Global port congestion hits 4-year high with 3.7M TEU idling

The Analysis

Pull the global berth queue up on a screen and the picture is hard to ignore: nearly 3.7 million TEU - about 11 percent of the entire container fleet - sitting idle waiting for a berth, the worst the industry has seen in four years. North Asia is the tightest region, and Kuehne+Nagel's own figures put Manila at roughly 4.7 days of waiting, Beira at 15.44 days, and Tema at about 7.75 days, while schedule reliability slipped to 37 percent in June with average delays around 3.6 days. If your containers move on the water right now, this is not a headline to bookmark - it is a planning problem sitting on your desk, and the shippers who treat it as one will be the ones still hitting their arrival dates when the queue lengthens.

The queue is the network, not a local bump

When 11 percent of the world's boxes are parked instead of moving, that is not one bad port having a rough month. It is the whole circulatory system of container shipping running slower than the cargo wants to flow. I have spent enough years watching port networks to know the difference between a regional hiccup and a structural clog, and this reads as the latter. Berth waits are the visible tip; underneath sits a tangle of late vessels, missed connections, and equipment stuck in the wrong place. A box waiting four days at Manila is not just losing four days. It is throwing off every schedule that assumed it would be somewhere else by now.

North Asia being the most congested region is the part that should sharpen your pencil if you source or sell into that corner of the world. The factories and the distribution centers there run on tight, predictable timelines, and a berth queue that lengthens by the week shreds the buffer that those timelines depend on. What looks like a far-away delay becomes your missed promotion, your empty shelf, your angry customer.

Why berth waits ripple further than they look

A vessel delayed at anchor does not simply arrive later. It arrives later into a system that had already sold that berth, that crane, and that train slot to the next ship. The delay pushes into the following week's sailings, and those push into the week after. Schedule reliability at 37 percent means fewer than four in ten vessels are keeping to plan - so the odds are better than even that the ship you booked is not where its published schedule says it is. Average delays of 3.6 days sound survivable until you stack them across a lane where you need three or four sequential calls to behave.

This is where the naive planner gets burned. They add a flat three days to every estimate and call it a buffer. But the variance is the enemy, not the average. A lane where delays swing from zero to fifteen days - look at Beira at 15.44 days - cannot be managed with a single number. You have to plan for the swing, not the mean.

Build the buffer around the swing, not the average

Take the figures at face value and plan against the worst credible case on each lane, not the published average. For North Asia inbound or outbound, assume a berth wait in the high-single digits to low-teens of days during the peak of this congestion, then work backward from your customer's hard date. If a promotion or a production line goes live on a fixed day, the container needs to be in place well before the queue is likely to clear, not after. The shippers who kept their promises through 2021 did exactly this: they treated the delay as a fact of the route, booked earlier, and accepted that some inventory would sit a little longer on the safe side.

A practical rule I give clients: separate your "must-arrive" cargo from your "nice-to-arrive" cargo and plan them with different buffers. The must-arrive box gets booked on the earliest credible sailing plus a swing buffer you can defend. The nice-to-arrive box can ride the cheaper, later option. Mixing the two onto one optimistic schedule is how a whole quarter goes wrong.

Reroute and transshipment as live tools, not last resorts

Rerouting used to be something you did after a crisis. In a 3.7 million TEU queue, it is a normal planning lever. If your habitual port is jammed, look at the neighboring gateways on the same coast that are clearing faster, or consider a transshipment hub where a feeder can pick up the box and slip it into a less crowded mainline call. Transshipment adds a handoff and a little time, but a clean hub with open berths can beat a direct call into a port where the anchorages are full.

The catch is that transshipment only helps if the hub itself is not congested. Check the hub's own berth wait before you commit, because a crowded transshipment point just moves your problem offshore. A good forwarder keeps a short list of hubs with spare capacity and will tell you honestly which ones are breathing. Use that list; do not discover it mid-crisis.

Booking and equipment strategy when slots are scarce

When reliability is 37 percent, the booking you hold is worth more than the rate you negotiated. Protect it. Book the earliest sailings you can defend, confirm the equipment early, and never assume a chassis or an empty will appear on the day you need it - congestion strands equipment exactly when everyone reaches for it at once. If your lane runs through a congested region, pre-position empties at the load point so the box is there when the cargo is ready, rather than waiting for a repositioning that the queue has already broken.

Build a relationship with a forwarder who can see multiple carriers' open slots, not just one. In a tight market the difference between making a sailing and missing it is often knowing that a competitor line has a berth open on a day your preferred line does not. Visibility across carriers is the single most useful thing you can buy when the network is this slow.

