Hydrogen vs EV: Who Wins?
Toyota / Hydrogen Advocates
“BEVs will capture only 30% of the market. Hydrogen is indispensable for heavy-duty transport and will scale once green hydrogen costs drop.”
Our Simulation
BEV reaches 60-85% of passenger sales by 2035. FCEV stays below 1%. Hydrogen survives only in industrial-adjacent heavy-duty corridors. Solid-state batteries eliminate hydrogen's last passenger car advantages.
What Actually Happened
Forward-looking forecast. Two independent runs (16 agents and 70 agents) converged on all 10 prediction dimensions - directionally identical conclusions despite 4.4x agent scaling.
What the Agents Said
Direct quotes from AI agents during the simulation - each with a unique persona, incentives, and behavioral logic.
“Hydrogen fuel cells are staggeringly dumb. The efficiency penalty means you need 3x more renewable electricity to drive the same distance. Why would you do that when batteries work?”
“Battery electric vehicles will capture only 30% of the market. The world needs multiple pathways. Hydrogen is indispensable for heavy-duty transport and industrial decarbonization.”
“Hydrogen works when you can treat fueling like a depot - same place, same time, same trucks. Random public fueling is a fantasy.”
“By 2035, BEV reaches 60-85% of new passenger sales. FCEV stays below 1%. The constraint is station utilization - without throughput, you can't finance the infrastructure.”
“We finance molecules with 10-year offtake contracts. Retail hydrogen stations without guaranteed throughput don't pass our investment committee.”
Agents in This Simulation
Each agent has a unique persona with distinct incentives, memories, and behavioral logic. They interact on simulated social platforms across 30 rounds.
Accuracy Scorecard (10 HITs, 0 PARTIALs, 0 MISSes)
Both runs predict BEV supermajority (60-85%) by 2035, FCEV below 1%
Both runs: FCEV <1% of passenger car sales through 2035
Both runs reject 'hydrogen dominates trucking' - predict 15-30% FCEV in long-haul
Both runs: hydrogen works at ports, freight hubs, fixed routes - not ubiquitous network
Multi-pathway hedge reduces risk but slows BEV learning velocity
DOE $1/kg target delayed to mid-2030s; $1-3/kg only in best regions by 2030
SSBs eliminate hydrogen's remaining passenger advantages (range, charging speed)
Steel, ammonia, maritime demand keeps hydrogen alive even if transport fails
Stable but asymmetric: BEV for 85%+ of transport, hydrogen for ~15% heavy-duty
5M+ public chargers vs ~1,200 hydrogen stations - gap widens through 2035
Key Metrics (Ground Truth)
“Corridor-and-Node Economics”
The simulation's most important finding isn't about who wins - it's about how hydrogen survives. Hydrogen doesn't die because industrial demand (steel, ammonia, maritime) keeps the molecule alive. But transport hydrogen becomes a 'piggyback tenant' on industrial infrastructure: it works at ports and freight hubs where hydrogen is already flowing for industrial use, and the refueling station can guarantee throughput. The moment you try to build a consumer hydrogen network independent of industrial anchor tenants, the economics collapse. This 'corridor-and-node' model is hydrogen's actual future - not the nationwide refueling network that advocates imagine, but a thin overlay on industrial hydrogen infrastructure at specific high-utilization nodes.