HMM Platinum Fire Puts 2,300 Boxes and a General Average Bill on Cargo Interests

The Analysis

The short version, from W.E. Cox and the ship's own advisories: the 5,593 teu HMM Platinum sailed from Kattupalli on India's east coast bound for Santos in Brazil carrying something in the order of 2,300 containers when fire broke out in a box off southern Madagascar. Salvage firm SMIT has been engaged under Lloyd's Open Form. W.E. Cox says damage could be substantial and a general average declaration is expected. The ship had first been aiming for Durban with arrival around 8 October; the latest advisory names Gqeberha, the port formerly called Port Elizabeth, as a possible refuge. Part of the cargo is under temperature control.

Now read those two numbers again. 5,593 teu is the vessel's nominal slot capacity. 2,300 is a count of boxes. Those are not the same unit, and not one headline I read said so. If the 2,300 were all 20ft they would occupy 2,300 teu, about four tenths of the ship. If a realistic sixty per cent of them were 40ft, the same manifest occupies nearer 3,700 teu, roughly two thirds of nominal. Nobody has published the manifest, and until somebody does, every downstream figure in this file that depends on utilisation is an estimate wearing a fact's clothes.

This note adds nothing to the firefighting. It takes apart the bill that is coming instead: how a general average contribution is actually computed, what proportion of cargo value you will be asked to put up front, in what form, how long it takes, why the reefer boxes are a separate emergency running on a separate clock, and the thing none of the coverage mentioned. For 2,300 boxes belonging to possibly a thousand different parties, the binding constraint was never the fire. It is whether every one of those consignees can produce a piece of paper quickly enough.

Start with the geography, because it sets the clock. Kattupalli sits north of Chennai on India's east coast; Santos is Brazil's biggest container port. Any service between the two rounds the south of Madagascar, then South Africa, then crosses the South Atlantic. The fire broke out in a box while the ship was still off southern Madagascar, which means she was days from anywhere useful and had very few places to put herself. That is why the refuge question matters more to cargo interests than the fire does: everything that follows, including where your box sits, for how long, at whose cost and under which country's surveyors, is decided by whichever port takes her in.

The first plan was Durban around 8 October. The later advisory names Gqeberha as a possible refuge. Three readings fit, and the advisory tells us which one won without saying why. Durban is the biggest container port on the South African coast and the natural first call coming west out of the Mozambique Channel; Gqeberha sits several hundred kilometres further southwest, nearer the Cape. A vessel with a box that has been alight does not get handed a berth inside a busy terminal quickly, and Durban has had its own berth queue this year, so availability may simply have moved. The second reading is the condition of the ship: if the salvor wants the seat of the fire cold and confirmed cold before committing her to a working harbour, a port with room to anchor and fewer neighbours is easier.

The third reading is arithmetic: by the time the call had to be made, Gqeberha may simply have been closer. Now note what none of this tells you, and what should worry any consignee with cargo aboard: Gqeberha does not handle anything approaching Durban's container volume, and reefer plug capacity at any port is finite and allocated. If several hundred boxes have to come off and sit, the number of live plugs and the yard behind them is your problem, because somebody berthed ahead of you took them first.

Next is what Lloyd's Open Form actually does to your property. LOF is no cure, no pay: SMIT is rewarded only to the extent it saves property at risk, and the size of that reward is decided later, in arbitration, measured against the value of what was saved. Two consequences for a cargo owner follow immediately. The salvor holds a maritime lien over the salved property, and a lien travels with the box rather than with whoever booked it. Nothing is released to anybody before the salvor's security is in place. So there are two securities in this file, not one: salvage security and general average security. Most importers budget for the second, and it is usually the first that blocks the delivery order.

One question usually eats more time than all of that arithmetic, and it comes first: how do you establish that your boxes are on board at all. A casualty notice gives you a ship and a voyage, not a manifest. Your box numbers sit with whoever issued your bill of lading, which for most interests on an India to Brazil service means a forwarder or a consolidator rather than any carrier's cargo desk. Ask today, ask in writing, and ask for the vessel load list rather than a verbal confirmation. There is a second snag buried in the same paperwork.

