Damage stability

Damage stability is the branch of naval architecture and maritime regulation that defines the ability of a ship to remain afloat, upright and operationally safe after one or more watertight compartments have been flooded as a result of collision, grounding, structural failure or other casualty. The field has been shaped by a sequence of disasters stretching from RMS Titanic in 1912 to El Faro in 2015, each exposing gaps in the prevailing regulatory framework. Two parallel regulatory philosophies have governed the subject: the older deterministic regime, which specifies permissible flooded lengths and minimum residual freeboard through direct geometric constraints, and the modern probabilistic regime, introduced into SOLAS II-1 via resolution MSC.216(82) adopted on 8 December 2006 and entering into force on 1 January 2009, which requires that the attained subdivision index A calculated across all credible damage scenarios equals or exceeds a required index R derived from ship length, ship type and number of persons on board. Special damage stability requirements apply under MARPOL Annex I for oil tankers, under MARPOL Annex II and the IBC Code for chemical tankers, under the IGC Code for gas carriers, and under the Polar Code for ships operating in polar waters. ShipCalculators.com provides a suite of tools covering all principal damage stability computations; a full listing is available at the ShipCalculators.com calculator catalogue.

Contents

Background and history

The Titanic disaster and the Bulkhead Committee

The sinking of RMS Titanic on 15 April 1912, which killed more than 1,500 people, placed watertight subdivision at the centre of international maritime policy for the first time. The ship had been designed in accordance with Board of Trade standards that required subdivision sufficient to keep the vessel afloat with any two adjacent compartments open to the sea. Titanic could in fact float with any four of her forward compartments flooded, but the iceberg opened five, and she sank in 2 hours and 40 minutes. The disaster drew immediate attention to the gap between the subdivision actually provided and the flooding actually encountered.

The British Board of Trade convened the Bulkhead Committee in 1912 and it reported in 1915 with a systematic method for calculating the length of individual compartments that a ship could have flooded without sinking. This method, known as the floodable-length approach, defined the floodable length at any point along a ship’s hull as the maximum length of compartment, centred at that point, which could be flooded without submerging the vessel to the margin line - a limit 76 mm (three inches in the contemporary imperial standard) below the freeboard deck at the ship’s side. The ratio of the permissible length of a compartment to its floodable length defined a factor of subdivision F, and minimum permissible values of F were set as a function of the ship’s length and the number of persons on board.

The first International Convention for the Safety of Life at Sea, signed in 1914, incorporated subdivision requirements based on these principles, but never entered into force due to the First World War.

SOLAS 1929 and the deterministic framework

The SOLAS 1929 Convention, which entered into force in 1933, established the first binding international deterministic damage stability regime. The floodable-length method was codified with explicit curves calculated from hydrostatic data using Bonjean curves - transverse sections of waterplane area at successive waterlines, from which cumulative displacement and centre of buoyancy could be derived as a function of draught. Passenger ships were classified by criterion numeral C, a function of ship length L in metres and the number of persons on board N, with three classes of subdivision standard - one-compartment, two-compartment and three-compartment - corresponding to ranges of C. The factor of subdivision F ranged from 1.0 at the least stringent end to 0.5 for the most stringent ships, meaning that permissible compartment length was at most half the floodable length.

The SOLAS 1948 and SOLAS 1960 Conventions refined the deterministic requirements progressively, tightening criteria for large passenger ships and extending coverage to cargo ships, but retained the same geometric logic: calculate floodable length at each station, apply factor of subdivision, and check that no actual compartment boundary exceeds the permissible length. The method assumed that flooding of a compartment was a discrete event with uniform permeability - the fraction of a compartment’s volume actually available to flood - and did not model the probability that a given collision would breach a given compartment. Permeability values were prescribed by compartment type: 0.95 for passenger spaces, 0.85 for machinery spaces, 0.60 for holds stowed with general cargo, and 0.95 for empty tanks or void spaces.

Post-Estonia developments and the Stockholm Agreement

The loss of the passenger ro-ro ferry MV Estonia on 28 September 1994, which killed 852 people in the Baltic Sea, exposed limitations specific to open-vehicle-deck vessels. The Estonia lost her bow visor in heavy weather, allowing water to enter the car deck. Because the car deck extended the full width and length of the ship without transverse watertight divisions, a relatively small volume of free water accumulated rapidly and destabilised the vessel. She capsized and sank in approximately 30 minutes. The investigation concluded that the deterministic SOLAS requirements, which did not account for water accumulation on a large open deck, were inadequate for ro-ro passenger ships.

Following the Estonia disaster, eight north-west European maritime administrations - Sweden, Denmark, Finland, Norway, Germany, Ireland, the Netherlands and the United Kingdom - adopted the Stockholm Agreement in 1996. The Agreement required ro-ro passenger ships operating in the area to demonstrate residual stability after flooding with a residual significant wave height (hs) of water on the car deck, representing the dynamic effect of waves entering the damaged vessel. The required hs depended on the actual sea area of operation, graded from 0 m for sheltered waters to 4.0 m for the open North Sea. This was the first regulatory instrument to require explicit modelling of wave-induced flooding dynamics rather than treating damaged stability as a purely static problem. The Stockholm Agreement applied as a regional supplement to SOLAS and remained in force until superseded by IMO’s revised ro-ro passenger ship provisions.

MSC.216(82) and the probabilistic transition

The limitations of the purely deterministic framework - particularly its failure to distinguish between damage scenarios of very different probability and its inability to reward subdivision arrangements that reduced the likelihood of catastrophic flooding - had been recognised by the IMO Maritime Safety Committee from the early 1990s. Work on a harmonised probabilistic method accelerated after the Estonia, drawing on theoretical frameworks developed by W. Wendel and later by the SLF (stability and load lines and fishing vessels) sub-committee.

