China-Europe Rail's Middle Corridor Tops 4,000 Trains as Tonnage Jumps 24.6%

The Analysis

Erenhot is the crossing nobody puts on a slide. It sits on the Mongolia border, north of the Gobi, where the standard-gauge Chinese rail meets the broad-gauge Mongolian and Russian rail and every wagon has to change wheels before it can keep going. Freight people call the route through here the middle corridor of the China-Europe rail network, and on 3 November it did something quiet but telling. The middle corridor passed 4,000 trains for the year. By 2 November the port had handled 3,993 trains and 3.902 million tonnes, up 4.0 percent in trains and 24.6 percent in tonnage year on year.

Stop and look at those two growth rates side by side, because that pair is the whole story. Trains up 4 percent. Tonnes up 24.6 percent. Those two numbers should move together, roughly, and they do not. When they split this far apart, something structural is happening on the trains themselves, and that is where the money, the risk and the bottleneck all live.

I have walked this border. Erenhot is a wheel-change town, and a wheel-change town has one throttle that matters. Everything about the 4,000-train number has to be read through that throttle, not through the headline count.

Let me start with the corridor, because people mix up two different legs and then argue about the wrong thing.

The main China-Europe rail artery runs through Kazakhstan, Russia and Belarus and lands at Malaszewicze in Poland, the fat pipe that carries the bulk of the volume toward northern Europe. Erenhot is not that pipe. Erenhot is the middle corridor, the Mongolia-facing gate, the one that runs up through Zamyn-Uud and Ulaanbaatar and then joins the trans-Siberian web further north. It is a different leg, a different set of customs authorities, a different set of winter problems, and a different destination mix. A lot of forwarders treat it as the backup door, the one you use when Malaszewicze sneezes. What the 4,000-train milestone tells you is that the backup door is no longer a backup. It is carrying real, scheduled, industrial freight, and the tonnage number proves it. Adding two routes this year, to a network of 76 services reaching more than 70 hubs in over 10 countries, is not a backup operation. That is a standing corridor.

Now the scissors gap, the 4 percent against the 24.6 percent, and why I care about it more than the round number.

Run the division. Divide 3.902 million tonnes by 3,993 trains and you get roughly 977 tonnes per train. That is the average payload per train this year. Now compare the growth: a 1.246 factor on tonnage divided by a 1.040 factor on trains gives about 1.198, which is close to a 20 percent jump in the average weight carried per train in a single year. Twenty percent. That is not noise and it is not rounding. Something changed about what a train carries. And there are two ways that happens, both of which you should understand, because they point in different directions for your business.

The first way is mix. The report says the exports are shifting from general cargo, the light stuff, textiles and small consumer goods and daily necessities, toward new energy vehicles, home appliances and machinery. A container full of stuffed toys and a container full of washing machines occupy the same slot on the same flat car, but they are nowhere near the same tonne. When the mix tilts from light consumer goods to vehicles and appliances, every train gets denser and the tonnage outruns the train count. So one reading of the scissors gap is a story about value climbing the ladder, light cargo giving way to heavy, higher-value industrial goods. That is the flattering version, and it is partly true.

The second way is load factor. A train can run long and light, half empty, moving for the schedule rather than for the payload. If in the past the corridor was running some empty or underloaded wagons to keep a published departure, and today those wagons are filled, then the tonnage lifts without the train count lifting much. This reading says the same train count is simply being used better. It is also plausible, and it is important, because it means the corridor has absorbed a big jump in real freight without needing a big jump in scheduled services. Efficiency, not capacity.

Which one is it. Honestly, probably both, and the split matters because they have opposite implications. If it is mix, your cost per tonne for light goods on this corridor is going up as heavy goods crowd the schedule, and you should care about getting your booking priority sorted. If it is load factor, the corridor has slack it has now used up, and the next leg of growth has to come from more trains, which the wheel-change throat may or may not be able to deliver. The two explanations look identical in a headline. They are completely different on the ground.

Here is where I stop summarizing and start telling you the thing the wire copy missed, and it is the hardest point in this piece. The scissors gap is not the interesting event. It is a symptom. The interesting point is that a tonnage outrunning train count means the binding constraint on this corridor is no longer the number of trains. It is the tonnes. And on a border like this one, tonnes are exactly the thing the gate is bad at.

