RESEARCH DOSSIER
Marine Cargo Packaging Stowage and Loss Prevention
Abstract. Marine cargo damage develops through interactions among product vulnerability, primary packaging, load-unit formation, cargo transport unit condition, inland handling, shipboard stowage, ocean motion, contractual performance, and insurance coverage. A cargo consignment can meet a container's nominal mass rating while remaining inadequately restrained internally; it can also be packed competently yet encounter a shipboard stack collapse caused by deficient lashings, stow-plan geometry, severe motions, or structural weakness elsewhere in the stack. These mechanisms must be distinguished when evaluating prevention measures and determining the significance of evidence for cargo insurance and carrier liability. This monograph synthesizes the 2014 IMO/ILO/UNECE Code of Practice for Packing of Cargo Transport Units, SOLAS requirements for verified gross mass and shipboard securing, container structural standards, marine cargo insurance wording, the Hague–Visby liability architecture, official accident investigations, and selected experimental and numerical research. It examines the evidentiary significance of the Ever Smart and CMA CGM G. Washington container losses and the timing of insurance attachment in the Australian Swashplate litigation. The 2026 International Maritime Dangerous Goods Code and mandatory container-loss reporting amendments are distinguished from advisory guidance and from revision proposals. The analysis establishes the need to assess packaging, in-container securing, container suitability, and shipboard securing as related but separately observable controls. It shows why compliance documents, survey evidence, and actual securing geometry must be reconstructed in sequence; why aggregate containers-lost-at-sea estimates cannot identify the effect of an individual packing intervention; and how prevention economics must account for both incident frequency and severity without mistaking illustration for observed causal evidence. The contribution is an original cross-disciplinary interpretation of established technical, legal, and empirical materials, rather than a new scientific theory or proprietary risk model.
Keywords: Marine cargo insurance; packaging adequacy; cargo transport units; CTU Code; cargo stowage; cargo securing; container loss prevention; verified gross mass; Cargo Securing Manual; inherent vice; Institute Cargo Clauses; Hague–Visby Rules; marine accident investigation; risk engineering; claims causation; transport economics.
Introduction
The scientific object of marine cargo packaging, stowage, and loss prevention is the physical preservation of consignments during successive transport operations and the allocation of the resulting economic loss when preservation fails. This object includes the packaging that protects the goods themselves, the methods by which packages and heavy items are restrained inside a cargo transport unit, the structural condition and loading of that unit, and the shipboard organization and securing of containers and other cargo units. It extends to handling at the places where responsibility passes between packers, road haulers, rail operators, terminals, ocean carriers, receivers, surveyors, cargo owners, and insurers.
The IMO/ILO/UNECE CTU Code supplies an unusually useful starting point because it covers the intermodal chain and addresses both packing and unpacking. The Code was endorsed in 2014 and is expressly nonmandatory at the international level. Its relevance to a specific claim depends on its technical content, the facts of the operation, and any contractual or domestic legal instrument that incorporates it. Treating the Code itself as a universal statute would misstate its authority. Separate mandatory requirements arise under SOLAS, the International Convention for Safe Containers, dangerous-goods instruments, and implementing law.
Marine cargo insurance introduces another set of questions. The insurer's coverage depends on the particular policy, insured interest, attachment and termination of risk, exclusions, causal sequence, and governing law. The contract of carriage introduces duties and exceptions of its own. A factual finding of insufficient packing therefore has at least three distinct possible meanings: a preventable engineering condition, a reason for denying or limiting carrier liability, and a circumstance engaging an insurance exclusion. Their legal and evidentiary consequences need not coincide.
The analytical method used here combines normative comparison, accident-process reconstruction, scholarly evidence appraisal, and transparent illustrative loss calculations. The source base prioritizes official IMO and government publications, published standards metadata, the UK legislative text, original accident-investigation summaries, insurer explanation of the 2009 Institute Cargo Clauses, contemporaneous appellate commentary, scientific articles, and the shipping industry's published loss series. Primary technical and statutory materials determine the stated obligation or guidance; secondary studies help explain mechanisms and constraints. No incident-rate regression, private claims dataset, field trial, or original lashing experiment is claimed.
The 2026 WSC container-loss report estimates 1,478 containers lost in 2025, compared with 576 in 2024, and links the 2025 total heavily to a single large incident. This is useful evidence of event concentration, not a direct estimate of the probability that an individual insured package suffers damage. Many cargo claims arise from moisture, handling, crushing, temperature deviation, theft, or inland transit without any container passing overboard. Conversely, one lost container can contain multiple consignments and generate liabilities beyond the insured invoice values.
Scientific Concepts and Historical Origins
The protected object and its interfaces
A package is the immediate material arrangement protecting an item or group of items against expected mechanical, climatic, handling, and contamination stresses. A palletized load unit combines packages for efficient handling and may transfer substantial stress through banding, stretch wrap, corner protection, skids, and the pallet structure. A cargo transport unit (CTU) is the transport enclosure or platform into which the load units are packed. Shipboard stowage arranges and secures containers or other cargo units within the vessel's structural and operational limits. The same word, stowage, can refer to placing goods inside a container or placing containers on board a ship, so technical reports should specify the location and operation.
Failure at one interface can pass forces to another. A crate that fractures may allow machinery to move and strike a container wall. A sound container may be exposed to extraordinary racking forces from the collapse of an adjacent shipboard stack. A container with an intact door seal may develop condensation because hygroscopic contents, timber, humid air, and temperature changes interact. An undamaged outer package may conceal an internal fracture. Each pathway produces different inspection evidence and potentially different contractual outcomes.
