What a Group A cargo is
Group A of the IMSBC Code is the set of solid bulk cargoes that may liquefy, or undergo dynamic separation, when carried at a moisture content above their transportable moisture limit (TML). They are a physical hazard, not a chemical one: a fine, moist, granular cargo such as iron ore fines, nickel ore, a mineral concentrate or bauxite fines can lose its strength under the ship’s motion and behave as a liquid, destroying stability. A Group A cargo may be loaded only if its actual moisture content is below its certified TML.
That single rule, moisture content below TML, is the whole of Group A safety in one line. Everything else, the tests, the certificates, the shipboard checks, exists to prove it before the cargo goes on board and to keep proving it through the voyage. The hazard is well documented and lethal: cargo liquefaction has been a leading cause of bulk-carrier loss of life this century.
The three groups: A, B and C
The IMSBC Code sorts every listed solid bulk cargo into three groups by its hazard:
- Group A cargoes may liquefy or dynamically separate above their TML. The hazard is physical and moisture-driven.
- Group B cargoes possess a chemical hazard: they may self-heat, emit toxic or flammable gas, deplete oxygen in an enclosed space, react dangerously with water, or present a dust-explosion risk. Direct reduced iron, coal (for its methane and self-heating), and ammonium-nitrate-based fertilizers sit here.
- Group C cargoes are neither liable to liquefy nor chemically hazardous: limestone, iron ore in lumps or pellets, sand.
So the answer to the common examination question, into which group does a cargo with a chemical hazard fall, is Group B. The groups are not exclusive: a cargo that both liquefies and carries a chemical hazard is classified Group A and B together, as several lead and zinc residues are. The schedules in the Code are also not a closed list; a cargo not listed must be assessed before it can be carried, treated below.
How liquefaction happens
A Group A cargo is a packed mass of fine particles with water in the spaces between them. At rest the particle-to-particle contact carries the load and gives the mass the strength of a solid. The ship’s motion breaks that down. Rolling, pitching and slamming apply repeated shear to the cargo, the grains rearrange and pack tighter, and the pore spaces shrink. In a fine, low-permeability cargo the trapped water cannot drain away fast enough, so the pore-water pressure climbs. As it climbs, the effective stress holding the grains together falls, and when the moisture is at or above the TML the pressure can rise until the grains carry almost no load at all. At that point the cargo flows like a liquid.
In the hold the result is a sequence that runs fast. The cargo surface goes fluid; a free liquid surface develops that shifts bodily with every roll and multiplies the heeling moment; the cargo migrates to one side; the ship takes a list that ballast cannot correct; and the list runs away to capsize. Many of these losses happened in minutes, often at night, with few or no survivors.
The geotechnics behind it explain why the hazard sits with fine cargoes and not coarse ones. A cargo’s strength comes from the effective stress, the contact force between grains, and that force is whatever total weight the grains carry minus the pressure in the pore water. Drive the pore pressure up and the effective stress, and so the strength, falls toward zero. A coarse, free-draining cargo cannot build pore pressure because the water runs out between the grains as fast as compaction squeezes it, so it never liquefies. A fine cargo with low permeability traps the water, lets the pressure build under the cyclic load of the seaway, and reaches the zero-strength condition once its moisture is high enough. That is why the schedules turn on particle size and moisture, and why a cargo that looks like dry rubble at the stockpile can still be a liquefaction risk if the fines fraction is wet.
Dynamic separation
A related mechanism was written into the Code after the bauxite losses. In dynamic separation the cargo does not liquefy uniformly through its depth. Instead the ship’s motion drives fine particles and water upward through the mass to form a slurry of water and fine solids on top of a still-solid bulk beneath. That surface slurry produces the same free surface effect as full liquefaction. Amendment 06-21 to the IMSBC Code, mandatory from 1 December 2023, revised the Group A definition to cover both liquefaction and dynamic separation, and added a formal definition of the latter. The mechanism is associated with bauxite fines and some coals.