Resolution MSC.216(82), adopted by the Maritime Safety Committee on 8 December 2006, amended SOLAS Chapter II-1 to introduce a unified probabilistic damage stability standard applicable to cargo ships of 80 m or more in length and to passenger ships regardless of length, with a transition date of 1 January 2009 for new ships. The amendment replaced the previous Part B (subdivision and stability of passenger ships) and Part B-1 (subdivision and damage stability of cargo ships) with a unified Part B that applies the same probabilistic index framework to both ship types, though with different required index values.

The transition to the probabilistic standard did not simply tighten existing requirements across the board. For some well-subdivided vessel types, particularly modern passenger ships that had been designed with generous stability margins, the new standard was easier to meet than the old one because it rewarded the actual arrangement of transverse and longitudinal bulkheads rather than requiring compliance with a universal minimum. For other vessel types with few transverse bulkheads or deep holds, the probabilistic standard identified vulnerabilities that the floodable-length approach had not captured, and led to design changes such as the addition of longitudinal centre-line bulkheads, inner-bottom raising or the subdivision of deep double-bottom tanks.

Fundamental concepts

Watertight integrity and flooding progression

Before either the deterministic or probabilistic framework can be applied, the ship must be understood as a collection of watertight volumes separated by structural boundaries. A watertight boundary - a bulkhead, deck, or shell plating - is one that will withstand the head of water applied by flooding without leaking. Weathertight boundaries withstand sea spray and wave action but are not designed to sustain hydrostatic pressure if fully submerged. This distinction is critical: a ventilation duct cover that is weathertight in normal sea conditions may admit large quantities of water when the ship is heeled after flooding, an effect known as downflooding.

Progressive flooding describes the process by which water admitted through an initial breach spreads from the first flooded space to adjacent spaces through openings in watertight or weathertight boundaries. Typical progressive flooding paths include: watertight doors left open or failing under hydrostatic load; cable transits, pipe penetrations, or air pipes not fitted with closing devices; ventilation ducts connecting upper spaces to lower holds; and access hatches to deep tanks fitted with weathertight but not watertight closures. SOLAS Part B Regulation 8 requires that the stability calculation consider downflooding through any openings that cannot be proven to be closed before flooding begins, using a conservative downflooding angle.

The damage stability calculation must therefore model not only the direct flooding of the breached compartment but also the progressive flooding through all connected spaces, and must check at each stage of this sequence that the ship maintains positive righting levers. The sequence is called the progressive flooding check or the intermediate flooding check, and in SOLAS terms it forms a separate requirement from the final equilibrium check embodied in the s-factor.

Residual freeboard and the damage waterline

When a compartment floods, the ship sinks deeper and may develop trim and list. The resulting waterline is the damage waterline, and the vertical distance from this waterline to the main deck edge (or to the margin line) is the residual freeboard. Positive residual freeboard is necessary to maintain reserve buoyancy against wave action in the damaged condition, but the amount required varies by regulatory regime.

For passenger ships under SOLAS II-1, the equilibrium waterline after flooding must not submerge the margin line (76 mm below the freeboard deck at side). For tankers under MARPOL Annex I, the waterline after hypothetical damage must not submerge the cargo oil tanks. For cargo ships, the margin-line criterion applies to each flooding scenario separately, and the damage margin-line clearance calculator provides a direct check. The residual freeboard also determines whether downflooding through open deck openings is a credible flooding path, and whether the weathertight integrity assumed in the stability booklet is achievable in practice.

Permeability

Compartment permeability μ is the fraction of a compartment’s gross enclosed volume that is actually available to be occupied by flood water. A cargo hold filled with baled cotton has a high permeability because most of the volume is air void; a hold filled with iron ore has a low permeability because the ore occupies most of the space. Permeability affects both the volume of water admitted (and hence the sinkage and trim) and the location of the free-surface moment of the flood water (and hence the loss of GM).

SOLAS II-1 prescribes permeabilities by space type for use in standard damage stability calculations. A machinery space, which contains engines, boilers, tanks and equipment that occupy a significant fraction of the hull, is assigned a permeability of 0.85. Passenger accommodation spaces are assigned 0.95, reflecting the large air volumes in cabins and corridors. Cargo holds stowed with general cargo or containers receive 0.70, reflecting a mix of cargo volumes and access voids. Void spaces, ballast tanks and trimming tanks that are empty at the time of damage receive 0.95. These standard values are conservative in many cases, and ship designers may use lower values if they can demonstrate the actual stowage arrangement, subject to administration approval. The compartment permeability calculator and the effective permeability for damage stability calculator support this assessment.

The deterministic regime

Floodable length and factor of subdivision

The deterministic method, which continues to apply to certain smaller cargo ships, tankers under MARPOL, and ships with unconventional subdivision arrangements not amenable to the index method, is based on three geometric quantities computed at successive stations along the ship’s length.

The floodable length FL at any station is the length of compartment, centred at that station, whose flooding would just submerge the margin line. The floodable length calculator implements this computation from hydrostatic input. The margin line is defined as a continuous line running parallel to the uppermost continuous deck edge, located 76 mm below that deck at the ship’s side. For passenger ships, a permeability of 0.95 applies to the flooded volume unless specific lower permeabilities are justified by the compartment’s contents. The damage margin-line clearance calculator checks that the margin line is not submerged for a specified flooding scenario.

