Seawater in; ammonia, methanol, e-diesel, or e-methane out — depending on which tab you click. This page is the pitch deck and the market study fused into one live model: move any number and the physics, the cost stack, and the verdict recompute.
Every serious ocean molecule starts as hydrogen from seawater. What you turn it into — ammonia, methanol, e-fuels, methane — is a question of what second ingredient you can get at sea (air is free, carbon is not) and what the ship at the pier will actually buy.
Ammonia is air + water + power — no carbon to ship. Methanol, e-diesel, and e-methane all need a carbon feedstock, and at sea that means a barge of CO₂, a power-hungry DAC plant, or lab-grade seawater capture. The fuel choice is really a carbon-logistics choice.
~250 Mt of bunker fuel a year; Singapore alone sold 56.8 Mt in 2025. Ships bunker at a handful of hubs, then run. Deep-sea orderbook: LNG first, methanol second, ammonia-ready third. You do not fuel a 15,000 TEU ship with 10,000 psi tanks — you sell it a liquid at the port it already calls at.
~1,500 Gulf of Mexico platforms need a decommissioning decision. Reuse the jacket, rebuild the topsides, plug into Port of Houston. UH ROICE has already published the reuse math — $4.76–8.44/kg at ~100 MW, before incentives. Pump Hydrogen brings the one thing every one of these plants must buy: the pump.
Pick a molecule. Eight presets below are the conversations you have already had — then move anything. The page is the spreadsheet.
Seawater is a rounding error. Electrolysis is the business. Carbon is a logistics problem. Firm power is the religion.
| Hub | 2025 volume | Why it matters for H2 / NH3 / MeOH |
|---|---|---|
| Singapore | 56.8 Mt | 25% of world bunker sales. Ammonia barge due 2027. Green corridor with Rotterdam. |
| Rotterdam / ARA | 9.8 + 7.9 Mt | EU ETS + FuelEU offtake. Port H2 pipeline live 2026. LH2 short-sea 2027. |
| Zhoushan | 8.0 Mt | First Chinese hub in the global top three. State-backed H2 ecosystem. |
| Fujairah | ~7.4M m³ | East-of-Suez price marker. Cape routing. |
| Busan / Korea | ~6 Mt class | Ammonia-ready fleet + national H2 strategy + e-methane buyer. |
| Houston / US Gulf | USG hub | Home water. Port Houston H2 truck-station grant with Linde. ~1,500 GOM platforms aging out. |
| Panama / Gibraltar | ~5 / ~4.6 Mt | Chokepoint bunkering. Diversion premium when Suez or Panama is stressed. |
First green hydrogen produced on a live gas platform, Eni Q13a-A, July 2026. Seawater → RO → electrolysis, wind-powered. The pilot class Pump Hydrogen sells pumps into.
268 m FPSO, southern Europe. 300 MW PEM, on-deck Haber–Bosch, ~55,000 t/yr H2 and ~300,000 t/yr green ammonia, offloaded by hose. FEED into 2026. The big play, at industrial scale.
First commercial e-methanol loop running: 52 MW electrolyzer, solar next door, biogenic CO2 trucked in, ~42,000 t/yr. Proof that the molecule business works — where the carbon is already standing.
Counter-evidence, honestly: the Netherlands paused new offshore H2 demos for five years because demand arrived slower than the pitch decks. Aberdeen's hydrogen buses — £556k new, £30k resale — are what happens when molecules and offtake are assumed. Both facts are in this model's defaults.
Mariner PX line: seawater RO feed, process liquid, ammonia transfer skids. $2–3.5 M first-article program, 14–18 months. This is the business that funds everything else.
OEM packages into a ROICE-class GOM reuse project or a US Gulf / North Sea bunkering node. Pump Hydrogen's name on the skid, not on the field development plan. Customer's capex.
Only with contracted offtake and firm power: consortium seat on one near-shore fuels node at Houston / Galveston. Shipyard + offtaker + power developer + Pump Hydrogen. The model above is this phase's homework.