Liquefaction is not an ordinary cargo shift
It helps to separate liquefaction from the more familiar dry-cargo shift, because they are different problems with different defences. A dry granular cargo such as grain or coarse coal can shift when the ship rolls beyond the cargo’s angle of repose, the steepest slope the dry material will hold, and the standard defences are trimming the surface level and, for grain, the specific stability and securing rules of the Grain Code. That is a one-time movement of a solid down a slope. Liquefaction is different in kind: the cargo does not slide as a solid, it loses its strength entirely and behaves as a fluid, so trimming and angle-of-repose thinking do not protect against it, and the free surface of the fluid keeps working against the ship roll after roll rather than settling at a new angle. A Group A cargo can be trimmed perfectly level and still liquefy if it is too wet, which is why moisture control, not surface trimming, is the governing defence. Conflating the two is a real and dangerous error, because the measures that contain a dry shift do nothing for a liquefying cargo.
It also helps to remember that a Group C classification is not a guarantee of an easy cargo. Group C means only that the cargo is neither liable to liquefy nor chemically hazardous; it can still be dense enough to overstress the structure if loaded carelessly, dusty, or prone to an angle-of-repose shift, so the group system addresses the two specific hazards of moisture and chemistry, not every risk a bulk cargo carries.
TML, FMP and moisture content
Three defined quantities govern Group A carriage, and the relationship between them is the heart of the subject.
Moisture content (MC) is the mass of water in the cargo as a percentage of the total wet mass, measured by drying representative samples.
Flow moisture point (FMP) is the moisture content, as a percentage of wet mass, at which a sample first begins to flow under defined agitation, the threshold where the cargo loses its strength.
Transportable moisture limit (TML) is the maximum moisture content treated as safe to carry. For the flow-table and penetration tests it is set at:
The ten percent margin allows for measurement scatter and for moisture picked up between testing and the end of the voyage. The Proctor-Fagerberg family of tests gives the TML directly from a moisture-density relationship without computing an FMP first.
The carriage rule follows from these definitions: a Group A cargo may be loaded only while MC is less than TML. If the measured moisture content reaches or exceeds the limit, loading must stop. The iron ore fines moisture calculator and the nickel ore flow moisture point calculator support the moisture side of this check for the two highest-risk cargoes.
Determining the TML: the test methods
The TML is set by an approved laboratory using one of the test methods in Appendix 2 of the Code. There are now six, not the three of the older editions, and the method matters because different methods can return a different TML for the same cargo:
- Flow table test. A sample is jolted on a dropping table and the moisture at which it first flows is read as the FMP, with TML at 90 percent of it. Suited to fine materials; the flow judgement is partly visual, which is its limitation.
- Penetration (bit) test. An oscillating platform applies energy and a penetration bit measures the depth it sinks, giving an objective FMP for mineral concentrates and coals up to around 25 mm; TML again at 90 percent.
- Proctor-Fagerberg test (PFC70). Adapted from soil compaction, a light hammer compacts the sample and the TML is taken at the moisture giving 70 percent saturation at the optimum moisture content. It suits most mineral concentrates up to about 5 mm and is unsuitable for coal.
- Modified Proctor-Fagerberg for iron ore fines (PFD80). Introduced after the 2009 Indian iron-ore-fines losses and mandatory from 1 January 2017, it uses a lighter hammer and sets TML at 80 percent saturation, reflecting that iron ore fines reach a higher saturation at the optimum than ordinary concentrates. It is the required method for iron ore fines.
- Modified Proctor-Fagerberg for coal, with a larger mould and a reconstitution step for coarse particles, plus criteria for whether a given coal is a Group A cargo at all.
- Modified Proctor-Fagerberg for bauxite, developed by the Global Bauxite Working Group after the Bulk Jupiter loss for bauxite’s particular drainage behaviour.
Ranked by the TML they tend to produce for the same sample, the penetration test is the most conservative and the modified Proctor-Fagerberg the least, so the choice of method has commercial weight: a higher certified TML lets a shipper load wetter cargo, which is exactly why the method is specified by the cargo schedule rather than left to choice.