The concept of ordinary incidents of transit is especially important to marine cargo cover. It denotes the routine stresses that packing and preparation must withstand under the applicable wording; the exact legal interpretation remains policy- and jurisdiction-dependent. Ordinary operational forces include acceleration and deceleration, vibration, lifting, stacking, imperfect roads, periodic vessel motions, and the handling reasonably expected from the agreed carriage. Severe weather may be an insured external peril and also a source of force that exposes a preexisting weakness. The existence of two causes demands analysis of their temporal and causal contribution.
Standardization and the expansion of responsibility
International containerization standardized interfaces for handling and transport but did not standardize every commodity's internal behavior. The ISO 668 container classification specifies dimensions and ratings for series 1 freight containers, while the ISO 1496-1 specification addresses design and testing of general-purpose containers. These are different propositions from the quality of a shipment's internal dunnage or the strength of a particular timber crate. A valid container certificate cannot establish that the cargo was correctly packed.
The 1972 Safe Containers Convention governs structural safety approval, plates, testing, and ongoing examinations for containers within its scope. It serves the international interchange of approved units and the prevention of structural failures. Ownership and examination of a CTU, the duties of its packer, verification of the declared weight, and the ship's stow-planning decisions may be distributed among different actors. The resulting chain is operationally integrated yet institutionally fragmented.
The 2014 CTU Code publication updated the 1997 intermodal packing guidelines and made the chain-of-responsibility approach more accessible. Its chapters distinguish information exchange, transport conditions, CTU suitability, cargo packing, dangerous goods, completion of packing, receipt and unpacking, and training. Annex 7 addresses packing and securing in greater technical detail, while annex 3 concerns condensation. The Code's informative companion material is separate from the Code itself. In historical terms, this development reflects recognition that a failure created before the sea voyage can become visible only on the vessel, on inland delivery, or after unpacking.
Current instruments and their normative status
The SOLAS gross-mass requirements entered into force on July 1, 2016. They require the shipper to obtain and communicate the verified gross mass of a packed container sufficiently early for stow planning, using the permitted whole-container weighing method or the approved summation method. A missing verified gross mass is a loading constraint, while a verified mass is a value needed by the master and terminal rather than a certification of all cargo-securing details.
The IMO CSS Code overview identifies ship motions and resulting accelerations as central sources of securing forces. National implementation provides operationally binding detail; for example, the UK's Cargo Securing Manual guidance explains that SOLAS chapter VI, regulation 5 requires cargo units to be stowed and secured in accordance with the approved manual, and assigns practical responsibility to the operator and master. The ship's manual is therefore an essential document in a stack-collapse investigation.
The IMDG Code 2024 Edition, including Amendment 42-24, became mandatory on January 1, 2026. It governs the particular packing, marking, documentation, segregation, and transport requirements for dangerous goods within its applicable scope. The same date brought mandatory lost-container reporting under SOLAS and MARPOL amendments. Reporting a loss improves safety information; it does not establish the precise insurance cause, cargo value, or packer's fault.
A 2026 UNECE working document describes a 2025 revision of the CTU Code. Its identification as a working document warrants examination of the multiorganization endorsement and effective-status record before treating the revision as a globally adopted successor. The analysis here uses the confirmed 2014 joint Code as the operative international reference and treats the later text as evidence of continuing standards development, without asserting general mandatory force.
Packaging Adequacy
Fitness for the journey and the goods
Packaging adequacy is a relationship between a consignment's vulnerability and the demands of the agreed journey. Its assessment starts with item geometry, center of gravity, fragility, liquid or powder characteristics, susceptibility to humidity or temperature, incompatibilities, hazardous classification, and handling instructions. It proceeds through the package's materials, cushioning, closure, compression strength, restraint points, stacking geometry, pallet configuration, and acceptance condition. Adequacy cannot be inferred from a package's appearance alone or from the absence of earlier claims involving superficially similar shipments.
The CTU Code's guidance on packing cargo into transport units implies a sequence of planning decisions: establish CTU suitability; know weight and dimensions; distribute cargo to respect load and axle limits where applicable; prevent movement through blocking, bracing, friction, and lashings; protect against sharp edges; preserve access and safe unpacking; and communicate hazards and special handling needs. The design must cover road, rail, terminal, and maritime exposures that actually occur. A scheme suited only to smooth crane lift or low acceleration on one segment can fail during braking, curving, shunting, rolling, or pitching.
Cargo restraint uses a load path. The cargo's inertia must be transmitted through its structural attachment points and restraint devices to reliable anchorages or load-bearing surfaces. Weak packaging or an unstable pallet may defeat an otherwise strong lashing. An unsupported load can concentrate forces on container sidewalls not designed to provide that form of restraint. Timber blocking, anti-slip material, frictional contact, banding, or inflatable dunnage can be effective only when their physical properties, contact area, anchoring, and arrangement are appropriate. The safety evaluation must examine the weakest element and the direction of the relevant force.
Jagelčák, Vrábel, and Pauliak (2013) examined dunnage-bag contact surfaces and compared measured restraining behavior with calculations used in development of CTU guidance. The experiment illuminates a specific mechanism: effective bag capacity depends on geometry and contact with the cargo, rather than nominal inflation pressure alone. It does not establish an average percentage reduction in marine claims for all bags, cargoes, and transport routes. The point generalizes cautiously to other devices: catalog ratings need to be matched to the installed geometry and dynamic loading direction.
Mass, center of gravity, and concentrated loads
Verified gross mass and safe internal weight distribution answer different questions. A compliant declared VGM can coexist with an excessively high, offset, or localized center of gravity; uneven floor loading; poor transverse balance; or inadequate restraint of dense equipment. For a container moving with an acceleration a, the idealized inertial force on a cargo component of mass m is F = ma. That basic mechanical relation motivates restraint design, but the relevant design accelerations vary with motion axis, transport mode, location, sea area, vessel response, and the applicable standard. A single acceleration coefficient applied uniformly to all legs is scientifically weak.
