A global network of smallholder co-ops feeding a highway is the wrong shape, and the reason is arithmetic. What travels along the network is the method, not the material.
The Silk Road was not a triumph of logistics over distance. It was a consequence of what silk is worth per kilogram.
Put freight at eight cents per tonne-kilometre and ask, for each cargo, how far it can go before the freight bill equals the entire value of the load.
Silk can go around the Earth eighteen times before freight eats it. Date palm residue makes it a hundred and eighty-eight kilometres. Between them is a factor of four thousand.
The Silk Road is a shape that only high value density can afford, and biomass is the defining case of the opposite. A network built to haul feedstock to a highway is a network whose cargo cannot reach the highway.
Take one tonne of fresh cashew apple — about 88% water, and fermenting within a day or two of picking — and juice and ferment it where it was grown:
A first-stage node turns a three-hundred-kilometre commodity into a seven-thousand-kilometre one, at no loss of value.
That is how the highway is reached. Not by hauling feedstock to the road — it cannot get there — but by making something road-worthy where the feedstock already is.
The textbook story about biomass says: plant capital scales as $Q^{0.65}$ but the collection radius grows as $\sqrt{Q}$, so total cost has a minimum and biorefineries are therefore small. It is a satisfying story and it is the one I expected to confirm.
Tested at cashew-apple densities, it does not hold:
There is no interior minimum. Cost falls monotonically to the edge of the resource, and even at a hundred-kilometre radius transport is under a fifth of cost. The famous biomass-radius result does not bind here, and this paper will not pretend it does.
What binds instead is two things the cost model does not contain. How much feedstock actually exists inside any radius — and perishability. Cashew apple ferments within a day, so its radius is set by hours:
No freight rate enters that calculation at all. The constraint is a clock.
G6's September brief for TBFCo states 32 million date palms in the Kingdom at roughly 20 kg/year each of surface fibre, fronds and seeds — currently waste with no competing use. That is 640 kt/year of residue.
Converted by a lignocellulosic route, the mass yield is the softest number in the chain, so it belongs in the brief as a band and never as a point:
The framing survives its own arithmetic: the residue stream is the same order as the existing business and plausibly several times it. What the brief should not do is quote a single litre figure, because the ends of the band differ by 1.8×, and a reader who checks will find that out faster than a reader who doesn't.
§§1–3 say the material cannot travel. §4 says the resource is real and local. Together they say the scattered-co-op network is not a logistics network, because there is nothing to haul along it.
IATA's global feedstock assessment reaches the same place from the other end: feedstock is regionally determined — agricultural residues in North America, sugar and starch in Brazil, over 80% of the world's palm oil in ASEAN, rice and wheat residues in India and China — and smallholders “lack resources, incentives, or organizational structures necessary to participate in formal biomass markets,” requiring intermediary hubs and cooperative models.
Process, catalyst, specification, certification and offtake — the chain that lets a litre made in Ceará and a litre made in Al-Ahsa satisfy the same CORSIA obligation. A licensing and standards network, with local conversion at every node, and nothing crossing an ocean that is not already dense.
∎G6's own two briefs are already built that way — cashew apple in Brazil, date palm residue in Saudi Arabia, the same chemistry class and a different local feedstock. This paper adds only the reason, in numbers, so that the next node can be chosen rather than found.
Seven gaps. The two that gate anything commercial:
No sustainability or ILUC accounting. Oil-palm-based SAF is restricted under EU rules on land-use-change grounds. Nothing here addresses whether any pathway discussed would certify under CORSIA, ReFuelEU or RenovaBio — and that is a gating question, not a detail.
The yield band is not sourced. 12–22% is plausible for lignocellulosic routes generally; no study of date palm residue specifically was consulted. The brief cites a 2026 UAE pyrolysis paper for feasibility, which is not the same thing as a yield, and the distinction is exactly the sort a counterparty's technical reviewer is paid to find.
Producing script: book6/wp127-no-silk-road-for-biomass-verify.py — 5 sections, 7 gaps. CITED: IATA, Global Feedstock Assessment for SAF Production, outlook to 2050; G6 internal briefs Date Palm Biomass: A Second Feedstock Track for Saudi SAF Production (TBFCo, September 2026) and Cajueiro SAF · Complete Market Landscape (August 2026). All prices are order-of-magnitude — see gap Q1.