Hydrogen-led fusion plant — power vs H₂ dispatch

What this shows: a low-LCOE fusion plant runs its reactor flat-out and splits the busbar between the grid and a baseload electrolyzer — making H₂ whenever power is cheap and diverting to the grid only during price spikes. Set the plant's cost, the H₂ price, the electrolyzer size, and the grid's price shape; the tool re-computes the grid LCOE after crediting H₂ revenue, the power/H₂ split, and NPV live.

$20$35$120
$1$3.5$6
20%90%100%
$20$40$90
0.2×1.0×2×
Model assumptions
  • Plant: 1 GWe-class D-T tokamak, ~1,000 MWe net, constant output, availability 0.85. "Plant LCOE" is the underlying levelized cost you set (a NOAK low-cost plant is the interesting regime).
  • Electrolyzer: alkaline, 52 kWh/kg → 19.2 kg H₂/MWh; capex $700/kW, O&M 3%/yr, CRF 8%/30 yr (incl. stack replacement in O&M).
  • H₂ logistics: $0.80/kg (compression + salt-cavern buffer + pipeline/transport), subtracted from the H₂ sale price.
  • Grid prices: an annual price-duration curve set by the mean and a volatility (VRE) slider; higher volatility deepens the cheap/negative tail and raises scarcity peaks. Negative-price surplus beyond the electrolyzer is curtailed.
  • Dispatch: sell to grid when price > H₂ floor (= (H₂ price − logistics) × 19 kg/MWh), else make H₂. Reactor/turbine never ramp; only the grid/electrolyzer split changes.
  • Metrics: the grid LCOE credits H₂ net revenue against total cost over grid MWh (a byproduct-credit view — it turns sharply negative once H₂ over-covers the plant, i.e. electricity becomes the byproduct). NPV = level annual profit × 12.4.
Grid LCOE — after H₂ credit
—
Plant NPV (30 yr)
—