US railroads log record 297,000 intermodal loads a week in August

I have been watching US rail for the better part of a decade, and the thing about a number like 297,000 loads a week is that it does not move in isolation. Behind it sits a pile of shippers who, a year ago, would have put that box on a truck and who this year put it on a train instead. The reasons are not mysterious. Diesel has been expensive enough that the fuel surcharge alone on a long-haul truck load is real money. Truck insurance premiums have climbed. And the driver pool is thin in places where it used to be deep. When those three push the same direction, the box slides onto the rail, and the volume shows up exactly where the AAR says it does.

So the first practical question for an importer or exporter is whether the domestic leg you are running on truck today should move to intermodal. Most of the people reading this are not moving freight by rail for the romance of it; they are moving it because a container landed at Los Angeles or Long Beach and has to reach a warehouse in Chicago or Dallas, or because product built in the Midwest has to reach a coastal port for export. That domestic leg is a decision made every single week, and the record volume is the cue to reopen it.

Let me put a concrete case on the table rather than hand-wave. Say you bring in two hundred FEU a month through Los Angeles and Long Beach and you need them in Chicago. Running that container over the road, from the port to the Chicago warehouse, you are looking at something around three thousand eight hundred dollars a box in a normal market, and more when capacity is short. Put the same box on the intermodal ramp, train to the Chicago-area ramp, and you are closer to two thousand six hundred. That is twelve hundred dollars a container in favour of rail. Two hundred containers a month times twelve hundred is two hundred and forty thousand dollars a month, call it close to three million a year, on one lane, before you have changed a single supplier or renegotiated a rate card.

The catch, and there is always a catch, is time. Rail is not instant. That same Los Angeles to Chicago move by truck is about three days door to door if the driver runs clean; by rail you are more like five, and sometimes six once you count terminal dwell at both ends. Two extra days in transit is the price of the saving. Whether that price is worth paying depends on what is in the box. To put a number on it, assume each FEU carries fifty thousand dollars of goods and your annual inventory carrying cost is eight percent. The extra two days tie up two hundred boxes for two days, which is four hundred box-days; at fifty thousand dollars and eight percent that is about eleven dollars a box-day, so the carrying cost of the slower rail move is roughly forty-four hundred dollars a month against the two hundred and forty thousand you save on line-haul. The time penalty is real but small, which is why the rail tilt makes sense for most of the book.

Now tie that to the calendar, because the AAR number lands in August and the freight that matters next is the holiday build. If your goods need to be on the shelf by early November for the late-November rush, and rail costs you two extra days in transit, then every container you route by rail has to leave the port two days earlier than the same container would by truck. That sounds trivial until you stack two hundred of them and realise the whole booking window has shifted two days left, and two days is exactly the slack that disappears when a vessel rolls or a terminal backs up. The record volume is the warning here: a network running at its strongest in eight years does not have spare slots to hand you in the last week of October. You lock space in early October or you gamble.

And this is where the second number matters more than people notice. Carloads at 235,000 a week and fifteen of twenty bulk groups up means the entire rail network is busy, not just the intermodal side. Grain, chemicals, plastics, metals, autos - the cars that carry those commodities share crews, yards, and locomotives with the trains that carry your containers. When bulk is hot, the network has less slack for intermodal, and intermodal is already at a record. So the peak you are worried about for holiday boxes is not a separate event from the bulk boom; they are the same tight network. I would not assume a flatcar will be sitting empty waiting for my phone call in week 44.

The cost of being late to book is the part people undercount. In a tight intermodal market, the railroad or the intermodal marketing company will sell you space, but the rate for a last-minute, unguaranteed box is not the rate you saw in September. Assume a peak premium of three hundred dollars a container on a box you did not pre-book. On two hundred containers that is sixty thousand dollars a month, pure premium, for the same train. And if the ramp is full and your box gets rolled, you do not get the train at all - you get a truck, at the three thousand eight hundred dollar rate, which is the two hundred and forty thousand you were trying to avoid, plus the premium you already paid. The hidden benefit of locking early is not the discount, it is the certainty that you are on the train and not on the expensive backup plan.