The IMO VGM guidance emphasizes the role of accurate weight in avoiding improper stacking and collapse. For packers, a gross figure must be supplemented by a distribution plan showing item masses, positions, loading order, and any heavy lift or eccentric-load measures. For the ship planner, the verified mass joins container dimensions, stack positions, deck or hold restrictions, lashing capability, and the ship's stability and motion profile. A transmission or software error between declared mass, terminal data, and final physical slot can compromise the entire workflow without changing the original weighbridge figure.
The same distinction applies to package strength. ISO container dimensions and ratings support intermodal compatibility, but general-container structural tests concern the design of the container. They do not certify that product cartons can carry the vertical load of adjacent cargo, resist uplift after a void develops, or tolerate repeated shock. Packaging specifications should explicitly state allowable stack compression, moisture exposure, acceptable unit load, and the anchorage method selected for the shipping configuration.
Moisture, condensation, and biologically sensitive cargo
Marine cargo damage frequently develops without a single dramatic casualty. Temperature gradients can cause water vapor within a closed CTU to condense on colder surfaces. Hygroscopic packaging and commodities absorb or release moisture as conditions change. A cargo may arrive with intact doors and seals but show water staining, corrosion, mold, paper deformation, adhesive degradation, or compromised packaging strength. CTU Code condensation-prevention guidance encourages attention to initial cargo moisture, material selection, ventilation suitability, container dryness, and voyage conditions.
Determining the source of moisture requires comparative evidence. Investigators should distinguish seawater ingress, rain during stuffing, condensation, wet timber or pallets, inherent moisture in the product, refrigeration condensate, and damage following delayed unpacking. Chloride testing, corrosion patterns, location of staining, container integrity, weather and temperature logs, and moisture records can assist, but individual observations are rarely conclusive without context. Desiccants and barrier films may mitigate defined vapor exposures; they cannot repair a leaking roof or compensate for exceptionally wet goods sealed inside an impermeable enclosure.
Temperature-sensitive cargo introduces another interface. A reefer's set point, supply and return-air records, airflow channels, stowage clearance, pre-cooling state, loading delays, and commodity respiration all matter. Adequate insulated packaging may become ineffective when airflow is obstructed. Conversely, a temperature excursion may reflect interrupted power supply or terminal handling rather than inadequate packing. The analytical boundary is the physical causal history, not an assumption that visible spoilage identifies a single responsible actor.
Dangerous goods, contamination, and closure controls
Dangerous goods require classification and packaging under applicable maritime requirements, including the mandatory 2026 IMDG Code. Classification, approved packaging, segregation, marking, and documentary declarations control both direct cargo loss and the possibility that a consignment initiates a fire, toxic release, reaction, or generalized casualty. Packaging inspection alone is insufficient when a hazard has been misdeclared or mixed with incompatible substances.
Ordinary cargo can also contaminate adjacent shipments through ruptured drums, spillage, dust, odors, biological residue, or incomplete cleaning of a CTU used previously for another commodity. Controls include pre-stuffing inspection, cleanliness records, packaging integrity checks, separation, suitable liners, and adequate protection at interfaces. A seal establishes a fact about closure integrity at the time of observation and possible unauthorized access; it is not conclusive evidence of appropriate initial packing, dryness, or internal restraint.
A completed packing operation should generate contemporaneous documentary evidence: CTU identity, inspection findings, photographs before door closure, cargo and package identification, packing and weight records, restraint plans, dangerous-goods declarations when needed, seal details, and any exceptions accepted with corrective action. These records become most valuable when they preserve sequence and responsibility. Retrospectively recreated photographs or generalized checklists have less ability to discriminate between competing explanations.
Stowage Failures
Failures inside cargo transport units
Internal cargo displacement often begins locally: a restraint loosens, friction decreases, a blocking member crushes, a pallet collapses, or an item rotates around an unsecured edge. The resulting load can exceed the strength of adjacent cargo or the restraint system, amplifying damage. Friction depends on material, surface condition, moisture, and contact force. Lashings carry tension and need suitable geometry and anchor points; a tie-down system may have a different mechanical function from a direct restraint. A restraint design should explicitly account for expected directions of acceleration and loss of contact.
Not every damaged crate indicates that it was the initiating failure. A container impact or abrupt terminal drop can break the crate first, and an earlier weak crate can permit cargo motion that damages the container. Reconstructing the direction of deformation, failure surfaces, restraint marks, abrasion, and the distribution of damaged packages helps determine temporal order. Packaging components collected after the accident should be compared with their specifications and the condition documented at origin.
The scientific literature offers mechanism-specific evidence rather than a single universally valid securing formula. In a 2024 experimental and numerical study, Li and colleagues reported that support and lashing force analysis could reproduce principal trends for a particular cylindrical cargo securing scheme. Pérez-Canosa and Orosa (2024) examined optimization of lashing arrangements for nonstandardized cargo. These studies support physically explicit design and sensitivity testing, while the tested equipment, geometries, boundary conditions, and model assumptions limit immediate extrapolation to container stacks, fragile consumer goods, and insurance loss frequency.
Shipboard stowage and dynamic loading
A container ship's deck stow experiences roll, pitch, heave, wind, hull bending, local vibration, and combinations of these forces. Deck position and stack geometry affect accelerations and the load transferred through corner posts, twistlocks, lashings, and adjacent units. A high stack with heavy mass in upper tiers can increase overturning and lashing demands. Container strength, the vessel's actual loading condition, metacentric height, expected route, stowage software, and cargo-securing manual must be interpreted together. A nominally approved stow plan can become inconsistent with reality if substitutions, late load changes, or incorrect verified weights are not incorporated.