The Proctor-Fagerberg family is worth understanding because it is the modern direction of travel. It borrows from soil mechanics the idea of a compaction curve: a sample is compacted at a series of moisture contents and its dry density measured, tracing a curve that peaks at an optimum moisture content where the packing is densest. The degree of saturation, the fraction of the pore space filled with water, is then read off, and the TML is defined as the moisture corresponding to a chosen saturation at that optimum, 70 percent for the standard test on ordinary concentrates and 80 percent for the modified test on iron ore fines. The difference matters because iron ore fines genuinely reach a higher saturation at their optimum than a typical concentrate, so applying the 70 percent figure to them would understate the safe limit; the 80 percent modified test was introduced precisely to correct that. The same logic produced the separate coal and bauxite versions, each tuned to the drainage behaviour of its material. The flow-table and penetration tests, by contrast, measure the flow threshold directly and apply the flat 90 percent margin, which tends to give a lower and more cautious figure. None of this is a free choice for the shipper or the laboratory: the cargo schedule names the method, and a TML obtained by the wrong method is not a valid certificate.
The can test
The can test is the simple shipboard screening check a ship’s officer can run on the quay, set out in the Code. The procedure is to half-fill a cylindrical can of about 0.5 to 1 litre with a representative sample, then bring the can down sharply onto a hard surface from a height of roughly 0.2 metres about 25 times at one to two second intervals, and examine the surface.
The interpretation is asymmetric, and the asymmetry is the point. If free moisture appears at the surface or the cargo goes fluid, that is a reliable warning that the moisture is probably above the TML, and loading should stop until the cargo is laboratory-tested. But a clean result does not prove the cargo is safe: a cargo can be over its TML and still pass the can test, particularly the higher-TML materials. The can test is a screen that can condemn a cargo but cannot clear it. The certified TML and moisture content, not the can test, are what permit loading.
Sampling and laboratory testing
A TML or moisture-content figure is only as good as the sample it came from, and sampling a heterogeneous stockpile is where much of the practical difficulty lies. The Code calls for representative sampling, taking many small increments spread across the stockpile or the loading stream and combining them, rather than a single grab that may miss the wet pockets. A wet layer at the base of a stockpile, water that has drained to one side, or rain that has soaked the surface can all leave the bulk moisture far higher than a careless sample suggests, which is why sampling during loading, not just before it, is part of good practice on a high-risk cargo.
The two certificates run on different clocks because they measure different things. The transportable moisture limit is a physical property of the material, so its certificate is valid for up to six months, subject to the cargo not changing in character. The moisture content is a condition that drifts with the weather and the handling, so it must be sampled and tested within seven days before loading begins, and retested if rain or snow falls in between. The laboratory must be one the competent authority of the load port recognises, using one of the approved Appendix 2 methods appropriate to the cargo. A common dispute on iron ore fines and nickel ore is the gap between a shipper’s certificate and an independent check sample drawn by the ship’s or owner’s surveyor; where the two disagree, the prudent course is to treat the higher moisture as the operative figure and stop loading until it is resolved.
Representative sampling follows recognised procedures that fix how many increments to take and how to combine and reduce them, so the laboratory works on a sample that genuinely reflects the bulk rather than a convenient corner of it. The principle is that a single grab is worthless on a heterogeneous stockpile: enough increments must be taken across the pile, and ideally across the loading stream as the cargo actually goes aboard, that local wet pockets are captured in proportion. Moisture content in particular is best checked from the cargo as loaded, because handling, the weather and drainage all move it between the stockpile and the hold. For the worst trades the practical answer is continuous oversight, a surveyor watching the loading, sampling through the operation, and empowered to call a stop, rather than a single certificate signed before the first grab touches the ship. The certificate is the legal record; the sampling discipline behind it is what makes the record mean anything.
Loading, the voyage, and liquefaction at sea
Hold preparation and loading for a Group A cargo are about keeping water out and the cargo level. The bilge wells are checked clear and the bilge system proved before loading, so that any water that does separate can be detected and, within limits, dealt with. After each hold is filled the cargo is trimmed reasonably level across the full breadth, which removes the steep faces and the void channels that let moisture migrate and that give a liquefied mass room to surge to one side. Loading is suspended in heavy rain, because cargo wetted on the belt or in the hold can cross its TML even if the stockpile figure was sound.
