BNSF Railway Launches RELOAD Program to Increase Empty Container Utilization and Reduce Drayage Miles

The Analysis

A box arrives in Chicago loaded. It is stripped in a warehouse off one of the interstates south of the city. Then it sits. Then somebody drives it back empty so that it can be filled again by somebody else. That return trip is the most expensive nothing in intermodal, and it has been the most expensive nothing in intermodal for every year I have been doing this. On 10 June BNSF gave it a name. RELOAD, which stands for Railway Empty Load Optimization And Distribution. The idea fits on a napkin. Take an import container that has just been emptied at an inland destination, look for an export booking near it, and hand the box straight across instead of sending it back toward the coast.

The railroad says the first deployments are in Chicago and Dallas-Fort Worth, and that participating shippers should see the cost of repositioning empties come down by 15 to 20 percent. I treat that range as a target somebody else has set rather than a number I have earned, and I will come back to why before the end of this. What makes the programme worth writing about is not the percentage. It is the sentence underneath it, which is that imports and exports pass each other on parallel drayage routes. That is true, it is true on particular roads I have driven, and it has been true in every market I have worked in.

This is a road problem wearing a software badge. One truck is running an empty back to a pool so the box can be loaded again. Another truck, on the same stretch of interstate going the other way, is running an empty out from that pool to a shipper who needs one. Both are moving air, and both are spending a driver's day doing it. RELOAD is the railroad using the fact that it can see both legs to notice that those two trucks were ever two trucks. That is worth something. How much, and to whom, is the rest of this piece.

Let me describe the day as it actually runs, because the inefficiency is invisible until you watch one box from end to end. An import container lands at the coast, goes onto a train, arrives at an inland ramp, and is drayed to a warehouse. The warehouse strips it over a day or two. Now the box is empty and it belongs to somebody who is not the warehouse, and it has to go somewhere. Usually it goes back toward the ramp or a container pool, and it waits there until an exporter books it, and then a second truck brings it back out to a shipper who may be fifteen miles from where it was emptied in the first place.

Triangulation is the obvious answer and it has always been the obvious answer. The box goes from the import strip point straight to a nearby export load, and the pointless run back to the pool disappears. What disappears with it is not only miles. One dray move comes off the board, one gate transaction stops happening at the ramp, one chassis day stops being billed, and one driver's day stops being spent on a load of air. Multiply that by the number of boxes a month and you have a real number, which is presumably what somebody at BNSF did before putting a name on it.

The reason it needed a programme is that it was never anybody's job. Coordinating it manually means two dispatchers in two companies, each of whom can see only their own leg, agreeing on a box, a place and an hour, on the phone, often the same morning. It happened when two people happened to know each other and it did not happen otherwise. A railroad that moves both the loaded import and the loaded export can see both legs on one screen, which is the whole argument for letting the network operator run the matching rather than leaving it to whoever happens to pick up the phone.

The two launch markets make sense for the same reason. Chicago and Dallas-Fort Worth are not the biggest inland points in America by accident. Chicago is where the east-west corridors cross the north-south ones, and Dallas-Fort Worth sits on the I-35 corridor running up from the border with a large consumer and industrial hinterland wrapped around it. Both have the density that makes a match statistically likely. Where volume is thin, a matching engine has nothing to match, and the saving stays on the page.

Let me put a number on it with every assumption written down. Suppose an operation moves 400 import containers a month into the Chicago market. Suppose each of those empties today runs an average of 45 miles before it finds its next load, and suppose the all-in cost of that repositioning, counting the driver's time and the chassis day, is 180 dollars a box. Both figures are mine and you should replace both with your own. On those assumptions the monthly bill is 72,000 dollars. Take 15 percent off and you save 10,800 dollars a month, which is a little under 130,000 dollars a year. Take 20 percent off and it is 14,400 a month, roughly 173,000 a year.

The time side matters more to me than the money, and it is usually left out of the conversation. Assume a dray move averages 45 miles at an effective 20 miles an hour once the gate is counted, so about two and a quarter hours per empty run. Cutting 15 to 20 percent of that mileage removes roughly seven to nine miles per box, which is twenty to twenty-seven minutes of running. Across 400 boxes that is 135 to 180 driver hours a month handed back, which works out at somewhere between seventeen and twenty-two eight-hour truck-days. Nobody pays for those days today, and they are exactly the days you discover you needed when a ramp queue appears in July.