Contract and risk clauses worth rewriting now

Most standard shipping contracts were written for a world where vessels arrived. In a 37-percent-reliability world they need a refresh. Push for clauses that account for berth-delay risk: a demurrage and detention clock that does not start ticking the moment the box is "available" on paper while the terminal is still jammed; a force-majeure or exceptional-circumstances definition that actually covers port congestion rather than only acts of god; and a service-level commitment measured on arrival at your door, not on the vessel's published schedule.

Also reconsider your inventory-ownership terms. If you are buying on terms that pass risk to you the moment goods are loaded, a six-week berth queue is your six-week problem. Where you have room to negotiate, shift the risk window so it does not open until the cargo is actually moving through a clear port. Small clause changes now are cheaper than a quarter of stockouts later.

A planning example, assumptions stated plainly

Assume an importer brings in about 200 FEU per month of goods on a North Asia lane, with a fixed retail launch date each season. Under normal conditions the cargo needs roughly 30 days of pipeline. With North Asia the tightest region and average delays near 3.6 days that swing much higher at the worst ports, plan the must-arrive share - say 60 percent of the volume - on a 45-day pipeline booked on the earliest credible sailings, and keep the remaining 40 percent on a flexible, cheaper lane. The explicit assumption is that the berth wait on the primary port behaves like the high-single-digit days seen across the region, not the 4.7 days of a single port. If the queue worsens, the 45-day buffer absorbs the swing; if it eases, the importer simply carries a little extra safety stock on the cheap side.

For an exporter on the Africa side, where Beira shows 15.44 days of waiting, the math is starker. A box that must catch a monthly sailing cannot afford to sit half a month at anchor. Plan the load to arrive at the port at least three weeks before the advertised cut-off, confirm the vessel is actually berthing on schedule, and have a fallback transshipment hub named in writing in case the direct call slips. The numbers are illustrative; the discipline is to size the buffer to the swing you can observe, not the average anyone publishes.

The pitfalls that quietly cost you the quarter

A few mistakes repeat themselves every congestion cycle. Trusting the published schedule as if reliability were still 80 percent is the big one - it turns a known risk into a surprise. Another is booking the cheapest sailing and hoping, which in a slow network means you are last in line when slots tighten. A third is forgetting that your supplier's "shipped" email means nothing until the vessel has berth and sailaway; a container sitting at a crammed terminal is not in transit no matter what the tracking says. And do not let demurrage and detention surprise you - under congestion those clocks are where margins go to die, so read the clause before the box does.

Watch the neighboring ports, not just your own

Congestion travels. A queue at one North Asia hub spills to the next as carriers reroute and as boxes chase open berths. So your planning should track the region, not a single facility. Keep a one-page scorecard of berth waits at the ports on your lanes and the ones adjacent to them, update it weekly, and let it drive which sailing you book. The shippers who sailed through 2021 were the ones reading the scorecard, not the ones reading last month's schedule.

What 3.7 million TEU means in fleet terms

It helps to translate that headline number into something you can feel. Roughly 3.7 million TEU waiting is about 11 percent of the boxes afloat on the water at any given moment. Think of the global fleet as a highway: one in nine trucks is parked on the shoulder. The cargo those boxes carry is not destroyed, but it is frozen in place, and the clock on every downstream commitment keeps running. For a shipper, that frozen share is the reason your "normal" transit quote no longer describes reality - the system simply has fewer boxes in motion per day than its schedule pretends.

The four-year high label matters too. This is not the pandemic-era spike, which came with blank sailings and a demand shock. This is a queue built on steady trade volume meeting ports that have not added berth capacity fast enough. That distinction should shape your planning: a demand-shock backlog can vanish when orders soften, but a capacity-bound queue lingers until concrete gets poured. Plan for the latter.

North Asia feels it first because the factories run lean

North Asia being the most congested region is not random. The manufacturing base there runs on lean inventories and tight handoffs, so a berth wait does not just delay a box - it starves a line that was counting on that box being in place yesterday. A delay that a bulk buyer in another region might absorb gets felt immediately on a production floor in the northeast of the continent. If your supply chain touches that region, assume the congestion lands on you faster and harder than the average figure suggests, and pad accordingly.

The practical move is to pull your replenishment trigger earlier in the cycle rather than waiting for the usual signal. When the queue is this long, the old "reorder when stock hits three weeks" rule leaves you exposed, because three weeks of pipeline can evaporate in a single bad berth wait. Move the trigger to five or six weeks and you regain the cushion the network took from you.