Where cargo moved under a house bill, the value of record an adjuster reaches for may be the value declared on the master bill rather than what your supplier wrote on the commercial invoice. Those two figures diverge often, and when they do you get a three-way argument about which arriving value governs, measured in weeks rather than days. Put the master bill, the house bill and your own invoice into one email thread on day one: one hour of work for a month of difference.

Now the arithmetic of general average, which is where the number either survives or does not. The rule is simple even if the practice is not: where a deliberate sacrifice or an extraordinary expenditure is made for the common safety, the cost is shared among the interests that came out alive, meaning ship, cargo and freight, each contributing in proportion to its own arriving value. Nobody publishes figures this early, so here is a working set and you should replace it with your own the moment the adjuster speaks. Assume the hull is worth eighteen million dollars, roughly where a secondhand ship of this size and vintage trades today, which is my assumption and not the adjuster's.

Assume the cargo is worth one hundred and fifteen million dollars: 2,300 boxes averaging fifty thousand dollars CIF, a plausible middle for India-to-Brazil general cargo and also an assumption. Assume freight at risk of two million, since only freight that would have gone to the bottom with everything else counts. Total arriving fund: one hundred and thirty-five million dollars. Now assume the general average expenditure, including the salvage award and everybody's costs, settles at fifteen million dollars. Serious LOF cases routinely settle in a band above ten per cent of salved value, so that is not a heroic assumption.

The general average rate then comes to roughly eleven per cent, and every interest pays eleven per cent of its own arriving value: about two million dollars from the shipowner, about twelve point eight million from cargo, about two hundred and twenty thousand from freight. Spread across the assumed cargo value, that is around five thousand five hundred dollars per box on the average fifty thousand dollar container.

That is the final number, and final numbers in a general average arrive years late. What is asked of you now is a different figure: a deposit or an undertaking at a provisional rate set high enough that the adjuster is never left short when the arithmetic is finished. Combined salvage and general average demands commonly open somewhere between fifteen and twenty-five per cent of the arriving value of the cargo; take twenty per cent as a working middle. On our fifty thousand dollar box, that is ten thousand dollars up front, before anybody touches the container. If the eventual contribution lands at five thousand five hundred, somebody refunds roughly four and a half thousand of that one day. Nobody will commit to the day, and if you are building a cash forecast, write it off.

Three ways to produce that ten thousand, and the choice is mostly made for you. Cash paid into the average adjuster's or their lawyers' client account always works, ties up working capital, and returns whenever the adjustment returns. An insurer's letter of undertaking is the normal route for anyone with cargo cover: the insurer promises to pay whatever is eventually found due, nothing is wired today, and the insurer will want the policy issued at one hundred and ten per cent of CIF value, because that is roughly what arriving value means in practice. A bank guarantee also works, costs a fee and consumes credit line. Put plainly: insured interests produce paperwork, uninsured interests produce cash and then wait.

Insurance is where this either runs clean or runs for months. Under the standard institute cargo clauses, general average contribution and salvage charges are insured perils, so a cargo insurer issues the undertaking and eventually pays the contribution. Three things break that. A missing policy, which means you fund everything yourself. A declared value struck at invoice only rather than CIF plus ten per cent, because the adjuster works from arriving value and invoice value can be short by exactly the slice you are asked to fund. And paperwork that takes weeks to assemble: bills of lading, commercial invoices, packing lists, the certificate of insurance, and later the delivery order, all original, all clean, all wanted at once. Collect them the week you hear about the casualty, not the week you want the container.

Run this down to the small end of the market, because that is where it bites hardest. Assume a Brazilian trading company with one 20ft of hardware at nine thousand dollars CIF on this ship. Twenty per cent is one thousand eight hundred dollars, in cash or guarantee, to take delivery of its own property at a port it never chose, and eighteen hundred on nine thousand is twenty per cent of the invoice simply to get through the gate. Now assume a second box on the same vessel belonging to a big importer with eight hundred thousand dollars of arriving value. The adjuster applies the same twenty per cent, which is one hundred and sixty thousand dollars, and nobody in that treasury loses a night's sleep. Identical rule, absurdly different pain. A flat percentage of arriving value is regressive in a way no one will ever win an argument about, and it is arithmetic worth knowing before you reassure a small customer that everything is under control.

