Origin storage, the buffer before the ship.
Tankage at the production hub that holds synthesised ammonia and methanol between a continuous plant and a discontinuous ship. It decouples a plant that runs every hour from vessels that call every few weeks.
Where a continuous plant meets a discontinuous ship.
The electrolyser and synthesis loops run around the clock. The vessels that carry the product north do not; they arrive in parcels, on a cadence measured in weeks. Origin storage is the tankage that absorbs the difference: refrigerated tanks for ammonia, ambient tanks for methanol, sized so the plant never has to throttle because a ship is late and a ship never waits at berth because the tank ran dry.
This is the mirror of the destination buffer, on the production side of the water. The destination node holds landed tonnes against downstream draw; this node holds freshly made tonnes against ship arrival. The two are the same physics with the demand inverted, which is why the storage cost ratios carry across and the refrigeration premium on ammonia shows up here as well.
What drives the node.
| Parameter | Value | Basis |
|---|---|---|
| Throughput, ammonia | 150 kT/yr | Dakhla build output P |
| Throughput, methanol | 500 kT/yr | Dakhla build output P |
| Buffer duration | ~14 days | Design basis, set by ship cadence and parcel size D |
| Ammonia tankage | ~8,500 m³ | 14 days at 150 kT/yr, refrigerated, 0.68 t/m³ D |
| Methanol tankage | ~24,300 m³ | 14 days at 500 kT/yr, ambient, 0.79 t/m³ D |
| Refrigeration premium | ~2.25× | Refrigerated ammonia vs ambient methanol, per m³ E |
What the node costs, indicatively.
On a like-for-like 14-day buffer at the same throughput, origin tankage lands at roughly the same order as the destination node, near $17M, dominated by the same refrigerated-ammonia premium. That figure is an analogy, not a sized result. In the Dakhla build it is not broken out as a line; it sits inside the on-site infrastructure and process-plant CAPEX rather than as an isolated node cost.
Real cost, not yet isolated.
Origin storage is the one node in the chain that is fully real in the project but not yet modelled as its own unit. The tankage exists in the Dakhla design and its cost is carried inside the build total, but it has not been pulled out, sized and stress-tested the way the transport, delivery and destination-storage nodes have.
Isolating this node honestly means committing to a buffer duration, and that number is a live design choice, not a default. A longer buffer protects the plant against late or missed ships but ties up tankage CAPEX and working capital in standing inventory; a shorter one is cheaper but raises the risk of a forced production cut when a vessel slips. Because that trade sits against the ship cadence in the corridor module, the node will be isolated once the loading cadence is fixed, rather than asserted now with a buffer number chosen to look tidy. Holding it as indicative is the disciplined position, not a gap left by accident.
Where the numbers come from.
Throughput (150 kT ammonia, 500 kT methanol): the Dakhla Phase 1 build outputs, project-grade.
Tank volumes (~8,500 m³ ammonia, ~24,300 m³ methanol): derived at a 14-day buffer and liquid densities of 0.68 t/m³ ammonia and 0.79 t/m³ methanol. Design-stage, and contingent on the buffer duration being confirmed.
Indicative CAPEX (~$17M): carried across by analogy to the destination storage node at the same throughput and buffer, using $900/m³ refrigerated and $400/m³ ambient. Not a sized origin result.
Currently inside the project CAPEX (Sim 3). Not isolated as its own model. Indicative, modelled by analogy, not achieved.
How this node stands alone.
Even held as indicative, the node has a clean interface. It takes continuous production on one side and releases parcels to the ship on the other, and it commits to neither the plant that fills it nor the vessel that empties it.