LA–Long Beach Rail Backlog Hits 35,688 Boxes; 10,562 Wait 9+ Days

The Analysis

Put the map on the table first. On October 26 the San Pedro Bay complex was holding 35,688 import containers waiting for on-dock rail service, and 10,562 of them had already been waiting nine days or longer. The two numbers belong together, because the first one describes a yard and the second one describes your invoice.

A third number tells you where the pain sits. Long-waiting boxes made up 41.4% of Los Angeles' rail inventory and 15.1% of Long Beach's. Same harbor, same railroads, same week, and a gap of nearly three times. September's average rail dwell came in at 7.19 days against 6.75 the month before. Half a day on the average is noise; what matters is that 10,562 boxes out of 35,688, which is 29.6% of the queue, have crossed the nine-day line, and that share is what decides whether your inland delivery date holds.

Then there is the date hanging over all of it. November 1 is when roughly 250 watchmen at eight terminals may walk out, and those eight terminals move about 85% of Long Beach cargo and about 33% of Los Angeles cargo. The labor risk sits on the side of the harbor with the short rail queue. The rail risk sits on the side with the smaller labor exposure. You cannot choose a terminal that avoids both, and that mismatch is the part of this story nobody is saying out loud.

Start with what the terminals are actually counting, because the phrase awaiting rail service is doing a lot of quiet work here. These are not boxes mislaid in a corner of the yard. They have been discharged off the vessel, they have passed the gate, the customs side has been touched at least once, and the booking says this box is meant to leave on steel rather than on a chassis. That instruction is what moves a container into a different queue, and it is that queue which is now 35,688 deep. A box waiting for a truck and a box waiting for a train sit in different systems, run on different clocks and are owned by different people, and only one of the two is in the news.

Why now, and why this size. Three pushes are arriving together, and any two of them on their own would be survivable. Importers have been pulling volume forward ahead of November 1, which fills the yards with boxes whose owners wanted them inland last week. The railroads have spent years running fewer, longer, tighter trains, a model that is superb when nothing disturbs it and thin when something does. The terminals have spent the same years consolidating into fewer and larger facilities, which is the same trade made out of concrete instead of steel: excellent throughput in the design case, very little slack in any other case. Then a strike authorization lands on top of all that and the slack is gone.

The 41.4% against 15.1% split is where the structure shows through, and it is not a report card on either port. Read it as a statement about where a box waits and in whose hands it is waiting. A high share of long waits means boxes have been handed into the rail system and are queued for a train that has not been built, or built and not pulled, or pulled and not yet put on the main line. A low share means the boxes are still close enough to the gate that a truck can simply take them away. Los Angeles is carrying the first condition this week and Long Beach the second. The distance between them is not effort. It is the shape of the track, the number of facilities feeding it, and how much of each port's volume is committed to intermodal in the first place.

Put November 1 next to that and the picture turns awkward. The eight terminals where watchmen could walk out handle about 85% of Long Beach cargo and about 33% of Los Angeles cargo. Labor exposure is concentrated on the side of the harbor whose rail queue is short. Rail exposure is concentrated on the side whose labor exposure is smaller. If you are moving freight through this complex you cannot pick a berth that escapes both, and the choice was made for you when the carrier assigned your vessel, weeks before either of these numbers existed. That is the sentence I would underline in any write-up of this week.

Who pays, and how much. The dividing line is not your commodity and it is not your customer, it is the last leg of your inland move. If your boxes are drayed to a building within an hour of the harbor, this backlog reaches you only through chassis supply and gate appointments, and both of those are irritating rather than expensive. If your boxes ride intermodal to Chicago, Dallas, Kansas City, Memphis or Columbus, then the whole 7.19-day dwell and the whole nine-day tail land on your delivery promise, and everything hanging off that promise moves with it: replenishment cycles, promotion calendars, retail compliance charges, and emergency air freight when a shelf goes empty. One harbor, two entirely different problems, separated by a single field in your booking.

