Wood Pellets: IMSBC Code Schedule and Carriage

Wood pellets are a Group B solid bulk cargo under the IMSBC Code , carried under two Bulk Cargo Shipping Names that split on whether the pellet contains additives or binders, and the hazard that has killed people is not fire but the cargo’s off-gassing, which depletes oxygen and releases carbon monoxide into every space that communicates with the hold.

The Code carries no entry called BIOMASS PELLETS. It carries WOOD PELLETS NOT CONTAINING ANY ADDITIVES AND/OR BINDERS, classified MHB(OH), and WOOD PELLETS CONTAINING ADDITIVES AND/OR BINDERS, classified MHB(WF), both Group B, both at a bulk density of 600 to 750 kg/m3 and a stowage factor of 1.33 to 1.67 m3/t. A cargo declared simply as biomass pellets has not been declared against the Code at all.

Two things about this cargo are commonly stated backwards. The Code imposes a mandatory numeric entry criterion, oxygen restored to 21 percent and carbon monoxide below 100 ppm, on cargo spaces and adjacent confined spaces alike, and that criterion is stricter than the 19.5 percent figure most confined-space training uses. And the Code imposes no prohibition whatever on ventilating a wood pellet hold: the sealed-hold practice is P&I loss-prevention guidance, and this article marks that boundary throughout.

What wood pellets are and how the Code describes them

A wood pellet is densified biofuel: sawdust, planer shavings and other wood waste from lumber manufacturing, fragmented, dried and extruded into a hard cylinder. The Code’s own Description, identical in both schedules, calls them light blond to dark brown in colour, very hard and not easily squashed, with a typical specific density between 1,100 and 1,700 kg/m3, and states that the raw material is compressed to approximately one third of its original volume.

That specific density is the density of the pellet material. The bulk density of the stow, 600 to 750 kg/m3, is roughly half of it, because the voids between cylinders are half the volume of the cargo. Confusing the two is the commonest arithmetic error on this cargo and it doubles the estimated tonnage.

Why the binder is the wood itself

No glue is needed for most wood. The heat and pressure of extrusion through a ring die soften the wood’s own lignin, which sets as the pellet cools and holds the fibers together. That fact governs the whole schedule structure: a pellet made without any added binder is still a pellet made of ruptured, chemically reactive wood cells, so the no-additive schedule is not the safe one. Both schedules carry the same oxidation hazard in the same words.

Milling and compression are what make the cargo reactive. Hammer milling to a flour and extrusion under pressure rupture the cell walls and expose the cytoplasm, resins, oils and fatty acids to air, so the reactive surface area per tonne is far higher in a pellet than in the log it came from.

Grades, and the specifications the trade actually contracts against

The Code gives one size band and one moisture range. The trade works to narrower numbers, and they come from two named documents.

ISO 17225-2:2021, Solid biofuels: Fuel specifications and classes, Part 2: Graded wood pellets, second edition published May 2021, cancels and replaces the 2014 first edition. It draws its permitted raw materials from ISO 17225-1:2021 and it expressly excludes thermally treated biomass pellets, including torrefied pellets, from its scope. That exclusion is why WOOD TORREFIED sits in its own IMSBC schedule with a different MHB set.

ENplus ST 1001:2022, version 8.0, published by Bioenergy Europe and DEPI, is the certification scheme the trade contracts against. Its Table 4 sets, for classes A1, A2 and B respectively, a mechanical durability of at least 98.0, 97.5 and 97.5 percent by weight to ISO 17831-1, moisture at most 10.0 percent as received to ISO 18134, ash at most 0.70, 1.20 and 2.00 percent dry to ISO 18122, fines below 3.15 mm at most 1.0 percent in bulk, and a net calorific value of at least 4.6 kWh/kg as received to ISO 18125, which is 16.5 MJ/kg.

Two of those numbers matter to the ship. The ENplus bulk density band is 600 to 750 kg/m3, which is numerically identical to the IMSBC schedule band, so a certified cargo and the Code agree on the figure that drives hold capacity. And the ENplus additive ceiling is 2.0 percent by weight, of which at most 1.8 percent may be added in production. A pellet certified at that level is a WOOD PELLETS CONTAINING ADDITIVES AND/OR BINDERS cargo under the Code, which is the point at which a quality specification becomes a declaration question.

Where the trade runs

The trade is a coal-substitution story. Power stations built to burn pulverized coal can be converted to burn pulverized wood pellets with limited modification, and the receiving end is usually a dedicated berth with enclosed conveyors and silo storage rather than a general bulk terminal .

Supply is concentrated in two regions. The United States is the largest exporting country, at 10.0 million metric tons in 2024 and 10.09 million in 2025 on USDA Foreign Agricultural Service figures, with production concentrated in the South and export capacity on the South Atlantic and Gulf coasts. Vietnam and Canada follow.

Demand is more concentrated still. The United Kingdom imported 9.641 million metric tons in 2024 and is the largest consuming country, having taken more than two thirds of global imports since 2012, with the United States supplying about three quarters of that volume. EU27 imports were 4.481 million metric tons in 2024 and 4.447 million in 2025 against consumption of about 22.7 million, the balance being produced inside the bloc. Japan and South Korea pull growing volumes under renewable-energy mandates.

The northern European short-sea trade out of Russia and the Baltic states closed to EU buyers in 2022. Council Regulation (EU) 2022/576 of 8 April 2022, the fifth sanctions package, amended Regulation (EU) No 833/2014 to insert Article 3i and Annex XXI, prohibiting the purchase, import or transfer of listed goods including wood products, with pre-existing contracts executable to 4 June 2022. Belarus had been covered a month earlier by Council Regulation (EU) 2022/355 of 2 March 2022. The volume was replaced by longer-haul North American and Southeast Asian supply, which lengthened the average voyage and therefore lengthened the time a sealed cargo sits off-gassing.

The IMSBC schedules for wood pellets

The Code carries two wood pellet schedules and they are identical in every field but one clause of the Hazard cell. Both are Group B, both carry no IMDG class, no subsidiary hazard and no UN number, and both give the same physical particulars. What separates them is the MHB code, OH against WF, and the reason for it is not the reason most sources give.