The permissible length PL at a station is the product of the floodable length and the factor of subdivision F, where F is derived from the criterion numeral C. For small ships where C is 23 or less, F = 1.0, meaning that a single compartment of full floodable length is permissible. For large passenger ships with high C values, F may be as low as 0.50, requiring transverse bulkheads at half-floodable-length intervals. The required subdivision index R calculator provides the R value under the probabilistic standard; the legacy floodable-length factor is computed separately within ship-specific stability booklets.

The compartment permeability μ is a critical input for both the deterministic and probabilistic regimes. SOLAS Part B prescribes standard values by compartment type, and actual permeability may be calculated as the ratio of the void volume to the total enclosed volume of a compartment. The compartment permeability calculator supports this calculation, and the effective permeability for damage stability calculator handles combined compartments.

Margin line and Bonjean curves

The margin line concept is derived from early naval architecture practice, where the freeboard deck was the uppermost continuous structural deck. The 76 mm clearance was chosen to provide a minimum safety margin against local wave action and listing after damage. If a flooding scenario causes the margin line to be submerged at any point, the compartment arrangement fails the deterministic standard regardless of residual stability.

Bonjean curves, named after the French naval architect A. N. Bonjean who developed the method in the 19th century, represent the cumulative cross-sectional area at each station as a function of draught. Integration of Bonjean areas over the ship’s length gives displacement at any draught and trim. Before digital computation, damage stability calculations required manual integration of these curves, a laborious process that typically required several days of draughtsman time for each loading condition. The advent of digital stability software in the 1970s and 1980s made it possible to evaluate hundreds of flooding scenarios automatically, which in turn made the probabilistic approach computationally feasible.

Special freeboard categories

Ships assigned reduced freeboard under the Load Line Convention - Type B-60 and Type B-100 ships, which receive 60% and 100% reductions from the tabular Type B freeboard respectively - are required to meet enhanced damage stability standards as a condition of the freeboard reduction. The rationale is that a reduced-freeboard ship has less reserve buoyancy and less margin before the deck edge enters the water after flooding. The load line article covers freeboard assignment in detail. The damage stability requirements for Type B-60 and B-100 ships include minimum residual freeboards after one-compartment and, for some ship types, two-compartment flooding, and minimum residual metacentric heights.

The probabilistic regime

Theoretical basis

The probabilistic damage stability framework is founded on the concept that a collision or grounding accident results in damage to a specific zone of the ship’s hull, and that the probability of any given zone being damaged depends on the geometry of that zone relative to the ship’s total subdivision length. The probability is treated as a frequency - the fraction of all real accidents that produce damage of a given extent - estimated from analysis of historical collision and grounding records. Survival after flooding depends on the residual stability characteristics of the flooded ship, expressed as a conditional probability of survival.

The attained subdivision index A is defined as the sum over all groups of adjacent compartments of the product:

A = Σ (pi × vi × si)

where pi is the probability that the damage zone contains compartment group i and no other flooded compartment is outside that zone, vi is the reduction factor for the vertical extent of damage, and si is the conditional probability of survival given that the damage zone corresponds to group i.

This formulation was introduced by the Danish naval architect J. K. Wendel and refined through IMO research projects in the 1970s and 1980s. The attained subdivision index A calculator implements the full index summation across all compartment groups and reference drafts.

Reference drafts

The index A is evaluated at three reference loading conditions, known as reference drafts, which represent different operational states of the ship:

ds is the deepest subdivision draught, corresponding to the fully loaded condition at the summer load line or deepest permissible draught for subdivision purposes.
dp is the partial subdivision draught, set at 60% of ds.
dl is the light service draught, corresponding to the lightest operating condition.

The overall attained index is computed as:

A = 0.4 × As + 0.4 × Ap + 0.2 × Al

The weighting reflects that ships spend more time near the loaded and partial draughts than at the light draught, but that all three conditions must be survivable. The multi-draft damage stability calculator evaluates A across all three reference draughts and combines them according to this weighting.

The p-factor (damage probability)

The p-factor represents the probability that a collision damages a longitudinal zone of length j starting at a distance from the aft end of the subdivision length Ls. The non-dimensional damage length zone is characterised by its position along the ship and its longitudinal extent, with separate distributions for transverse penetration depth, vertical extent above the keel, and the number of transverse zones breached.

The p-factor is computed from a set of empirical distribution functions derived from analysis of collision accident data. Key parameters are:

The normalised starting position of damage along Ls.
The normalised extent of damage along Ls.
The normalised transverse penetration into the ship, expressed as a fraction of the ship’s beam B, with the B/5 limit (one-fifth of the ship’s beam from the outer shell) being the standard transverse boundary for internal subdivision purposes.

The p-factor (damage probability) calculator calculates the contribution of each compartment group to the probability sum. The B/5 penetration assumption calculator verifies whether proposed longitudinal bulkheads comply with the B/5 line.

The s-factor (survival probability)

The s-factor is the conditional probability that the ship survives, given that a specific group of compartments is flooded. Survival is assessed in the static final flooding condition at equilibrium. The s-factor is a function of three stability parameters measured from the righting lever curve GZ in the flooded equilibrium condition:

GZmax, the maximum righting lever in metres.
Range, the range of positive righting levers in degrees beyond the equilibrium heel angle.
θe, the equilibrium heel angle after flooding.

The formulation assigns s = 0 (the ship capsizes with certainty) if GZmax is below 0.05 m, if Range is below 7°, or if θe exceeds 30° for single-compartment flooding (or 25° for simultaneous flooding of adjacent groups). The maximum value s = 1.0 is approached as GZmax and Range increase beyond specified intermediate values. Between the minimum and maximum thresholds, s varies linearly or as specified in the SOLAS II-1 tables. The s-factor (survival probability) calculator implements this calculation. The residual GM check calculator provides a quick screen of residual metacentric height in the flooded condition.