At sea the cargo is monitored as far as it can be. Rising water in the bilge wells, a developing list with no other cause, or a change in the ship’s roll, becoming sluggish or taking on a long, heavy period, can all be early signs that the cargo is moving toward a fluid state. If liquefaction is suspected, the master’s actions are aimed at reducing the energy going into the cargo and buying stability: alter course and reduce speed to ease the ship’s motion and take the seas in the easiest direction, avoid sharp helm and synchronous rolling, and keep the largest stability margin available by not slackening tanks unnecessarily. Pumping the bilges is not a cure and can be misleading, since the free water in the wells is a symptom, not the bulk of the problem in the cargo. The casualty record shows how little time these measures may buy, which is why the entire regime is built to keep an over-moist cargo off the ship in the first place rather than to manage it once it is aboard.
The principal Group A cargoes
The Group A schedules cover a long list, but a handful carry most of the risk and most of the trade:
- Iron ore fines (IOF), predominantly below 1 mm, with a dedicated Group A schedule and the mandatory modified Proctor-Fagerberg (PFD80) test. Coarse iron ore in lumps and pellets is Group C; the fines fraction is what triggers Group A.
- Nickel ore, the lateritic ores shipped mainly from Indonesia and the Philippines, high in clay and often loaded at high natural moisture, and the single most lethal liquefaction cargo by casualty record.
- Mineral concentrates from wet beneficiation: zinc, lead, copper, iron pyrite and similar sulphide concentrates, many of which are Group A and B together.
- Coal, a dual-hazard cargo that can be Group A and Group B at once. Under Amendment 07-23, coal is treated as a Group A cargo where its particle size distribution is dominated by fines below 5 mm; it is also Group B for self-heating and methane.
- Bauxite fines, reclassified to Group A after the Bulk Jupiter loss, distinguished from coarser Group C bauxite by particle size and drainage.
- Mill scale and mill scale fines from steelmaking, stored outdoors and often wet relative to their TML.
- Fluorspar and a growing set of fines and concentrates added at each amendment, from ground granulated blast-furnace slag to magnesite, dunite and crushed granodiorite fines under Amendment 07-23, and apatite concentrate and uncalcined phosphate rock fines coming in under Amendment 08-25.
Two of these deserve a closer look because they carry the bulk of the casualties. Iron ore fines became a regulated Group A cargo only after repeated losses exposed that the old Group C iron ore schedule did not fit fine, high-moisture material; the modern position splits coarse iron ore, which stays Group C, from fines, which are Group A, with goethite content and particle size deciding the line and the modified Proctor-Fagerberg test setting the limit. Nickel ore is the harder case operationally: the lateritic ores of Indonesia and the Philippines are mined and stockpiled in the open in the tropical wet season, are rich in clay, and arrive at the ship with a natural moisture that is frequently at or above any safe limit, so the cargo cannot be dried in any practical timescale and the only real control is to refuse it when it is too wet. Bauxite straddles the boundary differently again: most bauxite is a benign Group C cargo, but the fine fraction can dynamically separate, and the schedule split that followed the Bulk Jupiter loss is meant to catch exactly that fine, drainable material. Coal sits in two groups at once, a Group A liquefaction risk when it is fine and wet and a Group B chemical risk for self-heating and methane throughout, so a coal cargo can demand both moisture control and gas monitoring on the same voyage.
Casualties that shaped the rules
The Group A regime is, more than most of SOLAS, written from specific losses.
The Indian iron-ore-fines casualties of 2009 drove the iron ore fines schedule. The Black Rose capsized off Paradip on 9 September 2009 with about 25,000 tonnes of iron ore fines, killing the chief engineer as the crew abandoned, and together with the loss of the Asian Forest earlier that year prompted the Indian review that fed the modified Proctor-Fagerberg method and the Group A iron ore fines schedule made mandatory in 2017.
Nickel ore wrote the next chapter. The Vinalines Queen sank on 25 December 2011 carrying about 54,000 tonnes of nickel ore from Indonesia to China, with 22 of her 23 crew lost and a single survivor found days later on a raft. The Emerald Star foundered off the Philippines in October 2017 with nickel ore from Indonesia, losing 11 of 26 crew. Both losses fit the same pattern of ore loaded above its safe moisture and liquefying in a seaway.