There is a trap inside the rail saving that I have seen bite good operators, and it sits at the ends of the move, not the middle. Rail is cheap because it does the long line-haul on steel and leaves the first and last miles to a truck. But at the destination ramp your container still needs a chassis and a drayage truck to reach the warehouse, and chassis pools at some inland ramps run thin. If your box arrives and there is no chassis free, it dwells on the ramp, and ramp dwell turns into per-diem and demurrage that eats the twelve-hundred-dollar saving before the box ever moves. The driver shortage that pushed you to rail in the first place is the same shortage that makes the final drayage hard to book. So the rail decision is not finished when you book the train. You have to pre-arrange the chassis and the drayage at the receiving end, ideally with a committed drayage provider who holds chassis, or the back end of the move quietly costs you the front end's saving.

For the exporter side of the same question, the mirror image holds. If you are moving product from a Midwestern plant to a coastal port for export, rail to the port saves the same money and carries the same two-day cost, and the same chassis problem sits at the port ramp when the box comes off the train and waits for a truck to the terminal. The bulk number matters to you too, because if your input materials ride in rail cars - and a lot of industrial inputs do - fifteen of twenty categories up means your supplier's cars are busy, lead times stretch, and cars themselves can get short. I have seen a plant held up not by demand but by not being able to get an empty car when it needed one. Booking cars early, and watching demurrage at the origin, is the export-version of the same early-lock discipline.

One more reason the rail tilt makes sense this year beyond the bare rate is that it hedges the volatility you cannot see yet. Truck rates carry a fuel surcharge that moves with diesel week to week, and a hard winter storm can pull truck capacity out of a corridor for days with no warning. Rail burns less fuel per box and is far less exposed to a single storm taking the whole lane down. Insurance and driver costs are structural, not one-off, so the gap between truck and rail is not a spike that will fade; it is a spread that has widened and looks likely to stay wide through the fourth quarter. Putting more of your domestic volume on rail is, in plain terms, buying a steadier price and a steadier capacity for the part of the move you control.

None of this says rail is always the answer. For a short haul - under roughly five hundred miles - the drayage and terminal handling eat most of the train's saving, and truck wins on both cost and time. Rail also needs lead time you do not have on a same-day emergency; you cannot call an intermodal train the way you call a truck. And if your warehouse sits more than a drayage hop from the ramp, the last mile gets long enough to cancel the gain. The decision is a split, by distance and by SKU, not a religion. The mistake I see most is the opposite extreme - someone reads the rail record, switches everything to rail, and then discovers the chassis and the two-day slip at exactly the wrong moment.

One detail in the AAR print that people skim past: the 297,000 figure includes trailers, not only containers. A share of that volume is trailers on flatcars, which is a less efficient form of intermodal than containers on well cars, and it tells you the network is being used hard in every mode it has. For an importer the actionable part is the container share, because that is the box you already own; the trailer number just confirms the whole system is full. When even the less efficient option is running at record, the efficient option has even less slack than the headline suggests.

On picking the split, a simple threshold beats a committee. Take the two-day rail penalty and price it per SKU: if a SKU turns faster than the carrying cost of two extra days, keep it on truck; if it sits in inventory anyway, rail it. On the Chicago lane the carrying cost of the two days was about forty-four hundred dollars a month across the whole two hundred boxes, which is twenty-two dollars a box - trivial against the twelve-hundred-dollar line-haul save. So for most of the book the answer is rail, and only the genuinely time-critical tail stays on truck. The trap is letting the tail define the whole decision.

Free time at the ramp is the line that decides whether the saving is real. Most ramps give you a couple of days before per-diem starts, and per-diem on a container can run twenty to thirty dollars a day and climb. Two days of unplanned dwell at the destination, across two hundred boxes, is four hundred box-days at twenty-five dollars, or ten thousand dollars a month, which is small against the saving but real, and it is the number that disappears the moment the chassis is pre-arranged. The point is not the size, it is that it is avoidable, and avoidable cost you leave on the table is cost you chose to pay.

The twenty commodity groups the AAR breaks out are worth a look because fifteen of them up is a broad signal, not a single hot sector. Grain, coal, chemicals, plastics, metals, motor vehicles, stone and glass, petroleum - when most of those rise together, industrial demand is broad and the cars that carry them crowd the same yards as your containers. For an exporter buying US-made input, that broad strength also means input is moving and prices are firm, so the rail car you need for your own outbound is competing with fifteen busy categories. Booking early is how you jump that queue.