Required index R

The required subdivision index R is determined by ship type and by two parameters: the subdivision length Ls and the number of persons N (for passenger ships). For cargo ships, R depends solely on Ls:

R = 1 − 1 ÷ (1 + Ls ÷ 100 + Ls² ÷ 11,000)

This formula means that shorter cargo ships require lower R values - a ship of 80 m length requires R approximately 0.58, while a ship of 200 m requires R approximately 0.71. For passenger ships, a contribution from the number of persons on board N is added, increasing R beyond the cargo-ship value. The required subdivision index R (cargo) calculator computes R for cargo ships, and the attained versus required index calculator verifies compliance. The full required subdivision index R calculator covers passenger ships.

Vertical extent factor v

The vertical extent factor v accounts for the probability that damage extends upward from the keel to a given height. For ships with multiple decks, damage that penetrates only to the tank top or a low deck will flood fewer volumes than damage reaching the upper decks. The v-factor reduces the credit given to a subdivision arrangement when the damage penetration assumption would not actually reach a particular deck. For ships without horizontal subdivisions in the relevant zone, v = 1 and makes no modification to the p × s product.

For ships with double bottoms and double side shells, the v-factor can be less than 1.0 for damage scenarios limited in vertical extent, which increases the effective p × v × s contribution of these scenarios relative to full-depth damage. This reflects the statistical observation that minor groundings and side-swipe collisions often produce damage confined to the outer shell without penetrating the inner structure. However, the SOLAS probability tables do not currently distinguish between structural arrangements in the v-factor computation; they use a generic distribution of vertical damage extent from accident data. This is one area where the coupling between structural design and damage stability assessment remains incomplete.

Damage penetration assumptions

The standard transverse penetration depth for the p-factor calculation is bounded at B/5 from the outer shell, or 11.5 m, whichever is less. This value was derived from analysis of historical collision data collected by various maritime research institutes and approved by the IMO through the SLF sub-committee process. The B/5 boundary defines the position inboard of which longitudinal bulkheads are assumed not to be breached in a standard damage scenario. A longitudinal bulkhead located exactly at the B/5 line is assumed to be at the boundary of the standard damage zone.

Designers use the B/5 constraint to protect cargo tanks in tankers or to isolate engine rooms from cargo holds. A centre-line bulkhead in a tanker lies well within the B/5 boundary and is assumed to be breached in any damage of standard transverse extent, providing no subdivision credit for preventing flooding of both port and starboard cargo tanks simultaneously. A longitudinal bulkhead outside the B/5 line - a wing tank boundary for example - is credited as an effective separation because the standard damage cannot reach it. The B/5 penetration assumption calculator checks whether a proposed internal longitudinal boundary provides subdivision credit under this assumption.

For damage extending beyond B/5 in severe collisions or ramming scenarios, the SOLAS framework does not require compliance; the standard is calibrated to provide survival in the most statistically probable damage scenarios, not in every possible scenario.

Cross-flooding and equalisation

When flooding is asymmetric - for example, when only a port-side compartment is breached - the ship develops a list. SOLAS allows cross-flooding arrangements (pipes, ducts, or valves connecting equivalent compartments on opposite sides of the ship) to reduce the equilibrium list by transferring water to a symmetric compartment. Two time limits govern whether cross-flooding can be assumed in the stability calculation:

If cross-flooding equalises within 10 minutes, the flooding of both sides may be assumed instantaneous in the s-factor calculation, reducing the effective heel and potentially improving the survival probability.
If cross-flooding takes between 10 and 60 minutes, a time-dependent reduction factor applies.
If cross-flooding takes longer than 60 minutes, no credit may be taken for it in the damaged stability assessment.

The cross-flooding equalisation time calculator estimates equalisation time from pipe diameter, pipe length, head difference and compartment volumes.

For water ingress through a breached shell plate or piping failure, the flooding time - simple orifice estimate calculator provides a first-approximation of flooding duration.

Special regulatory regimes

Oil tanker damage stability under MARPOL Annex I

The MARPOL Convention Annex I imposes damage stability requirements on oil tankers that operate in parallel with, but are distinct from, SOLAS subdivision requirements. The MARPOL standard addresses the probability of oil outflow following collision or grounding, as the primary pollution risk is the release of cargo from breached tanks.

Regulation 23 of MARPOL Annex I specifies that oil tankers of 150 gross tons and above must meet prescribed damage stability standards in a hypothetical damage scenario. The damage extent is defined in terms of ship length L (transverse penetration b ≥ B/5 or 11.5 m, whichever is less; longitudinal extent l ≥ L/3 or 14.5 m, whichever is less; vertical penetration from the keel upward through the full depth). The ship must float after this damage at a position of equilibrium, list and trim that keep all oil-laden compartments above the waterline, with certain specified residual stability requirements.

The hypothetical oil outflow parameter Oc for collision damage and Os for stranding damage are calculated from the location and volume of oil tanks likely to be breached. These parameters drive the concept of mean outflow, which is the expected volume of oil released averaged across all statistically likely damage positions. MARPOL Annex I Regulation 31, adopted through MEPC.117(52) and subsequent amendments, requires that mean outflow does not exceed a value proportional to the ship’s cargo capacity.

Wing tanks in tankers serve a dual function: they act as double-hull protection for the cargo tanks and, if flooded, can generate a list that is addressed by cross-flooding arrangements between symmetric wing tanks. The progressive flooding criterion requires that the ship remain stable as flooding spreads through any opening in the damaged condition.