The IMO CSS Code identifies transverse accelerations and ship motions as a central concern. The ship-specific Cargo Securing Manual requirement links those concerns to approved arrangements and operational responsibility. Compliance must be judged against the manual applicable to the actual ship and voyage; a generic diagram or an unrelated vessel's load table offers little probative value.
Ever Smart: interacting failures and documentary inconsistency
The UK's Marine Accident Investigation Branch reported that the container ship Ever Smart sustained a stack collapse on October 30, 2017, while crossing the North Pacific. Of 151 containers in the affected stow, 42 were lost overboard and 34 damaged. Its official accident findings associated the event with severe motion, likely pitching and hull vibration, and cargo arrangements that were inconsistent with the ship's securing manual. The report also raised the possibility that lashings were improperly secured.
This case illustrates the importance of separating weather as a trigger from the load path's available margin. Heavy seas mattered, but inadequate distribution or application of securing equipment can change a stack's susceptibility under the same motions. A documentary comparison between intended plan, actual bay configuration, approved manual, and physical securing condition is therefore more informative than the statement that the vessel experienced bad weather. MAIB's findings are technical safety findings; they do not by themselves adjudicate a particular cargo insurer's obligation or determine the liability of every participant.
Ever Smart also demonstrates that shipboard loss of entire containers may arise despite some cargo inside the lost units having been packed correctly. It would be inaccurate to infer a failure of carton, pallet, or internal dunnage merely because the CTU was lost overboard. The appropriate prevention comparison is between interventions able to change the failure path: securing-plan consistency, control of weight distribution, lashing maintenance and verification, and operational response to motion.
CMA CGM G. Washington: structural strength and interacting bays
In its report on CMA CGM G. Washington, MAIB reconstructed a January 20, 2018 event in which unexpected large rolling during heavy North Pacific seas preceded collapse in three bays. The official MAIB summary records 137 containers lost overboard and 85 damaged. Investigators identified diminished strength of nonstandard 53-foot containers, inaccurate weight declarations, mis-stowage, and loose lashings among the relevant risk factors. The attributed initiating mechanisms differed among bays.
This distinction matters for causal attribution. The failure of one stack can mechanically damage another, so proximity is not proof of shared initiating cause. The presence of an overweight or incorrectly positioned container in a failed bay matters differently from generalized suspicion that some units may have inaccurate weights. The actual weights, physical stack positions, structural types, dynamic loads, and lashing condition need to be evaluated by bay and by the most plausible sequence. A finding made for one bay cannot simply be copied across the entire casualty.
Comparison of the two MAIB investigations supports a robust but limited conclusion: major losses may result from the interaction of several ordinary safeguards under demanding motions, and their causal reconstruction requires contemporaneous stowage and securing evidence. It does not establish a population-wide fraction of container losses attributable to one device, one container type, or one company's personnel.
Marine cargo insurance and the packing exclusion
Institute Cargo Clauses (A), (B), and (C), dated January 1, 2009, commonly distinguish insufficient or unsuitable packing and preparation from other insured transit occurrences. The insurer's explanation of clause 4.3 states that the wording treats container stowage as packing and addresses whether the deficient packing was performed by the assured or its employees, or before the attachment of insurance. Independent contractors receive specific treatment in the wording. The actual policy, endorsements, assured identity, agency relationships, and timing are decisive; the clauses should not be assumed to govern every international shipment.
This exclusion engages a factual inquiry distinct from container-stack engineering. A failure to restrain a helicopter inside its CTU may be analyzed under clause 4.3; a stack collapse caused by shipboard lashings may be a different physical and contractual event. Where several circumstances combine, investigators must identify the actual source and timing of the relevant insufficiency and distinguish it from an external peril. The UK Marine Insurance Act section 55 states a general rule of proximate causation for insured losses subject to the policy and also addresses delay, ordinary leakage, wear, inherent vice, and related exclusions. Its precise application depends on the governing legal framework and policy drafting.
The appellate litigation in Swashplate Pty Ltd v. Liberty Mutual Insurance Company provides a concrete illustration of timing. The case involved damage to a helicopter transported from the United States to Australia and a dispute over insufficient packing and the inception of cover. An account of the Full Federal Court's 2020 judgment records that the initial denial was overturned on appeal, with attention to a five-day static-cover extension and the construction of the voyage policy. The central scientific lesson is that identifying when packing occurred is incomplete without determining when the insurance attached under the entire agreement. The appellate outcome, rather than the overruled first-instance conclusion, governs this illustrative case.
Carrier responsibility and the independent insurance contract
Where applicable, the Hague–Visby Rules require carriers to exercise due diligence to make the vessel seaworthy and to care for goods through loading, handling, stowage, carriage, and discharge, subject to enumerated rights and exceptions. The UK's Carriage of Goods by Sea Act 1971 reproduces these rules in its schedule. Article IV(2)(n) recognizes insufficiency of packing among the listed defenses; Article III(2) addresses the carrier's duty of care. The scope, time period, documentary trigger, and defenses depend on the jurisdiction, the bill of lading or other contract, and applicable enactment.
A cargo insurer's payment does not prove carrier negligence; nor does the existence of a carrier defense automatically exclude first-party cargo cover. Insurance pays according to the insured's policy and may be followed by recourse against responsible third parties under applicable subrogation principles. A cargo claim file should accordingly preserve evidence suitable for both coverage decisions and potential recovery, without treating them as the same legal determination.
