Vincent Thomas Bridge Shuts 7 November for 16 Months on the LA Drayage Route

The Analysis

San Pedro, Terminal Island, Front Street, Harry Bridges Boulevard. Four names in one closure notice, and the only one that matters is the bridge they are all standing around. Caltrans will fully close the Vincent Thomas Bridge from 7 November 2026 for about sixteen months to replace the deck, on a $750m project with reopening targeted for early March 2028. The bridge links San Pedro with Terminal Island and carries roughly 53,000 vehicles a day.

Put the two truck numbers next to each other and they say something the fact sheet does not say out loud. About 8.8% of the traffic is heavy trucks, and around 4,700 of the daily truck moves are port-related. Eight point eight per cent of 53,000 is roughly 4,660. That is the same population. For heavy vehicles this is not a municipal crossing that happens to carry some port traffic; it is port infrastructure with a nice view. Those 4,700 moves now go via Front Street and Harry Bridges Boulevard.

And the yard was already tightening before anyone closed anything. Truck dwell at the port ran 3.18 days in September against 2.95 in August. When a crossing shuts, my first question is never the rate. It is which node jams, what it costs to go around that node, and whether the road is still worth running at all. I have spent my working life moving boxes over borders and across railways that change gauge in the middle of the night, so I read this like a map and not like a press release. Every corridor has a story, and a pothole. This pothole is sixteen months deep.

The facts first, and I will keep them exactly as they were published. Caltrans will fully close the Vincent Thomas Bridge from 7 November 2026 for about sixteen months to replace the deck. The project carries a $750m price tag and a reopening target of early March 2028. The bridge connects San Pedro with Terminal Island and carries roughly 53,000 vehicles a day, about 8.8% of it heavy trucks, and around 4,700 of the daily truck moves are port-related. During the closure those moves are routed via Front Street and Harry Bridges Boulevard. Before any of this started, truck dwell at the port was already going the wrong way: 3.18 days in September against 2.95 in August. That is the announcement plus the surrounding data. Everything below is the part a fact sheet does not print.

Why now, and why all the way shut rather than in pieces. A deck replacement is not maintenance you can do in night windows across three years. It is the kind of job that decides whether a structure keeps carrying container loads at all, and doing it in slices costs more calendar time than closing it once. Caltrans has chosen the sixteen-month full closure over years of narrowed lanes, and the arithmetic behind that choice is the part worth reading. Seven November 2026 plus sixteen months lands in early March 2028. The published reopening date is not a target with slack behind it; it is the same arithmetic printed as a date. So the reopening date is the schedule itself, and a slip in the deck work moves the reopening one day for one day. Treat early March 2028 as the best case that survives only if nothing goes wrong, not as a date you can build inventory around.

Put that on a calendar next to the port's own rhythm and you get the shape of the next two years. The closure starts on 7 November, late enough that the year-end import push is already landing and early enough that nobody has finished building peak inventory. Then come the months that decide it. The first six to eight weeks after 7 November will be the messy part: signal timings on the detour that were tuned for a different traffic mix, drivers running a route they have never run, and gate systems built on the assumption that a truck shows up inside its window.

Call the new pattern settled around January 2027. After that you get a steady state that lasts until the next disruption, and there are two of them inside this window: the late-summer 2027 peak build, and the pre-Lunar New Year push in early 2028, which lands about six weeks before the reopening target. The bad case is not the closure itself. The bad case is a two-month slip that puts reopening into the 2028 spring restock, with the same 4,700 daily moves still on surface streets.

Who eats it, and how much. Begin with the drayage carriers and the owner-operators, because they feel it first and they are the ones who decide whether capacity stays. An owner-operator is paid by the move, not by the hour. If a detour costs twenty-five minutes a turn and a driver was running two and a half turns a day, that is roughly an hour of paid work gone, and an hour at San Pedro rates is most of a day's margin.

Drivers do not absorb that, they leave for other work, and they leave faster than replacements arrive. Next are the terminals on Terminal Island. Their gates were sized for trucks arriving on schedule and their yards were already holding 3.18 days of dwell in September against 2.95 in August. A yard already that tight does not absorb late trucks; it converts them into dwell. Then come the importers, and here the split is geography rather than size. If your boxes land on Terminal Island and your DC is in the Inland Empire, the bridge sits on your critical path every single day.

If your boxes land on the Long Beach side or leave by on-dock rail, it does not. Then there is the group nobody writes about: exporters and empties. When truck capacity tightens, import pulls pay better than export pulls and empty repositions pay worse than both, so export bookings and empty returns are the first things a carrier quietly stops chasing.