The two Bulk Cargo Shipping Names

• WOOD PELLETS NOT CONTAINING ANY ADDITIVES AND/OR BINDERS. MHB(OH), other hazards. The Hazard cell says the pellets may ferment over time if the moisture content is over 15 percent, generating asphyxiating and flammable gases, “but gas concentrations do not reach flammable levels. This cargo has a low fire risk.”
• WOOD PELLETS CONTAINING ADDITIVES AND/OR BINDERS. MHB(WF), solids that evolve flammable gas when wet. The same fermentation sentence ends differently: those gases “may cause spontaneous combustion.”

So the split is about whether the fermentation gases reach a flammable concentration, not about additives reacting with water. Both cargoes ferment above 15 percent moisture. Only one is judged capable of reaching a flammable atmosphere from it, and the formal MHB(WF) criterion sits at IMSBC section 9.2.3.4, which turns on a UN Manual of Tests and Criteria gas-evolution test rather than on a chemical argument about binders.

Everything else in the two Hazard cells is word for word the same, and the first sentence is the one that kills: shipments are subject to oxidation leading to depletion of oxygen and increase of carbon monoxide and carbon dioxide in cargo and communicating spaces. Both cells also carry a third sentence the trade often forgets, that handling of wood pellets may cause dust to develop, with a risk of explosion at high dust concentration.

Hazard classification and the MHB framework

Group B is defined at IMSBC section 1.7 as cargoes which possess a chemical hazard which could give rise to a dangerous situation on a ship. That is the whole definition, and it is what puts wood pellets alongside coal , fishmeal , petroleum coke and sulphur rather than alongside a Group A concentrate or an inert Group C mineral.

Materials hazardous only in bulk is the sub-classification the Code applies at section 9.2.3 to a Group B cargo that is not a dangerous good in packaged form. The table at 9.2.3.1.5 gives seven codes: CB combustible solids, SH self-heating solids, WF solids that evolve flammable gas when wet, WT solids that evolve toxic gas when wet, TX toxic solids, CR corrosive solids, and OH other hazards, with OH defined at 9.2.3.1.4 as the residual category for a hazard the other six do not describe.

Wood pellets sit in OH or WF and neither schedule carries SH, despite the cargo self-heating. That looks wrong until the hazard is read: the dominant documented killing mechanism on this cargo is atmospheric rather than thermal, and OH is the category the Code has for a hazard that does not fit the other boxes.

Physical particulars, field by field

Identical on both schedules, from Appendix 1 of the Code as amended by resolution MSC.539(107):

Note the size band against the trade. ENplus permits only 6 plus or minus 1 mm and 8 plus or minus 1 mm diameters and lengths from 3.15 to 40 mm, so a certified industrial pellet sits inside the Code’s diameter band and can run longer than the Code’s length band. Neither figure is a compliance threshold; both are descriptive.

The operational fields are shorter than a practitioner expects:

Three of those cells read “No special requirements”, which means most of what a competent operator does on a pellet voyage is loss-prevention practice and contract rather than Code obligation. Two impose positive duties: Weather precautions and Precautions. The segregation cell is the one the article-length treatments of this cargo usually miss, and it routes the cargo into IMSBC section 9.3 as though it were an IMDG class 4.1 flammable solid .

Which fields bind and which are informative

IMSBC section 1.4.2 makes the “Description”, “Characteristics other than Hazard classification”, “Hazard” and “Emergency procedures” text of an individual schedule recommendatory rather than mandatory. So the bulk density, the stowage factor, the angle of repose and the size in the table above are informative. Section 1.2.1 reinforces it: the schedules’ properties are given only for guidance, and before loading it is essential to obtain current valid information from the shipper.

The mandatory part of the Characteristics table is the Hazard classification block: Class, Subsidiary hazard(s), MHB and Group. Those four cells bind. So does the Precautions cell, which is written with shall throughout.

Amendment history, and what the current set does not change

The two schedules are eleven years old, not new, and their history is worth stating because it settles the declaration question.

Amendment 07-23 is the set in force, mandatory since 1 January 2025, and it governs every voyage to 31 December 2026. Amendment 08-25 was adopted on 26 June 2025, may be applied voluntarily in whole or in part from 1 January 2026, and becomes mandatory on 1 January 2027. Neither wood pellet schedule changes, so a fixture concluded today and a fixture concluded in 2027 are on the same text. The amendment cycle is what a charterer needs to watch on other cargoes, not on this one.

The declaration trap: which schedule your cargo goes under

The two schedules describe a cargo that looks identical at the hatch, so the declaration is the only thing that tells the ship which one applies, and the consequence of getting it wrong is a fire risk assessment made on the wrong basis. This is the practitioner question the schedule structure creates and neither schedule answers.

The obsolete single entry, and why it still appears

A declaration reading simply WOOD PELLETS names a Bulk Cargo Shipping Name that has not existed since 1 January 2017, when amendment 03-15 entered into force. It survives in charter party pro formas, in stowage plan templates, in terminal documentation and in company forms that were written before 2015 and never revised. Treat it as an incomplete declaration rather than a wrong one: the shipper has not yet said whether the cargo is MHB(OH) or MHB(WF).

The fix is a redeclaration before loading, not an assumption. Assuming the no-additive schedule because most pellets contain no added binder is the wrong way round: it is the assumption that removes the spontaneous combustion warning from the ship’s risk picture.

What “additives and/or binders” captures

The Code does not define the phrase, and the two Descriptions differ by one clause: the additive schedule says the raw material is extruded “using appropriate additives and/or binders”. ENplus supplies the commercial line that the trade actually works to, capping additives at 2.0 percent by weight of which at most 1.8 percent may be added in production and 0.2 percent post-production, for example as a coating oil.

That gives a usable test for a fixture. A pellet certified to ENplus with a declared in-production additive is the WF cargo. A pellet certified with none is the OH cargo. A pellet with no certificate is whatever the shipper declares it to be, and the ship’s recourse is the shipper’s cargo declaration and the certificate accompanying it.

