Agriculture & Future Food Systems · Article

Integrated pilots and comparison boundaries

Design a bounded production pilot around a defined food function, comparable alternatives and measurements that show whether the proposed connections actually work.

Preparation and related reading

Shared Foundations

This article helps you

Define the question and boundary of an integrated pilot.
Compare useful output on a consistent basis.
Separate observed integration from modelled expansion and remaining evidence gaps.

A pilot needs a question

A proposed installation connects microbial production with a cultivated feed supply and recovered utilities. A diagram can make the connections look complete before a single interface has been tested. Choose one decision first: for example, can a specified recovered stream reliably meet a defined receiving requirement?

A pilot is a bounded operating test, not simply a small copy of a hoped-for factory. State its process, scale, period, inputs, intended output and uncertainty. Its value comes from what it can establish under relevant conditions.

Draw the comparison boundary

Draw the included operations for both the integrated option and its alternative. Feed preparation, transport, water treatment, electricity, recovery, cleaning and waste may all affect the result. A connection inside the chosen boundary is an internal transfer, not a new external resource each time it circulates.

Select an alternative that can deliver the intended function: a separate fermentation operation, an established cultivated ingredient or another suitable process. Apply compatible quality and output requirements. A comparison with an unsuitable or differently processed product answers a different question.

Keep excluded activities explicit. A limited pilot can measure a particular interface well while leaving the broader life-cycle or supply-chain comparison for later. Detailed whole-system assessments have their primary home in Food Systems, Resources & Restoration.

Define the useful output

Kilograms of wet broth, dry biomass, a purified protein and a ready ingredient cannot be treated as equivalent outputs. Decide what the intended application needs and how it will be verified. State moisture basis, composition method, recovery and relevant functionality.

Our worked comparison assumes a shared useful protein basis solely to illustrate accounting. In a real comparison, equal analytical protein mass would still not prove equal digestibility, amino-acid quality, safety, taste or culinary performance.

Include rejected or off-specification material in the operating record, but do not count it as meeting the defined food output. A recovery loss changes resources per useful output even if the reactor produces the same gross biomass.

Separate observation from scenario

Record what the pilot actually measured, what was estimated and what came from another study. Measurement method, uncertainty, operating state and input variability remain attached to the result. A short favourable interval should not silently become a year of reliable operation.

A photovoltaic microbial-protein study compares prospective pathways using literature inputs. A later variable-power study extends prospective designs with scaling and learning assumptions. They can guide questions about integration, but their scenarios are not measurements of this pilot.

A manufacturer mycoprotein submission describes integration with a biorefinery and identified return streams. It supplies a process example, not independently audited evidence that every proposed connection is safe, zero-waste or economical. Evaluate the specific interface and its receiving conditions.

Test the connections under relevant conditions

A coupled operation depends on the timing, composition and reliability of its streams. Observe representative source and receiving conditions, including start-up, cleaning and interruption where relevant. Define how unsuitable material is withheld and how each process meets its requirements when shared supply is unavailable.

Check delivery rather than only an equipment command. Relevant evidence may include actual temperatures, flow, composition, recovered product and outside make-up inputs. Monitoring needs to distinguish a missing observation from a measured zero.

Technology-readiness guidance separates evidence of maturity from the aspiration to integrate components. For a food-production pilot, those engineering checks must sit alongside product-specific food evidence. The diagram connects operating observations and a defined output to a bounded next decision.

Make an integrated comparison answerable

A conceptual comparison, not evidence that the two illustrated products are nutritionally equivalent or a completed life-cycle assessment.

Read the diagram as text

Two alternatives are compared only after their useful output and included observations are defined. Remaining gaps then guide a bounded decision.

Parts

Integrated option A: Specify interfaces, operating period and included inputs.
Shared useful-output basis: Quality and function must actually be equivalent.
Suitable alternative B: Apply compatible product and process boundaries.
Measured comparison: Outputs, outside inputs, losses and uncertainty.
Next bounded decision: Expand only what the evidence and remaining checks support.

Connections

Integrated option A → Shared useful-output basis: define output
Suitable alternative B → Shared useful-output basis: match function
Integrated option A → Measured comparison: observe operation
Suitable alternative B → Measured comparison: compare fairly
Shared useful-output basis → Measured comparison: consistent denominator
Measured comparison → Next bounded decision: resolve the gap

Decide what the pilot justifies

Before expansion, identify which measured result supports the decision and which uncertainty could reverse it. A next phase may need longer operation, another input condition, better recovery, an interface redesign or a safety assessment. No single universal test duration establishes every requirement.

Compare energy, feedstocks, water, nutrients, waste, land and infrastructure on the stated output basis. Integration may reduce one outside input while increasing another. Do not count an avoided-product credit twice or assume that moving a waste stream makes it disappear.

A useful conclusion names the connection that worked and the limits of that observation. It can recommend an appropriate next test while leaving future cost, scale and market claims at their actual evidence stage.

Worked examples

Choose a common denominator

Two invented intervals produce material assumed to meet the same usable protein specification. A recovers 20 kg dry product with 50% protein using 160 kWh electricity; B recovers 30 kg with 25% protein using 120 kWh. Is B better on electricity per kilogram of that protein?

A protein=20×0.50=10 kg;electricity intensity=160/10=16 kWh/kg protein.
B protein=30×0.25=7.5 kg;electricity intensity=120/7.5=16 kWh/kg protein.
Per kilogram dry product the figures are 8 and 4 kWh/kg, but that is a different output basis.

The stipulated protein-based electricity intensities are equal. Product equivalence is assumed for this teaching case; these values do not establish full efficiency, nutrition or environmental performance.

A wider period includes start-up

A hypothetical test records 160 kWh during productive operation plus 40 kWh for start-up within the chosen period. It recovers 10 kg of the defined output. What measured electricity intensity covers that period?

Included electricity=160+40=200 kWh.
Intensity=200/10=20 kWh/kg output.
Using only productive-operation electricity would give 160/10=16 kWh/kg on a narrower boundary.

20 kWh/kg covers the stipulated period including start-up. It is not a measured commercial-year intensity; other resources and uncertainty require separate accounts.

Check your understanding

Are equal kilograms of wet broth and dry ingredient an equivalent comparison?
Show answer and reasoning
No.
Moisture, composition, recovery and intended function differ.
Can a demonstrated connection justify every future scaling and cost assumption?
Show answer and reasoning
No.
The test establishes only the conditions and endpoints it actually covers.
Should an unmeasured discharge be entered as zero?
Show answer and reasoning
No.
Missing information is uncertainty, not evidence of no flow.

Key ideas

Choose a pilot question and a consistent useful-output basis.
Measure actual interfaces and keep external inputs and losses visible.
Let the evidence define the next decision and the limits of expansion.

References

Sources supporting the explanations and examples in this article.