Two more figures come out of the same arithmetic. Four hundred boxes each making one fewer trip through a ramp gate is about 400 fewer gate transactions a month at a place where the gate is the bottleneck. And if each matched box saves a day of chassis time, that is 400 chassis-days a month, or roughly thirteen chassis freed up on any given day. I am not going to pretend those numbers are precise. They are the right order of magnitude, they are derived from assumptions you can check, and they tell you where to look.

Before any of that means anything, write down today's number. This is the step that gets skipped and it is the step that decides whether you can prove anything at all in August. For each ramp you use in Chicago and Dallas-Fort Worth, record the average empty miles per box, the average hours per empty move including the gate wait, and the monthly repositioning cost. Split it by ramp, because the ramps are not the same. Without that baseline, 15 percent is a claim made by somebody else about somebody else's boxes, and you will have no way to say whether it happened to yours.

Then be honest about which kind of shipper you are. If you both import and export through the same inland market, this is aimed directly at you: you are the person whose import box can walk across the road into your own export load, and you are the one who collects the miles, the hours and the chassis day. If you only import, the direct saving is not yours. What you get is thinner and still real: fewer empty trucks circulating around the ramps you depend on, shorter gate queues, and better chassis availability on the days you need one. Say which of those two you are before you sign up, because the first group will see a number on a report and the second group will mostly feel it in the yard.

For the two-way shipper, this changes a decision that used to be boring. Which inland ramp you route an import through was a question of drayage radius and rail price. Now it is also a question of whether empties get absorbed near that ramp. A ramp sitting in the middle of a cluster of exporters makes your boxes disappear into someone else's export booking. A ramp out in the middle of nowhere sends every empty back to the coast. That second ramp may still be the right answer on price, but it is no longer the same answer, and the difference should show up on the route sheet as a column rather than as a memory.

It reaches back to the coast as well, which is where it gets interesting for those of us who plan routes rather than ramps. Suppose you are choosing between landing a Chicago-bound box on the West Coast at Los Angeles or up in the Pacific Northwest. Assume the rail leg from the Pacific Northwest to Chicago runs four days and the one from Los Angeles runs five, and replace both with your own rail schedule because mine is an assumption. On those numbers the northern road is a day better and it also tends to have more slack at the gate. If the empties from that box are going to be absorbed by an exporter near the Chicago ramp, the rail day matters twice, because the box starts earning again a day sooner instead of sitting.

The port-area congestion claim deserves a careful reading too. Fewer empty movements around the port and the ramp does mean fewer gate transactions and less yard shuffling, and that is a genuine improvement. But it lands where the movements were, which is at the inland ramp and around it, not necessarily at your door. If your pain is at the marine terminal rather than at the inland ramp, this programme touches it indirectly at best. Do not budget for a congestion improvement at a port you do not operate through, and do not promise your customers a day back that actually belongs to somebody else's yard.

Now the operating rhythm, which is where pilots like this one usually die. Gates have hours and appointment windows fill in the morning for the next day. A matched box has to arrive at the export loader inside that window, not approximately inside it. Chassis have to be sitting under the box at the strip point on the day the export load is planned, and they usually are not, because the chassis follows the truck and not the box. Ask your drayage provider, your warehouse and the ramp the same three questions and compare the answers: what time does the gate open, how far ahead does an appointment have to be made, and where does the chassis actually sit overnight. Those three answers disagree more often than anyone expects, and the disagreement is measured in days.

The hardest condition is the timing of the match itself. Triangulation only works if there is an export load that needs a box, near the place where yours was emptied, inside the days when yours is available. An exporter who needs a box next Tuesday is no use to a container that came empty on Monday afternoon and has four days of free time left. Before you commit volume, decide what counts as a match: I would write it down as an export pickup within 25 miles of the strip point and a load inside four days of the box being emptied. Those are my numbers, chosen so that the match is genuinely cheaper than the old way. Anything looser and you have simply moved the empty run and called it a saving.

Seasonality will decide whether that condition is ever met. If your exports are thin in the first quarter and heavy in the third, the match rate will collapse in the first quarter no matter how good the engine is, and your empty repositioning cost will go straight back to where it was. Plan the pilot for a period when both sides of your business are running. Running it in a quiet month and concluding that the programme does not work is the same mistake as running it in a busy month and assuming it always will.