Chemical tanker damage stability under MARPOL Annex II and the IBC Code

Ships carrying noxious liquid substances in bulk under the IBC Code are categorised by hazard level into three ship types: Type I (highest hazard), Type II (moderate hazard) and Type III (lower hazard). The damage stability standard is most stringent for Type I ships, which must survive flooding of two adjacent compartments, and least stringent for Type III ships, which require only one-compartment survival. Type II ships must meet a two-compartment standard in certain length ranges.

The damage extent for chemical tankers under MARPOL Annex II is similar to the MARPOL Annex I oil tanker standard but with different penetration parameters for the more hazardous product categories. The residual stability criteria after flooding require a minimum righting lever curve with a maximum GZ of at least 0.10 m over a range of at least 20° and an area under the GZ curve from the equilibrium position to the downflooding angle of at least 0.0175 m·rad. These thresholds are higher than those applied to general cargo ships, reflecting the greater consequence of loss of the vessel or cargo.

Gas carrier damage stability under the IGC Code

Gas carriers carrying liquefied gases in bulk are subject to the International Code for the Construction and Equipment of Ships Carrying Liquefied Gases in Bulk (IGC Code), which prescribes damage stability standards similar to the IBC Code. Ships carrying flammable gases must survive two-compartment flooding when the flammable cargo is in tanks within the damaged zone, with the equilibrium condition maintaining cargo containment above the waterline. Pressure vessel tanks that are structurally independent of the ship’s structure may be credited with greater flooding resistance than gravity tanks.

Polar Code damage stability

The Polar Code, made mandatory under both SOLAS and MARPOL, imposes supplementary damage stability requirements on ships operating in Arctic and Antarctic waters. Category A ships (designed for operation in at least medium first-year ice, which may include old ice inclusions) must meet enhanced two-compartment flooding standards in the polar bow and midbody regions, where ice impact is most probable. Category B ships (designed for operation in thin first-year ice, which may include old ice inclusions) must meet enhanced one-compartment flooding standards.

The justification for the enhanced polar requirements is that ice damage is typically more extensive longitudinally than collision damage, that rescue in polar regions is limited and may take many hours or days, and that the consequences of sinking in ice-covered waters are more severe than in temperate seas. The damage extent assumptions for polar ships reflect this by using larger transverse penetration factors than the standard SOLAS values.

Ro-ro passenger ships and dynamic stability

The treatment of water on a ro-ro passenger ship car deck has evolved substantially since the Estonia disaster. The current SOLAS framework, as amended by resolution MSC.421(98) adopted in 2017 and addressing aspects of dynamic damage stability, requires ro-ro passenger ships to demonstrate survivability after flooding with an equivalent residual significant wave height on the car deck. The vehicle deck is modelled as a flooded compartment with a characteristic permeability, and residual stability is assessed with the additional free-surface moment generated by the water on deck.

The critical parameter is the heel angle at which water begins to ship over the lower edge of the vehicle deck opening in a damaged ship. If this angle is reached before adequate reserve stability has been demonstrated, the ship fails the criterion. Cross-flood ducts in the car deck, combined with water drainage arrangement, can reduce the equilibrium heel and improve compliance. The dynamic aspects of wave-induced flooding remain an active area of research and are addressed in the IMO second-generation intact stability criteria framework, which shares methodology with the damage stability assessment.

Ship types and subdivision geometry

Passenger ships

Passenger ships are subject to the most stringent subdivision requirements in SOLAS, because the large number of persons on board and the difficulty of evacuation from a large vessel demand that the ship remain stable and afloat for an extended period after damage. The required index R for passenger ships increases with the number of persons N in addition to the subdivision length, reaching values above 0.90 for large cruise ships.

SOLAS requires that passenger ships meet the A ≥ R standard at each reference draught individually, not only in the combined weighted index. The deepest subdivision draught ds corresponds to the maximum draught at which the ship normally operates, accounting for ice accretion and fuel stores. The light service draught dl may correspond to an arrival condition with consumed stores and fuel.

Vertical subdivision in passenger ships - watertight decks or partial tank tops within cargo or machinery spaces - can contribute to the v-factor and improve the attained index. Passengers and crew above the bulkhead deck are implicitly assumed to be in safe locations during a damage scenario, but the progressive flooding criterion ensures that staircases and lift trunks cannot transmit flooding to undamaged zones.

Cargo ships

Cargo ships of 80 m or more in length are subject to the probabilistic standard under SOLAS II-1. Ships below 80 m may apply the deterministic method or, at the flag state’s discretion, the probabilistic method. The subdivision length Ls is the distance measured at the deepest subdivision draught between the forward and after limiting bulkheads enclosing all watertight compartments.

Hold permeability in cargo ships depends strongly on the nature of the cargo and packing arrangement. The SOLAS II-1 standard prescribes a permeability of 0.70 for general cargo holds, 0.95 for empty or ballast holds, and 0.60 for holds stowed with bulk grain or similar dense cargo. The compartment permeability calculator allows derivation of actual permeability from compartment void volume data.

Bulk carriers

Bulk carriers present specific damage stability challenges because their holds are large, poorly subdivided spaces with high permeability when empty (0.95) but lower permeability when filled with dense cargo. SOLAS Chapter XII applies additional structural requirements to bulk carriers, including double-skin side construction for certain ships, which incidentally improves damage resistance. A flooded hold of a bulk carrier represents a very large free-surface moment that can severely reduce residual GM even when the ship does not sink.

Container ships

Container ships have large holds with relatively small intermediate heights between tank tops and hatch covers. The cargo itself has significant permeability (typically 0.60 to 0.70 depending on the fill of containerised goods), and the stacked containers above the hatch covers may destabilise the ship through wind heeling if the ship develops a list after flooding. The attained index calculation for a large post-panamax container ship typically involves thousands of compartment group combinations, requiring automated stability software.