Think about what happens at Erenhot. China runs standard gauge, 1,435 millimetres. Mongolia and Russia run broad gauge, 1,520. No train crosses this border. Wagons cross it, either lifted container by container from a Chinese platform onto a broad-gauge platform, or with the whole bogie swapped under the car at a wheel-change hall. That hall has a capacity measured in units per day, and every unit it handles is a chassis, not a tonne. Now here is the twist. A switch from light general cargo to EVs and appliances does not just add weight inside the same box. It often adds forbiddingly heavy, unusually large, or dangerous-goods-classified items that are slower to handle.

A loaded EV or a pallet of lithium-battery appliances is not the same lift as a bag of textiles. Heavy, high-value, sometimes hazmat cargo takes longer per unit through the hall, needs more inspection, and ties up more flat cars. So the same 4,000 trains are now carrying a payload that the gate handles more slowly per unit of weight, and the tonnage jumps while the train count barely moves. The scissors gap is, in part, a picture of the gate being asked to do a heavier, harder job with the same wheels and the same halls.

That has a direct cost consequence that nobody prints. If the throat is measured in units and the freight is getting heavier and more demanding per unit, then the queue time per train can rise even when the train count is flat. The corridor can look calm on a train-per-day chart and be working much harder underneath it. For a shipper, that is the risk: the published schedule stays reassuring while the real dwell at the border inches up. Keep an eye on transit time, not on train counts. The train count is a marketing number. Transit time is the number that lands on your customer's desk.

Now step back and think about what this means for the shipper who does not read railway statistics for a living. The middle corridor is advertised as the crowded main line's relief valve, and a relief valve that is quietly working harder than its chart suggests is the most dangerous kind, because you switch to it precisely when the main line is under stress, and you switch to it at the exact moment it is least able to absorb you.

There is a second overlooked point, and it is about the map, not the arithmetic. The middle corridor is not a second version of the Malaszewicze route. It is a genuinely different leg, going up through Mongolia, and it does not serve the same destinations in the same way. That means two important things. It is closer, in purely geographic terms, for the north China export base and for a slice of the western network, so for the right origin it can be the shortest way out, not merely the fallback.

And it feeds a different set of European hubs, which for a given consignee can be either better or worse, and that is the part people miss. A forwarder who reflexively books the middle corridor because the main line is congested may hand the load to a route that reaches a hub far from the eventual customer, and the last-mile cost eats the transit-time gain. Do not treat the corridors as interchangeable. Treat them as two different doorways into Europe, and pick the door by where the goods are actually going.

One more thin layer on the map, and it needs a careful word. The port says it runs 76 services. A service in that list can mean a route that sails, in the rail sense, on a regular published schedule, or it can mean a lane that was opened and tested and is not yet a dependable weekly drumbeat. The difference matters enormously to you, because a dependable service is a service you can plan a factory around, and a trial lane is a service you can only use for cargo that has slack in its schedule. When you read 76, ask the next question. How many of those 76 run on a firm published frequency week after week, and how many are on-demand or seasonal. The corridor's real capacity is the first number, not the second. I would rather ship on thirty steady services than on seventy-six that come and go with the season.

Now the geometry of the alternatives, because the obvious move is to say that when one corridor tightens, shift to another.

The main door through Malaszewicze is the workhorse, and it moves the bulk of the volume toward Poland and the northern European hubs. It is also the one most exposed to the weather of geopolitics and border politics, which is why the middle corridor has drawn interest in the first place. The middle corridor through Erenhot is shorter for some northern origins and dodges some of the political risk on the classic route. The sea-land hybrid through Kazakhstan and the Caspian, sometimes called the trans-Caspian route, is the third leg. Each has a different cost and a different clock.

Shifting volume from one to another is not a click. It is a re-book, a new set of customs filings at a new border, a new transit time to promise, and, often, a new set of counterparties at the handoff. The cost of the shift is not a fee. It is the lead time you have to re-price and the promises you have to re-cut. And everyone reads the same news, so the moment one corridor tightens, the others fill, which means your alternative is only an alternative while it is still underused.