Then there is the clock on the box itself. An import container has free time attached to it, and an export booking has a window attached to it, and the two have to line up. If the box sits for two weeks waiting for an export booking that never quite arrives, you have not saved a repositioning cost, you have created a detention bill and blocked a parking space. The failure mode is quiet and it shows up on a statement a month later. Set a rule before the pilot: if a matched box has not been loaded by the fourth day, it goes back to the old arrangement and the match is recorded as a miss. A programme that cannot fail visibly cannot be measured.

Box condition and responsibility at handover are the other place this gets uncomfortable. When the box goes from your strip point to somebody else's loading dock, somebody has to accept it, and somebody has to own whatever damage is found when it arrives. That has to be settled among the warehouse, the drayage provider and whoever is loading it, before the first box moves, in writing, in one sentence each. I am not going to go through the paperwork here because that is not my desk. I am saying that the operational handover needs one named person on each side who can say yes, and that the conversation has to happen in June rather than in the middle of the first week.

I would also keep the 15 to 20 percent where it belongs, which is as an expectation attached to participating shippers in the two launch markets during initial deployment. It is not a universal, it is not a guarantee, and it is not a number you should put into a budget before you have a baseline of your own. Treat it the way you would treat a transit time a carrier quotes you: useful as a benchmark, useless as a promise. The saving you can bank is the one you measure against your own miles and your own hours, on your own ramps, with your own volume mix.

So run it small and run it measurable. Pick one ramp, in one of the two launch markets, on one lane where you have genuine two-way volume. Route at least 50 import boxes a month through the matching process so there is enough to see a pattern, and no more, so that a bad month does not take a customer with it. Eight weeks is long enough to see whether the match rate holds and short enough that you still remember what the old arrangement cost. Keep the rest of your volume on the existing arrangement so you have a control group, which is the only way to know whether the change did anything.

Set the thresholds before the first box moves, not after the first month disappoints you. A match counts only if the pickup is inside 25 miles and the load happens inside four days. The match rate has to clear 60 percent on those terms or the pilot gets re-scoped rather than extended. Empty miles per box should come down at least 12 percent against the June baseline and hours per empty move at least 10 percent. And no export booking may miss its cut-off because of a handover, which is the one number that matters more than all the savings, because a missed export sailing costs more than a year of empty miles.

Have the conversation with your drayage provider early, because their incentives are not automatically yours. If this works, they lose miles, and miles are what they bill for. Some will quietly prefer the old arrangement. Ask them directly what they would need from it: guaranteed volume elsewhere, a share of the saving, or simply a lane they would rather have. A drayage provider who thinks the programme is being done to them will find a hundred small reasons it does not work, and every one of those reasons will look like an operational problem rather than an attitude.

The warehouse end needs the same attention. Somebody has to load the box, and loading a box that arrived this morning is not the same as loading one that was ordered from a pool three days ago. Dock scheduling has to change, and the receiving team has to know that a matched box may turn up with two hours of notice. Ask the warehouse manager what notice they actually need, write it down, and put it into the match rule. A pilot that works on the railroad's screen and fails on a loading dock has not worked.

What I would not do is put the most time-sensitive export customer on this in month one. Pick an export lane where a two-day slip is survivable, learn where the handover actually breaks, and then move the important cargo across once the pattern is boring. Every new arrangement looks fine for the first fortnight, because everyone is paying attention. The fourth week is when you find out what it really is.

If it does work, it changes more than a cost line. An import box that can be absorbed inland is worth more than one that has to be walked back to the coast, and that changes how you compare coasts, how you compare ramps, and eventually how you bid for inland business. It also makes the border routes more interesting: a box that comes up from Laredo and gets emptied in Dallas sits in a market with a lot of export appetite around it, and that is worth knowing before you decide where the next lane goes. I am not going to tell you the programme will work. I am telling you that the question it asks is the right one, and that the answer is measurable from a desk in Chicago this month.

Road problems are solved by looking at the road. For twenty years the empty run has been treated as a cost of doing business, the way a flat tyre is treated as a cost of driving, and it is not. It is a route with nothing on it, driven by somebody who is paid by the mile, and it has survived because no single party could see both ends of it. Now one of them can. Go and measure it against your own numbers before you decide what it is worth, and if the number is real, take it.

↑ Back to the key points

— By Ömer Kaya