SAFETY TECHNOLOGY

Bulk carriers: bulkheads and survivability

Captain Dennis Barber is currently
principal consultant with Marine &
Risk Consultants. The observations in
this article (which add a new view to
many words that have appeared in this
journal from naval architects on this
difficult subject) are based on
experiences both as mariner and
superintendent. He has also studied
probabilistic assessment of hulls and
circumstances rather than historical Fig 1. The assumption, 'single compartment': ILLC66 interpretation.
facts, which, as he suggests, are hard
to come by when they are on the
ocean floor. This article represents
one part of the problem; it also points
to the law, in the form of the Load Line
Convention, as being adequate, if period. Mechanisms by which this is possible
interpreted correctly. Without a have been examined and one in particular
holistic approach, bulk carrier failures should cause concern. A large proportion of
and losses will continue, the author the current bulk carrier fleet remains
alleges, especially in the high-tensile- relatively unaffected by recent improvements
steel hulls that have dominated since in internal construction introduced by both
the mid-1980s. IMO and IACS. Reference is made to the
requirements of SOLAS Chapter XII and
ULK carriers have been an important IACS URS 19 regarding strengthening
B issue for the international shipping
community for the majority of the past
bulkheads to withstand a likely head of water
and associated dynamic forces, in the event of
decade. Problems with ships disappearing at a compartment flooding. These apply to new Fig 1a. Asymmetric flooding due to failure of
sea were causing unease as early as 1980. bulk carriers of single side-skin construction. ballast tank trunkings in side shell failure.
The International Maritime Organisation Existing ships rely on another part of
(IMO) commissioned a formal safety Chapter XII and URS18, which requires only
assessment of bulk carriers (FSA BC) in the bulkhead between Nos 1 and 2 holds to be
1999, and at the same time, other formal strengthened in the same way. Furthermore,
safety assessments were carried out and were ships with double side-skins are completely between the keel, and the deck between
reported to IMO. exempt. recognised longitudinal subdivisions), the
In its deliberations, participants in those The efforts of both IMO and IACS to date assumptions shown in Fig 1 and Fig 1a cannot
FSA studies analysed extensive historical data are commendable but it is believed that these be sustained. Furthermore, because the
on bulk carrier losses, but were faced with a measures fall short of what is needed to connections between top and lower ballast
fundamental problem. Bulk carrier losses, by prevent continuing losses of bulk carriers. tanks pass through cargo holds (usually in
their very nature, leave little or no evidence, This is because these measures do not cross- way of the bulkheads), if the side shell is
and much of what is recorded is inevitably the refer to the requirements of the International damaged, the trunkways may flood, and
result of speculation, some of it well Convention on Load Lines 1966 (ICLL66) for adjacent ballast tanks will flood, in addition
informed, some little more than journalistic reduced freeboard ships that require single- to the original assumed cargo hold. This is
surmise. It is notable that few records include compartment survivability. reinforced by fact (Fig 1b). Asymmetric
any information on internal integrity of the It has been assumed that a ship's capability flooding is apparent in the list: this ship later
ships that are lost. to survive has been calculated for each bulk sank.
This is not surprising, many cases have carrier assigned a reduced B-type freeboard This does not take any account of dynamic
either disappeared without trace or - (B-60). However, many existing ships cannot forces due to sloshing within the
understandably - survivors did not witness be guaranteed to withstand such a flooding compartments. It is recognised that with
what happened below deck. However, in event. many lighter cargoes, cargo would contribute
assessing events leading to a ship's sinking it In most dry bulk carriers, a distinction is to the strength of the bulkhead (Fig 2a), and
is correct to say - at least in the case of larger made between 'dry' hold bulkheads and those
ships that comprise the majority of suspected forming the boundaries of 'wet' or ballast
Fig 1b. Asymmetric flooding of ballast tanks.
cases - a single compartment flooding should holds. Prior to IACS' unified requirements
Hold (No 4) and starboard side ballast tank(s)
not be sufficient to sink a ship. Therefore, referred to above, each classification society were broached through side-shell failure. The
after the initiating event, there must have applied its own rules regarding the strength of ship later sank.
been subsequent events that allowed structures such as bulkheads. Depending on
progressive flooding to occur. the society concerned, an average dry
For a ship to sink so rapidly, that in many bulkhead would be constructed to withstand a
cases, not even a distress message was static head of water between 50% and 75%
possible, the influx of water must have been head. A figure of around 60% is typical. If
extraordinarily massive, and in a very short this head is superimposed on a general
arrangement or capacity plan, lower values
that are possible below the load waterline will
Note: Due to the considered importance of this subject, a be seen.
somewhat similar article to this will appear in Seaways, The conclusion must therefore be that if a
the journal of The Nautical Institute. The views presented ship suffers flooding of any single
here are solely those of the author and do not reflect those compartment (remembering that in the load
of any company or organisation. line sense, a compartment includes all