Note the different logic the Code applies next door. SUGARCANE BIOMASS PELLETS is drafted so that up to 2 percent of oxide-based mineral additives such as calcium, magnesium and aluminium oxides stays inside the no-additive assumption, on a single schedule. The wood pellet family splits instead. Two adjacent cargoes, two different threshold structures, and no cross-reference between them.

Shipper declaration, documentation and the master’s authority

IMSBC section 4, titled “Assessment of acceptability of consignments for safe shipment”, places eighteen items on the shipper at 4.2.2. For a Group B cargo the load-bearing ones are the Bulk Cargo Shipping Name at .1, the group at .2, the total quantity at .5, the stowage factor at .6, the bulk density where SOLAS regulation XII/10 applies at .7, trimming at .8, and whether the cargo is harmful to the marine environment under MARPOL Annex V appendix I at .17.

What a Group B non-liquefying cargo is relieved of is item .10, the moisture content and transportable moisture limit certificate, which applies “in the case of a group A cargo”. So there is no TML certificate and no transportable moisture limit declaration on a wood pellet cargo, and no liquefaction risk assessment. That relief is worth stating because the 15 percent fermentation threshold in the Hazard cell looks like a moisture limit and is not one: it is a hazard description, not a declared or certified value.

Section 4 contains no power for the master to refuse a cargo. The routes that do exist are IMSBC section 5.1.3, the right to require the cargo be trimmed level where there is any concern about stability; SOLAS regulation VI/7.5, the right to suspend loading where the ship’s structural limits would be exceeded, with notification to the port State; the master’s overriding authority under ISM Code paragraph 5.2; and the contract of carriage. The master’s right to refuse cargo is real and it is assembled from those four, not read off section 4.

Where a consignment does not match any schedule, the route is IMSBC section 1.3 through the competent authority of the load port, and for a cargo that may present Group A or Group B hazards the three-authority procedure at 1.3.1.1 that the industry calls a tripartite agreement . Neither is needed on wood pellets, which have had their own schedules since 2015.

Off-gassing: why the cargo consumes oxygen and makes carbon monoxide

Fresh wood pellets consume the oxygen around them and release carbon monoxide, carbon dioxide, methane and lighter volatile organic compounds, and the two effects together are what make the atmosphere lethal. The driver is auto-oxidation, not combustion and not primarily microbial action in a dry cargo. The UK Health and Safety Executive attributes the carbon monoxide to an auto-oxidation process, in particular oxidation of the fatty acids in wood.

Carbon monoxide blocks oxygen transport by binding haemoglobin in preference to oxygen. An oxygen-depleted atmosphere starves the body directly. A pellet-affected space can be simultaneously below any safe-entry oxygen figure and far above any toxic carbon monoxide threshold, which is why testing for one gas alone has killed people.

Age, species and temperature

The rate is not a constant and the schedule gives no way to predict it. HSE Safety Bulletin OPSTD 3-2012, issued 5 November 2012, records that pellets produce more carbon monoxide within the first six weeks after manufacture, that the rate rises with temperature, and that pellets made from pine produce more than spruce because pine contains more unsaturated fatty acids. The bulletin also names available oxygen, exposed surface area and mechanical abrasion as influences.

The practical reading is blunt. The dangerous cargo is the one that was made last and loaded freshest, in warm weather, in the first days of the voyage. Neither Bulk Cargo Shipping Name distinguishes a parcel pelletised last week from one that has aged two months in a ventilated silo, and no cell of either schedule requires the age to be declared.

Abrasion compounds it. Every transfer point, free fall and grab bite breaks a fraction of the cargo into fines, and fines have more surface per tonne than whole pellets. A cargo handled roughly at the load port arrives at sea more reactive than the same cargo handled gently.

Measured concentrations from instrumented voyages

The numbers come from one study and it has not been superseded in eighteen years. Svedberg, Samuelsson and Melin instrumented five vessels carrying Lodgepole pine pellets from British Columbia to Sweden on voyages of 7 to 9 weeks, and published in the Annals of Occupational Hygiene 52(4):259-266 in 2008.

Three findings from that paper matter more than the headline ranges.

The adjacent staircases were nearly as bad as the holds. The study sampled the staircase beside the hatch on three of the five vessels and found concentrations lower than in the holds but still at very high levels, which the authors read as evidence of fairly free passage of air between the two spaces. That is the instrumented confirmation of the migration the casualty record shows.

The atmosphere becomes dangerous within a week. On one vessel headspace carbon monoxide reached 900 ppm after seven days, with oxygen settling to a steady state near 10 percent after about three weeks. A short voyage on a fresh cargo is not a low-exposure voyage.

Natural ventilation does not clear a space. A decay test on one vessel, monitoring carbon monoxide at 2.5-minute intervals for two hours after the deck access door was opened, gave 0.23 air exchanges per hour, from which about 26 hours would be needed to bring the space down to 35 ppm. That single figure is the quantitative reason an open door is not an entry control.

Communicating spaces: where people actually die

People do not die in the hold of a wood pellet ship. They die in the forepeak, the bow thruster room, the lashing locker, the access stairway and the duct space, in compartments that look and feel like normal ship spaces and are not associated with the cargo at all. Every documented marine fatality on this cargo has that shape, and it is the single most important carriage fact about it.

Resolution MSC.581(110), the current IMO recommendations for entering enclosed spaces, now supplies the vocabulary for it. A connected space shares air with another space by a permanent or temporary route, including a manual door even where the door is watertight, and is to be treated as containing a hazardous atmosphere until testing proves otherwise. An adjacent space shares only a boundary, with no openings whatsoever. Section 10.3.2 of the same resolution describes the pellet-ship geometry precisely: where cargo-space doors, ventilation trunks and pipework run into forecastle workspaces, store rooms, windlass hydraulics and bow thruster rooms, those working spaces should be identified as connected spaces.

On that vocabulary, every wood pellet casualty was in a connected space, not an adjacent one. Section 10.3.3 of the resolution lists the cargoes that have caused fatal accidents by asphyxiation, explosion and fire, and item .3 is wood chips and wood pellets.

The migration paths, from the casualty findings

The three documented gas paths were each a defect a deliberate pre-loading inspection could have found.