The interaction can be illustrated without presuming a result. If a machine shifts because of an insecure origin blocking arrangement, the packer's records, the date of packing, and the policy wording are central. If it is crushed when a terminal vehicle drops the container, handling records and the actual transit peril become central. If the entire CTU is lost in a vessel-stack collapse, the lashing plan and shipboard casualty evidence dominate the physical reconstruction. The relevant insurance and carriage contracts must then be applied separately to each established sequence.
Mitigation Effectiveness
Prevention controls and mechanisms
A prevention measure is effective when it changes a defined causal probability, severity, or exposure under conditions comparable to those being evaluated. A packing photograph is primarily an evidentiary control; its presence may encourage compliance, but it cannot physically restrain unsecured cargo. A certified lashing arrangement may increase resistance to displacement; its value depends on proper installation, force direction, inspection, and voyage conditions. Better forecasts and routing can alter exposure to sea states but cannot guarantee that a structurally deficient stack remains intact. Effective prevention therefore combines physical controls with verification and operational decisions.
At the product and package level, interventions include protective engineering design, vibration and drop resistance appropriate to the goods, moisture controls, corrosion protection, tamper evidence, and accurate dangerous-goods classification. At the CTU level they include condition inspection, structurally appropriate floor loads, center-of-gravity management, blocking, lashings, separation, and safe unpacking. At the ship level they include accurate VGM transfer, stowage planning consistent with the approved manual, lashing inspection and maintenance, stack-limit validation, and motion-aware operations. A single audit score should not conceal an unaddressed catastrophic failure mode.
| Control level | Loss pathway addressed | Direct observable evidence | Principal residual limitation |
|---|---|---|---|
| Commodity and package | Crush, shock, moisture, contamination | Tested specifications, packing records, condition photographs | Tests and actual route conditions may differ |
| Pallet or skid | Load-unit collapse, shift, concentrated pressure | Unit-load geometry, banding, marked masses, inspection | Pallet integrity may deteriorate during transit |
| Internal CTU securing | Sliding, tipping, puncture, door impact | Securing plan, restraint specification, pre-closure photographs | Unknown installation quality or dynamic load |
| CTU structural condition | Door, wall, floor, and frame failure | CSC status, inspection and maintenance records | Hidden defects or abnormal stacking stress |
| Terminal and inland leg | Shock, mis-handling, water exposure | Handling and interchange records, location times | Incomplete third-party process evidence |
| Shipboard stowage | Stack collapse, lashing failure, overboard loss | VGM, final plan, Cargo Securing Manual, lashing records | Nonlinear coupled motion and cascading effects |
| Cargo insurance and claims | Financial loss, disputes, delayed recovery | Policy wording, attachment time, surveys, notice | Contract-specific exclusions and causation |
The table is an analytical classification of established risk controls and evidentiary objects. It is not a validated scorecard or a newly proposed standard. For each consignment, the implementation record should identify which control was actually tested, by whom, at what stage, and against which applicable specifications.
Cost and expected-loss reasoning
Let p0 denote the probability of a defined damage event without an intervention, p1 the comparable probability with it, L0 and L1 the corresponding conditional economic severities, and C the intervention's all-in cost. A simple decision comparison is:
Expected loss reduction = p0 × L0 − p1 × L1.
Net expected benefit = expected loss reduction − C.
These identities are ordinary expected-value arithmetic; they do not constitute a new actuarial model. They require consistent cargo and exposure definitions, clear treatment of indirect losses, and explicit assumptions about dependence between frequency and severity. A prevention measure may reduce severity while leaving frequency broadly unchanged, or improve detection while leaving physical risk unchanged. Insurance deductibles, claim-handling expense, salvage, third-party recovery, and business interruption can shift how the benefit is distributed between cargo owners, carriers, insurers, and society.
For an illustration only, suppose a shipment with a potential loss of $200,000 has an assumed damage-event probability of 0.8% before a control and 0.5% after it, with equal conditional severity. The assumed annualized expected reduction for one such exposure would be $600. If the control costs $250 per exposure, the illustrative net expected benefit is $350 before administrative, capital, or broader social adjustments. The numbers are hypotheses for explaining arithmetic; they are not estimates drawn from shipping data, a measured treatment effect, or an underwriting recommendation. If the underlying probability difference lacks identification, the apparently precise dollar result carries no empirical authority.
A sensitivity analysis should vary at least the baseline frequency, severity tail, probability reduction, control cost, routing, cargo type, and claim attribution. For high-value machinery, even a small severity reduction may matter. For low-value bulk commodities, compliance costs may dominate a narrowly defined private expected saving but remain justified by worker safety, catastrophic third-party exposure, or law. The correct scope of economic evaluation is therefore tied to the actor and decision being studied.
What global container-loss statistics can establish
The 2026 World Shipping Council estimate reports approximately 1,478 containers lost out of about 280 million transported in 2025, a quotient of roughly 0.000528%, reported by WSC as about 0.0005%. The comparable published 2024 figures are 576 and approximately 250 million, yielding about 0.000230%. The report says a single 2025 event accounted for 640 containers, approximately 43% of the annual total. WSC's members represent approximately 90% of global container-vessel capacity, and its estimate extrapolates to the remaining capacity using an assumed comparable experience.
These figures indicate the small scale of overboard container losses relative to container movements and the importance of rare multi-container casualties. They do not measure the frequency of broken packaged goods, internal shifting, wet damage, reefer spoilage, or cargo insurance claims. Nor does the apparent change from 2024 to 2025 establish that industry packing practice deteriorated; one major incident strongly affects the difference. The denominator is container movements, not unique insured items, voyages, bill-of-lading consignments, container-years, or monetary sums insured. Comparisons across years need stable scope and a transparent account of extrapolation.