Stage inventory upstream, not only at your door

Most shippers keep their buffer at the destination - safety stock in the DC. In a slow network, that single cushion is fragile, because every box feeding it faces the same queue. A stronger posture is to stage buffer at two points: some safe stock near the source or at a regional transshipment hub, and some at the door. The upstream stock lets you ride out a berth wait without a shelf going empty, because you are not dependent on the next sailing alone. This costs a little more to hold, but it turns a six-week berth queue from an emergency into a scheduling note.

The trade-off is real and worth stating plainly: more positioned inventory means more tied-up capital. Size it against the cost of a stockout on your must-arrive lines, not against a generic efficiency target. The lines where a miss burns a customer relationship deserve the upstream buffer; the lines where a delay is merely annoying do not.

Tell your customer the truth about the date

When the queue lengthens, the instinct is to promise the old date and hope. Resist it. A customer told early that arrival will slip by a week, with a reason they can repeat to their own boss, is a customer who plans around you. A customer told at the last minute that the box is still at anchor is a customer shopping for a new supplier. The congestion is industry-wide and visible; using it as an excuse reads as incompetence only when you spring it late.

Put the revised window in writing as soon as you see the berth wait building on your lane, and update it weekly. Make the number conservative so you beat it more often than you miss it. In a 37-percent-reliability market, a shipper who is consistently early against a honest date earns more trust than one who is occasionally on time against a fantasy.

The equipment trap inside a slow network

Congestion does not only strand vessels; it strands the boxes and chassis those vessels would have carried onward. A terminal jammed with imports has no room to turn boxes around, so empties pile up where they are not needed and vanish where they are. On a congested lane, never assume the equipment will be where your cargo is. Pre-position, confirm in writing, and treat the empty as a booked resource with its own lead time, not as something that appears by magic on load day.

The same trap bites on chassis at the destination. When boxes finally clear a jammed terminal, everyone reaches for a chassis at once, and the pool runs dry. If your destination yard is tight, arrange chassis provision separately and early, because the last shipper to the yard is the one hand-loading onto a truck bed.

Read the weekly queue like a weather report

Treat the berth-wait figures as a forecast you check every week, not a headline you read once. The numbers move - Manila at 4.7 days this reading may be 6 next, Beira at 15.44 may ease, a quiet hub may suddenly clog as carriers divert. The shippers who sailed through 2021 were the ones who watched the report and rebooked ahead of the storm, not the ones who trusted a schedule printed months earlier. Make the Friday queue check a standing item, and let it drive next week's bookings.

Size your response to your own exposure

Not every shipper needs the same buffer. A company moving a hundred boxes a year on one lane should not build the same war room as one moving ten thousand across a dozen ports. Size the response to where the money actually is: your must-arrive, high-value, customer-facing lines deserve the full treatment - the swing buffer, the transshipment fallback, the rewritten clauses. Your low-value, flexible lines can ride the cheap option and absorb the odd delay. The mistake is spreading the effort evenly, which either over-spends on cargo that does not matter or under-protects the cargo that does.

The data you should be demanding from your forwarder

In a slow network, visibility is not a nice-to-have; it is the tool. Ask your forwarder for a weekly read on berth waits at every port on your lanes, for open-slot visibility across at least three carriers, and for a written transit commitment that names the gateway and the rail ramp rather than a vague "West Coast." If your forwarder cannot produce those, that is itself a finding - you are flying blind through the exact fog where others are navigating by instrument. The shippers who come through this queue with their dates intact are the ones whose forwarder treated the delay as data, not as an excuse.

Do not let the queue become your excuse

Inside your own building, the congestion can become a convenient blanket explanation for every late delivery, and that is its own danger. A delay caused by a 15-day berth wait at Beira is real; a delay caused by your team booking late and then blaming the queue is not. Keep the two separate. Track, for each late box, whether the cause was the network or your own process, and fix the part that is yours. The shippers who weaken during a slow period are rarely the ones the ocean beat - they are the ones who stopped disciplining their own bookings because the ocean made a good scapegoat.

The queue will be blamed for a lot this year. Make sure it is only blamed for the delays it actually caused, and the rest stays on your side of the line where you can still do something about it.

This is a long tide, not a single wave

I would be careful about calling this a one-quarter problem. The fleet is full, the capacity that was added during the boom is now being absorbed by demand that has not relented, and the berths that would relieve the pressure take years to build - the very kind of investment Coos Bay's grant represents on the U.S. side. So plan for the queue to be a feature of the map for a while, not a blip. The network is not broken; it is simply running hot, and a hot network rewards the shipper who planned the swing instead of the one who trusted the average. The boxes will move. The question is whether yours moves on a date you can name.

↑ Back to the key points

— By Tomás Ferreira