The electrification shift, projected: three scenarios to 2035.
Every figure in this piece is a published, sourced estimate, not a review.solar analytical model. Three scenario tracks run throughout: Low (IEA Stated Policies Scenario, BNEF Economic Transition Scenario), Base (IEA Announced Pledges Scenario, BNEF base case), and High (IEA Net Zero by 2050, BNEF Net Zero Scenario), drawn from the IEA World Energy Outlook 2025 and BloombergNEF's 2026 Electric Vehicle Outlook. Historical data to 2025 comes from the IEA Global EV Outlook 2026, the IEA-PVPS Snapshot 2025, and BloombergNEF's May 2026 storage update. This is a comparative framework assembled from named primary sources, not investment or financial advice.
The global EV fleet more than triples by 2030 even in the cautious scenario, and the gap between scenarios keeps widening after that.
The global EV fleet stood at 80 million vehicles at the end of 2025. By 2030 the three tracks diverge sharply: 245 million under the Low scenario, 295 million Base, 380 million High. By 2035 the spread widens further: 510 million Low, 525 million Base, 620 million High. Annual sales share of new passenger cars follows the same shape: from 25% in 2025 to 42 to 62% by 2030 and 52 to 73% by 2035, depending entirely on which policy and cost trajectory actually plays out.
Chart data
| Series | Year | Value |
|---|---|---|
| Low (STEPS) | 2025 | 80M |
| Low (STEPS) | 2030 | 245M |
| Low (STEPS) | 2035 | 510M |
| Base (APS) | 2025 | 80M |
| Base (APS) | 2030 | 295M |
| Base (APS) | 2035 | 525M |
| High (NZE) | 2025 | 80M |
| High (NZE) | 2030 | 380M |
| High (NZE) | 2035 | 620M |
Regionally, the base-case trajectory has China at 78% new-vehicle EV share by 2030 and 88% by 2035 (already 55% in 2025), Europe climbing from 28% to 55% then 72%, and the United States moving from 10% to just 25% then 28%, a track BNEF attributes directly to the rollback of fuel-economy standards and the removal of the US$7,500 Section 30D tax credit, projecting only 24% of the US passenger fleet electric by 2040. India and Southeast Asia move fastest in relative terms: India from 8% to 32% to 58%, Southeast Asia from 20% to 38% to 55%. Norway, already at 96%, is treated as a standalone outlier rather than a leading indicator: its VAT exemption, toll privileges and charging density produced saturation conditions no other market had replicated by 2025, putting it five to seven years ahead of the European average and fifteen-plus years ahead of the US under base-case assumptions.
StorageAnnual battery storage additions nearly double by 2030 in every scenario, even the cautious one.
Annual grid-scale BESS additions reached 112 GW in 2025. The three tracks project 180 GW (Low), 205 GW (Base) or 230 GW (High) by 2030, and 210, 220 or 265 GW respectively by 2035, a flattening curve in the back half of the decade as the fastest-growing early markets mature. The context worth holding onto: this segment went from negligible to 112 GW of annual additions in roughly a decade, reaching the 100 GW-a-year milestone in four years, against eight years for solar and fifteen for wind from the same 10 GW starting base (BNEF, May 2026).
SolarCumulative solar capacity roughly triples by 2035, and the low scenario still adds more than the entire installed base of 2024.
Global cumulative solar PV closed 2025 at 2.85 TW. By 2030 the range runs 5.3 TW (Low) to 5.8 TW (Base) to 6.7 TW (High); by 2035, 6.7, 7.55 and 8.65 TW respectively. Even the Low scenario's 2035 figure, 6.7 TW, is roughly three times the entire cumulative installed base that existed at the end of 2023.
Chart data
| Series | 2025 actual (2.85 TW, off this scale) | Low | Low-High range base | High |
|---|---|---|---|---|
| 2030 | 5.3 TW | 5.3 TW | 5.8 TW | 6.7 TW |
| 2035 | 6.7 TW | 6.7 TW | 7.55 TW | 8.65 TW |
EV charging is the fastest-growing slice of new electricity demand, and it's still smaller than data centres.
Total global electricity demand runs from roughly 29,000 TWh in 2025 to a base case of 39,500 TWh by 2035. EV charging demand grows fastest in relative terms: from about 280 TWh in 2025 to 1,130 TWh (2030 base) to 2,110 TWh (2035 base), a 7.5-fold increase that still represents only around 5.3% of total projected 2035 demand. Industrial electrification and buildings/heat pumps stay the larger absolute drivers throughout. Data centre and AI demand is smaller in absolute terms but growing on a similarly steep curve, from roughly 500 TWh in 2025 to 1,700 TWh by 2035 base case, competing for the same grid capacity and critical minerals that EV and storage deployment need.
The EV-as-appliance economyVehicle-to-grid moves from pilot to early commercial deployment around 2027, and the dispatchable pool it creates dwarfs today's entire battery fleet.
V2G-capable vehicles are projected to grow from 5 million in 2025 to 100 million by 2030 and 180 million by 2035, with bidirectional charger deployment trailing behind at 42 million and 170 million respectively. DNV projects V2G-dispatched capacity will exceed 10% of total global grid storage by 2029, with the market shifting from pilot to early commercial phase from around 2027. The scale is worth sitting with: 180 million V2G-capable vehicles in 2035, at a conservative 40 kWh average usable capacity, is a theoretical dispatchable pool of 7,200 GWh, roughly 23 times the entire global grid-scale battery fleet installed at the end of 2021.