Bauxite produced the most consequential single casualty. The Bulk Jupiter sank off Vietnam on 2 January 2015 with 46,400 tonnes of bauxite, losing 18 of her 19 crew, going from general alarm to gone in under twenty minutes. The investigation concluded the cause was dynamic separation rather than classical liquefaction, and the response, through the Global Bauxite Working Group, produced the split into a Group A bauxite fines schedule and a revised Group C bauxite schedule, mandatory from 1 January 2021, and the dynamic-separation definition added to the Code in 2023.
The pattern across these losses is consistent and is what makes the cargo so dangerous. The cargoes that liquefy are high-volume, low-value bulks shipped from ports with limited testing infrastructure, often in the rainy season, on a commercial pressure to load and sail; the failure is sudden and total rather than progressive, so the crew get little warning and the ship can be lost before a distress message is complete; and the casualties cluster on a few origins and a few cargoes, the Indian and Brazilian iron ore fines, the Indonesian and Philippine nickel ore, the tropical bauxite. Each rule change, the iron ore fines schedule of 2017, the bauxite split of 2021, the dynamic-separation definition of 2023, traces directly to one or more of these losses, which is why the regulatory history reads as a list of ships. The lesson the industry keeps relearning is that the only reliable defence is to keep an over-moist cargo off the ship at the load port, because nothing aboard reliably saves a ship once the cargo has gone fluid.
The regulatory framework
The IMSBC Code is the International Maritime Solid Bulk Cargoes Code, adopted by IMO Resolution MSC.268(85) and made mandatory under the SOLAS Convention, Chapter VI for the carriage of cargoes and Chapter VII for dangerous goods in solid bulk form, from 1 January 2011. It replaced the older recommendatory Code of Safe Practice for Solid Bulk Cargoes. The Code is amended on a two-year cycle, and naming the amendment in force matters at survey: Amendment 06-21 (MSC.500(105)) became mandatory on 1 December 2023, Amendment 07-23 (MSC.539(107)) on 1 January 2025, and Amendment 08-25 (MSC.575(110)) applies voluntarily from 1 January 2026 and becomes mandatory on 1 January 2027.
Each of the recent amendments touched Group A in a specific way, and a practitioner needs the right one. Amendment 03-15 (MSC.393(95)), mandatory from 1 January 2017, created the dedicated Group A iron ore fines schedule and the modified Proctor-Fagerberg test for it, the direct product of the 2009 Indian losses. Amendment 05-19 (MSC.462(101)), mandatory from 1 January 2021, split bauxite into the Group A bauxite fines schedule and the revised Group C bauxite schedule and added the modified Proctor-Fagerberg test for bauxite, the product of the Bulk Jupiter investigation. Amendment 06-21 (MSC.500(105)) widened the Group A definition from liquefaction alone to liquefaction or dynamic separation and defined the latter. Amendment 07-23 (MSC.539(107)) added fourteen new cargo schedules, several of them Group A fines and concentrates, formalised the particle-size basis for treating coal as Group A, and made the bulk density a required item in the shipper’s declaration. Amendment 08-25 (MSC.575(110)) adds further Group A schedules, including apatite concentrate and uncalcined phosphate rock fines, for mandatory use from 2027. The through-line is that the Group A list grows at every cycle, so a cargo that was unlisted or Group C on a past voyage may be a scheduled Group A cargo on the next.
Under SOLAS Chapter VI the shipper must give the master cargo information in writing in good time before loading, and for a Group A cargo that means the TML certificate and the moisture content certificate. The master has not just the right but the duty to refuse loading where the information is missing or invalid, where the declared moisture content reaches the TML, or where the visible state of the cargo contradicts the paperwork. The competent authority of the port of loading approves the laboratories and the test procedures for its jurisdiction. Compliance is also checked downstream: port state control officers under the Paris and Tokyo memoranda examine the cargo documentation and the loading condition, and a missing or invalid Group A declaration, or a cargo loaded above its TML, is a detainable deficiency. The flag state and the ship’s classification society sit behind that, the flag through the statutory regime that gives the Code its force and the class society through the approval of the loading manual and the structural limits the cargo must respect.