Here is the part I have not seen in any coverage, and it is the one that decides the size of the bill. Everybody is counting miles. A detour of four or five miles is not what hurts. What hurts is that drayage capacity in San Pedro Bay is not measured in trucks, it is measured in turns per truck per day, and a surface-street detour attacks turns twice: once through distance, and again through variance. Front Street and Harry Bridges Boulevard are city streets with signals, intersections, school runs and driveways. A freeway ramp gives you a predictable four minutes. Four miles of signalised street gives you a mean of maybe twelve minutes and a spread that runs to forty on a bad afternoon.

Terminal gates, appointment windows and chassis pools are all built around the mean. None of them are built around the spread. Every extra minute of spread has to be paid for somewhere, and the place it gets paid is the yard, because a truck that arrives thirty minutes late does not arrive thirty minutes late once. It pushes the next move back, and the next one, and by Friday that truck has run one fewer turn this week and the box it should have pulled is still sitting in the yard adding to dwell.

Do that arithmetic with the published numbers and with assumptions I will state plainly, so you can argue with them rather than take them. Assume the 4,700 port-related truck moves a day are produced by trucks running 2.5 turns a day; that is my assumption, not a published figure, and if your carrier gives you a different number, redo this with theirs. Four thousand seven hundred moves at two and a half turns means about 1,880 trucks working the bridge in a day. Assume the detour adds 25 minutes per turn; also my assumption, and the honest range is 15 to 40 minutes depending on the hour.

Four thousand seven hundred moves times 25 minutes is 117,500 minutes a day, which is about 1,958 hours. Assume a truck has ten productive hours in it. That is roughly 196 truck-days of capacity gone every day, about 10% of the working fleet. Buying that back does not require a better route. It requires 196 more trucks with drivers, in a market short of both for years, or you accept roughly 490 fewer moves a day, which is about 10% of daily port drayage throughput. Nobody puts that number in a fact sheet.

Now connect it to the one trend figure in the story. Dwell moved from 2.95 days in August to 3.18 days in September. That is 0.23 of a day, about five and a half hours, close to 8%. It happened before anyone closed anything, so it is the baseline the closure lands on, not a symptom of the closure. Dwell matters here because it is the variable that turns lost truck capacity into money. A box sitting 3.18 days is occupying a slot in a yard short of slots, and it is eating into free time.

If the closure pushes dwell from 3.18 to, say, 3.6 days, the boxes that sat comfortably inside free time start arriving at the edge of it and the ones already at the edge go over. There is a second-order effect people miss as well: higher dwell means a fuller yard, a fuller yard means longer internal moves and more rehandles, more rehandles mean slower gates, and slower gates mean fewer turns. It is a loop that feeds itself, and the closure is the nudge that starts it.

Look at the map before you accept the published detour as the whole story. Terminal Island does not have one way in. There is the Long Beach-side crossing, the Gerald Desmond Bridge, and there is a lift bridge on the Wilmington side. The fact that Caltrans is publishing Front Street and Harry Bridges Boulevard tells you something on its own: the alternatives are constrained, whether by weight, by geometry, by lift openings or by what the streets around them can carry. Do not take my map for it, take the port's traffic management plan and Caltrans's, and ask your drayage carrier which crossing they actually intend to use and at what hours.

That last question is worth asking twice. A residential detour through San Pedro generates complaints about noise and dust inside a few weeks, and the answer to complaints is usually a restriction on night operation. Night and weekend gates are the obvious release valve for a squeeze like this. If that valve is partly shut by community pressure, the squeeze is worse than the arithmetic above suggests.

There is a node inside this that almost never appears in a bridge story, and it is the chassis pool. Drayage in San Pedro Bay runs on shared chassis, and chassis supply obeys the same arithmetic as trucks: how many boxes one chassis can serve in a week depends on how fast it comes back. Take ten per cent of daily turn capacity out of the system and the chassis fleet effectively shrinks in proportion, because boxes sit longer on wheels. In practice it surfaces at the worst moment, a Friday afternoon before a long weekend, and it surfaces first at the off-dock yards rather than on terminal, because the pool protects the terminals and squeezes everybody else. If your carrier runs its own chassis this is a smaller problem for you. If you rely on the pool, ask what the recovery plan is, and ask it in October rather than in January.

Where I could be wrong, and what would flip it. Take the opposite view seriously for a minute. Fifty-three thousand vehicles a day includes a great deal of traffic that has nothing to do with the port, and closing a through route can leave the remaining network less congested than before rather than more. If import volumes into San Pedro Bay soften through 2027, if the terminals improve appointment compliance on their own, or if a real share of boxes leaves by on-dock rail instead of by truck, then twenty-five minutes a turn is absorbed inside the slack the system already carries and this becomes a story about a bridge rather than a story about supply chains.

There is also a version where Caltrans gets signal priority and a dedicated truck lane onto the detour, and the per-turn cost falls to ten minutes, which halves the capacity loss. The conditions I would watch are concrete. Dwell below about 2.5 days means the yard can absorb it. A drayage fleet that is not shrinking through the winter means the turns are still economic. And a detour travel time published by the port week by week means someone is managing it rather than announcing it. If none of those three are true by the end of January 2027, plan for the full ten per cent for the whole sixteen months.

