Former Atlantic Sapphire CEO:
Could in-sea closed facilities be a better option?

In his LinkedIn post, Andreassen writes that a year ago he compared the CAPEX (capital expenditure) of listed, land-based salmon producers with full production cycles with their actual harvest volumes. Now he is doing it again.

"There is still a significant gap between nominal capacity and actual production," he writes.

The companies

Andreassen has set up figures for the last 12 months up to and including Q2 2026:

• Atlantic Sapphire harvested 5,452 tonnes of head-on gutted salmon (HOG) – 64% of the upper limit of 8,500 tonnes in the company's revised capacity range for Phase 1. Implicit CAPEX was reduced from $62 to $50 per kg.

• Salmon Evolution harvested 5,300 tonnes – 67% of the nominal capacity of 7,900 tonnes in Phase 1. Implicit CAPEX was reduced from $44 to $35 per kg.

• Proximar harvested 2,118 tonnes – 40% of nominal capacity. Implicit CAPEX fell from $200 to $63 per kg.

• Nordic Aqua Partners harvested 3,116 tonnes – 39% of the total capacity of 8,000 tonnes in stages 1 and 2. Implied CAPEX was reduced from $83 to $49 per kg. Production in Q2 corresponds to 68% of nominal capacity on an annual basis, although available capacity in the tanks in phase 2 contributed positively in the quarter.

• Gigante Salmon harvested 1,928 tonnes – 12% of the nominal capacity of 16,000 tonnes. Implicit CAPEX was approximately $70 per kg. The plant is still in an early ramp-up phase.

• Andfjord Salmon had no harvests in HOG in either of the two rolling 12-month periods. The company's previous harvest of market-sized fish was in 2023. Sales of post-smolt in 2026 are excluded because they are not directly comparable.

“All of the companies in production increased their harvest volume over the past 12 months. Salmon Evolution is now below the CAPEX benchmark of around $40 per kg that is often used. The others are showing improvement, but most are still significantly below nominal production capacity,” he writes.

He points out that nominal capacity, available tank capacity, and mature production capacity are not necessarily the same, especially in a ramp-up phase.

Closed facilities

And he says that he believes another interesting comparison is closed facilities at sea.

He lists a hypothetical example in which he assumes that a 10,000 m³ system costs about US $10 million – equivalent to about one-sixth of the tank volume in Salmon Evolution’s phase 1. With Salmon Evolution’s current production per cubic metre, such a system would, according to Andreassen, produce about 880 tonnes of HOG, which corresponds to a CAPEX of about $11 per kg. At nominal productivity, he writes, production would be about 1,320 tonnes of HOG, which corresponds to about $8 per kg.

"That is about a third of Salmon Evolution's current implicit CAPEX per kilogram of actual production," he writes.

Andreassen emphasises that the scope included in the CAPEX calculation must be comparable.

"Moorings, feed and oxygen systems, water treatment, vessels, onshore infrastructure, and working capital can significantly affect the calculation," he writes.

He then asks the question, "Can closed offshore facilities deliver many of the benefits of land-based production, but with significantly lower capital intensity?"

"I share these numbers because I believe in this type of production. But installed capacity alone does not prove that the economy works. Stable biology, repeatable harvest volumes and competitive costs do. Reality must come before theory," he writes.