Tankers

Oil tankers addressed under SOLAS II-1 are generally well-subdivided by virtue of the cargo segregation requirements of MARPOL Annex I, which divides the cargo length into a series of cargo tank blocks with wing tanks and a double bottom. This subdivision frequently provides an attained index well above the required value for a cargo ship of equivalent length. However, the longitudinal bulkheads between wing tanks and centre tanks lie within the B/5 penetration boundary in many tanker designs, particularly narrower vessels, meaning that a standard damage scenario floods both the wing and centre tank sections in the same longitudinal zone. This reduces the subdivision credit and may require that additional transverse bulkheads be fitted within the cargo length.

The oil tanker and chemical tanker articles cover the structural arrangements in more detail. The interaction between the MARPOL Annex I stability standard and the SOLAS II-1 standard means that tanker designers must satisfy both sets of criteria simultaneously, and the more stringent of the two governs in practice.

General cargo ships and multipurpose vessels

General cargo ships and multipurpose vessels present varied subdivision geometries depending on the number of holds, the arrangement of tweendeck spaces and the disposition of ballast tanks. Tweendecks, which are intermediate partial decks within a hold, divide the vertical extent of the flooding space and can contribute to the v-factor if they are structurally watertight. However, tweendecks are frequently fitted with large hatches or non-watertight doors for cargo access, which negates any subdivision credit. The v-factor can only be claimed for a horizontal subdivision if all openings in it are either closed watertight or located above the damage waterline in the flooded condition.

Multipurpose vessels that carry both general cargo and containers or roll-on-roll-off cargo in the same voyage face the challenge that the permeability of the hold changes with the cargo type. A hold filled with loose bulk cargo has a permeability of 0.60, while the same hold loaded with containers has a permeability of 0.70. The stability booklet must specify the applicable permeability for each loading condition, and the damage stability calculation must use the permeability that gives the worst result for the criterion being checked - generally the higher permeability, which produces greater sinkage and free-surface moment.

Regulatory compliance and classification

Flag state and classification society roles

Damage stability compliance is primarily a flag state obligation under SOLAS: the flag administration is responsible for ensuring that ships flying its flag are designed, constructed and equipped in accordance with SOLAS II-1. In practice, most flag states delegate the technical review and approval of stability calculations to a recognised organisation, which in the maritime context means a member society of IACS. The classification society reviews the designer’s stability booklet, the subdivision drawings and the damage stability calculation output, and if satisfied issues an endorsement that the damage stability requirements are met.

The stability booklet, required by SOLAS III/2-1 and SOLAS Chapter II, must contain the full damage stability calculation for each loading condition covered by the certificate, including the p-factor, s-factor and v-factor contributions of every compartment group evaluated, the A and R values at each reference draught, and the margin-line clearances. It must also contain guidance on the watertight closure status assumed in each calculation (which doors, hatches, and vents are closed) and instructions to the master about maintaining these closures in service.

Port state control officers under the port state control regime may inspect the stability booklet and verify that the master and officers understand the damage stability limits applicable to the current loading condition. A ship operating in a loading condition not covered by the approved booklet, or with incorrect understanding of the watertight closure requirements, is liable to detainment. The principal MoUs, including the Paris MoU for European and north Atlantic waters and the Tokyo MoU for Asia-Pacific, record damage stability deficiencies as a category of port state inspection finding.

New ship versus existing ship applicability

The MSC.216(82) probabilistic standard applies to ships whose keels were laid on or after 1 January 2009. Ships with keel-laying dates before this date were subject to the previous deterministic requirements under SOLAS Part B-1 (for cargo ships) or Part B (for passenger ships). Passenger ships built before 1 January 2009 that have since been substantially modified may be required to demonstrate compliance with the new standard, depending on the nature and extent of the modification and the guidance of the relevant flag state.

The practical consequence is that a significant portion of the world fleet, particularly cargo ships built before 2009, continues to operate under the deterministic standard throughout their service life. As these ships age out of the fleet and are replaced, the proportion of the cargo fleet under the probabilistic standard increases. Classification society surveys at each special survey periodically verify that the as-built subdivision remains consistent with the approved damage stability calculations, and that no structural modifications have been made that would invalidate the calculation.

Equivalences and alternative compliance

SOLAS Regulation II-1/55 permits administrations to accept alternative arrangements or materials if they provide a level of safety not inferior to that achieved by the standard requirements. In the context of damage stability, this provision has been used to allow novel subdivision arrangements - for example, combined transverse-and-longitudinal subdivision schemes in wide-body vessels - to be assessed by a probabilistic analysis rather than by strict conformance with the penetration assumptions. This process requires flag state approval and, in many cases, IACS classification society concurrence.

The use of model testing and time-domain numerical simulation as evidence of equivalent safety in damage stability has been explored in research but is not yet fully codified in SOLAS. IMO has issued guidance circulars on the interpretation of specific regulations, and the SDC (Ship Design and Construction) sub-committee, which inherited the work of the former SLF sub-committee, continues to develop unified interpretations.

Notable casualties

Estonia (1994)

The Estonia sank in the Baltic Sea on 28 September 1994 with the loss of 852 lives, the second worst peacetime maritime disaster in European history. The bow visor, a hinged forward door sealing the car deck ramp, failed in rough weather with significant wave heights of approximately 4 to 5 m and separated from the ship. Water flooded the car deck, which was undivided by transverse watertight bulkheads for its full length, consistent with the ro-ro ship design of the period. The free-surface effect of the accumulating water reduced metacentric height rapidly and the ship heeled beyond 40° within 20 to 30 minutes of the visor loss. The Estonia’s casualty directly drove development of the Stockholm Agreement and the subsequent IMO revisions to ro-ro damage stability, and remains the reference event for all ro-ro car-deck flooding analysis.