THE NAVAL ARCHITECT OCTOBER 2002 9

 SAFETY TECHNOLOGY

perhaps it is such an assumption that has led

to the position that exists. Perhaps it is not
generally understood that cargoes of greater
density do not generally contact the bulkhead
at all. But for these ships concerned, there is
no caveat issued regarding the density of
cargo to be carried. The unsupported
bulkhead is totally reliant on its own strength
to remain intact. This can been seen in Fig 2b.
Fig 3 shows the presumed reality. After an
initiating event of side shell or hatch failure,
the hold will flood. When the water level
Fig 2a. Holds fully loaded with low-density Fig 2b. Holds loaded with high-density cargo.
reaches the level at which the bulkhead
cargo. Multiple compartment flooding may result.
strength reaches its limit, the bulkhead, if
unsupported by cargo, will fail (Fig 4).
Subsequently, bulkheads will continue to fail
and more holds will flood until a watertight
bulkhead at the ballast hold is intercepted This phenomenon has been verified by a way' vessel much earlier than may be allowed
(shown as bold green in Fig 3), by which time review of close-quarter encounters in the for by a watch officer of this 'giving way'
the buoyancy forward will have been almost Dover Strait routeing system. The net effect vessel. Being the larger target reduces the
completely destroyed. The hull structure of of this will be a greater risk of damage to aft possibility of a near miss and increases the
the largest ships will be under severe stress sections in collisions that have occurred possibility of a hit. The consequences of
and may well fail at this point, as did the because of interaction. That is, when ships being struck abaft No 1 hold will almost
Mineral Dampier after she was punctured in a manoeuvre too close during overtaking, or by certainly be the sinking of the bulk carrier.
collision. errors in turning calculations when intending Apart from the high risk to life, in a restricted
As indicated in Mineral Dampier's case, to pass astern of another ship. channel the wreck will represent a sizeable
another aspect in which rules and regulations The Mineral Dampier/Hanjin Madras underwater obstruction that could, in extreme
fail, is in the protection of existing ships collision followed a similar pattern when the cases, close or severely restrict access to ports
against the hazards and mitigation of damage two ships encountered each other in a crossing and threaten other ships that might strike the
that occurs to the primary barrier away from situation. On that occasion, the encounter was wreckage.
No 1 hold. The strengthening of the bulkhead in open waters. In a traffic separation scheme There are high-risk areas associated with
between 1 and 2 holds will do nothing to the likelihood of such an encounter is greater. collision risk in traffic separation schemes:
protect the ship from a failure in any hold Bulk carriers in such circumstances are
abaft that position, unless it is a hold designed seriously at risk because of the subdivision • where there is a high density of crossing
to accommodate a full head of ballast. In inadequacies described above. Existing ships traffic
modern ships, such holds are normally with only hold 1/hold 2 bulkhead strengthened • where a scheme encounters a corner, such
restricted to only one. There are a number of and loaded to their marks with dense cargo will as at Basurelle in the English Channel. The
known cases where side shell has failed in No be most at risk. larger ship, will inevitably have a deeper
3 or 4 holds. It can be shown that a slow moving bulk draught and be slower. Faster ships will
carrier is potentially more likely to be victim of therefore constantly overtake her. When she
Collision scenarios a collision during overtaking than a faster ship, encounters a corner she is fully entitled to turn
At the time of the 1966 Load Line such as container ship, which is more likely, - if necessary across the intended course of the
Convention, the most likely zone in which because of its speed to be overtaking. overtaking vessel. She is likely to have little
ships were expected to suffer damage due to Furthermore, fine lines of faster ships make choice because of the need to stay in deep
collision was forward. Since 1967 however, them a projectile with greater penetrating water. She is then dependent on the
the introduction of traffic separation schemes power than the bluff-bowed bulk carrier or overtaking ship being alert to the change and
has changed the way in which ships encounter tanker. Bulk carriers in particular are therefore making the necessary alteration to her own
each other. highly susceptible to side penetration if a course. Reference to the reports in the MARS
In congested waters, it will more often be close-quarter situation develops into a scheme of the Nautical Institute and to the
the case than not that ships will be negotiating collision. Furthermore, damage is more likely Belcher research would imply that such
a traffic separation scheme, which means to be in the after part of the hull. All the more awareness is lacking in many cases
chances of meeting another ship head on, or reason to urgently review the policies of • where a scheme converges on a special
nearly so, are virtually eliminated. Instead, leaving older ships unprotected. route, such as the deep-water route at
close-quarters situations that do develop will It can also be shown that a larger bulk carrier Sandettie in the Dover Strait. The tendency of
be more associated with same direction in a collision will result in more catastrophic larger, deep-draught ships is to cut across the
travel, ie, overtaking or crossing at a broad consequences. Being proportionately less headings of other ships navigating too close to
angle such as ferries crossing a strait. manoeuvrable, it is at the mercy of the 'giving the separation zone; they have little other

Fig 3. The reality. 'Single compartment' ILLC66 interpretation. The bold green lines are watertight bulkheads.