• On the AMIRANTE, the door between the forepeak and the forward cargo hold had three closing hinges and only two were engaged.
• On the LADY IRINA, gas reached the bow area through a small gap at the rim of an inspection cover on the ventilation duct running from the cargo hold through the forecastle. The path was traced afterwards by a classification society engaged by the owner.
• On the CORINA, an opening 2,100 mm by 1,150 mm in the forward bulkhead of cargo hold no. 1 was closed only by planks slotted into guide bars, which is not a gastight barrier.

None of these is a large opening, and none needs to be. Carbon monoxide has a relative density of 0.97, recorded in MSC.581(110), so it is very slightly lighter than air and distributes through a space rather than pooling cleanly at the deck or the deckhead. A single sample point at one level can miss it.

The documented casualty record

The three casualties in Danish waters were brought together by Frank Huess Hedlund and Oessur Jarleivson Hilduberg in a 2017 chapter for IntechOpen. None was investigated by the Danish Maritime Accident Investigation Board: the investigating bodies were a police force and two foreign flag States.

AMIRANTE, IMO 7425334, St Vincent and the Grenadines flag, 4,083 GT, carrying 2,600 tonnes of wood pellets from Riga to Copenhagen. On 15 July 2009, in the Baltic Sea about 16 km north of Bornholm, an able seaman and a motorman entered the forepeak and died. Autopsy carboxyhaemoglobin was 52 and 60 percent. Carbon monoxide measured 100 ppm in the forepeak about 40 minutes after ventilation was switched off, with oxygen normal. Investigated by Bornholm Police as a closed criminal case file.

LADY IRINA, IMO 9137038, Netherlands flag, 3,323 GT, on passage from Arkhangelsk. On 13 July 2014 the chief engineer entered the forecastle compartment and the bow thruster room beneath it and died; the chief officer and two seamen were injured. The ship diverted to Fredericia, Denmark, and berthed at 2250, after the casualty. The forecastle door had been shut at least 26 hours and there was no mechanical ventilation to the bow area. Readings taken after 36 hours closed gave 690 ppm carbon monoxide in the forecastle and 555 ppm in the bow thruster room, and an unventilated store holding the fixed carbon dioxide fire-extinguishing bottles read above 2,000 ppm carbon monoxide, which was the instrument’s ceiling rather than a measurement. Investigated by the Dutch Safety Board.

CORINA, IMO 8908545, Poland flag, 5,796 GT. On 28 April 2015 at Hanstholm, Denmark, discharging 1,921 tonnes of a 6,744 tonne cargo, an able seaman entered the lashing equipment room in the forepeak and died, with autopsy carboxyhaemoglobin at 60 percent. Four crew members and a Danish port officer were hospitalised. Carbon monoxide measured 66 ppm at the entry door and 366 ppm at lower levels, both after mechanical ventilation had already been started. The shipper had declared the cargo correctly, as Group B MHB with an entry warning, and the entry regime failed anyway. Investigated by the Polish State Maritime Accident Investigation Commission as final report WIM 13/1.5, Warsaw, 2016.

Two earlier port casualties predate that set. A stevedore died at Rotterdam in 2002 during pellet discharge, reported in the Svedberg paper with several others injured. At Helsingborg, Sweden, in November 2006, a seaman was killed and a stevedore seriously injured after entering an unventilated stairway beside a hold discharging British Columbia pellets; the stevedore reached 43.8 percent carboxyhaemoglobin on hospital admission and was left with permanent neurological damage, and several rescue workers were slightly injured. No pre-entry test had been taken, contrary to the vessel’s usual practice. That casualty is what prompted the Svedberg study.

The pattern did not stop in 2015. On 25 July 2025, at Berth MB5 in Daesan Port, Republic of Korea, the Marshall Islands flag bulk carrier AVRA lost her Bosun during discharge. He entered the cargo hold no. 4 aft access hatch space, a space connected to the hold, without complying with the company’s enclosed-space entry procedures, and was found collapsed at about 1033. The casualty is under investigation by the Republic of the Marshall Islands Maritime Administrator. It happened ten years after the CORINA, nine years after SOLAS regulation XI-1/7 came into force, and a month after MSC.581(110) was adopted.

The rescuer toll

Untrained rescue entry turns one fatality into several. At Helsingborg the injured included rescue workers. On the CORINA five people were hospitalised after one man collapsed, among them a port officer who was not part of the crew. In Ireland in November 2010 a householder died in a 7 tonne pellet store and his wife and another man needed hospital treatment after attempting to reach him.

The Code addresses this directly. IMSBC section 3.2.6 restricts emergency entry into a cargo space to trained personnel wearing self-contained breathing apparatus and protective clothing, always under the supervision of a responsible officer. It is written with shall.

Enclosed-space entry: the Code’s own criterion

The IMSBC wood pellet schedules impose a mandatory, numeric, two-gas entry test, and it is stricter than the figure most confined-space training uses. The Precautions cell of both schedules reads:

Entry of personnel into cargo and adjacent confined spaces shall not be permitted until tests have been carried out and it has been established that the oxygen content and carbon monoxide levels have been restored to the following levels: oxygen 21% and carbon monoxide < 100 ppm. If these conditions are not met, additional ventilation shall be applied to the cargo hold or adjacent confined spaces and re-measuring shall be conducted after a suitable interval.

The same cell adds that an oxygen and carbon monoxide meter shall be worn and activated by all crew when entering cargo and adjacent enclosed spaces. That is a mandatory personal-monitoring requirement written into the cargo schedule, and it is separate from the pre-entry test.

Two things follow that the general enclosed space entry regime does not supply. The criterion covers adjacent confined spaces as well as the hold, so the forepeak and the bow thruster room are inside it. And the remedy for a failed test is named: ventilate, then re-measure. A sealed-hold policy with no provision for ventilating before entry does not comply with its own schedule.

The instruments the ship must carry

SOLAS regulation XI-1/7, introduced by resolution MSC.380(94), adopted 21 November 2014 and in force from 1 July 2016, requires every ship to which SOLAS chapter I applies to carry an appropriate portable atmosphere testing instrument or instruments. As a minimum these shall be capable of measuring concentrations of oxygen, flammable gases or vapours, hydrogen sulphide and carbon monoxide before entry into enclosed spaces, and suitable means shall be provided for calibration.

