroeder.bsky.social · 24/09/2026

I'll be watching the EIA form 923 generation data, which is published with a two-month lag. October data should be published around Christmas, and each month from then on will tell us something interesting. Of course, Fervo could also choose to tell us more.

140

Dan Schroeder @dvschroeder.bsky.social · 24/09/2026

Congratulations to Fervo Energy for achieving first power at its Cape Station enhanced geothermal complex in Beaver County, Utah! 🔌💡🌋 Now we watch to see how their electricity production develops over time. (h/t @petemassie.bsky.social) fervoenergy.com/fervo-energy...

fervoenergy.com

Fervo Energy Achieves First Power at Cape Station, a Landmark Moment for the Future of Enhanced Geothermal Systems - Fervo Energy

HOUSTON, September 24, 2026 – Fervo Energy (Nasdaq: FRVO), a global pioneer of next-generation geothermal energy, today announced that Cape Station, its

2181

Dan Schroeder @dvschroeder.bsky.social · 24/09/2026

Valar Atomics, would-be maker of helium-cooled microreactors, now says it's diversifying into "uranium TRISO-fuel processing to feed its reactors, water desalination, steel manufacturing, critical mineral refining — and data centers." 🔌💡☢️ www.sltrib.com/news/environ...

sltrib.com

2 big data centers are coming to Utah’s coal country. How they’ll be powered has residents ‘up in arms.’

“It’s all descending, these data centers,” said Cindy O’Neil, who tracks the progress of the projects. “They’re setting these things within a couple of miles of homes.”

142

Dan Schroeder @dvschroeder.bsky.social · 24/09/2026

I suspect imports will also be an important contributor.

010

Dan Schroeder @dvschroeder.bsky.social · 23/09/2026

Sorry, I should have provided a link for the IEA numbers. See the first chart under "Regional outlook" here: www.iea.org/reports/ener... I also apologize for missing the words "and planned" in the caption to the hexagon map at your site. That oughta account for most of the discrepancy.

iea.org

010

Dan Schroeder @dvschroeder.bsky.social · 23/09/2026

Thank you! Though I remain puzzled. That report suggests that average utilization rates are well above 50%. How, then, should I interpret the rest of the factor-of-7 discrepancy between your numbers and IEA's estimated US total? Are you including data centers that aren't yet operational?

100

Dan Schroeder @dvschroeder.bsky.social · 23/09/2026

IEA says that in 2025, data centers used about 5% of US electricity, which would average only about 25 GW.

000

Dan Schroeder @dvschroeder.bsky.social · 23/09/2026

OK, so putting emissions aside for a moment, it says the list of data centers comes from the S&P 451 Research Datacenter KnowledgeBase. What do you know about how we should interpret the numbers in that list? For instance, I count 360 hexagons for 180 GW total capacity, but what does that mean?

201

Dan Schroeder @dvschroeder.bsky.social · 23/09/2026

Very nice! Could you clarify what (ahem) data set you're using for state-level data center capacities?

110

Dan Schroeder @dvschroeder.bsky.social · 23/09/2026

Lots of problems with electric school buses in New York and elsewhere. Makes me wonder whether there's meaningful competition among vendors. (gift link) www.nytimes.com/2026/09/23/n...

nytimes.com

Why Some Schools Are Hitting the Brakes on Electric Buses

100

Reposted by Dan Schroeder

Zeke Hausfather @zekehausfather.com · 22/09/2026

The world has made real progress in steering away from some of the darkest climate futures where global emissions doubled (or more) this century. But as David Wallace-Wells and I note in today's NYT, we are not out of the woods yet, and climate uncertainty is not our friend: www.nytimes.com/2026...

417774

Dan Schroeder @dvschroeder.bsky.social · 21/09/2026

Looks to me like Oregon's electricity consumption was pretty flat from ~1997 through ~2018. I should think their marginal electricity source would be gas, which they could have reduced quite a bit without the data center growth. But I don't know the details, nor do I follow Oregon politics.

Chart from https://usenergydata.org/electricity-demand/ showing annual electricity consumption in Oregon from 1990 through 2024. From the late 1990s through 2018, consumption remained pretty flat, fluctuating between a low of 45.2 TWh and a high of 50.3 TWh. Starting in 2019, consumption increased every year, reaching 59.7 TWh in 2024.

Map from https://physics.weber.edu/schroeder/energy/PowerPlantsMap.html showing power plants in the states of Oregon and Washington (and some of Idaho), represented as dots sized by net generation in 2024, each colored by its principal energy source. Hydropower dominates, but there are significant amounts of gas generation throughout the region, plus significant wind, one nuclear plant, and one medium-sized coal plant (which shut down at the end of 2025 but will, according to plan, be replaced with more gas).

Chart from https://bpierpont.github.io/state-generation/ showing trailing 12-month electricity generation in Oregon, broken down by energy source, from January 2002 through March 2026. Hydropower dominates, though with significant fluctuations over the years. Gas generation has grown over this time period and has exceeded 30% of the total in recent years. Wind has also grown and is now about 15%, while solar has reached about 5%. Coal once provided about 9% of the electricity generated in Oregon, but was phased out by 2021.

030