Where a cargo is not listed in the Code at all, it cannot simply be loaded. The competent authority of the loading port assesses it, and if it may present a Group A or B hazard a tripartite agreement is concluded between the competent authorities of the loading port, the discharge port and the flag state, setting the conditions of carriage as a bespoke schedule, a copy of which the master must hold before loading.
Documentation and loading controls
The certificates carry time limits that reflect what they measure. The moisture content must be sampled and tested no more than seven days before loading begins, because moisture is a condition that changes; if rain or snow falls between sampling and loading the declaration is void and the cargo must be retested. The TML certificate, measuring a more stable physical property, is valid for up to six months. The shipper’s cargo declaration must state the bulk cargo shipping name, the IMSBC group, the TML and the actual moisture content, and, since Amendment 07-23, the bulk density for a cargo loaded on a bulk carrier of 150 metres or more, to meet SOLAS Chapter XII.
On board, the master verifies that the declared moisture content is below the TML and that the certificates are current and from a recognised laboratory, runs can tests on samples as a cross-check, inspects the stockpiles and rejects visibly wet or slurry-like material, and for the highest-risk trades appoints an independent surveyor. After each hold is loaded the cargo is trimmed reasonably level across the full width, which cuts the void space and the channels that let moisture migrate and reduces the room for a cargo shift. None of these steps is optional for a Group A cargo, and skipping the trim is itself a violation.
The shipper carries the legal responsibility for the declaration, but that does not transfer the risk off the ship. A false or careless certificate does not make a wet cargo safe, and the master who loads on it still loses the ship if the cargo liquefies. This is why the master’s independent checks matter even when the paperwork is in order: the can test, the visual inspection of the stockpile and the loading stream, the watch on the weather, and the readiness to stop. P&I clubs advise their members that a declared group should not be taken on trust where the physical evidence contradicts it, and that a cargo presented as Group C but visibly wet and fine should be treated as a potential Group A cargo until proven otherwise. The cost of a wrong refusal is delay and a commercial dispute; the cost of a wrong acceptance has repeatedly been the ship and her crew.
Why a bulk carrier is so vulnerable
The consequences of liquefaction are so severe partly because of how a bulk carrier is built and loaded. A liquefied cargo creates a free liquid surface across the full breadth of a large hold, and the free surface effect of so wide a body of fluid cuts the effective metacentric height sharply, so the ship loses stability margin even before any cargo physically shifts. As the fluid mass then migrates to the low side it adds a real heeling moment on top of the lost stability, and the two combine into a list the ship cannot recover from. The detail of the free surface and stability is in metacentric height.
A bulk carrier also has little reserve to absorb this. Loaded down to her marks with a dense cargo, she has modest freeboard, large single holds with no internal subdivision to limit the spread of a fluid mass, and a hull form that rolls readily. Once the cargo is fluid there is no quick fix: the water cannot be drained from the bulk in any useful time, the cargo cannot be re-stabilised at sea, and ballast adjustments cannot offset a heeling moment that grows with every roll. The combination of a sudden onset, a wide free surface, a heavy off-centre mass and a ship with little margin is why these casualties are so often total and so often fatal, and why the entire Group A regime is preventive: the safe place to stop a liquefaction casualty is the load port, not the open sea.
Limitations
This article states the IMSBC Code framework for Group A cargoes as it stands through mid-2026; the Code is amended every two years and the schedule for a specific cargo, including its group, its required test method and its carriage conditions, must be read from the edition in force for the voyage. The TML and moisture-content figures are cargo-specific and are set by an approved laboratory for the actual material, not by any general rule, and the can test is a screening check that can condemn but never clear a cargo. The figures and dates here, the 90 percent FMP factor, the seven-day and six-month certificate limits, the 1 January 2025 currency of Amendment 07-23, are the general position and should be confirmed against the current Code text, the cargo schedule and the competent authority’s requirements for the specific load port.
See also
Calculators
Related wiki articles