Four points a pellet ship should read off that regulation. It names four gases, not two, and carbon dioxide is not among them, which is a real gap on an organic cargo. It specifies no number of instruments; the “at least two sets” figure comes from MSC.581(110) paragraph 7.3 and is a recommendation. It is a pre-entry test made from outside the space, so a personal detector worn inside does not satisfy it, a distinction MSC.1/Circ.1477 of 9 June 2014 makes explicit in its selection guidance. And the calibration clause is mandatory, which is what makes an uncalibrated detector a deficiency against a convention requirement and therefore within the scope of port state control under SOLAS regulation I/19.

Range matters as much as calibration. Measured hold concentrations run to 14,650 ppm, and in the LADY IRINA casualty the reading in the carbon dioxide bottle store was recorded only as above 2,000 ppm because the instrument had no more scale. A detector that pegs tells the user the atmosphere is bad and nothing about how bad, which is exactly the information needed to decide whether another hour of ventilation is enough.

The recommendatory layer, and the footnote that points at a revoked instrument

Both Precautions cells carry a footnote referring the reader to the Revised recommendations for entering enclosed spaces aboard ships, resolution A.1050(27). A.1050(27) has been superseded. The current instrument is resolution MSC.581(110), adopted on 27 June 2025, whose operative paragraph 3 invites the Assembly to revoke A.1050(27). The Code text predates it, so a ship that follows the footnote is following a superseded resolution.

MSC.581(110) differs from its predecessor on exactly the points this cargo turns on. Its paragraph 7.4 sets entry criteria of 20.9 percent oxygen by volume, carbon dioxide below 0.5 percent by volume which is 5,000 ppm, below 1 percent of the lower flammable limit where flammables are possible, and below 50 percent of the occupational exposure limit of any toxic vapour or gas, with a note that national requirements may differ. Section 3.1 requires a ship-specific Enclosed Space Register listing every enclosed space with its connected and adjacent spaces, its hazards and how the atmosphere may change with the cargo carried.

Section 10.5 is written for organic cargoes and reads as though it were written for this one. Paragraph 10.5.4 warns that existing oxygen depletion tables may be inadequate because oxygen may remain at 17 to 14 percent when carbon dioxide has reached 4 percent or more. Paragraph 10.5.6 states that death caused by carbon dioxide can precede death resulting from oxygen deficiency in organic types of cargoes. Paragraph 10.5.7 says ships carrying organic solid cargoes in bulk should continue to test for carbon dioxide before entry and frequently thereafter.

Appendix 5 of the resolution carries the danger tables, derived from submissions CCC 9/8/3, CCC 9/INF.9 and CCC 9/INF.10. It records 35 ppm as the workplace exposure limit reference for carbon monoxide, 400 ppm and 800 ppm as intermediate danger levels, 1,200 ppm as immediately dangerous to life or health, and the observation that oxygen may still read 17 to 14 percent when carbon monoxide has reached 1,200 ppm. It adds that death from carbon monoxide can precede death from oxygen deficiency in organic types of cargo.

Exposure limits, and which jurisdiction each belongs to

An occupational exposure limit is set by a labour authority, not by IMO, so there is no single global number and an article that gives one is naming a national rule as though it were the world.

The 19.5 percent oxygen figure is the one to be careful with, because it is the number most confined-space training teaches and it is 1.5 points below what this cargo’s own schedule requires. Entering a pellet-affected space at 19.5 percent means entering below both the mandatory schedule criterion and the IMO recommendation.

Carboxyhaemoglobin is the measure that appears in the casualty reports rather than in any rule. Svedberg gives normal saturation at 0.3 to 0.8 percent and smokers at 3 to 8 percent, and records that most people who die of carbon monoxide poisoning have above 60 percent. Clinical outcome correlates poorly with saturation, and the three autopsy values in the marine record are 52, 60 and 60 percent.

Ventilation: what the Code requires and what the clubs advise

The IMSBC Code says nothing about ventilating a wood pellet hold. The Ventilation cell of both schedules is one sentence, it concerns adjacent spaces, and it is permissive:

Ventilation of enclosed spaces adjacent to a cargo hold before entry may be necessary even if these spaces are apparently sealed from the cargo hold.

The Carriage cell reads “No special requirements”. There is no prohibition on ventilating the holds anywhere in either schedule, and there is no mandate to ventilate them either.

The proof that this is deliberate sits in the next schedule along. SUGARCANE BIOMASS PELLETS, inserted by the same amendment set that last touched the wood pellet Descriptions, has a Ventilation cell that begins: cargo spaces carrying this cargo shall not be ventilated during the voyage. The Code knows exactly how to write a no-ventilation mandate. It wrote one there and chose not to write one here.

The sealed-hold practice, and its three separate rationales

The prevailing carriage practice is nonetheless to seal the holds, and it rests on three different arguments from three different sources, none of them an IMO requirement.

The fire argument is the P&I one. Skuld, in Wood pellets and risk of fire published on 2 May 2019 with Brookes Bell, advises that it is generally advisable to carry wood pellets without applying any ventilation, so that any self-heating tendency and its effects are self-limiting: as the oxygen in the headspace is consumed, the oxidation slows for want of an oxidiser. The same guidance recommends voluntary oxygen, carbon monoxide and flammable-gas records similar to those kept on coal, and sealing all holds on smoke, carbon monoxide above 100 ppm or high temperatures.

The condensation argument comes from bulk pellet safety data sheets, which warn against force-ventilating where the ambient air temperature exceeds the pile temperature or where ambient humidity is high, because the result is wetting of the cargo surface. That risk is highest when discharging into a low ambient temperature.

The cargo-quality argument is the Code’s own. The Weather precautions cell requires the cargo to be kept as dry as practicable, and a pellet that takes on free water swells, loses durability and moves from chemical oxidation toward the harder fermentation regime above 15 percent moisture.

Extending the fire argument to the toxic-gas case, and saying that ventilating the holds would sustain the carbon monoxide production, is a reasonable inference and it is not what Skuld says. Treat it as reasoning rather than as guidance, and note the tension the schedule itself creates: the Precautions cell requires additional ventilation to be applied to the cargo hold where the entry criteria are not met. Sealed at sea and ventilated before entry is the reconciliation, and it needs to be written into the procedure rather than assumed.

