The defining infrastructure project of the century. Cheap solar, cheap batteries, an electrified everything — and a deployment problem the size of every grid on earth.
Scope13 slides · global
FrameTech ready, deployment lagging
Horizon2026 — 2050
MoodRealistic optimism
02 · The setup
Eighty percent of human energy is still combustion.
Despite a decade of headlines, fossil fuels — coal, oil, gas — supply roughly four-fifths of primary energy globally. The transition is real, but it is a multi-decade rebuild of every grid, every car fleet, every furnace.
Global primary energy mix · 2024 est.
Fossil share of primary energy
~79%
Coal, oil, and gas. Down from 87% in 2000 — a real but glacial shift.
Annual clean energy investment
$2.0T
Now nearly 2× annual fossil capex. The capital flywheel has flipped.
Years of buildout ahead
25–30
Even on aggressive deployment, full primary-energy decarbonization is a multi-decade project.
03 · The cost curves
The economics already won.
Solar modules, wind turbines, lithium-ion cells: each on a learning curve that has crushed cost faster than almost any forecast. The transition's hardest argument has dissolved.
Cost decline · 2014 = 100
Utility solar LCOE since 2010
−90%
Now the cheapest source of bulk electricity ever priced.
Onshore wind LCOE since 2010
−70%
Bigger turbines, bigger rotors, capacity factors above 50% in best sites.
Lithium-ion pack price since 2010
−90%
From ~$1,200/kWh to under $115. The single biggest enabler of EVs and grid storage.
04 · Electrify everything
The strategy is one word: electrons.
Replace every combustion engine with a motor, every furnace with a heat pump, every flame with resistive or inductive heat. Electrification roughly triples energy efficiency for the same end-use service.
Vehicles
3×
EV drivetrain efficiency vs. internal combustion. Tank-to-wheel — and global EV share is now ~22% of new car sales.
Heat (homes)
3–4×
Heat pump COP vs. resistive heating. Even cold-climate models now outperform gas furnaces.
Industry
~30%
Of industrial heat is below 200°C — already in reach of electric or heat pump tech today.
Trucks & rail
2030s
Battery-electric and catenary trucks crossing TCO breakeven this decade.
The flow
05 · The grid problem
When the wind stops, what then?
The grid was built for steady, dispatchable thermal plants. Solar peaks at noon. Wind is intermittent. Demand peaks at evening. Bridging this gap — flexibility — is the central engineering challenge.
Daily generation vs. demand · illustrative
Five flexibility levers
Storage — batteries, pumped hydro, thermal
Transmission — long-haul HVDC connecting weather zones
Overbuild & curtail — cheaper than storage at the margin
Permitting time · US transmission line
10+ yrs
The grid problem is half engineering, half paperwork. Lines, not panels, are the bottleneck.
06 · Battery storage
Stationary storage went from niche to commodity.
Lithium iron phosphate (LFP) chemistry — cheap, safe, abundant — is now the workhorse of grid-scale storage. Deployment is doubling every two years.
Global grid battery capacity · 2024
~180 GW
From under 5 GW in 2018. The fastest-scaling new asset class on the grid.
LFP cell pack price · 2024
$95/kWh
Below the $100 threshold once considered the EV inflection. Storage TCO scales accordingly.
Typical duration deployed today
2 — 4 hr
Right-sized for the evening solar gap. The economics work; longer duration is a different game.
2030 forecast capacity
~1.5 TW
8× growth in 6 years. China is ~50% of installs; Texas is the single largest market by state.
SOURCES · BNEF, IEA, EMBER · LFP = LITHIUM IRON PHOSPHATE
07 · Long-duration storage
The 100-hour problem is still open.
Lithium handles hours. But weeks of low wind, or seasonal mismatches, demand storage that lithium can't economically supply. A research frontier with several promising chemistries — none yet at commodity scale.
Iron-air
Form Energy's bet. Uses oxidation/reduction of iron — abundant, cheap, low energy density. Theoretical cost ~$20/kWh, 100-hour discharge. pilot
100hr
Pumped hydro
Old, proven, ~95% of world storage today. Geographically limited, slow to permit. Closed-loop projects expanding. mature
~80%
Round-trip efficiency
Thermal & flow
Molten salts, hot rocks, vanadium flow batteries. Modular, scalable, but immature economics. The wild-card category. research
10+ yrs
To scale
08 · Hydrogen
Useful, but not for everything.