Insurance is the quiet part of the truck-cost story. Truck liability premiums have risen to where some carriers have exited lanes rather than price them, and that removes capacity in exactly the corridors where intermodal competes. A railroad does not reprice by the load the way a trucker feels fuel and insurance; its cost curve is smoother, which is why the rate you lock with the railroad is more stable through the quarter than any truck quote you get. Stability is a feature when your holiday plan cannot absorb a rate that doubles in week 44.

The chassis pool is worth understanding because it is the lever you actually pull. At many ramps the chassis are owned by a pool operator, not the railroad, and during a peak the pool runs short because every shipper wants one at once. If you have a committed drayage provider who holds chassis, your box gets one; if you are calling spot, your box waits. This is the single most common reason a rail move that looked cheap on paper ends up costing more than the truck it replaced. The fix is boring and reliable: arrange the chassis before you celebrate the rate.

To show the rail tilt is not only a long-haul story, take a medium lane like Oakland to Dallas, about half the distance of Los Angeles to Chicago. Truck there might run twenty-six hundred dollars a box, rail around nineteen hundred, a seven-hundred-dollar save. Two hundred boxes a month is still a hundred and forty thousand dollars a year, and the transit gap shrinks to a day. Even at medium distance the math favours rail for the slow share, which is why the record is network-wide and not just on the coast-to-coast lanes. The short-haul exception stays real, but it is a smaller slice than people assume.

The last piece is the relationship, not the rate. A railroad or intermodal marketing company that gives you a written guaranteed-space rate for the peak is worth more than a cheaper open quote you can be rolled off. Ask specifically for space on your lanes in weeks 42 through 47, name the destination ramp, and get the free time in the same document. The carrier would rather lock your volume now than fill it with spot later, so the conversation in early October is one you can win. Wait until the network is full and the only thing left is the premium.

So here is what I would actually do this week if I were booking this freight. The first move, and it is the one that costs nothing to start, is to open the intermodal space conversation with your railroad or intermodal marketing company for the October through November build, and to ask for a written rate with guaranteed space rather than an open quote you can be rolled off. Get the number for your specific lanes - Los Angeles to Chicago, Oakland to Dallas, whatever you run - and pin the per-container saving against truck so the finance team sees it in dollars. By the first week of October, commit the volume you are confident about, the slow and bulky share, and leave the fast SKUs on truck where the two days would hurt. At the same destination ramp, line up a drayage provider who holds chassis and confirm the free time at the ramp in writing, because that is the line where the saving leaks out.

For anything moving in rail cars rather than containers, book the cars now and track demurrage at origin so a tight network does not strand your input. And keep one truck lane warm as the backup, priced and ready, so a rolled train does not become a stock-out. The backup is not wasted money; it is the insurance that lets you push the rest of the book to rail without fear, and fear is what makes people book late.

The pitfall list is short but real. Do not assume the September rate holds in week 44; get it guaranteed or budget the three-hundred-dollar premium. Do not count the rail saving until the chassis and drayage at the far end are booked, or the box sits and the per-diem climbs. Do not put time-critical goods on rail and then promise a customer a truck-speed date. Do not switch the whole book at once; prove the split on one lane for one cycle, watch the dwell and the demurrage, then widen it. And watch the bulk number as your early-warning signal - when fifteen of twenty categories are up, the network is full, and full networks do not forgive a late booking.

I have watched enough peak seasons to know the pattern. The people who lock intermodal space in early October sail through; the people who wait for the October rate card to confirm what they already suspected are the ones calling me in week 44 asking whether a truck is available at any price. The AAR print is the confirmation, not the news. The record has already been set; the only question left is whether your boxes are on the train or on the expensive plan nobody wanted.

作者 Leo

• By the first week of October, commit intermodal space for the October-November holiday build on your confirmed slow-and-bulky lane volume and get the per-container rate with guaranteed space in writing.
• By 4 October, lock a drayage provider that holds chassis at the destination ramp and confirm the ramp free time in writing, so dwell does not erase the rail saving.
• By 11 October, split the book by SKU: route the slow, bulky share (assume the 200 FEU/month Chicago lane) to rail at a 1,200 dollar per FEU saving, keep fast SKUs on truck.
• For rail-car movements, book cars now and track origin demurrage weekly; do not let a full network strand input materials.
• Keep one truck lane priced and ready as backup so a rolled train costs at most the 300 dollar per FEU peak premium, not the 1,200 dollar per FEU truck gap.
• Recheck the AAR bulk signal weekly; when 15 of 20 commodity groups stay up, treat the network as full and pre-book, never chase space in week 44.