CLOSEDFISHCAGE Results: What the Prototype Tests Actually Showed

The most revealing detail in the CLOSEDFISHCAGE final report is a limit: the weather did not allow the team to test waves higher than 0.5 metres during the reported wave measurements. More exposed conditions were investigated with a numerical model. That distinction is essential when reading the project’s positive account of its prototype.

A useful result is not a promise that a design will work everywhere. It is a finding tied to an apparatus, a test and a set of conditions. This guide separates those parts of the historical record.

What was tested in the water?

The consortium’s final report describes a 12-metre-diameter prototype tested at Toft, Brønnøysund, from September 2010 to June 2011. It had a closed canvas enclosure, an intake supplied from 60 metres below the surface, water-quality monitoring and alarms.

The report records a current test reaching 1.0 m/s without a distinct visible change in the underwater canvas shape. It also describes roughly 100 salmon kept for about six months as an indication test. These are observations from the reported trial, not specifications for a present-day product.

The original project description explains why enclosure shape and water control belong together. The results raise a narrower question: which part of that integrated system has evidence behind it, and how far does the evidence extend?

Three kinds of evidence that should not be merged

Reading a prototype report
Evidence What it can contribute What still needs checking
Physical trial Observed behaviour of the tested configuration under the recorded conditions. Duration, loading, instrumentation, uncertainty and whether the proposed application lies within that range.
Numerical model An estimate of behaviour under defined inputs, including conditions difficult to reproduce during a trial. Assumptions, calibration, validation data and sensitivity to uncertain inputs.
Economic scenario A calculation connecting design choices and operating assumptions to estimated cost. Prices, production cycle, stocking assumptions, capital treatment and whether the comparison uses equivalent boundaries.

Consider a hypothetical review of a new deployment. An engineer receives a simulation for a particular wave condition and a photograph showing an intact prototype. The photograph does not validate every input to the simulation, and the simulation does not establish that the photographed trial experienced the modelled sea state. The two may support each other, but the connection has to be demonstrated.

The same principle applies to the fish trial. Observing fish in a prototype can identify promising performance and practical problems. It does not, by itself, settle questions about every stocking level, season, site or failure mode. Those are additional comparisons, not details that can be inferred from the word “validated”.

Where the supporting documents fit

The final report identifies D1.1 and D1.2 as public requirement documents, while D2.1 and D2.2 concern prototype development and are marked confidential. That classification is part of the historical report; it does not mean that downloadable copies of every deliverable survive here.

For a reader tracing an old citation, the document identifier matters as much as its subject. “The cage report” could mean requirements, a design description or a trial result. Record the deliverable number, its version if available, and whether you have read the document itself or only a summary describing it. This prevents a secondary description from gradually turning into an apparent direct observation.

The CORDIS results overview is a shorter entry point. The detailed final report is the better place to follow the distinction between project objectives, completed work and the consortium’s interpretation of potential impact.

How to use these results now

Use the project as a worked example of an engineering evidence chain. Start with a requirement, identify the component or process intended to meet it, locate the relevant test, and check the conditions attached to the result. Where the chain stops at a model or a summary, say so.

For an actual purchase or design decision, historical evidence would need to be connected to the exact equipment and current site assessment under consideration. Changing the dimensions, loading or operating arrangements changes the question. This archive supplies research context, not an operational approval.

This is an independent reading of a completed research project. Test observations are attributed to the consortium’s report; no new tests have been performed by this publication.