Supply Chain Action Points
Read this first — the conclusion, and the moves to make:
- Before the first export movement, split your transport model so a railroad tariff and a trade tariff sit in different lines, and confirm which one your September 1 carrier notice actually was.
- Track the Illinois River gauge weekly against your own barge index; if the gauge is still falling when you commit volume, price the rate at plus 13 to 15 percent rather than at base.
- For volume you expect to divert to rail, budget the secondary premium at the current $638 a car and confirm with two carriers how many shuttle cars you could actually get during the build before you rely on it.
- Lock what diesel exposure you can: at $5.599 a gallon, up 49.9 percent, ask for a fuel-surcharge cap on trucking contracts for the next 90 days and model a further 15 percent move.
- Set a diversion trigger by tonnage and date: if more than a quarter of your harvest volume is still unpriced within two weeks of the build, secure rail or truck cover in writing, because the secondary market is where the price gets set.
The USDA September 3 Grain Transportation Report shows broad upward pressure: the Illinois River barge index rose 13.6% week-on-week, rail shuttle secondary premiums averaged +$638 per car, and US on-highway diesel was up 49.9% year-on-year at $5.599/gal. BNSF, Union Pacific and Canadian National also implemented soybean tariff changes effective Sep 1. Shippers face higher baseline transport costs into the harvest build.
The Analysis
The USDA Grain Transportation Report for September 3 landed with three numbers pointing the same way: the Illinois River barge index up 13.6 percent week on week, rail shuttle secondary premiums averaging plus $638 a car, and US on-highway diesel at $5.599 a gallon, up 49.9 percent year on year. BNSF, Union Pacific and Canadian National put soybean rate changes in place on September 1.
Two of those three numbers are seasonal and one of them is not, and telling them apart is the whole job. The barge line is a river-level story as much as a demand story. The diesel line is fuel. The rail premium is a capacity statement, and it is the one that tells you what the other two are about to do to you.
If you are moving grain into the harvest build, this is the week the transport budget you set in spring stops being a plan and starts being a wish.
Start with why the river index moves the way it does, because the number looks like demand and mostly is not. The Illinois River barge rate spikes when water drops: less draft means fewer tonnes per barge, the same fixed barge cost spreads over less cargo, and the per-tonne rate climbs before anyone has decided anything. A 13.6 percent weekly jump is a gauge reading wearing a demand costume. The diesel line is simpler and worse, up nearly half year on year, and it hits every mode at once because every mode burns it. The rail shuttle premium is the interesting one: a secondary-market premium of $638 a car is what you pay when primary shuttle capacity is already committed and you are buying what is left. That is not fuel and not the weather. That is capacity.
Put the three together and the mechanism is plain. This is the harvest build, the window when grain moves to the elevator and the elevator moves it out, and it collides with a fuel market that ran up all year and a river doing what rivers do in a dry stretch. The three modes are supposed to take turns absorbing the volume. In practice they fail together, and that is what the reported numbers hide. When the river gets expensive, cargo diverts to rail. When primary shuttle is full, the diverted cargo lands in the secondary market, which is exactly where that $638 premium gets set. Barge tightness and rail premium are the same event wearing two hats.
From the corridor it looks like this. Grain export runs on three legs and the handoffs between them are where the money leaks. The river terminal takes barge and truck, the shuttle loader takes unit trains, the export elevator takes all three, and when one leg tightens the queue does not disappear, it relocates. I have stood at a river terminal in a low-water week watching trucks wait because the barges could not be loaded to full draft and the outbound slot got cut. That trucking bill showed up in no barge index. It showed up on the shipper's own books three weeks later, under a line nobody watches.
Who takes the hit, in order. The grain exporter moving soybeans and corn to the Gulf is first, because they sit at the end of all three legs and absorb every premium. The river-dependent shipper with no rail siding is next and worst off, because they cannot divert and they cannot negotiate, so they pay the barge rate and they wait. The shuttle-served elevator is better protected through primary capacity, at least until the secondary premium tells a different story. The truck-only local shipper is barely touched by the barge index but fully exposed to the diesel line, which applies whether the haul is 20 miles or 200. The group that is genuinely fine is anything on a fixed rail contract with the rate already locked, and that is a shrinking club.
Timing is where the modes differ and where most plans go wrong. Diesel moves immediately, it is in the price of the next load with no lag worth modelling. The barge rate moves fast too, spot pricing, days not weeks, which is why a 13.6 percent weekly move is real and not noise. Rail is the lagged one. Contract rates were set months ago and will not move, but the secondary premium you pay to get on when primary is full moves weekly, and that number decides whether the diverted barge cargo actually gets carried. So you get a fast squeeze from fuel and river, and a slower, stickier one from rail, arriving in sequence across the build rather than all at once.
The part the harvest headlines skip is the terminology trap and the second-order queue. When a rail carrier files a soybean tariff change, that is a freight rate and charge schedule, a railroad tariff, and it has nothing to do with a trade tariff or a duty. Read the two as the same thing and you will misprice your landed cost, because they belong in different lines of your model. The second half is the queue. Every time the river pushes cargo onto rail and the secondary premium rises, the next river shipper watches the rail option get more expensive too, diverts back, or waits. The modes stop being substitutable exactly when you need them to substitute, and that coupling is invisible in any single index.
Run the arithmetic with clear assumptions. Take an exporter moving 500,000 bushels of soybeans to the Gulf in the harvest window by barge. A bushel of soybeans is 60 pounds, so 500,000 bushels is about 15,000 short tons. Assume a base barge rate of $30 a ton, so $450,000 on the barge leg. A 13.6 percent jump takes it to $34.08 a ton, or $511,200, an extra $61,200 on one harvest movement. Now assume half that volume diverts to rail because the river cannot take full draft. If the secondary premium is $638 a car and a shuttle car holds about 110 tons, that is roughly $5.80 a ton, so 7,500 tons adds about $43,500 in premium on the diverted half, before the diesel-driven trucking increase at the terminal. Under these assumptions the harvest alone carries over $100,000 of avoidable cost that was nowhere in the budget set in spring.
What turns it. Rain on the Illinois basin ahead of the build peak lifts the river, drafts recover, and the barge index gives back most of the 13.6 percent as fast as it took it. A softer diesel print does the same to the fuel line across all three modes at once. The one condition that reverses everything is export demand: if Gulf soybean demand weakens, elevators stop bidding for capacity and every premium above base collapses, river, rail and truck together. Watch the river gauge and the export line, not the headline harvest number, and you will know which way your cost is going about two weeks before it reaches an invoice. The trap in a build like this is treating a river-level spike as a permanent reset and locking next year's transport budget to a number that was born in a dry month.
↑ Back to the key points