Mid-September port congestion: Kaohsiung 2.4d, Manila 2.7d, Rotterdam 1.5d amid strikes and

Everstream's port congestion report of 16 September 2026 is a map of where your cargo is currently sitting and waiting, and if you move anything through Asia, the Gulf, or Europe, the wait times on this list should already be in your quote. The numbers are not trivia; they are the buffer you must add before you promise a delivery date, and most forwarders are still quoting last month's assumptions against this month's terminals.

The headline numbers from the 16 September 2026 report. Kaohsiung is at 2.4 days of waiting, with a 19-hour suspension on 2 to 3 September from storm Saudel. Manila is 2.7 days. Salalah in Oman is 3.3 days, and its yard density ran up to 73 percent. Colombo is 2.3 days. Mundra came in at 1.38 days after a 13-day empty-yard strike that ended 11 September. Nhava Sheva is 1.29 days. For an importer-exporter these are not trivia; they are the buffer you must add before you promise a delivery date.

Europe on the same report. Valencia is 2.5 days. Le Havre is 2.3 days, with a wind suspension on 9 September. Hamburg is 1.9 days, slowed by construction. Rotterdam is 1.5 days but with yards at critical status. Genoa is 2 days. The pattern is a spread of one-and-a-half to two-and-a-half-day waits across the main EU gateways, which means no single European port is a safe haven and each adds its own flavour of delay to your transit plan.

The Americas. Santos in Brazil is 3.3 days, driven by weather. Vancouver is interesting: berth wait is only 1.0 day, but rail dwell is 3.3 days, so your container can clear the ship fast and then sit on the ground waiting for a train. That distinction matters because most people only watch the berth number and miss the inland clog that follows. If your cargo moves beyond the coast into the Canadian interior, the rail dwell, not the berth, is your real clock.

What a wait day actually costs you. A day of port waiting is not just a day; it is a day of demurrage exposure, a day your inventory is not earning, and a day your customer's promised window shrinks. On a 40-foot box of goods valued at, say, 60,000 dollars, a day of delay at a 6 percent annual carrying cost is about 10 dollars, which sounds small until you multiply across a peak season where hundreds of boxes each lose two or three days. The cost is in the aggregate and in the service-level misses, not in any single container.

Salalah is the one I would watch hardest, because the 73 percent yard density is the warning light. High yard density means the terminal is running out of space to stack boxes, which precedes empty-container shortages and extended reliefs. When a yard hits the seventies, the terminal starts restricting what it accepts and how long you can keep a box there. For cargo routing through the Gulf or transshipping at Salalah, that density reading is your leading signal: before the wait climbs, the yard fills, and you should be rerouting before the fill shows up as a delay on your shipment.

Single-port dependency is the quiet risk the report exposes. If your Europe-bound cargo funnels through Valencia because that is where your carrier calls, a 2.5-day wait there is your only plan. The same goes for leaning entirely on Salalah or Colombo in the Indian Ocean lane. The takeaway from the data is to avoid single-port dependency and to watch yard density for empty-container and relief windows. Build a second gateway into your routing so a wind suspension at Le Havre or a storm at Kaohsiung does not become a missed delivery.

Here is a worked example on the Gulf lane. Suppose you ship from the Indian subcontinent to North Europe and normally transship at Colombo, now at 2.3 days, or Salalah, now at 3.3 days with a 73 percent yard. If Salalah's density keeps climbing, your transshipment wait can jump from 3.3 to five or six days with little notice. Pre-booking an alternative transshipment at, say, a less dense port, or splitting the volume across two hubs, turns a single point of failure into a managed spread. The extra planning cost is trivial next to a week of stranded cargo during a peak.

The buffer is the simplest defence and most people under-do it. Add berth-wait buffer to every shipment touching Kaohsiung, Manila, Salalah, Colombo, or Santos, which sit between 2.3 and 3.3 days. That means quoting the customer a delivery window that already includes the wait, not one you hope to beat. The forwarders who quote the optimistic number and then miss it lose the account; the ones who quote the realistic window with the buffer baked in keep it. The buffer is not pessimism; it is respect for the data in front of you.

Relief windows are the flip side. Ports move in cycles, and a yard that is critical this week may ease next week as vessels clear. Watching yard density and berth wait weekly lets you catch the window when it opens and push volume through the port that just loosened. Rotterdam at 1.5 days but yards critical is a place where the window can close fast, so ship when you can, not when the schedule is convenient. The operator who times the relief window saves the same money the one who ignores it loses.