The timing is worse than the level, and this is where importers habitually mis-plan. Half a day on the average reads like nothing. But 29.6% of the queue is sitting past nine days, and the tail is what shows up at your distribution center, not the average. A box that goes on a train today still needs the line-haul run after the yard, and a box that sat nine days before it moved arrives roughly a week and a half later than the transit assumption you quoted. Rebuild safety stock against the average and you will be short by about a week in the second half of November, and you will hear about it from a customer rather than from a report.

Here is the part the headline does not contain, and it is where the money actually goes. A container in the on-dock rail queue is sitting in a gap between two contracts. The ocean carrier's promise ended when the box was discharged and made available. The railroad's promise begins when it accepts the box onto a train. For the days in between, nobody has promised you anything at all. Your equipment free time and your terminal free time keep running anyway, because both clocks start at discharge and neither one pauses for a missing train. You cannot claim delay against the carrier, you cannot claim it against the railroad, and you cannot claim it against the terminal. Seven days of waiting with no promise attached to it is seven days of demurrage and per diem on your account with nobody to send the invoice to. That is the expensive part of this story, and the figure 35,688 does not contain it.

The only instrument that reaches into that gap directly is free time, and the only moment you can get more of it is before the box ships. Four or five days is a common allowance on this trade and it is nowhere near enough when the rail queue is running past nine. Ask for fourteen days on the equipment and fourteen on terminal storage, in writing, at booking, on the specific lane you use, and read a refusal as information about how much of this risk the carrier plans to leave on your side of the table. Asking after the box has landed is a courtesy request, and courtesy requests get declined. Asking at renewal is a term, and terms hold.

The question that sits next to that one, and that almost nobody asks, is whether you can influence which terminal you get. Sometimes you can. Some service contracts name a terminal or at least a terminal operator, and some carriers will honour a written preference for a fee or against a volume commitment. Most importers never raise it, because until a week like this one nobody thinks of the discharge terminal as a variable at all. It is a variable, and on the numbers above it carries roughly a three-times difference in your odds of landing in the nine-day bucket. If you move enough volume to be worth a phone call, make it before the next contract cycle rather than during the next backlog.

Run the arithmetic on your own book, with the assumptions out on the table where everybody can see them. Assume you ship 30 containers a month by intermodal from San Pedro Bay to an inland ramp. Assume the harbor's own proportions hold for you, so about 30% of them, nine boxes, land in the nine-day-plus tail. Assume those boxes wait five days past your free time, which is what nine days looks like measured against a four-day allowance. Assume detention and demurrage at 195 dollars per box per day, terminal storage and chassis at 45 dollars per box per day, and cargo worth 60,000 dollars a box carrying 20% a year in financing, insurance and obsolescence, which works out to 32.88 dollars a box per day. Nine boxes times five days times 272.88 dollars is 12,280 dollars a month, and spread across your 30 boxes that is about 409 dollars a box of pure waiting.

Set that against the alternative and the comparison hands you a rule you can actually use. Suppose pulling a box to a harbor-area facility, breaking it down and trucking the freight to the same inland point costs 900 dollars more than the rail move you booked. The clock you are buying off costs 272.88 dollars a day. The break-even is 900 divided by 272.88, which is about 3.3 days. Any box you expect to sit longer than three and a half days should go over the road, and any box you expect to move inside three days should stay on steel. That one line converts a statistic about 35,688 containers into a decision somebody can make on Monday morning, and that conversion is the only good reason to read a statistic this size.

One piece of floor discipline makes the rule usable. Ask your broker or your drayage partner for a daily extract of every box you have in the queue, showing discharge date, last status change and free-time expiry, and reconcile it against your own calendar once a day while this is running. Two things come out of it. You catch the boxes that are about to cross free time, which is most of the point. And you catch the boxes that are not waiting for a train at all, the ones stuck on a rail billing that was never accepted, an embargo at the destination ramp, or a hold that nobody escalated. In a queue this size some share of the boxes are always stuck for a reason that has nothing to do with capacity, and those are by far the cheapest ones to fix.