Hydrogen is the Swiss Army knife people are tempted to use as a hammer. The honest case is narrow: industry, aviation, shipping, fertilizer. Not cars. Not home heat.
Long-haul aviation — via synthetic fuels (e-kerosene)
Shipping — ammonia or methanol bunker fuel
Where hydrogen loses
Passenger cars — 3× round-trip energy loss vs. battery EV
Home heating — heat pumps win on cost and efficiency
Grid balancing < 12 hr — batteries cheaper
Light trucks — battery TCO crossing over
Green H₂ cost target · 2030
$2/kg
Today's cost (electrolysis)
$5–7
Global H₂ demand · 2024
~95 Mt
Of which is "green"
<1%
09 · Nuclear
The dispatchable clean baseload.
Existing reactors run, life-extended, deliver carbon-free electrons 90%+ of the time. New construction is hard. SMRs are the bet. Fusion is the long shot.
Existing fleet
~440 reactors globally provide ~9% of electricity. Average capacity factor: 92%. License extensions to 80 years are now routine in the US.
2,500
TWh/yr · clean baseload
SMRs (next 10 yrs)
Small modular reactors: 50–300 MW, factory-built, designed for rapid siting near data centers and industrial loads. NuScale, X-Energy, Kairos, BWX leading.
~2030
First commercial deployments
Fusion (long term)
NIF achieved net energy gain (Q>1) in 2022. Commonwealth, Helion, TAE racing toward demonstration plants. Realistic commercial timeline: 2040s+.
2040+
Commercial earliest case
10 · Carbon removal
We will need to suck CO₂ out of the sky.
Even on a near-perfect transition, residual emissions from agriculture, aviation and cement remain. Net-zero pathways assume gigatonnes per year of CO₂ removal by mid-century. Today: thousands of tonnes. The gap is six orders of magnitude.
Direct air capture cost · today
$600/t
Down from $1,200 a decade ago. Climeworks, Heirloom, 1PointFive scaling first plants.
Required by 2050 (IPCC pathways)
5–10 Gt
Per year. Today's removal capacity: ~0.01 Mt. A factor-of-1,000,000 industry to build.
Target cost
$100/t
The threshold for affordable, scaled removal. Plausible by 2035 with learning curves.
The CDR portfolio
DAC — direct air capture, geologic storage. Pure but energy hungry.
BECCS — bioenergy + CCS. Land use trade-offs.
Enhanced weathering — crushed silicate rock on fields.
Oil reshaped the 20th century. Lithium, copper, nickel, rare earths and silicon-wafer fabs will reshape the 21st. Every transition mineral has a single dominant processor — usually China.
Refined lithium
~65%
China's share of global processing. Australia digs it; China refines it.
Cobalt mining
~70%
DRC origin. ~80% of refining → China.
Polysilicon for PV
~80%
Of world production in China, mostly Xinjiang.
Rare earth refining
~85%
Magnets for wind turbines and EV motors. Effectively a single-source supply chain.
Friend-shoring response
The US Inflation Reduction Act, EU Critical Raw Materials Act, Japan's KSM strategy, Australia's "Future Made" all push the same playbook: subsidize domestic processing, partner with allies, decouple selectively from China.
The honest constraint
Decoupling is expensive and slow. Mines take 10–15 years from discovery to production. Refining requires capital, environmental permitting, and a workforce that has migrated overseas. The mineral race may define the 2030s.
12 · Honest assessment
99%
of the technology to decarbonize is commercially available today.
The hard part isn't invention. It's permitting, transmission, supply chains, workforce, and political will. The transition is now an execution problem.
What's working
Solar + storage on commodity learning curves
EVs hitting price parity in major markets
$2T/yr clean energy capex flywheel
China driving manufacturing scale
What's stuck
Transmission permitting (10+ yr cycles)
Industrial heat > 500°C
Aviation, heavy shipping fuels
Long-duration storage economics
Carbon removal at scale
What it needs
Permitting reform — fast
Long-distance HVDC networks
Mineral supply diversification
Skilled trades workforce 5×
Carbon prices that bite
13 · Closing & references
The fastest infrastructure rebuild in history. Not fast enough.
We have the tools. The cost curves cooperate. The remaining work is political, logistical, and physical — every grid, every fleet, every furnace, in 25 years.