Europe specifics deserve their own line. Valencia and Le Havre at 2.3 to 2.5 days mean your Spain and France bound cargo needs extra days in the quote, and Le Havre's wind suspension shows weather can add a zero-notice day. Rotterdam's critical yards are a structural constraint, not a blip, so treat its 1.5-day berth wait as fragile. Hamburg's construction slowdown is predictable and plannable, which is the good kind of delay because you can schedule around it. Read each port's reason, because the remedy differs by cause.

For the Americas, separate the two clocks at Vancouver. Berth wait of 1.0 day is excellent, but rail dwell of 3.3 days means your box sits on the terminal after discharge waiting for a train slot. If your destination is the Canadian interior or the US via that corridor, plan the rail dwell as the real transit component, not the berth. Santos at 3.3 days from weather is the opposite profile: the ship side is the clog, and you manage it with earlier booking and weather contingency rather than inland planning.

For an importer-exporter the congestion hits both ends. Your export out of Asia sits at Kaohsiung or Manila; your import into Europe sits at Valencia or Rotterdam. The same report should drive both your outbound quote and your inbound replenishment timing. Most teams run these as separate silos and miss that the congestion signal is one dataset. Pull the report into both the export desk and the procurement desk so the left hand prices the wait the right hand is already absorbing.

Common mistakes. Watching only the berth number and ignoring rail dwell. Treating a 73 percent yard as someone else's problem. Routing everything through one port because the carrier favours it. Quoting the optimistic transit and eating the miss. Not separating weather suspension from structural slowdown when you plan the remedy. Each of these turns a manageable wait into a stockout or a penalty, and none of them require genius to avoid, only the discipline of reading the report before you book.

My read on the 16 September data is that this is a normal-but-lumpy congestion picture, not a systemic collapse, which is exactly why it is manageable if you respect it. The waits are real, the yard densities are the early warning, and the single-port dependency is the avoidable trap. Build the buffer, watch the density, keep a second gateway, and time the relief window. The cargo that moves on a plan built from this report arrives; the cargo booked on last month's assumptions is the cargo that waits.

Make this report a recurring input rather than a one-time read. Congestion is a moving target, and the 16 September snapshot is already history by the time you act on it. Set a weekly pull of the latest port congestion numbers and fold them into your quoting model so the buffer you add today reflects this week's yard, not last month's. The operators who systematise this are the ones whose quotes stay honest through a volatile season while their competitors scramble to explain why the shipment is late. A fifteen-minute weekly habit beats a quarterly surprise every time.

For the Asia export desk, the lesson from Kaohsiung and Manila is direct. Kaohsiung at 2.4 days with a recent storm suspension means your south-China and Taiwan-bound cargo needs the wait baked into the promise, and a repeat weather event will not announce itself. Manila at 2.7 days is a structural queue, not a one-off, so stop quoting it as if it were a one-day port. The practical move is to pre-build the buffer and then communicate it as the real window, so when the storm hits you are within promise instead of behind it. The desk that quotes the optimistic number is the desk that spends the quarter apologising.

For the Gulf and Indian Ocean lane, Salalah and Colombo are the ones to watch, and Mundra is the one to trust cautiously. Salalah at 3.3 days with a 73 percent yard is the loudest warning on the page; route around it or pre-book the alternative before the density bites. Colombo at 2.3 days is calmer but still a wait you must quote. Mundra at 1.38 days looks like the relief, but remember it just came off a 13-day empty-yard strike that ended 11 September, so the recovery is fresh and fragile; do not assume it stays smooth. Diversify across these hubs so a single one's bad week does not become your missed delivery.

For Europe, read each gateway by its cause. Valencia at 2.5 days and Le Havre at 2.3 days with a wind suspension are weather-and-volume waits you pad for. Rotterdam at 1.5 days but yards critical is the one to fear most, because critical yards mean the berth number understates the real dwell; plan it as fragile and ship when the window is open. Hamburg at 1.9 days from construction is the plannable one, so schedule around it and stop treating it as a surprise. Genoa at 2 days rounds out a picture where no EU port is a free pass, and your routing should carry a second gateway by default.

For the Americas, the two profiles could not be more different. Santos at 3.3 days from weather is a ship-side clog, so earlier booking and a weather contingency in the plan are your tools; there is little inland magic to work. Vancouver is the inverse: berth wait of 1.0 day is excellent, but rail dwell of 3.3 days means your box sits after discharge waiting for a train. If your cargo moves into the Canadian interior or via that corridor into the US, the rail dwell, not the berth, is the clock you manage. Plan the two separately, because treating Vancouver as a one-day port will quietly add three days you never quoted.