THE NAVAL ARCHITECT OCTOBER 2002 11

SAFETY TECHNOLOGY

structure to comply on paper but not in reality

is inadequate to achieve the principal aim of
the ICLL - to prevent the sinking scenario that
will lead to loss of life.
Recommendations to improve the
effectiveness of primary barriers (side shell,
hatch covers) in preventing seas entering the
hull should continue to be developed, but the
possibility of such barriers being penetrated -
especially in the collision situation, must still
be considered. Adequate contingency must be
in place to prevent catastrophe. The
regulations contained within ICLL66 in this
respect are currently adequate. Some re-
wording may help clarification, but if ships
are built to comply with those standards their
chances of survival will be much higher than
currently appears to be the case for bulk
carriers carrying dense cargoes.

Available options
A measure of safeguard would be achievable
if the existing ships that remain vulnerable
were only to carry lighter density cargoes. By
shoring up the bulkheads, cargo itself would
prevent their collapse (Fig 2a). Unsupported
bulkheads at full load drafts must be avoided.
If, on the other hand, it could be proved that
the bulkheads on existing ships were strong
enough to resist the forces outlined above,
then there is no reason why they should not
continue to carry the more dense cargoes. Age
alone must not be the deciding factor. Many
older ships are built to a much higher strength
standard than more modern optimised designs.
New class rules implemented by IACS do
address this bulkhead issue for new ships, but
existing ships are not required to achieve the
same standards. The same applies to the
requirements contained in SOLAS Chapter
XII. Both new and existing ships sail in the
Fig 4. The dramatic results of bulkhead collapse during a ballasting trial on a new ship. Note that same oceans where they will be subjected to
water on the far side is probably still in place because, being a 'dry' hold, it is not equipped with the common influences attacking the hull from
ballast pumping suctions. the outside. Quite apart from the fact that
degradation will make the existing ship more
susceptible to failure, it cannot be regarded as
more expendable than the new ship. The idea
that vulnerable ships pre-1998 will eventually
choice. This increases the risk of these large pressures, and these ships remain at risk. They disappear is unacceptable. Disappearances at
targets falling victim to the faster projectiles represent a sizeable proportion of the world bulk sea cannot be used as an alternative to a
overtaking them. carrier fleet. demolition policy. The seafarers manning
Many of the same ships, however, have been older ships possess lives that are equally as
Conclusion assigned reduced freeboards under the ILLC66, valuable as those of their more fortunate
These situations described represent a despite this weakness that appears to contravene colleagues aboard newer ones.
significant risk to bulk carriers more than to any regulation 1 of that convention, which requires There can be no legitimate reasoning,
other type of ship, with the possible exception that ships shall be of adequate strength for the therefore, that allows greater risks to exist on
of general cargo ships. The progression of freeboards assigned. The single-compartment one ship above another, and indeed the
flooding events, however, is much more likely flooding survivability requirement of the ICLL66 does not do that. But the discussion
to be catastrophic in bulk carriers carrying high- reduced B freeboard must therefore be above should show that such risk on bulk
density cargo. Casualty data records for bulk considered to mean that the internal structure of carriers is unacceptably high - a situation that
carriers do not list bulkheads as having failed, a ship is strong enough to withstand all the appears inadvertently to have arisen under
for the understandable reasons given earlier. forces to which it will be exposed in the current interpretations of the convention.
However, risks are significant and there can be circumstances envisaged by ICLL66 - including Clearly such interpretations are wrong and
few other explanations for the rapid sinkings - flooding. The stated consideration in regulation must be corrected. Delegation of technical
now considered a trademark of the bulk carrier. 1 therefore, that conformity with the tasks to classification societies will almost
If the bulkheads were stronger there is little requirements of a classification society certainly continue but it should be
doubt that it would buy time. In some losses, recognised by the administration, must surely be remembered that IACS - the main
the extended time to sink suggests that questioned. Under such controls, ships have representative body for class - does not govern
bulkheads did hold, at least for a period been built that cannot stay intact internally when all the shipping in the world. That
sufficient to enable evacuation. However, flooding occurs. responsibility falls to the administrations that
existing ships are not required by class rules to ICLL66 does not concern itself with the cargo make up IMO. They alone, in the form of
have bulkheads - other than the one between carried (dense cargoes pose a greater risk), only IMO, can bring about the changes necessary
holds 1 and 2 - built to withstand such water the ability of the ship to stay afloat. For a ship's to make the corrections.

12 THE NAVAL ARCHITECT OCTOBER 2002