Whichever way the hold policy falls, the rule for the connected spaces does not move. They are force-ventilated and re-tested before every entry, because that is where people go.

The coal comparison, and the mental model that gets it backwards

Coal is the cargo most crews reach for as the analogue, and the two differ on both the gas and the paperwork.

Two corrections to the usual telling. The coal default after the first 24 hours is also a closed hold, so “coal is ventilated and pellets are sealed” is wrong as a general statement; and for coals liable to self-heat the spaces are closed immediately, with only natural surface ventilation limited to the minimum time needed to remove methane. The general rule behind both cargoes is IMSBC section 3.5.6: where a cargo may heat spontaneously, ventilation other than surface ventilation shall not be applied, and on no account shall air be directed into the body of the cargo.

The real contrast is the paperwork. Coal has a monitoring and ventilation procedure written into its schedule appendix. Wood pellets have none, and their Carriage cell reads “No special requirements”. That is the strongest argument for monitoring a pellet voyage as though it were mandatory: the Code’s silence is not a finding that the cargo is safe, it is the absence of a rule.

Self-heating, fermentation and fire

Self-heating is the secondary hazard on wood pellets and the Code says so in as many words on the no-additive schedule: gas concentrations do not reach flammable levels and the cargo has a low fire risk. The atmospheric hazard usually becomes lethal long before the thermal hazard becomes a fire.

Three heat paths contribute. Chemical oxidation of cellulose, lignin and fatty acids produces the same heat as it produces carbon monoxide. Fermentation adds a microbial path where moisture supports it, which the Code puts above 15 percent moisture content. And a deep, compacted, sealed stow retains heat in its core faster than it conducts away.

The 15 percent line is where the two schedules part. On the additive cargo, fermentation gases may cause spontaneous combustion; on the no-additive cargo they do not reach flammable levels. That is a difference in fire outcome from the same trigger, and it is why moisture exclusion is a safety control on this cargo rather than only a quality one.

The schedule’s own fire response is short and it is the same on both entries: batten down, use the ship’s fixed fire-fighting installation if fitted, exclusion of air may be sufficient to control the fire, and extinguish with carbon dioxide, foam or water. Note that water is permitted, which is not the case on every self-heating cargo, and that the fixed gas fire-extinguishing system is the first-named tool. Special emergency equipment is self-contained breathing apparatus and combined or individual oxygen and carbon monoxide meters, which the Code says should be available rather than shall.

Neither schedule imposes an age limit or a cure period before loading. The Code does impose one on a comparable cargo: CHARCOAL must be weathered for not less than 13 days before shipment, with a manufacturer or shipper statement. Any pellet ageing requirement on a fixture is therefore a contractual or terminal standard, and the ship should read it as such rather than as a Code obligation it can rely on.

Gas and temperature monitoring during the voyage

Neither wood pellet schedule requires any monitoring during the voyage. The regime that competent operators run is loss-prevention practice, and the casualty record is the reason it exists.

The parameters are oxygen and carbon monoxide, which are the two the schedule’s entry criterion names, with carbon dioxide added on the authority of MSC.581(110) paragraphs 7.4.2 and 10.5.7 for an organic cargo. Readings are taken from sealed sampling points on each hold without entering it, most frequently in the first week when off-gassing peaks, and the time and sampling point are logged with the value.

The trend is the signal, not the number. A steady high carbon monoxide reading in a sealed pellet hold is expected and is not by itself evidence of anything wrong. A sharp rising trend, or a temperature climb where the ship can measure temperature, points at active self-heating. That is the opposite of the instinct a crew brings from an inert cargo, and it is the point most easily lost.

The connected spaces are sampled too, and this is the part the casualty record argues for. A clean hold log says nothing about the bow thruster room, and the Weather precautions cell of both schedules states plainly that there is a high risk of renewed oxygen depletion and carbon monoxide formation in previously ventilated adjacent spaces after closure of the hatch covers. A space cleared once does not stay cleared.

Temperature monitoring runs in parallel where the ship has the means. Older bulk carriers often lack fixed hold-temperature probes, in which case Skuld’s guidance is that temperature can be taken through sounding pipes, with the caveat that a sounding pipe reading says little about temperatures elsewhere in the hold. Newer designs increasingly carry fixed sensors. The logs go to the operator and the shipper for trend analysis and are the record that demonstrates the master discharged the duty of care if a casualty inquiry follows.

Hold preparation and the gas-tightness inspection

Hold preparation for wood pellets follows the general cargo hold preparation standards with two pellet-specific emphases: dryness, and the gas-tightness of every boundary between the hold and a space a person enters.

The schedule’s own requirement is minimal. Hold cleanliness reads “Clean and dry as relevant to the hazards of the cargo”, which is a lower bar than the grain-clean standard many pellet charters specify, so a hold cleanliness certificate on this cargo is answering the contract rather than the Code. Residues from a previous cargo matter for contamination and for claims; moisture matters for safety, because a wet hold accelerates both swelling and fermentation.

Bilge wells are cleaned and tested clear so that water reaching the bottom of the stow can be detected and pumped, and so that the bilge system does not become a gas path into the engine room.

The gas-tightness check is the preparation step the casualty record argues for most strongly, and no schedule requires it. Before loading, the crew identifies and where possible seals the paths by which off-gas can reach an adjacent compartment: hatch cover seals, access trunk and duct covers, inspection-cover rims, and any bulkhead penetration between the hold and a forepeak, bow thruster room, store or stairway. The three known casualty paths were a door hinge, a duct inspection-cover rim and a plank barrier. Where a barrier cannot be made gastight, the space behind it is designated as connected to the cargo, signed, and brought into the entry regime for the whole voyage.

Hatch cover weathertightness testing serves both directions here. It keeps seawater out of a cargo the Code requires to be kept as dry as practicable, and it keeps off-gas inside the hold rather than venting onto deck where it can be drawn into accommodation or machinery intakes. Note that weathertight is not gastight, and the schedule asks for neither: this is practice built on the hazard, not on the cell.

