Learn · Articles
Modular or centralised is the wrong question
"Modular or centralised?" is the wrong question. (1/6)
01 · The frame
The wrong question
Both models have failed publicly, and expensively. Anyone pointing that out is right — and the lesson the sector is drawing from it is wrong.
The mistake was never which side you picked. It was that this got asked first.
Architecture is an output. It is what you arrive at after the work, not the decision you open with. Three things decide it.
02 · What decides it
Three things decide the architecture
How much biowaste there actually is, and in how many separate places. What quality it is, because that sets conversion and conversion sets everything downstream. And what the business environment does to the economics — legal framework, labour and energy cost, and what that market will really pay for protein and frass.
Quality is the one that gets underrated. Bioconversion runs around 22% on a high-grade stream and about 12% on a poor one — and at 12% the advantage is gone entirely. Same equipment, same species, same operator.
Work those three first and the answer is usually not ambiguous.
03 · The waste
What the waste decides
Tens of sources at 5–15 tonnes a day each, spread across a region? Decentralised and modular almost certainly wins. You cannot truck scattered wet waste to one plant and still have a business.
One source above 30 tonnes a day with a real gate fee attached? A single centralised automated plant almost certainly wins. Modularity buys you very little there, and automation pays for itself.
Same technology. Opposite answers. The waste decided, not the ideology.
04 · The line
Facility style, not a size limit
That is also the honest use of the 20 tonnes-a-day line I keep coming back to: it indicates facility style, not a size limit. Below it, smaller and modular tends to fit. Above it, automation starts earning its keep.
And whichever side you land on, one discipline does not move: CAPEX efficiency per tonne per day of capacity. Modular is no excuse for a bad number. Automated is no guarantee of a good one. That figure is comparable across both architectures, and it is the one to ask any vendor for.
05 · The order
Reverse the order
Reverse the order — pick the architecture, then go looking for waste to fill it — and the chain is depressingly predictable. You size for the plan instead of the stream. The stream cannot feed it. So you buy substrate. Now you are a protein producer rather than a waste processor, competing with soymeal on price and on carbon, where published life-cycle figures put insect protein at 3–35.5 kg CO2eq per kg of protein against 0.9–2.2 for soymeal. That comparison is fatal — and it is the right comparison to make of a protein business. It is simply not a comparison of a waste treatment plant.
The design follows the waste and the economics, not an ideology.
06 · Over to you
What decided it
In the last project you saw, what actually decided the architecture — the waste, or the pitch deck?
Keep exploring
Related topics and pages
- BSF is a waste business — why the waste cost and the local buyer decide whether a case works
- CAPEX in BSF production: What you actually pay for — where the capacity chain and the climate control spend really go
- What Biowaste Actually Works: The Stream Selection Hierarchy — how biowaste streams split into the ones that work and the traps
Read next
- Where BSF Wins (a case with numbers) — a warm-climate case that puts a per-ton number on the waste-handling route
- Where BSF Loses (a case with numbers) — the five gaps that flip a BSF case and the per-ton number that decides a build
- Four Ways to Handle Biowaste — Where Each One Actually Wins — where biogas, composting and landfill each fit, and where distributed BSF processing wins instead
- Climate control, not biology, is the breakthrough — why the order of operations matters: the waste and economics first, then the environment
Next step
Let the waste decide
Manna can start from the real biowaste volume, its quality and the local economics before any facility architecture is chosen.