Chart data
| Series | Year | Value |
|---|---|---|
| V2G-capable vehicles | 2025 | 5M |
| V2G-capable vehicles | 2027 | 18M |
| V2G-capable vehicles | 2030 | 100M |
| V2G-capable vehicles | 2035 | 180M |
Six deployment modes are converging on the same underlying asset. Vehicle-to-home lets an EV battery power household load during peak-tariff windows or outages, functionally a stationary home battery with 40 to 100-plus kWh usable. Solar self-consumption pairing times smart charging to midday rooftop PV surplus, worth an estimated US$800 to 2,000 a year per household. Virtual power plants aggregate EV batteries as a coordinated grid asset, a 1,000-vehicle fleet at 60 kWh comparable to a small grid-scale storage project. Frequency and demand response uses millisecond-response smart charging, with one Nature Communications study finding 12 to 43% V2G participation could cover the world's entire short-term storage need by 2030. Second-life deployment repurposes retired EV packs at 70 to 80% original capacity for commercial and industrial storage, with the IEA projecting second-life applications reaching 10% of global battery demand by 2030. Fleet electrification turns commercial and transit fleets, overnight-charged and daytime-exportable, into a grid asset in their own right, with over 35% of new EV chargers projected V2G-compatible by 2030.
What could break thisSix factors sit underneath every number above, and none of them are fully priced into the scenarios.
Critical mineral supply. The IEA projects lithium demand growing sevenfold by 2035 under net-zero scenarios, with copper supply shortfalls flagged from the second half of this decade. A 2025 Nature analysis found US-allied production capacity covering only about 5 million EV batteries cumulatively by 2032, against a minimum requirement of 10 million or more. Probability of material impact on deployment timelines: moderate to high.
Sodium-ion commercialisation. BNEF projects sodium-ion could cut global lithium demand by roughly 40% by 2035 if commercialisation accelerates; CATL and BYD both had Na-ion cells in production as of 2025. If its cost curve mirrors early LFP learning rates, the mineral constraint above narrows materially. Probability of material impact on lithium demand: moderate.
AI and data centre demand. Global data centre capacity is projected to reach 226 GW by 2030 at roughly 15% compound annual growth, with AI servers alone consuming an estimated 306 TWh that year, a fivefold increase from 2024. This demand competes directly with electrification investment for grid capacity and critical minerals, a non-linear interaction the EV and storage projections above don't fully net out. Probability of material grid interaction impact: high.
US policy reversal. BNEF's 2026 EV Outlook projects a 19% drop in US EV sales in 2026 specifically, driven by the fuel-economy rollback and the removal of the Section 30D credit, with only 24% of the US fleet projected electric by 2040, well below every other major market. This impact is already in progress and is reflected in the US figures above, not a future risk.
Emerging-market leapfrog pace. Southeast Asia already exceeded US EV penetration in 2025. As Chinese-manufactured EVs approach sub-US$15,000 price points, adoption in India, ASEAN, Latin America and sub-Saharan Africa could outrun every scenario modelled here, the same bypass-rather-than-replace dynamic that let mobile phones skip landlines. Scenario models built from developed-market analogies tend to underestimate this. Probability of material upside deviation from the base case: moderate to high.
Duck-curve intensification. As solar crosses 30 to 50% of midday generation in leading markets, evening ramp requirements grow non-linearly, both the demand signal for storage and the risk of grid instability if storage deployment lags behind it. South Australia, California, Germany and China have all shown this dynamic at scale; it's now appearing in India, the Netherlands and Chile. The inverter-storage-EV stack becomes the critical grid-stabilisation layer by roughly 2030 on current trajectories. Probability of material grid stress events in leading markets: moderate to high.
My readThe scenario spread isn't noise. It's the actual size of the decision still in front of policymakers.
What I'd hold onto from this table isn't any single 2035 number, it's how much the Low and High scenarios still disagree by 2035 despite starting from an identical 2025 base. A 108-point gap in EV fleet size (510 million versus 620 million) and a 21-point gap in sales share (52% versus 73%) aren't measurement uncertainty. They're the modelled range of outcomes from policy choices that haven't been made yet, chiefly whether more governments follow the EU's fleet-standard model or the US's now-reversed tax-credit model, and whether mineral supply or sodium-ion commercialisation resolves first. The V2G numbers are the ones I'd watch closest over the next three years: a market moving from pilot to early-commercial phase around 2027, sitting on a dispatchable pool that could dwarf the entire existing grid-battery fleet by the mid-2030s, is the kind of structural shift that tends to get modelled too conservatively right up until it isn't.
All projections above are third-party published scenario estimates. review.solar draws no independent forecast and states no view on which track is most likely to occur.
Sources14 references
- IEA Global EV Outlook 2026 (via Virta)
- BloombergNEF: Electric Vehicle Outlook 2026
- IEA-PVPS: Snapshot of Global PV Markets 2025
- IEA World Energy Outlook 2025 (via TaiyangNews)
- BloombergNEF: Energy storage enters the 100 GW era (May 2026)
- BloombergNEF: Energy storage outlook, via Utility Dive (Oct 2025)
- IEA: Global electricity demand surge forecast (via BOE Report)
- DNV: Energy Transition Outlook 2025
- Econ Market Research: Vehicle-to-grid technology market (2026)
- BCG: EV strategies in the US, Europe and China (Sep 2025)
- Reuters / OFV: Norway's new car sales were 96% electric (Jan 2026)
- IEA: Global Critical Minerals Outlook 2025
- Ember: The EV leapfrog: how emerging markets are driving a global EV boom (Dec 2025)
- IEA: World Energy Outlook 2025
Read our ownership, funding and editorial disclosures.
Stay across the benchmark.
New scorecards, score changes and market signals, published as they happen.