Access doors to connected spaces are labelled with the hazard and access-controlled. The LADY IRINA’s forecastle access door already carried a sign reading danger, low oxygen level, and the chief engineer died behind it, which is the limit of what signage alone achieves.

Loading, trimming and hold capacity

Loading is by enclosed shore conveyor and shiploader with dust suppression , feeding the cargo into the hold with as little free fall and breakage as the terminal arrangement allows. Breakage costs twice: fines off-gas faster because they carry more surface per tonne, and airborne dust is the explosion hazard both Hazard cells name.

Trimming, and the crossover into the grain regime

The Loading cell says trim in accordance with sections 4 and 5 of the Code, and that is not a trimming exemption. Which part of section 5 applies is decided by the angle of repose, and for wood pellets the route is unusual.

Both wood pellet Bulk Cargo Shipping Names appear on the Appendix 3 paragraph 1.1 list of cargoes that are non-cohesive when dry, alongside SUGARCANE BIOMASS PELLETS and WOOD TORREFIED. Most mineral schedules fall instead to the paragraph 1.3 default, which treats an unlisted cargo as cohesive and takes the quantitative provisions out of play. Here they are in play.

Section 5.4.2 makes the angle of repose decide which provisions apply, and the schedule gives approximately 30 degrees, which sits exactly on the boundary:

• Section 5.4.3, at or below 30 degrees: these cargoes, which flow freely like grain, shall be carried according to the provisions applicable to the stowage of grain cargoes, footnoted to SOLAS chapter VI and the International Code for the Safe Carriage of Grain in Bulk, resolution MSC.23(59). The bulk density is taken into account in determining the scantlings and securing arrangements of divisions and bin bulkheads, and the stability effect of free cargo surfaces.
• Section 5.4.4, above 30 and up to 35 degrees: trim so that the vertical distance between the highest and lowest levels of the cargo surface does not exceed one tenth of the breadth, with a maximum of 1.5 m, or load with trimming equipment approved by the competent authority.

Appendix 3 paragraph 1.2 requires the angle of repose to be determined before loading is completed, precisely so this question is settled before the ship sails. Saying that the cargo self-trims and therefore needs nothing is the wrong conclusion: at approximately 30 degrees it flows freely like grain, which is what puts it into the grain provisions rather than out of any provision. Section 5.1.3 also stands unchanged, giving the master the right to require the cargo be trimmed level wherever there is any concern about stability, and the self-trimming bulk carrier hull form does not displace it.

Weather and water exclusion

Two Weather precautions obligations bind during loading and both are written with shall. The cargo shall not be handled during precipitation. And during handling, all non-working hatches of the cargo spaces into which the cargo is loaded, or is to be loaded, shall be closed. Neither is advisory, and both belong in the loading plan and in the ship-shore agreement under the BLU Code .

Hold capacity, and whether the ship cubes out

The relationship between bulk density and stowage factor decides how a pellet cargo fills the ship, and the two are reciprocals once unit-converted. A hold of net volume 12,000 m3 at the mid-range 675 kg/m3 holds about 8,100 tonnes; a five-hold handysize of 60,000 m3 total holds about 40,500 tonnes, well inside the deadweight of a ship that size.

The density that enters the calculation is the as-stowed bulk density, not the loose poured density at the conveyor and certainly not the 1,100 to 1,700 kg/m3 specific density of the pellet material. The relationship also breaks where the stow is not uniform. broken-stowage at the hold ends and around structure removes a few percent of usable volume the simple form ignores; settling over the voyage shows up as a drop in the cargo surface rather than as extra capacity, because the mass was fixed at loading; and moisture uptake invalidates the range entirely, since swollen pellets occupy more volume and disintegrated ones can pack denser. A draught survey at each end remains the quantity of record.

Closing up

The most dangerous moment of the loading evolution for people is the closing-up. Off-gassing begins as soon as the hold is sealed, and the connected spaces start to fill. Spaces the crew walked through an hour earlier are on the way to becoming lethal, and the Code says so in the Weather precautions cell: there is a high risk of renewed oxygen depletion and carbon monoxide formation in previously ventilated adjacent spaces after closure of the hatch covers.

The entry regime is in force from that moment until the cargo is discharged and the spaces are proven clear by test. Before the ship sails, the declaration and stowage plan are signed off, the monitoring log is opened, and the access control and signage are set.

Discharge, residues and post-discharge entry

Discharge is preferably by enclosed continuous unloader delivering into shore silos, which limits both dust and the escape of carbon monoxide into the working area. Grab discharge breaks more pellets, raises more dust and releases more gas where stevedores work. Two of the six documented casualties happened during or around discharge, and one of the dead and one of the injured were shore workers rather than crew, so the hazard briefing owed to visiting stevedores is not a formality.

The hazard does not end when the hold looks empty. Residual pellets, fines and dust in the hold corners, the hopper and the unloader continue to oxidise, and the schedule’s entry criterion of 21 percent oxygen and carbon monoxide below 100 ppm does not stop applying because the bulk of the cargo has gone. Entry for inspection or cleaning after discharge is enclosed-space entry, tested and ventilated, and the same applies to the connected spaces, which do not clear themselves when the hold does.

Post-discharge cleaning removes fines and dust before the next cargo. Pellet residue is combustible and dusty, so cleaning is a fire and dust-explosion control as well as a contamination control. Where the next cargo is moisture-sensitive or food-grade, the holds are washed, dried and inspected to the standard that cargo demands, with attention to drying because residual wash water reintroduces the moisture problem for any remaining fines. Whether the residues are harmful to the marine environment for MARPOL Annex V purposes is an item the shipper declares under IMSBC 4.2.2.17, and it governs what may be washed overboard and where.

The sibling wood and biomass schedules, compared

Wood pellets sit in a family of six Code entries that share the oxygen-depletion mechanism at different intensities and diverge sharply on fire. The figures below are from the 07-23 consolidated text.

Three of those rows carry a trap for a ship that alternates between them.