This distinction matters for prevention evaluation. An insurer could report fewer claims after introducing packing inspections while simultaneously facing different cargo composition, routes, deductibles, reporting thresholds, and weather. A carrier may reduce overboard container losses without reducing minor inland damage. A study of mitigation effectiveness must specify its outcome and the exposure at risk, assess competing changes, and, where possible, compare appropriately matched treated and untreated shipments or a credible policy change.
Underwriting, incentives, and contractual practice
Insurers can use engineering surveys, loss histories, packing specifications, commodity descriptions, routes, survey warranties or conditions, and loss-prevention requirements to characterize cargo risk. A useful control record is specific: the class of goods, CTU type, defined loading process, named securing standard, inspection checkpoint, and the evidence necessary to confirm compliance. Generic statements that goods were packed to international standards may fail to identify which standard, version, and requirement actually governed the shipment.
An incentive for better protection can involve underwriting acceptance, coverage conditions, a deductible, loss-sensitive pricing, or contractual recourse. Each mechanism changes behavior only to the extent that the party making the physical decision experiences or values its consequences. A freight forwarder may choose the packer while the cargo owner buys cover; the terminal may control final stow slot while the shipper declares weight; a carrier may maintain lashings while a third-party contractor has performed the origin blocking. Risk engineering therefore benefits from identifying decision rights and evidence-access rights in addition to nominal insurance interests.
The sound economic conclusion is conditional: verified controls should influence expected-loss assessment when there is a plausible and evidenced connection to the exposure, but their exact premium or reserve effect requires suitable claims data and actuarial judgment. This monograph makes no claim of a universal premium discount for compliant packing and offers no individualized insurance advice.
Research Evidence Quality and Sensitivity
Source authority, version control, and source scope
Different kinds of evidence answer different questions. The IMO's CTU Code explains recommended safe practice; the SOLAS VGM amendment imposes a verified-weight obligation; the approved Cargo Securing Manual controls a ship-specific operation through the applicable legal implementation; ISO metadata identifies the scope and current status of published standards; an insurance policy defines contractual cover; a statute describes the relevant national legal framework; an accident report reconstructs a particular event; and a journal paper tests a specified scientific proposition. Their authority is not interchangeable.
The available ISO descriptions confirm the scope and currency of ISO 1496-1:2013, including published amendments, and ISO 668:2020. The proprietary full standards text was not reproduced or independently re-tested here, so quantitative design limits not present in the publisher's accessible descriptions are not asserted as if directly verified. The scientific literature is used with the same discipline: a reported mechanical experiment is evidence about its stated apparatus and conditions, not a proof of marine claim causation across commodities.
For current compliance discussion, dates are essential. The 2014 CTU Code is a jointly endorsed advisory document. SOLAS VGM has applied since July 2016. The 2024 IMDG amendment entered mandatory application in January 2026; the lost-container reporting requirement also commenced in January 2026. A revision described in a UNECE 2026 working-document listing cannot be treated as a universally binding new CTU Code merely because it bears the words 2025 revision. This edition- and status-specific record is part of the scientific result.
Reconstructing a cargo casualty
A complete physical inquiry begins with the condition at origin, cargo description, measured weight, packaging and securing plan, CTU suitability, and time-stamped closure evidence. It proceeds through handoffs, intermediate inspections, terminal operations, actual ship slot, stowage-plan revisions, VGM transmissions, securing checks, and environmental and voyage records. The final stage covers discovery time, photographs before alteration, surveys, sample provenance, salvage, and unloading or unpacking actions. Chain-of-custody gaps should be identified rather than silently filled by assumptions.
The relevant hypotheses may include insufficiency of original packing, latent product defect, inherent vice, rough handling, water ingress, condensation, reefer failure, erroneous mass information, actual shipboard stowage contrary to plan, failed lashings, container structural failure, unusually severe motions, and collapse propagated from neighboring stacks. Each hypothesis should be evaluated for a physical mechanism, predicted traces, counterevidence, timing, and competing explanation. The best-supported account need not identify one exclusive cause; multiple conditions can be necessary or contributory.
Claims evaluation adds temporal and contractual evidence: the insured's interest, the insured transit, date and time of attachment, clause edition, endorsements, exclusions, warranties and conditions, governing law, bills of lading, carrier terms, notices, and any rights of recovery. The Swashplate appellate account illustrates the risk of deciding the packing exclusion before interpreting the start of coverage as a whole. A survey finding that packing was inadequate is a causal input; it is not automatically the complete legal answer.
Scientific literature and external validity
Kaup and coauthors' review of container-ship loading discusses cargo damage and losses and proposes attention to the loading process. Experimental work on dunnage contact geometry, measured lashing forces, and nonstandard cargo securing optimization contributes distinct mechanical perspectives. These publications are complementary but address different specimens, methods, outcomes, and conditions. It would be a category error to pool their technical results as if they estimated the same insurance effect.
Accident-investigation reports have comparatively rich causal detail but are selected by occurrence and investigative jurisdiction. Their findings can identify failure mechanisms and institutional weaknesses with considerable specificity. They cannot directly provide a population denominator for all safely completed voyages or measure the share of losses prevented by one intervention. Insurance claims files are broader in some respects but also selected by policy coverage, deductibles, reporting behavior, and insurer practices; they may underrepresent uninsured damage or near misses.
The WSC survey methodology brings a further limitation: capacity-based participation and extrapolation are appropriate for an industry estimate but rely on assumptions for nonparticipating operators. Year-on-year changes mix exposure, weather, fleet, routes, incident clustering, and reporting. Mandatory reporting from 2026 may affect the observed future series independently of actual changes in risk. An analysis that compares pre- and post-2026 figures should address this potential break in observation.