Costa Concordia (2012)

The Costa Concordia ran aground off Isola del Giglio, Italy, on 13 January 2012, killing 32 people. The ship struck a rock pinnacle that breached the hull across at least five watertight compartments, well in excess of what even the most stringent passenger ship subdivision standard was designed to accommodate. Investigations noted that watertight door closures were slow and that crew familiarity with the progressive flooding sequence was limited. The Costa Concordia’s casualty led IMO to examine the adequacy of watertight door drill requirements and the instructions given to crew about maintaining watertight integrity during an evolving flooding event. A revised circular on enhanced safety assessment for large passenger ships emerged from the review. The ship was eventually re-floated in a complex salvage operation in 2014, the largest salvage operation ever attempted.

Sewol (2014)

The South Korean ro-ro passenger ferry Sewol capsized on 16 April 2014, killing 304 people, predominantly secondary school students on a school trip. The immediate cause was a sharp turn combined with the effects of unauthorised conversion and excessive additional top-weight that had reduced the ship’s effective metacentric height. Although the Sewol’s sinking was fundamentally an intact stability failure rather than a damage stability failure, the subsequent flooding of the car deck as the vessel listed contributed to the speed of sinking. Investigation findings on the dangers of reduced GM in ro-ro vessels reinforced ongoing IMO work on second-generation intact stability criteria and the treatment of wave-induced heel effects in damage scenarios.

El Faro (2015)

The US-flagged container ship El Faro sank in Hurricane Joaquin on 1 October 2015, with the loss of 33 lives, in the Atlantic Ocean near the Bahamas. The National Transportation Safety Board investigation determined that flooding entered the ship through open ventilation scuttles on the 2nd deck that had not been secured before the ship entered the hurricane. El Faro illustrates a category of damage stability casualty driven not by structural failure but by the failure to maintain watertight integrity through openings susceptible to flooding in heavy weather - a problem that cannot be addressed by the geometric arrangement of watertight bulkheads but requires operational controls and drills. The casualty was also associated with downflooding through a ventilation duct connected to a flooded hold, which progressively admitted water to adjacent holds.

The El Faro investigation resulted in recommendations for more explicit training requirements regarding the closure of weathertight openings before entering heavy weather, the provision of more reliable sea-state information to masters, and a review of voyage planning obligations for vessels operating near tropical cyclones.

MSC.510(105) and recent developments

2022 amendments entering force January 2024

Resolution MSC.510(105), adopted by the Maritime Safety Committee at its 105th session in April 2022, introduced amendments to SOLAS II-1 that entered into force on 1 January 2024. The amendments addressed a number of practical issues identified since the 2009 entry into force of the probabilistic standard, including:

Clarification of the treatment of inner-bottom tanks and double-bottom tanks in the computation of the transverse penetration factor.
Revised criteria for the treatment of damage to side casings - structural sponsons extending beyond the main hull envelope that affect the effective beam for penetration calculations.
Revised guidance on the treatment of damage to wing tanks adjacent to double-bottom tanks, relevant to tankers that retain deterministic-equivalent arrangements.
Amendments to the flooding time requirements for cross-flooding arrangements, clarifying the 10-minute and 60-minute boundaries.

Second-generation intact stability criteria

Concurrent IMO work on the second-generation intact stability criteria, which addresses dynamic stability phenomena such as parametric rolling, dead ship condition and excessive acceleration, shares mathematical methodology with damage stability assessment. The revised criteria introduced the concept of direct stability assessment using numerical simulation, and the same time-domain simulation tools used for intact stability can be applied to assess survivability in damage stability scenarios with wave action, yielding a more realistic picture of capsizing risk than the static s-factor formulation.

Numerical flooding simulation

The static s-factor formulation assumes that flooding reaches equilibrium before capsizing occurs, which is not always true. Modern classification societies and specialised software vendors have developed time-domain flooding simulation tools that model the transient flooding process through breach openings, cross-flooding ducts and progressive flooding paths. These tools can capture the effect of ship rolling during flooding, the time to reach equilibrium, and the probability that transient heel angles exceed survivable limits before equilibrium is reached. IMO has begun incorporating requirements for numerical flooding simulation in specific rule interpretations, particularly for novel ship types and for verification of complex cross-flooding arrangements.

IACS common structural rules and damage stability

The International Association of Classification Societies (IACS) Common Structural Rules for Bulk Carriers and Oil Tankers specify structural scantlings that influence the extent of damage in a collision. The interaction between structural strength and damage extent is not explicitly modelled in the standard SOLAS probabilistic framework, which uses empirical damage extent distributions derived from accident data and does not distinguish between different structural configurations. Research programmes funded by IACS member societies and the European Commission have proposed coupled structural-stability assessment methods that would allow credit for reinforced collision barriers in the p-factor computation.

Damage stability in the design process

Damage stability compliance must be addressed from the earliest stages of the preliminary design process, before the arrangement of main transverse bulkheads is fixed. The number and position of watertight transverse bulkheads is one of the most fundamental decisions in a ship’s general arrangement, and moving a bulkhead by even a few metres in the final design can have a significant effect on the attained index. A ship designed without considering the damage stability requirements from the outset may require costly structural modifications late in the design process, or may accept a reduced cargo capacity to provide additional subdivision length.

The design process for damage stability typically involves the following iterative sequence: define the preliminary general arrangement with bulkhead positions; compute the p-factors for each compartment group defined by the bulkhead layout; estimate the s-factors for each group from approximate hydrostatic data; compute the attained index A and compare with R; if A < R, adjust the bulkhead positions or add longitudinal subdivisions and repeat. The attained subdivision index A calculator and the multi-draft damage stability calculator support this iterative process at the early design stage before full stability software models are available.

