Maersk Doubles Durban Congestion Fee to $900 a Box for Far East Imports from November

The Analysis

Maersk has published a revision to its Congestion Fee Destination on imports into Durban from Far East Asia, the Indian Subcontinent and the Middle East. Effective 1 November 2026 for non-regulated origins and Vietnam, and 8 November 2026 for South Korea and Taiwan, a 20ft container moves from 250 dollars to 450, and a 40ft from 500 to 900. Per box. On top of ocean freight.

Start with the headline, because the headline is wrong by twenty per cent. 450 divided by 250 is 1.8. So is 900 divided by 500. Nothing doubled. Both tiers went up exactly eighty per cent, and if your costing sheet says the congestion fee doubled so we add a hundred per cent, you have quoted your customer a hundred dollars too much on every 40ft for a number Maersk never published. Write 1.8. Then write the increment on its own line: plus 200 dollars on a 20ft, plus 400 on a 40ft. That second pair is the only version your accounts department will ever see.

The interesting number is not the eighty per cent. It is the calendar. This fee did not exist in August. Maersk created it in September at 250 and 500, blamed disruption at the port of Durban, and repriced it upward before the original tariff had aged two months. A brand new line item that is revised upward inside sixty days is not cost recovery. It is a probe. What follows takes the 400 dollars apart: what kind of money this actually is, who pays it under your incoterms, when it lands in your books, what it does to a CIF Durban quotation, and the two thresholds past which the sensible move stops being a phone call to your carrier and becomes a change of routing or a change of contract.

Read the scope precisely, because this is where importers lose the most money. The notice covers imports into Durban from three origin blocks: Far East Asia, the Indian Subcontinent, the Middle East. Note what it does not say. Nothing about exports out of Durban, nothing about Richards Bay or Cape Town, nothing about cargo whose bill of lading reads Durban while the real consignee sits in Johannesburg, Germiston or Kempton Park, and nothing about cargo transshipped onward to neighbours. That fourth group is a large slice of South African inland traffic, and it will be invoiced at Durban because that is where the box touches the quay. Work on the assumption that the fee attaches at the port of discharge, not at the place your customer takes delivery, and keep working on it until somebody at the carrier puts the opposite in writing.

Two effective dates, seven days apart, same fee, same port. Vietnam sits with the non-regulated group on 1 November; South Korea and Taiwan get 8 November. The regulated and non-regulated distinction comes from filing regimes on those trades, where tariff changes take effect on their own notice clock rather than on the operator's preferred one. Whatever the mechanism, the operational consequence is plain: your November arrivals straddle both dates, and geography decides which one you get. A Taiwan-origin booking and a Shenzhen-origin booking sailing in the same week can carry different fee dates. If your team reconciles documents at month end, one of those two will be wrong. Fix it at the booking line, not in the ledger.

Now the legal character of the money, which decides who bears it. The name answers it: Congestion Fee Destination. It is levied per container on the destination side and stacked above the ocean freight line. Whether the shipper prepays or the consignee pays on collection is a commercial arrangement, not a legal one, and that arrangement decides who feels the increase. Under FOB, everything after loading belongs to the buyer, so on an FOB Ningbo shipment the 400 dollars is your South African customer's problem and your only obligation is not to be vague about it in the quote. Under CIF or CFR it depends entirely on what your sale contract says. The incoterm itself puts most post-shipment costs on the buyer, but plenty of CIF contracts are drafted to include everything except duty, and plenty of commercial teams have sold delivered-to-door dressed up as CIF for years. If your paper is one of those, the 400 dollars is yours. Pull the clause before you pull the freight number.

Look at the internal arithmetic next, because it tells you more than the headline does. A 20ft pays 450, a 40ft pays 900. That is exactly 450 dollars per TEU, flat across both equipment types. Three things follow from a perfectly TEU-neutral charge. Consolidating two 20ft into one 40ft saves you nothing at all on this line, since you were paying 900 either way, so the saving in any consolidation case has to come out of freight and terminal handling, not out of this fee, and if somebody in your team is building that slide, take it off the desk.

A 40ft High Cube pays the same 900 as a standard 40ft although it swallows appreciably more cargo, so high-cube-heavy shippers are quietly cross-subsidised by volume-light ones. Nothing in the tariff distinguishes a 40ft of textiles from a 40ft of consumer electronics worth five times as much, because the fee is value-blind, which means the lower your invoice value per box, the heavier this single line becomes as a share of your sales.