Sensitivity tests and competing explanations
Sensitivity analysis is needed whenever a conclusion depends on uncertain facts or assumptions. For a packaging-insufficiency hypothesis, vary the plausible transit accelerations, restraint installation quality, friction, cargo mass, and timing of the damage. For shipboard stack collapse, compare recorded and planned loading, alternative weight declarations, vessel motions, cargo-securing manual limits, and the possible initiating unit. For moisture damage, examine alternative water sources and the lag between delivery and inspection. For coverage, test the consequence of alternative attachment times and causation findings under the actual clause set.
An instructive example is a consignment arriving with damaged internal supports after a rough sea passage. The intuitive claim that heavy weather alone caused the damage must be tested against packing photographs, support specifications, shock indicators, securing-device marks, and signs of gradual fretting. The competing claim that poor packing alone caused the loss must likewise be tested against the severity and timing of extraordinary external forces. Without adequate records, a range of plausible explanations remains; forcing a single narrative does not improve the scientific quality of the result.
The evidence ledger accompanying this monograph records source-specific scope and access limitations. It does not imply that a publisher abstract constitutes full access to a proprietary standard, that a law-firm explanation replaces the court's full reasons, or that an official accident summary permits all calculations reported in its underlying technical annexes to be independently reproduced. These are material boundaries on interpretation, recorded for future revision if fuller source material becomes available.
Integrated Scientific Synthesis
The decisive distinctions
The first integrated conclusion is that marine cargo preservation involves multiple distinct physical control locations. Product packaging protects against damage to the goods; in-container securing keeps cargo and packages in a safe position relative to the CTU; container structural fitness governs the enclosure; shipboard securing governs the relationship among container stacks, vessel structure, and the motions of the voyage. A verified gross weight helps plan the latter two operations but does not substitute for internal restraint. A Cargo Securing Manual addresses the shipboard arrangement but does not individually certify every package loaded at origin.
The second conclusion concerns the boundaries of rules and evidence. The CTU Code establishes an international expert baseline for practice, while SOLAS, CSC, the IMDG Code, implementing statutes, and ship-specific approved arrangements supply different forms of legal or operational authority. Their common objective of safer transport does not make their legal effects identical. Likewise, compliance with one instrument is informative only about the matters that instrument actually governs.
The third conclusion concerns causation and insurance. Damage and the proximate insured cause must be reconstructed before policy exclusions and carriage defenses can be applied appropriately. The technical vocabulary of insufficient packing, stowage failure, or inherent vice must be connected to evidence about the origin and operation of the failure. Coverage under Institute Cargo Clauses, carrier responsibility under an applicable goods-by-sea regime, and subrogated recovery are separate legal inquiries with interacting facts.
The fourth conclusion concerns organizational fragmentation. A shipment may be packed by one contractor, documented by a forwarder, weighed or verified by another participant, handled by terminals, assigned a final ship slot by shore planners, secured by the vessel's personnel, and insured by a party remote from the physical decisions. Effective prevention depends on information and control moving across these interfaces. The quality of documented handoffs matters as much as an isolated equipment specification when reconstructing failures.
The fifth conclusion concerns empirical evaluation. Large container-loss events are rare relative to total movements but can dominate annual totals. A fall in lost-at-sea containers is not automatically a fall in insured package damage or a measured benefit from improved packing. The causal impact of a specific intervention requires comparable exposures, a defined outcome, a credible comparison, and transparent uncertainty. Official investigations offer detailed causal process evidence; fleet statistics offer scale; experiments test mechanisms; claims portfolios may reveal financial outcomes. Their combined use produces a better understanding only when the limits of each source remain explicit.
Implications for science and research practice
For Marine and Aviation Insurance, the findings clarify how packing exclusions, insured transit, carriage obligations, and survey causation interact in the maritime cargo domain. For Risk Engineering and Loss Prevention, they locate intervention effectiveness at the physical interfaces where risk is created or controlled and describe what verification evidence is needed to attribute a result. These are two disciplinary memberships of one canonical dossier; the subject remains a single scientific object.
The monograph's original analytical contribution lies in bringing technical documentation, failure reconstruction, contract interpretation, and exposure statistics into one source-specific argument. This synthesis is descriptive and evaluative. It does not assert a new law of cargo risk, propose an unvalidated risk taxonomy as a proprietary doctrine, or assign numerical causal reductions unsupported by empirical designs.
Future evidence collection should prioritize paired origin and destination photographs for the same consignment, digital securing plans with actual installation verification, versioned stowage and VGM event histories, and linked survey and claim outcomes. Research designs should separate minor insured cargo damage from catastrophic overboard events, distinguish pre- and post-2026 reporting regimes, and evaluate controls at the level where they actually operate. Longitudinal loss datasets incorporating exposure units and non-loss observations would improve understanding of mitigation effectiveness substantially.
Conclusion
Marine cargo packaging and stowage are a coordinated chain of physical decisions whose performance depends on the goods, the packaging and securing scheme, the condition of the transport unit, the terminal and inland operations, and the ship's actual stowage and motion environment. The chain's vulnerability frequently lies at interfaces that are invisible in an ordinary bill of lading or insurance declaration.
The CTU Code provides the principal global advisory reference for packing cargo transport units, while compulsory VGM, container structural rules, shipboard Cargo Securing Manuals, and dangerous-goods requirements address particular safety functions. Correct source classification and edition control are indispensable: the presence of a cited standard does not establish that every provision was mandatory or complied with in a given operation.
