John Bistline
@bistline.bsky.social
Energy systems modeling, economics, policy | IPCC, NCA, Stanford/CMU alum | Views my own
Economics
30%
Engineering
28%
Santa's wearing shorts this year. Christmas 2025 is forecast to be the warmest mean temperature across the Contiguous U.S. in the past half century.
h/t my colleague Erik Smith for the analysis: 2025 temps from GFS 06z, 1975-2024 from ERA5, area-weighted (lat) and averaged over the CONUS.
h/t my colleague Erik Smith for the analysis: 2025 temps from GFS 06z, 1975-2024 from ERA5, area-weighted (lat) and averaged over the CONUS.
The Clean Heat Climate Opportunity — The 2035 Initiative at UC Santa Barbara
A Near-Term Roadmap for Electrifying Low- and Medium-Temperature Industrial Heat
www.2035initiative.com
Check out the full report here from UCSB researchers (including @leahstokes.bsky.social): www.2035initiative.com/clean-manufa...
The EMF 37 study also discusses opportunities and challenges in this space: www.sciencedirect.com/science/arti...
The EMF 37 study also discusses opportunities and challenges in this space: www.sciencedirect.com/science/arti...
Reposted by Leah Stokes
"Hard-to-abate" is sometimes code for "we haven't looked closely yet." This great new report shows opportunities for clean heat (low/medium-temp industrial process heat) through electrification, which also has public health benefits.
Solar wait times in interconnection queues are increasing, too.
Reposted by Paul Lehmann, Juan Cole, Paul Nightingale
Wind developers: "We're ready to build."
Grid: "Please hold. Your call is very important to us. Estimated wait time... 8 to 12 years."
Grid: "Please hold. Your call is very important to us. Estimated wait time... 8 to 12 years."
South Australia has an impressive amount of wind and rooftop solar. It also relies on gas and imports, which range from 2% of daily demand (Nov. 17 on the left) to >80% (Jun. 26 on the right). This great chart is courtesy of my colleague Mike Caravaggio.
Bottom line: System planning extends beyond minimizing costs. If we want models to inform planning and policy, non-economic deployment limits need to be explicit, transparent, and stress-tested.
Check out our new EPRI and RFF report here: www.epri.com/research/pro...
Check out our new EPRI and RFF report here: www.epri.com/research/pro...
Why this matters: When land or deployment is constrained, the system substitutes.
More solar instead of wind, more gas or CCS in some regions, higher costs, different transmission needs. Pathways diverge.
More solar instead of wind, more gas or CCS in some regions, higher costs, different transmission needs. Pathways diverge.
Land use is a great example.
Wind is abundant, but setbacks, federal lands, and local ordinances dramatically shrink what's developable.
Our new EPRI analysis shows this vividly: Layering land restrictions and local ordinances shrinks land far more than many models assume.
Wind is abundant, but setbacks, federal lands, and local ordinances dramatically shrink what's developable.
Our new EPRI analysis shows this vividly: Layering land restrictions and local ordinances shrinks land far more than many models assume.
One common workaround: proxy constraints.
Instead of pretending everything is deployable everywhere, models can limit annual build rates, restrict land availability, or cap technologies, testing how sensitive pathways are to real-world bottlenecks.
Instead of pretending everything is deployable everywhere, models can limit annual build rates, restrict land availability, or cap technologies, testing how sensitive pathways are to real-world bottlenecks.
Most energy models are good at technology costs, fuel substitution, and system value. They're much weaker at things like:
- Where projects are allowed
- How fast infrastructure can scale
- Whether communities say "yes"
Ignoring these can bias results toward unrealistically smooth buildouts.
- Where projects are allowed
- How fast infrastructure can scale
- Whether communities say "yes"
Ignoring these can bias results toward unrealistically smooth buildouts.
#ModelingMonday: A key modeling assumption with big consequences is non-economic deployment constraints.
Permitting, land access, local ordinances, supply chains, social acceptance. These shape what can get built but are tricky to model.
Permitting, land access, local ordinances, supply chains, social acceptance. These shape what can get built but are tricky to model.
EPRI Home
The Electric Power Research Institute (EPRI) conducts research, development, and demonstration projects for the benefit of the public in the United States and internationally. As an independent, nonpr...
www.epri.com
The full report is here: www.epri.com/research/pro...