Timing is next, and it is where free money sits. Far East to Durban on the direct South Africa loops runs roughly twenty-five to thirty-five days depending on the service and the calls before the Cape; Indian Subcontinent origins are nearer eighteen to twenty-five; Middle East origins shorter again, say fifteen to twenty. A box closing in Shanghai this week lands in Durban in the first half of November, comfortably inside the window. So there is exactly one question worth asking your Maersk desk today: what triggers the new number?

Some operators trigger destination surcharges on estimated arrival at the port, some on bill of lading date, a few on booking or equipment release. If the trigger is arrival, cargo you loaded three weeks ago arrives to the new fee and no amount of manoeuvring changes it. If it is booking date, then the forty or fifty bookings you can still push through this week are worth up to 400 dollars each. Cheapest return on a single email anywhere in this file.

Now the question anybody holding a calculator should ask: does the number survive contact with arithmetic? I have no access to Maersk's congestion ledger, so here are my assumptions, stated so anyone with better data can correct them. Assume a mid-sized ship deployed on the Far East to South Africa trade costs in the region of twenty to thirty thousand dollars a day all-in at today's charter levels. Assume it discharges something like six to eight hundred import units at Durban on a call. At 900 per 40ft and 450 per 20ft, a call putting ashore around seven hundred units at an average of 750 dollars each brings in something over half a million dollars.

Half a million is roughly twenty ship-days of charter cost. Durban congestion is real and it is bad, but it is not twenty ship-days on every call bad, and even if it were, the honest way to charge it would be a variable line that moves with berth delay, not a flat per-box number that does not. A flat per-box charge that ignores how long the ship actually waited is not a reimbursement of anything. It is congestion rent, and its level tells you how much leverage importers have, not how bad the port is.

Why now, and why in this form. Three things line up. The seasonal one: South African retail builds stock through November and December for the summer and Christmas trading peak, which means containers have to move and alternatives are thin; the textbooks call that inelastic demand and the textbooks are right. The tactical one: a congestion fee collected at destination does not show up in the ocean freight number that procurement teams benchmark against published indices or against indexed service contracts. Move the freight rate and every index moves and a dozen renegotiation doors open at once.

Move a destination fee invented six weeks ago and almost nothing else stirs. The procedural one: the split date tells you the operator is threading its own filing calendars across different regulatory regimes, which takes planning weeks ahead, which means this was decided before anyone could know what Durban's berth productivity would look like in November. Put those together and the honest reading is that this is not a reaction to conditions at the quay this week. It is a yield decision taken with a clear view of what importers will tolerate in the fourth quarter.

Layer it, because three groups holding the same headline take three different kinds of pain. The exporter quoting CFR or CIF Durban out of Shenzhen, Nhava Sheva or Jebel Ali takes margin damage: those rates were quoted two months ago and can only be re-priced if the contract allows it. The FOB exporter takes no cash hit at all but picks up a different problem, because his buyer is about to discover 400 dollars he never budgeted and will ask whether anyone knew; answer honestly and early with the tariff reference and it stays commercial, answer vaguely and it becomes a claim. The importer whose volumes split between Durban door delivery and Durban-to-Johannesburg inland turns has the least visible exposure, because the trucking and rail legs were priced for the year already, the congestion fee arrives with no dedicated budget line, and normally nobody notices until the January landing-cost report prints and somebody asks where fourteen points went.

Let us put real numbers on it, with every assumption written down so you can swap in your own. Assume one importer running twelve 40ft high cubes a month into Durban from the Far East. Assume average invoice value of sixty thousand dollars per box, roughly what a mixed general-merchandise 40HQ carries. Assume gross margin before freight of twelve per cent and an all-in ocean freight cost before the congestion fee of two thousand eight hundred dollars per box. The fee moves from 500 to 900, so plus 400 dollars a box.

Four hundred on sixty thousand is 0.67 per cent of sales. Four hundred against gross profit of twelve per cent, which is seven thousand two hundred dollars, is 5.6 per cent of the profit on that box. Twelve boxes a month is four thousand eight hundred dollars a month, and fifty-seven thousand six hundred across twelve months at flat volumes. As a share of the freight bill it is fourteen per cent: your freight was two thousand eight hundred and it became three thousand two hundred, which is fourteen points added to your largest single logistics line without anybody touching the base rate.

Mix six 20ft a month into the same book at plus two hundred each and the monthly increment goes to six thousand dollars, about seventy-two thousand over a year. Now pick your cure. To hold twelve per cent you need either a price rise of 0.67 per cent or a freight cut of fourteen per cent. One of those you can do before lunch, and the other one you cannot do at all.