SUGARCANE BIOMASS PELLETS is the stricter cargo, not the looser one. Its Ventilation cell prohibits ventilating the cargo spaces during the voyage, its Carriage cell requires the hatches to be weathertight against the ingress of water, and its Precautions cell adds ignition-source control over hold lighting, requiring hot halogen lamps to be kept away and their fuses removed or secured while the cargo is present. Its Description also permits up to 2 percent of oxide-based mineral additives inside the no-additive assumption, which is a different threshold logic from the wood pellet split.

WOOD TORREFIED changes the oxygen number. The entry criterion is 20.7 percent rather than 21 percent, which is a trap for a crew that has memorised one figure. Torrefaction, roasting the biomass to a coal-like product, moves the hazard toward combustibility and self-heating, which is why the MHB set is CB, SH and CR and why ISO 17225-2:2021 excludes thermally treated pellets from its scope.

WOODCHIPS runs the opposite way on moisture. Its off-gassing hazard is stated as carbon dioxide rather than carbon monoxide, its entry test is oxygen only at 20.7 percent, and its fire risk is low at 15 percent moisture or above and increases as the cargo dries, to the point where dry chips are readily combustible and can ignite by friction. Its schedule also carries a line harder than anything in the pellet schedules: a condition with complete depletion of oxygen may be present in less than 48 hours.

There is no schedule called PULP WOOD CHIPS. PULP WOOD is routed by the Appendix 4 index to the Wood products General schedule, alongside LOGS, TIMBER, SAW LOGS and ROUNDWOOD, which applies only to wood products loaded and discharged in bulk by such means as elevator or grab.

Port state control and the regional overlays

The IMSBC Code is mandatory through SOLAS regulation VI/1-2, which requires the carriage of solid bulk cargoes other than grain to comply with its relevant provisions, and SOLAS chapter VI is the chapter a port state control officer works from. What that means on a pellet ship is narrower than the length of this article suggests, because so much of the good practice above is not a rule.

The items that are surveyable against an instrument are the portable atmosphere testing instrument and its calibration means under SOLAS regulation XI-1/7; the record of the two-monthly enclosed space entry and rescue drill under SOLAS regulation III/19.3.3 and III/19.5, introduced by resolution MSC.350(92) and in force from 1 January 2015; the shipper’s cargo declaration and its content against IMSBC section 4; the entry procedures in the safety management system, which the ISM Code brings within audit through paragraphs 6.4, 7 and 12.3; and the loading plan and its agreement with the terminal under SOLAS regulation VI/7.

Beyond the IMO baseline, the overlays are labour rather than maritime. The occupational exposure limit that applies to a person working in a pellet-affected space is set by the flag State’s national law and, in port, by the law of the place of work. That is why the carbon monoxide limit in the table above ranges from 20 ppm in the European Union to 50 ppm in the United States for the same gas and the same eight hours. A company standard that adopts the strictest of the set avoids having to answer the question per port.

The trade overlay is the sanctions position on Russian and Belarusian origin, under Council Regulation (EU) 2022/576 and Council Regulation (EU) 2022/355, later narrowed by Regulation (EU) 2022/1269 to permit third-country transfers only of fuel wood under CN heading 4401 and charcoal under CN 4402. That is an origin question for the charterer and the bill of lading , not a carriage question for the master, but it decides which load ports are available.

Limitations

This article is a carriage and regulatory reference, not a substitute for the IMSBC Code text, the shipper’s declaration, or a competent person’s risk assessment of the specific consignment and ship. The schedule parameters quoted here are the published values, and under IMSBC section 1.4.2 the physical particulars are informative rather than mandatory; the values that govern a given cargo come from the declaration and the certificate accompanying it.

The boundary between Code obligation and industry practice is marked throughout because it is unusually wide on this cargo. Three of the wood pellet schedule’s operational cells read “No special requirements”, and the sealed-hold practice, the voyage monitoring regime, the pre-loading gas-tightness inspection and any pellet ageing limit are all loss-prevention practice or contract rather than Code requirements. Nothing here should be read as asserting that the Code requires them.

The off-gassing rate depends on species, age, temperature, moisture and handling in ways no single figure captures. The concentrations quoted are the measured ranges from five specific voyages in one 2008 study, not predictions for any cargo. That study has not been superseded for shipboard measurement, which is a statement about the state of the literature rather than about the reliability of the numbers for a different trade lane, a different species or a different voyage length.

The casualty record here is the documented public record from a peer-reviewed paper, an academic synthesis of police and flag-State files, and a national regulator’s safety bulletin. It is not exhaustive, and the absence of an incident from this account does not make a space or a practice safe. The AVRA casualty of 25 July 2025 is under investigation and only its outline facts are stated; the investigation report should be read before anything more is relied on.

The exposure limits tabulated are national and regional and they change. The oxygen and carbon monoxide entry criterion in the schedule is the figure that binds a ship carrying this cargo, and it is the only one of the set that does.

The hold capacity relationship gives a planning estimate that ignores broken stowage and assumes a uniform, dry, undegraded cargo. It is not a stability calculation, and intact and damage stability, the grain provisions that section 5.4.3 routes this cargo into, load-line and freeboard, and the structural limits of the bulk carrier are separate determinations it does not address.

Above all, no calculation and no schedule lookup makes an enclosed space safe. The recurring cause of death on this cargo is entry into a space that was not recognised as communicating with the hold, tested for the wrong gas or not at all, and ventilated by an open door rather than a fan. The controlling safeguards are physical and procedural: identify every connected space before loading, test both oxygen and carbon monoxide immediately before every entry, force-ventilate, and never enter to rescue without self-contained-breathing-apparatus and a tended lifeline.

Related Articles

• The IMSBC Code
• IMSBC Group B Cargoes
• Materials Hazardous Only in Bulk
• Enclosed Space Entry
• Marine Confined Space Entry and Tank Inspection
• Marine Cargo Hold Ventilation
• Coal: IMSBC Code Schedule and Carriage
• Cargo Hold Preparation Standards
• Angle of Repose of Solid Bulk Cargoes
• Bulk Carrier
• Petroleum Coke: IMSBC Code Schedule and Carriage
• Brown Coal Briquettes: IMSBC Code Schedule and Carriage
• Fishmeal: IMSBC Code Schedule and Carriage