The sensitivity of A to bulkhead position is not uniform along the ship’s length. Bulkheads bounding the machinery space often control because the machinery space has a prescribed permeability of 0.85 and relatively low residual stability due to the mass of installed machinery. Adding a second transverse bulkhead within or adjacent to the machinery space can substantially improve the attained index by splitting the high-consequence machinery space flooding scenario into two smaller scenarios of lower combined probability.

For passenger ships, the addition of a centre-line longitudinal bulkhead in accommodation decks can improve s-factors by limiting the asymmetric flooding that generates large equilibrium heel angles. However, centre-line bulkheads within the standard B/5 penetration zone do not reduce p-factors and provide no credit against the probability of flooding; their benefit is entirely in the survival probability of an already-flooded scenario. This asymmetry in the treatment of longitudinal and transverse subdivision sometimes leads to counterintuitive results where a more subdivided arrangement provides lower A because the additional boundaries create more damage zone combinations of intermediate probability and low survival probability.

Calculations and verification

Tools for practitioners

The principal computations required for probabilistic damage stability verification are:

Calculation of the p-factor for each compartment group from the normalised position and extent of the group within the subdivision length. The p-factor (damage probability) calculator provides this output.
Calculation of the s-factor from the flooded waterline geometry, including GZmax, range and equilibrium heel. The s-factor (survival probability) calculator implements the SOLAS II-1 formula with the applicable thresholds. The residual GM check calculator is used for a rapid screen before the full s-factor evaluation.
Summation of the p × v × s products across all compartment groups at each reference draft to obtain As, Ap, and Al, and combination to give the overall A. The attained subdivision index A calculator and the multi-draft damage stability calculator perform this summation.
Comparison of A against R, where R is obtained from the required subdivision index R (cargo) calculator.
Verification of the margin-line criterion for each flooding scenario using the damage margin-line clearance calculator.
If cross-flooding arrangements are provided, verification of equalisation time using the cross-flooding equalisation time calculator.

In practice, full probabilistic damage stability verification for a large ship involves hundreds to thousands of compartment group combinations and is performed using dedicated stability software such as NAPA, Maxsurf Stability or equivalent programs. The calculators above are suited to individual scenario checks, educational use, and audit verification of software outputs.

Relationship with hydrostatics and intact stability

Damage stability assessment is downstream of the ship’s hydrostatics and Bonjean data, which define the displacement, centre of buoyancy, metacentric radius and waterplane area at all draughts and trims. The metacentric height in the damaged condition (GM damaged) differs from the intact value because the addition of flood water reduces freeboard, changes trim, and generates a free-surface effect in the flooded compartment. The trim and list article covers the geometry of the flooded waterplane.

The relationship between block coefficient and subdivision is indirect but significant: a fine-hulled ship with low block coefficient has narrower compartments for a given beam, which reduces the probability that a longitudinal bulkhead falls within the B/5 penetration boundary. Hull form design choices therefore have consequences for damage stability compliance that must be evaluated early in the design spiral.

References

International Maritime Organization. SOLAS - International Convention for the Safety of Life at Sea, Consolidated Edition 2020. IMO Publishing, London, 2020.
IMO Resolution MSC.216(82). Adoption of amendments to the International Convention for the Safety of Life at Sea, 1974. Maritime Safety Committee, 8 December 2006.
IMO Resolution MSC.510(105). Adoption of amendments to the International Convention for the Safety of Life at Sea, 1974 (SOLAS Chapter II-1). Maritime Safety Committee, April 2022.
IMO Resolution MSC.421(98). Amendments to the International Convention for the Safety of Life at Sea, 1974 (relating to ro-ro passenger ships). Maritime Safety Committee, June 2017.
International Maritime Organization. MARPOL - International Convention for the Prevention of Pollution from Ships, Consolidated Edition 2017. IMO Publishing, London, 2017.
International Maritime Organization. International Code for the Construction and Equipment of Ships Carrying Dangerous Chemicals in Bulk (IBC Code). 2007 Edition. IMO Publishing, London.
Joint Accident Investigation Commission of Estonia, Finland and Sweden. Final Report on the Capsizing on 28 September 1994 in the Baltic Sea of the Ro-Ro Passenger Vessel MV Estonia. 1997.
Italian Ministry of Infrastructure and Transport. Costa Concordia Casualty Report. Direzione Generale per le Investigazioni Ferroviarie e Marittime, 2013.
National Transportation Safety Board. Sinking of US Cargo Vessel El Faro, Atlantic Ocean, Northeast of Acklins and Crooked Islands, Bahamas, October 1, 2015. NTSB/MAR-17/01. 2017.
Ministry of Ocean and Fisheries, Republic of Korea. Final Investigation Report: Sinking of the Passenger-Ro-Ro Vessel Sewol. 2014.
Bulkhead Committee. Report of the Committee Appointed by the Board of Trade to Consider the Standards of Subdivision of Merchant Steamships. HMSO, London, 1915.
Wendel, K. “Subdivision of ships.” Jahrbuch der Schiffbautechnischen Gesellschaft, 1960. (Original probabilistic framework development.)
Papanikolaou, A. D. (ed.). Risk-Based Ship Design: Methods, Tools and Applications. Springer, Berlin, 2009.
IMO SLF Sub-Committee. Development of a revised SOLAS Chapter II-1 (Parts A, B and B-1): Explanatory Notes to the SOLAS Chapter II-1 Subdivision and Damage Stability Regulations. SLF 49/INF.3, 2006.