The Ever Smart and CMA CGM G. Washington investigations demonstrate how heavy seas, stack geometry, mass declarations, container strength, and lashing arrangements may combine in large casualties. They also show the necessity of comparing the final physical configuration with the approved planning record. Attributing every loss in such an event to internal packaging would misrepresent the available causal evidence.
Marine cargo insurance requires a distinct legal analysis. The relevant policy, period of cover, packing exclusion, proximate cause, and governing law determine how physical facts translate into an indemnity decision. Carrier liability introduces a separate framework of duties and defenses. The Swashplate appeal shows the material importance of interpreting attachment and extensions before concluding that a packing defect falls within an exclusion.
Loss prevention works through specific physical and procedural mechanisms. Its success must be measured against defined exposure and outcome data, with attention to cargo mix, route, weather, claim thresholds, and exceptional incidents. The WSC figures are valuable for understanding the rarity and concentration of overboard losses, but cannot establish the effectiveness of a universal packaging intervention.
The scientific result is an integrated, evidence-bounded explanation of why prevention, loss causation, and insurance allocation must be evaluated together while preserving their distinct evidentiary and legal questions. The next advances in this field will come from better longitudinal incident data, verifiable control records, and explicit comparison of interventions across comparable transport exposures.
References
International Maritime Organization, International Labour Organization, and United Nations Economic Commission for Europe. (2014). IMO/ILO/UNECE Code of Practice for Packing of Cargo Transport Units (CTU Code), MSC.1/Circ.1497. IMO official CTU Code.
United Nations Economic Commission for Europe. (2014). IMO/ILO/UNECE Code of Practice for Packing of Cargo Transport Units. UNECE Code documentation.
United Nations Economic Commission for Europe. (2026). IMO/ILO/UNECE Code of Practice for Packing of Cargo Transport Units (CTU Code)—2025 revision, ECE/TRANS/WP.24/2026/8, working-document record. UNECE 2026 working document.
International Maritime Organization. (2014–2016). Verification of the Gross Mass of a Packed Container, SOLAS VI/2 amendment MSC.380(94), with MSC.1/Circ.1475. IMO VGM requirements.
International Maritime Organization. (1991, as amended). Code of Safe Practice for Cargo Stowage and Securing, resolution A.714(17). IMO CSS Code.
United Kingdom Maritime and Coastguard Agency. (2024). Annex C: Stowage and Securing—Cargo Securing Manual, MGN 699 (M). UK CSM guidance.
International Maritime Organization. (1972, as amended). International Convention for Safe Containers (CSC). IMO CSC overview.
International Organization for Standardization. (2013, including published amendments). ISO 1496-1:2013: Series 1 Freight Containers—Specification and Testing—Part 1. ISO 1496-1.
International Organization for Standardization. (2020, including published amendment). ISO 668:2020: Series 1 Freight Containers—Classification, Dimensions and Ratings. ISO 668.
International Maritime Organization. (2024). International Maritime Dangerous Goods Code, Amendment 42-24, mandatory January 1, 2026. IMO IMDG Code.
International Maritime Organization. (2026). Raft of Shipping Rules in Force from 1 January 2026, including SOLAS V/31 and V/32 lost-container reporting. IMO 2026 rule summary.
United Kingdom. (1971). Carriage of Goods by Sea Act 1971, Schedule: Hague Rules as amended by the Brussels Protocol 1968. Official legislative text.
United Kingdom. (1906, as amended). Marine Insurance Act 1906, section 55: Included and Excluded Losses. Official section 55.
Marine Accident Investigation Branch. (2020). Loss of Cargo Containers Overboard from Container Ship Ever Smart, Accident Investigation Report 14/2020, incident October 30, 2017. MAIB Ever Smart report.
Marine Accident Investigation Branch. (2020). Loss of Cargo Containers Overboard from Container Ship CMA CGM G. Washington, Accident Investigation Report 2/2020, incident January 20, 2018. MAIB Washington report.
World Shipping Council. (2026, June 25). Containers Lost at Sea Report: 2026 Update, estimates for calendar year 2025 and comparative 2024 data. WSC 2026 report summary.
If P&C Insurance. (n.d.). Packing of Goods—Not Just Pretty Wrapping, commentary on clause 4.3 of Institute Cargo Clauses A/B/C (2009). Insurer clause commentary.
Norton White. (2020, August 28). Swashplate Pty Ltd v Liberty Mutual Insurance Company Trading as Liberty International Underwriters [2020] FCAFC 137, appellate-case commentary. Appeal case analysis.
Jagelčák, J., Vrábel, J., and Pauliak, M. (2013). “The Measurement of Contact Surface of Dunnage Bag Used for Cargo Securing in Different Gaps Between Cargo.” Transport and Communications, 1(1), 4–7. DOI: 10.26552/tac.C.2013.1.2. Journal article.
Li, M., Wang, G., Liu, K., Lu, Y., and Wang, J. (2024). “Experimental and Numerical Analysis of Supporting Forces and Lashing Forces in a Ship Cargo Securing Scheme.” Journal of Marine Science and Engineering, 12(1), 158. DOI: 10.3390/jmse12010158. Peer-reviewed article.
Kaup, M., Łozowicka, D., Baszak, K., Ślączka, W., and Kalbarczyk-Jedynak, A. (2022). “Review of the Container Ship Loading Model—Cause Analysis of Cargo Damage and/or Loss.” Polish Maritime Research, 29(4). DOI: 10.2478/pomr-2022-0041. Journal article.
Pérez-Canosa, J. M., and Orosa, J. A. (2024). “A New Methodology for Optimization of Lashing Lines in the Securing Arrangement of Non-Standardized Cargo on Ships.” Applied Sciences, 14(23), 11442. DOI: 10.3390/app142311442. Peer-reviewed article.