The inventory implication is the part teams forget. A longer port wait is not just a scheduling note; it is more days of stock sitting offshore, which means more safety stock on the water or more risk of a shelf going empty. For an importer, add the congestion buffer to the reorder point, not just to the delivery promise, or you will protect the date and still run out. The exporters who only fix the quote and ignore the replenishment math are the ones whose customers feel the gap even when the shipment arrives inside the promised window. Treat the wait as inventory, because that is what it is.

Customer communication is cheaper than a missed window. When you quote the realistic buffered date, say it is the buffered date and why, rather than springing a delay after the fact. Buyers plan their own shelves and promotions around your promise; a date you padded honestly is a date they can build on, while a date you missed is a relationship cost no apology repairs. The forwarders and traders who turn the congestion data into a proactive note, we are quoting X because Salalah is at 73 percent, keep the account; the ones who go quiet until the late call lose it.

Use the data as leverage in carrier and terminal negotiation. When Rotterdam yards are critical and you know it from the weekly pull, you can ask the carrier for the less-congested call or a different gateway before the congestion shows up as your delay. When a port is loosening, you can push volume through it and maybe win a rate, because the terminal wants the box moved. The congestion report is not just a defensive tool; it is a buying signal if you read it weekly and act while the window is open. The operators who treat it as someone else's problem pay the spread; the ones who trade it keep it.

A second worked example on the Vancouver rail dwell makes the cost real. Suppose a 40-foot box of goods worth 50,000 dollars sits an extra 3.3 days of rail dwell beyond the berth. At a 6 percent annual carrying cost that is about 27 dollars a day, so roughly 90 dollars of float per box, which sounds trivial until you run a hundred boxes a month through that corridor and the dwell stretches during a rail disruption. Then it is 9,000 dollars a month of invisible carry, plus the penalty of a missed inland appointment. The berth number looked fine; the rail dwell was the cost, and nobody had quoted it. Build the dwell into the quote and the problem becomes visible and manageable.

Look at the numbers as a routing map rather than a complaint list. Every port on the 16 September report tells you where to add time and where to find slack, and the disciplined operator uses both. Kaohsiung and Manila tell you to pad the Asia exit; Salalah and Colombo tell you to diversify the Indian Ocean; Valencia and Le Havre tell you to pad the EU entry; Rotterdam tells you to watch the yard; Vancouver tells you to separate rail from berth. Read together, they are not a story of crisis but a set of dials you turn per lane. The cargo that moves on a plan built from those dials arrives; the cargo booked on last month's calm is the cargo that waits.

For the importer bringing goods into these ports, the same data drives your replenishment, not just your quote. If your European DC relies on Rotterdam and the yard is critical, your safety stock math has to assume the occasional extra dwell, or your pick faces gaps. If your North American flow runs through Vancouver, the rail dwell is the number that decides whether the box reaches the shelf on time, and your inbound plan should carry it explicitly. The mistake is treating congestion as the forwarder's problem; it is your inventory problem the moment the box is late, and the report is the early warning you already pay nothing for.

Close the loop by making congestion review a standing weekly habit owned by one person. Whoever runs your quoting or your replenishment should spend fifteen minutes every Monday pulling the latest port numbers, noting which gateway moved, and adjusting the buffers and the gateway choices for the week. This is not a project with a finish line; it is a small recurring discipline that keeps your promises honest through a volatile market. The operators who systematise it stop apologising for delays they could have seen coming, and that reliability is what their customers remember when the next lane gets loud.

One more lens worth applying is the carrier schedule itself. The reported waits are terminal-side, but the vessel or train that feeds the terminal has its own reliability, and the two compound. A carrier that arrives on a day the yard is at 73 percent lands you in the longest queue, while the same carrier a few days later may sail through. The operator who watches both the sailing schedule and the port density can shift a booking by days to land in the slack, often at no extra freight cost. Congestion is rarely uniform across a week; the skill is arriving when the yard is empty, not when it is full. This is the same logic as timing the relief window, applied at the booking stage rather than after the fact. It costs nothing but attention, and it is the difference between quoting a buffer you actually need and quoting one you could have avoided.

• Add berth-wait buffer to every shipment touching Kaohsiung, Manila, Salalah, Colombo, or Santos (2.3 to 3.3 days).
• Avoid single-port dependency on Salalah and Colombo; pre-book alternative transshipment.
• Watch yard-density readings (Salalah hit 73%) as your leading signal for empty-container and relief windows.
• For Europe, build extra days into Valencia and Le Havre quotes and track Rotterdam yard-critical status.
• On Vancouver, monitor rail dwell (3.3d) separately from berth wait (1.0d) when planning North American inland moves.