There's an interactive site here: interactive.epri.com/wattworks/p/1
There's an interactive site here: interactive.epri.com/wattworks/p/1
AI looks great on multiple-choice... but ask it to explain something and the wheels come off. EPRI's new benchmarking shows a 27-point accuracy drop when moving from MCQs to open-ended questions. Apparently the electric sector is more complex than the International Math Olympiad.
Yes, electricity demand in Texas is growing rapidly relative to the past, but very unlikely to be "doubling in the next ten years" rapidly.
www.ercot.com
ERCOT's implementation of HB 5066 meant that for the 2024 planning cycle it accepted TSP "officer letter" loads as reasonable when building the forecast: www.ercot.com/files/docs/2...
Part of this is growth in requests from large loads, esp. data centers: www.ercot.com/files/docs/2...
But the very large jump between 23-24 is partially due to HB 5066 requiring ERCOT to consider load forecasts provided by utilities/TSPs, incl. loads that don’t have signed IX agreements.
But the very large jump between 23-24 is partially due to HB 5066 requiring ERCOT to consider load forecasts provided by utilities/TSPs, incl. loads that don’t have signed IX agreements.
I sometimes refer to ERCOT load projections as "Loch Ness Monster" curves for multiple reasons.
Reposted by Richard S.J. Tol
Batteries aren't the future; they're in the grids many states are building now. In a few years, storage has gone from a rounding error to double-digit capacity shares in AZ, NV, CA, TX, and others. The energy storage map is being redrawn a lot faster than people think.
A first look into congestion pricing in the United States: PM2.5 impacts after six months of New York City cordon pricing
npj Clean Air - A first look into congestion pricing in the United States: PM2.5 impacts after six months of New York City cordon pricing
www.nature.com
More here: www.nature.com/articles/s44...
It's interesting to see benefits extend far outside the target zone, although they happened on a smaller scale.
It's interesting to see benefits extend far outside the target zone, although they happened on a smaller scale.
Wow... NYC's congestion pricing has led to big drops in PM2.5 and real improvements in local air quality.
More here: speed2power.epri.com
I really like this table that compares the main powering configurations and their tradeoffs on speed, reliability, emissions, and customer impacts in one place.
I really like this table that compares the main powering configurations and their tradeoffs on speed, reliability, emissions, and customer impacts in one place.
Everyone's arguing about how much power AI data centers will use.
EPRI's new Speed to Power report is about something more fundamental: how fast that power can be delivered while making data centers grid resources (flexible load, faster interconnections, cleaner mix).
EPRI's new Speed to Power report is about something more fundamental: how fast that power can be delivered while making data centers grid resources (flexible load, faster interconnections, cleaner mix).
Reposted by John Bistline
Ken Caldeira
@kencaldeira.com
· 18d
Successful technologies decrease in cost on average by about 20% or so for each doubling of cumulative capacity, but past departures from this rate are not good predictors of future departures from this rate.
Skillful prediction requires technoeconomic analysis, and not just a rearview mirror.
Skillful prediction requires technoeconomic analysis, and not just a rearview mirror.
Variability of technology learning rates
Climate and energy policy analysts and researchers often forecast the cost of low-carbon energy technologies using Wright’s model of technological inn…
www.sciencedirect.com
Check out this great new paper by Angelo Carlino, @aliciawongel.bsky.social, Lei Duan, @edgarvirguezr.bsky.social, @stevenjdavis.bsky.social, @energymorgan.bsky.social, and @kencaldeira.com here: www.sciencedirect.com/science/arti...
We love to say "costs fall X% every time capacity doubles." This paper looks at 87 technologies and says: actually, no. Past learning rates are not reliable predictors of future learning.
Households face annual costs of roughly $400–$900 from climate change—mainly from disasters, higher insurance, and energy costs—with lower-income families and certain regions hit disproportionately, from @kclausing.bsky.social, Knittel, and @cwolfram.bsky.social www.nber.org/papers/w34525
There wasn't gas with CCS added in this period. I often pair this historical figure with a panel on model projections, and some of those forward-looking scenarios have CCS-equipped capacity.