Here is what every report I read left out, and it is the thing that decides when any of those 2,300 boxes actually moves. Coverage led with the fire, with photographs of smoke, with the change of refuge. Not one of them led with the security queue. A carrier and an average adjuster release cargo against security, and neither is in the habit of splitting a release so that one box walks while its neighbour's paperwork is still missing. The release therefore moves at the speed of the slowest document in the pile. So across 2,300 containers belonging to perhaps a thousand different parties, spanning Indian exporters, Brazilian importers, forwarders, consolidators and a scattering of insurers, the binding constraint is not whether the fire is out.

It is whether a thousand consignees can each produce the right instrument, and note how unevenly difficult that is. For the large importer, twenty per cent costs one signature from an insurer who already holds the premium. For the small one with no open policy, it means finding cash or finding a bank that will issue a guarantee against a facility they may not have, neither of which can be arranged on a Friday afternoon. The security threshold is where this casualty gets expensive, and almost nobody wrote the word down.

Two clocks started running, and they do not run at the same speed. The paper clock: notification from your forwarder, confirming which box numbers are actually yours, notice to insurers, the adjuster fixing a provisional rate, arranging the instrument, serving it, getting release. With a clean manifest and a switched-on broker that is realistically seven to fourteen days. With Indian-origin manifest data quality, several layers of sub-booking and one uninsured consignee in the chain, thirty days is routine and sixty is not unusual. The cargo clock runs faster, and the thermometer governs it. Part of this cargo is temperature controlled. Assume a chilled consignment starts to fail its specification somewhere between forty-eight and seventy-two hours after power is lost, and that plenty of buyers will reject it afterwards on documents even if the contents look fine, because cold chain integrity is a paper claim as much as a physical one.

Assume reefer plugs at the refuge port cost, on my own working assumption, forty to ninety dollars a unit a day plus monitoring, so a month alongside adds roughly eighteen hundred dollars for that one box on top of everything else. Write one line down from this: ten days of paper delay burns the margin on most chilled cargo, and no general average adjustment has ever refunded a rot write-off.

Now argue against yourself, which is what a decent adjuster does and most cargo interests do not. Scenario one: the damage is confined to a handful of boxes, the ship is repaired and completes to Santos, everything arrives late but arrives, and the general average comes in well below eleven per cent. Your deposit is then largely refunded in due course and the true cost is delay, not money. Nothing to be done about that beyond planning contingency in future. Scenario two: she cannot continue and cargo transships.

That adds handling, monitoring, a second port's storage tariffs and quite possibly a second general average act. Rare, expensive, and a good reason to have your insurer's surveyor looking at the boxes you are interested in now rather than after they have been restowed twice. Scenario three, which surfaces in every casualty: general average should never have been declared, the hull underwriters should swallow it, this is simply owners making cargo fund their losses. There is a genuine professional argument about how readily the York-Antwerp rules are invoked.

You will not win it this month, and spending your energy there is worth strictly less than having your security ready the day it is demanded. Scenario four: damage is minor, no general average is declared, and all of the above costs you a week of administrative attention. Even then, being the account with papers ready is how you get called first out of a thousand-box queue.

One more number belongs in front of anyone reading this as shipping news, because it is cheap and it is only available beforehand: cargo insurance. An open policy costs a small fraction of a per cent of insured value. This event asks you for twenty per cent of one shipment, up front, today. Set those two rates side by side and the purchasing decision is not a close one. If you are reading this with boxes on the water and no cover, that is the most expensive sentence in this file, because cover cannot be arranged now for cargo already in distress: insurance is sold against risk that has not happened yet, not against the ship that is on fire.

So where is the line worth locking? Below twenty per cent of arriving value and with insurance behind you, sign the average bond, serve the undertaking and get your box moving, because the cost of arguing exceeds the cost of agreeing. Above twenty-five per cent of arriving value, or above the gross profit you will make on the whole shipment, stop wiring and start arguing: at that point paying for an independent valuation of what actually arrived can be cheaper than accepting what the adjuster assumed it was worth. Everything else is waiting, and waiting is the expensive part. Nobody has ever bought cargo insurance because a ship arrived on time. And if anyone tells you this is a story about a fire at sea, offer them the cargo interest's version instead: it is a story about whose bank can print a guarantee before somebody else's cargo rots.

↑ Back to the key points

— By Vivian Zhao