Run a number on your own book, with every assumption written down so you can argue with them. Assume you import 40 containers a month through San Pedro Bay and all of them truck over the bridge. Assume your drayage carrier passes on a detour surcharge of $75 a container; nobody has published a rate, and my guess is it lands between $40 and $150 depending on where your box sits and how tight the lane is. Forty containers at $75 is $3,000 a month, or $36,000 across sixteen months.

Assume the closure also adds 0.4 days to dwell, taking you from 3.18 to about 3.58 days, and assume free time is four days with demurrage at $200 a container a day after that; both are assumptions, check both against your own terminal tariff. If a quarter of your boxes now spend one day past free time, that is 10 containers at $200, or $2,000 a month. Add a third line most people forget, which is safety stock. If door-to-DC transit becomes less predictable and the spread widens by half a day, you are carrying roughly an extra day of pipeline inventory.

At a landed value of $30,000 a container and 8% annual carrying cost, one extra day across 40 containers is about $263 a month. So the visible bill is the surcharge and the invisible bill is demurrage plus inventory, and the invisible one is the one nobody invoices you for.

What I would actually do, in the order it has to start. Before 7 November, pull forward everything you can, and get containers that have been sitting past two days out of the terminals while the bridge is still open, because every box you clear now is a box not competing for scarce truck capacity later. Get your drayage carrier to put the detour in writing: the route, the added minutes a turn, the surcharge, and what happens if the detour gets slower. Ask the terminals you use what appointment availability looks like from November and whether they are adding night or weekend windows, and if they are, book into them now rather than after the queue forms.

Size your exposure with the turn arithmetic above rather than with a per-container surcharge, because the surcharge is the part you can negotiate and the lost turns are the part you cannot. And set a trigger: if dwell at your terminals passes 3.5 days, or carrier on-time gate performance drops more than ten points, shift to the alternative rather than waiting for it to settle.

The alternatives, and the trap inside each one. Moving volume to another port is the obvious answer and it is usually mispriced. Oakland, Houston, the Pacific North-west and the East Coast via Panama all take West Coast cargo, but you are not swapping a bridge for a bridge; you are swapping drayage for inland rail or long-haul trucking, and the inland leg is most of the cost. Compare total landed cost to your DC door, container by container, with a realistic rail transit and a realistic drayage cost at the far end, and include the two or three days of variability you take on.

How the ocean rate itself behaves from here is Vivian's beat rather than mine, but do not price a port switch without it. Moving boxes out by on-dock rail avoids the bridge entirely and is the cleanest fix if your box is already on Terminal Island and your destination is inland; the trap is that on-dock rail is tight in peak and you will be competing with everyone who has just had the same idea. Building more inventory is the third option and the one people reach for first; the trap is that you are buying certainty with working capital at the exact moment the closure is also raising your dwell and your demurrage.

And one trap that catches everybody: do not assume the day the bridge reopens is the day things go back to normal. A sixteen-month-old detour leaves habits behind, some drivers will not come back, and a reopened bridge arrives with its own punch-list closures. Give yourself four to six weeks of transition either side of March 2028.

There is a timing point here that has nothing to do with traffic. Most 2027 drayage and inland contracts get negotiated in the first half of the year, which means the closure will be six to eight months old when you sit down to talk money, and by then your carrier will have real data on what the detour costs them. That is a reasonable basis for a surcharge and an unreasonable basis for a blank cheque. What I would put into the contract is not a fixed number but a mechanism: a published detour time, a quarterly review against it, and a surcharge that steps down when the published time improves. And if you can get it, take a commitment on turns rather than on moves, because a carrier who promises you moves can meet that promise by running trucks longer, while a carrier who promises turns has to actually solve the route.

Does all of this mean you should leave San Pedro Bay? For most importers, no, and I would rather say that plainly than build a case for drama. Sixteen months of a constrained crossing is a real cost, but it is a cost on one leg of one route, and ports are sticky for good reasons: your forwarder's relationships, your drayage rates, your DC geography, your rail options, and the plain fact that every other port has its own constraint you have not priced yet. There are three cases where leaving is genuinely worth costing out. One case is where your volumes are small enough that you have no drayage leverage at San Pedro and you are the last call a carrier makes when capacity is tight.

Another is where your DC is inland and east of the Rockies, in which case you were already paying for a long inland leg and a different port changes the arithmetic more than it changes the risk. The third is where your cargo is time-critical enough that a ten per cent hit to daily turn capacity breaks a commitment, a retail reset with a hard date, or a promotional build that cannot slide. In all three, cost the move as a two-year decision rather than as a way to dodge sixteen months.

↑ Back to the key points

— By Ömer Kaya