One detail belongs in your file before you let anyone quote you a recovery date. The nine-day bucket is an observation, not a billing status. Your terminal and your railroad will each tell you something slightly different about the moment a box becomes on rail, and the two systems reconcile on their own schedule rather than on yours, which means the day you stop paying is not necessarily the day the queue shrinks. Ask both sides, in writing, which event ends your clock: discharge, rail acceptance, or train departure. Importers who skip this question end up arguing about a week of charges after the fact, and they lose those arguments because nobody wrote the trigger down. Ten minutes of clarification before the box lands is worth more than a week of correspondence afterwards.

Talk to your customers on a specific date rather than in general terms, and do it before they come to you. If your inland delivery quote rests on a transit assumption built in September, the 29.6% tail is now sitting inside that assumption and your quoted dates are wrong by roughly a week for about one box in three. Pick the accounts where a missed date triggers a chargeback or a compliance penalty, give them the revised inland date this week with the number attached, and offer the transload option at cost where the account is worth protecting. A date that moves early gets forgiven. A date that is simply missed does not, and that difference is worth more than most of the freight savings in this piece.

Now try to break the conclusion, because a conclusion you cannot break is not worth much. If the November 1 walkout does not happen and import volume rolls over after the peak, the queue drains over two to three weeks and the right behavior is to wait, pay the detention and not buy transload capacity you will never need. If the walkout happens and runs longer than a week, the eight terminals carrying 85% of Long Beach cargo are the ones that stop, local drayage replaces rail as the binding constraint, and the right behavior flips completely: pull what you can out of Long Beach before the first shift and stage it off-dock.

If your freight is local dray to begin with, this backlog barely touches you and your time is better spent on chassis and gate appointments. The variable that flips the answer is not the size of the queue. It is which side of the harbor your carrier chose, and whether your box leaves on wheels or on steel.

The escape routes deserve the same honesty you would give a vendor. Redirecting discharge to the Pacific Northwest, Houston or Savannah does not buy you an empty ramp, it buys you a different queue, and it costs five to eight days of additional ocean transit to find out how long that queue is. For promotion-tied or high-value freight that trade is usually wrong, and for routine replenishment it is sometimes right. Transloading near the harbor buys you control and costs you the seal: you take on a cargo handling claim, you need a properly documented or bonded facility, you inherit whatever wood packaging and labeling problems the sealed box was hiding, and you now need truck capacity in a market that is tight in late October too.

Carrying more inventory is the cleanest alternative and the most expensive in cash, since one extra week of safety stock at 30 boxes a month and 60,000 dollars a box ties up roughly 415,000 dollars, against which a monthly waiting bill of 12,280 dollars looks like a bargain. Most importers will end up waiting. Knowing that in advance is better than discovering it after paying for a restructuring that saves less than it costs.

Zoom out once, because that is the only way this stops being an October problem. Ten years ago the question of where an Asian import landed had something close to an automatic answer, and the automatic answer was this harbor, because the intermodal machine behind it was the densest thing anybody had built. What has changed since is not the capacity of San Pedro Bay. It is the number of other doors. East coast and Gulf coast terminals have taken share, inland ramps have been built out, and importers who got burned in the congestion of 2021 and 2022 now keep a second and a third routing alive as policy rather than as an emergency measure.

This harbor was a default and is now one option among several, and an importer whose network has only one door is carrying a risk that has nothing to do with 35,688 boxes. The queue is a pulse. The redistribution of entrances is the turn, and the turn is what should shape the contracts you sign next year.

So end where the money is. The number worth watching this week is not 35,688. It is your own count of boxes that have crossed free time, with the crossing date written next to each one. Build that list by Friday, price every box against the 3.3-day rule, move the ones that fail it, and put the rest on a daily watch. Keep the same list through November, because the strike date will come and go, the queue will drain and refill, and the only thing that consistently separates importers who absorb this from importers who pay for it is whether they knew which boxes were bleeding on any given morning. Everything else in this story is weather, and weather passes.

↑ Back to the key points

— By Tomás Ferreira