Before choosing, count the boxes that are not yours to re-price. Most November arrivals into Durban are not spot wins; they sit under fixed-price annual supply agreements with retail accounts, and where those agreements read delivered Durban including all charges except customs duty, the 400 dollars belongs to you until renewal day. That count is your true exposure. It will be smaller than your total November book but larger than you expect, precisely because nobody reads the destination-charges clause while there is no destination charge. So do this in order: list the November loadings, tag each line with its incoterm, tag each again with whether the sale agreement caps destination charges, then multiply only the exposed lines by 400. That figure, not twelve times four hundred, goes into your management report.

Two thresholds are worth writing into a note and pinning to the wall. The contract threshold first: when destination surcharges reach one and a half per cent of invoice value on a standard box, stop treating them as noise you pass along and reopen the price. Our sixty thousand dollar box now carries 900 dollars of congestion fee, which is exactly one and a half per cent. You are not approaching that line, you are standing on it. Anything above sixty thousand of value per 40HQ sits under it, anything below thirty thousand is above two and a half per cent and should already have been re-priced last week.

The routing threshold is less obvious. Nobody reroutes to save 400 dollars: moving a gate from Durban to Maputo and hauling to Johannesburg adds, on my own working assumption, twenty-five hundred to thirty-five hundred dollars a container, plus border time and a new clearing agent, which swamps the fee several times over. So the routing trigger cannot be the fee. It has to be the dwell. Ten days past free time at roughly ninety dollars a day per 40ft costs you another congestion fee, and twenty days costs you two.

If your boxes regularly sit past that, the delay is the bigger invoice and Maputo becomes a genuine comparison rather than a conference anecdote. Pull the actual average days past free time on your last ten arrivals and put those two figures side by side; the rest of the debate is noise.

Now argue against yourself, which is the step most pricing meetings skip. Scenario one: another mainline carrier publishes an import congestion fee at Durban before December. This is then structural. Assume it stays, write 900 into your 2027 landed-cost base and stop re-litigating it monthly, because the level moved and it will not move back. Scenario two: nobody follows within six weeks. Maersk then has a choice between absorbing most of the 900 inside service-contract discounts for bigger accounts, which is exactly what happened to a long list of joyfully announced surcharges before this one, or rolling the tariff back towards 500. Either way, the importer who quietly asked for congestion-fee-inclusive pricing at the last renewal gets relief and never tells anyone, while the importer who did not ask keeps paying 900 and spends eighteen months wondering why his competitor's quote is tighter.

Scenario three: the port improves, genuinely and measurably, before Christmas. Anyone who has watched this trade for a decade can tell you which way that tree grows. Rate it unlikely and treat 900 as your planning floor. There is one more objection I hear every week: absorb the 400 to protect the account. Fine, absorb it, but write down that you handed over 5.6 per cent of unit profit on that lane twelve times a month, and let whoever owns that decision sign for it.

Here is the part none of the coverage carried, so here it is in plain language: the fee is about six weeks old. Every wire story led with Maersk doubling the congestion fee. Not one of them led with Maersk inventing the congestion fee in September at 250 and 500 dollars and repricing it before a single full month of it had been billed. The difference between doubling and day-one pricing is the entire story. A surcharge with twenty years of history behind it behaves like a cost line and moves slowly. A surcharge with six weeks behind it is a market probe, and what it finds out this month decides your landed cost next year.

Which means the thing to watch is not Durban's berth productivity. It is everyone else's tariffs. If three carriers follow inside a month, Durban has been permanently repriced for importers and you budget for it. If nobody follows, Maersk's own commercial team quietly hands most of it back to whoever asks, in the shape of a freight rate that looks competitive and quietly contains the fee. The port did not decide whether you pay 900 dollars. Every other importer's behaviour did.

One last piece of arithmetic for anyone still reading this as a port story. Six weeks ago this line item did not exist. Today it costs eighty per cent more than on the day it was born, it is larger than most people's inland haulage bill inside Gauteng, and it was set by a company with every incentive to find out whether you bother to check. Congestion surcharges work like hotel minibars: nobody looks at the price until checkout, and by then somebody has already drunk the water. Your lock date is 15 December 2026. Either the rest of the trade has matched by then, in which case you write 900 dollars per 40ft into next year's budget and stop discussing it, or nobody has, and you walk into January asking for the whole thing back inside a freight rate you can actually see. Put both branches in the calendar now, because the second one only pays out if you remembered to ask.

↑ Back to the key points

— By Vivian Zhao