Let me put numbers on it, assumptions stated, because a figure without assumptions is a rumour. Suppose you move a modest programme, sixty containers a month, from a north China origin to a central European destination, and suppose the rail transit runs around eighteen days door to door when the border is calm. If a heavier payload mix adds two days of dwell at the wheel-change hall, that is two days on every shipment in the programme, which compounds if you are running a just-in-time line.

Over sixty containers, two days each is a hundred and twenty container-days of inventory float per month. Value those containers at forty thousand dollars of goods each and carry inventory at twenty percent a year, and the extra dwell costs you on the order of 120 times 40,000 times 0.20 divided by 365, call it about 2,600 dollars a month, before you count the expedite premium on the one or two containers that actually matter or the overtime when the delayed wave lands at the warehouse.

That is the number for a small programme. Scale it to a serious shipper and the same two days becomes a line item worth negotiating over. The point is not the exact figure. The point is that two days of border dwell is not a rounding error on a rail programme, and the scissors gap is exactly the signal that dwell is under pressure.

Here is the counterfactual, because a one-sided story is a sales pitch. Suppose the payload growth is mostly mix and mostly benign, heavy goods that the hall handles fine because they are containerised and standard, and suppose the corridor keeps adding scheduled services and keeps the throat ahead of the volume. Then the 24.6 percent is a pure good-news story and the middle corridor quietly becomes a first-choice trunk for a wider band of cargo. I would not rule that out, and if you are a forwarder it is the case you want. The case that should worry you is the other one. Tonnage keeps climbing, the wheel-change halls do not, dwell creeps up, and the corridor that looked like spare capacity six months ago becomes a congested door of its own, just as everyone has rerouted onto it. Corridors have a habit of getting popular right before they get slow, and the tonnage number is the first place that slowdown shows up.

So what do you actually do with this, as a shipper or a forwarder. You do not abandon rail and you do not take the 4,000 number at face value. You do the dull things that protect the promise, and you do them before the winter peak, not after.

Track transit time, not train counts, on the lanes you actually use. Set a simple rule: if the median door-to-door on your middle-corridor lane slips by more than a day against your baseline for two straight weeks, raise it with the operator and re-promise your own customers before they find out from the late box. The train count will not warn you. Dwell will.

Ask your operator one direct question about the 76. Of those services, how many run on a firm weekly frequency, and which of them touch the hub your cargo actually needs. Book your firm volume on the steady ones and treat the trial lanes as opportunistic capacity only. Do not build a production schedule on a lane that might not sail next month. A published frequency is a promise you can plan a factory around. A lane that exists on paper is a lane you can only use for cargo that has slack in its schedule, and the moment your customer needs a date you can hold, it stops being a lane at all.

Qualify the corridor by cargo type, not by headline. If you are on light general cargo, understand that as heavy industrial goods crowd the schedule you may increasingly be the one who gets bumped, and get your booking priority and your cut-off discipline tight. If you are on EVs or appliances, understand that your cargo is exactly the freight that strains the wheel-change hall, so build in the extra dwell now rather than discovering it in a service failure.

Keep two doors open and price both. Register with an operator or agent on the main Malaszewicze route and on the middle corridor, even if the bulk of your volume flows one way this quarter, because a corridor you have never booked is a corridor you cannot switch to in a hurry. And remember the trans-Caspian leg as a third valve, thin but real, for the cargo whose schedule can take it.

Fix your contracts for a reroute. If your supply agreements assume a certain rail transit, insert a clause that lets you reroute across corridors and adjusts the promised date for documented border congestion, with the liability shared rather than dumped on you. The word from this corridor is that tonnage is outrunning trains, and that is precisely the condition under which transit promises get broken.

Do those things and the middle corridor becomes a route you use on purpose, chosen for a reason, not a door you stumbled through because the main line was busy. Skip them and you will learn the difference between a train count and a transit time the expensive way, at the wheel-change hall, on a winter night, watching your box sit while the snow comes down and the line ahead does not move. Every corridor has a story, and this one has a pothole. The pothole is that the tonnes are winning the race against the trains, and nobody put that on the milestone banner.

↑ Back to the key points

— By Ömer Kaya