Rotterdam strike on Sept 4 halts truck handling at key terminals

The Analysis

Put the map of Rotterdam on the table and the first thing that stops being true is the word port. What carries that name is not one node. It is a string of terminals along forty kilometres of water, each with its own owner, its own gate, its own labour agreement and its own idea of what a working day looks like.

On 4 September the labour side of that string stands down for a shift. APM Terminals MVII suspends truck access from late morning to late evening, RST and RSC stop handling between 10:00 and 20:00, ECT Delta and Euromax close, while RWG and the depots stay open.

What follows is about why that sentence is not a port closure and why, for one class of cargo, it is worse than one, which cargo that is, and what ten years of consolidation did to make ten hours cost a week.

Four of the deep-sea terminals on that string shut their gates for ten hours. Two did not. The depots, which is where the empty boxes live, stayed open the whole day. Read the list again and notice what kind of list it is. It is not a list of capacity. It is a list of gates. Every terminal named had quay cranes that could have kept working. What stopped was the interface between the terminal and the road.

Rotterdam moves a large share of its hinterland cargo by barge and by rail, and that share is the reason the port can absorb a bad day without shelves emptying. A truck-only action therefore does not hit the port evenly. It hits the cargo that uses road because the lot is too small for a barge, the lead time too short for a train, or the delivery point too far from a barge terminal to survive the double handling. Retail and consumer goods on short replenishment cycles, cold chain, automotive aftermarket, anything sold on a promise of a named day. Bulk, chemicals and the big contracted flows on barge barely notice a ten-hour gate closure, because their boxes were never going through that gate.

RWG staying open reads like redundancy, and on paper it is. In practice the surviving terminal is not a substitute for the closed one, for four reasons that have nothing to do with quay capacity. Its appointment book is already committed. Its barge and rail connections are sized for its own hinterland book, not a neighbour's. Its yard is organised around its own vessel calls, which do not move. And its gate is staffed for a shift that was planned weeks ago. Redundancy in a port is not spare crane hours. It is spare appointment slots, and those are not transferable at ten in the morning.

The congestion the notice predicts before and after is not incidental. An action with published hours is a scheduled event, and scheduled events produce a surge on both sides of the window. Trucks that would have run at eleven run at eight. Trucks that would have run at six run at nine. The volume does not disappear, it is compressed into a shorter period either side of the closure, and a gate with a fixed hourly throughput cannot compress with it. The measurable loss is not ten hours. It is ten hours plus the queue in front and the queue behind.

Here is what a notice of this kind cannot tell you, and it is where most plans go wrong. During a scheduled action the binding constraint is not crane capacity and not yard capacity. It is appointment inventory, meaning the number of bookable truck slots per hour at one terminal, in one direction, for one container size. Those slots sit in a booking system with quotas, and a quota does not expand because the terminal next door went dark. A shipper who re-slots to the open terminal finds this out the way everyone finds it out: the terminal has room, the yard has room, and there is no slot until Thursday. Watch for that failure on 3 September, the day before, because that is when the re-slotting happens and it is the day the real damage is done.

Scale separates the field cleanly. A shipper moving a hundred boxes a week has a named contact at two terminals, a standing slot allocation and a fallback that was negotiated rather than improvised; 4 September costs it a few phone calls and a re-sequenced yard plan. A shipper moving eight boxes a week has none of that. It has a forwarder, a booking reference and whatever slot the system hands it, and its boxes are the ones that land at the back of the queue when the queue is allocated by whoever calls first. The same ten hours is an irritation to one and a lost week to the other.

Direction matters too, and asymmetrically. Import cargo has a demurrage clock and a customer waiting. Export cargo has a cut-off, which is a harder wall. A missed cut-off does not delay a delivery by a day, it delays it by a sailing, and on a weekly service a sailing is a week. Exporters out of the Netherlands and the German hinterland are exposed to that wall on 4 September even though the story gets written as a story about importers.

Step back, because this shape is older than the strike. Rotterdam spent twenty-five years doing what every major port did, which is consolidating. Scattered small terminals became a handful of very large ones on reclaimed land, each with one gate, one automation stack, one labour agreement and one operating company. That bought the thing the market was paying for: quay productivity, and the ability to turn a very large ship inside a window that fits a proforma. What it spent was gate redundancy. On a normal day nobody can see the trade. On 4 September it is the whole story, and it will be the whole story again the next time anyone on that string puts a gate down, for whatever reason.

Whether this is a pulse or a turn depends on one thing the notice does not say, which is what the action is about. A one-day stoppage over hours is a pulse. It clears, the queue drains, and by the following Tuesday the port is a port again. If the dispute moves from hours to shift patterns it stops being a pulse, because shift patterns are what an automated terminal is built around, and a terminal that cannot run its planned shifts cannot run its planned proforma. Nothing in the source tells us which of the two this is. That absence is the reason to plan for the tail rather than for the day.

Take a small arithmetic, with every figure below set by me and none of them taken from the notice. Assume an importer bringing in forty FEU a week through Rotterdam, sixty per cent of it truck-borne, which is twenty-four FEU across five working days, about five a day. Assume the closure plus the queue either side costs two effective working days of gate throughput. That is ten to twelve FEU displaced, and they do not vanish, they arrive two to three days late into a warehouse that was not expecting a gap. With twelve days of cover on the affected lines nothing happens, the stock absorbs it. With four days of cover the importer is out on the fifth day, and ten hours has turned into a conversation with a customer. The arithmetic says the discriminator is inventory cover, not freight cost, which is why a plan built entirely around re-routing containers is answering a question nobody asked.

Now the counterfactual, because it is the useful test of any plan. Suppose nothing is done in advance. Slots are not re-booked, SKUs are not ranked, empties are not pulled forward. What follows is not a uniform two-day slip. What follows is that the cargo with the better forwarder relationship takes the surviving slots and the cargo without one takes the following week, so the loss is not spread across the book, it is concentrated on the smallest shippers and the newest lanes. That is a queue-allocation failure wearing the costume of a capacity failure, and it is the one worth designing against.

The trap that catches most people is subtler than a missed cut-off. The depots stayed open, so empty returns look like a non-issue and get left alone. But an empty return is the front half of an export booking. If returns pile up on 4 September because the laden gate was the bottleneck and everyone deferred, then the export cut-off on 8 September fails for a reason that has nothing to do with 8 September. Pull the empties forward before the gate, not after it.

A port is a promise that the sea and the road will meet on schedule. Ten hours is a small thing to break that promise with, and the terminals will be running again before the week is out. What does not come back on schedule is the plan the shipper built on the assumption that the promise holds. Keep the map on the wall and mark which of your gates has a second one behind it. That is the only line on it that changes what you do on the third of September.

↑ Back to the key points

— By Tomás Ferreira