Analysis of Incidents

Discover a comprehensive analysis of maritime incidents reported by SIGTTO members, featuring insights from terminal managers and ship operators.

This study explores accident frequency, near-miss events, and effective safety measures in gas tanker operations, providing valuable data and recommendations for enhancing maritime safety.

The Data-Base of Incidents

Among SIGTTO members who were addressed as part of this study in 1995, a high proportion of terminal managers and ship operators responded. Reports from ship operators helped to provide a range of terminal experience, sometimes from non-member terminals.

The LNG trade involves about 150 million m³ of LNG transported each year and about 3 000 ship loading and discharging operations in the same period See SIGTTOs LNG Log.x. Returned questionnaires cover the LNG business as follows: from within the world LNG fleet, replies representing approximately 50 % of the fleet were received from ship operators. In addition, terminal managers representing approximately 1/2 annual port calls responded. This information ts supplemented by details from technical journals. Accordingly, while some significant accidents may not been known, it is virtually certain that all major accidents for this trade have been recorded.

Information on the LPG (and similar) trades is unsatisfactory for the provision of reliable statistics. SIGTTO membership only covers about 50% of these trades: this tends towards larger terminals and ships. Accordingly, the number of ship loadings and dischargings captured within the study (circa 2 000) is low in contrast with the full nature and extent of business as reflected in Table 1.

In the paragraphs below, an indication is given of the overall frequency of incidents. This data should be reasonably reliable for significant accidents in the LNG trades. However, because of limited data, particularly for the smaller ships and terminals, the information can only be used with great caution for the LPG trades. Indeed, once ship-size falls below say 30 000 m³ the data-range should be considered incomplete.

It should be noted that the accident data-base, in Appendix 1, is very small in comparison to the estimated number of port calls in the period (38 accidents in about 500 000 port calls over 13 years). Also, it is known that numerous extra accidents could have happened – particularly when considering the subject matter of some incidents already recorded. Accordingly, it should be appreciated that the addition of even a few new accidents to the data-base would make the frequency rates much worse than those recorded. For example, by breaking down the 38 accidents into tonnage groups it is found that in the same period 24 accidents are noted against 6 000 port calls for larger ships over 30 000 m³. Hence a much higher incident rate 1s attributed to these carriers.

As mentioned above, the data-base lists 38 significant accidents. These are all the accidents of this type and degree of seriousness known to SIGTTO In Table 6, attached to the end of this section, an indication of accident seriousness is shown by background highlighting in the Report column Accidents shown with a dark background are considered major. There are 15 such major incidents recorded. Although there are a few accidents recorded prior to 1982, the statistics which follow assume a 13 year coverage period.

For the larger ships (> 30 000 ms) 24 accidents are recorded: in the smaller tonnage group only 13 are known – one accident is terminal specific In 1994 the world’s gas carrier fleet totalled about 910 ships. These can be grouped approximately in accordance with Table 1.

Table 2 lists the 38 accidents according to identified hazards and subdivides them according to cargo type. The table has the disadvantage of showing all known data in a similar light even although it is known that there is considerable lack of data for some trades. This anomaly ts corrected in Table 3 (but in this case the data-base is much smaller).

Table 3 provides similar data to Table 2 but in this case the analysis has been restricted just to major accidents. This table – although sparse in detail – may provide a better balance between hazard-areas in the different trades so providing better comparison of risk across the hazard range (this is because most major accidents are known to SIGTTO).

Table 4 has been developed from an overview of the accident data in Tables 2, 3 and 6. As noted, it is based on world-wide risks and actual risks on a port by port basis may well change. However, the table is considered of use to terminal management by providing a prioritised list of the major hazard areas which should be addressed in each marine terminal.

Frequency of Accident or Near-Miss

Strikings at jetties are very infrequent occurrences. Only two cases are recorded, both with respect to small ships, and of these one is considered major. This event involved a barge loading liquefied gas.

Ships Striking Gas Carriers Alongside

It is probable that smaller ships are more likely than larger ships to endanger the cargo connection in this way. This is asserted on the basis that small ships are more likely to be berthed in areas exposed to passing traffic – large ships are usually berthed in more remote areas.

Also a small ship is more likely to be moved along a Jetty (or swept from it) in the case of collision.

Considering the data in Table 4 it is concluded that the risk of break-out (15,750) is high – it identifies a very real risk which should be addressed. As covered in Accident Prevention The Use of Hoses & Hard-Arms at Marine Terminals Handling Liquefied Gas“Safe Berths, Mooring & Operating Procedures”, ship’s moorings are the primary means of defence in protecting the cargo connection. As shown in this section, ship break-outs are one of the principle hazards. It is strongly recommended that terminal managers thoroughly review mooring operations and clearly establish that moorings are adequate.

It is concluded that few berths are immune from the break-out problem. Although the risk may be small – it is not insignificant. Table 6 suggests that every year about two ships break-out from their berth.

Hoses

Problems with hoses bursting are only infrequently reported and before conclusions from the data base can be reached it is essential to know that the whole trade has been reasonably well covered. This cannot be said and therefore the data must be treated with caution.

However, from the accident data, it 1s clear that great care must be taken when using hoses for liquefied gas transfer. The use of hoses has contributed to many deaths. This issue has affected the ammonia trade more than any other.

Generally the use of hard-arms is to be preferred, however, at marine terminals where the economies of scale are less apparent, hoses can be used to advantage In such cases great care must be taken over hose quality.

Hard-arms

The survey shows that hard-arms provide a much safer operating environment than hoses.

It is the norm for hard-arms to be used in the LNG trades. This is because of the essential security required, the substantial size of the facilities and the high throughput of the business generally warrant such caution to be taken. Here, the use of hard-arms is also a necessity because of the extremely low temperature at which liquid methane is handled.

QCDCs

QCDCs are sometimes fitted for very special reasons, however, they introduce significant risks in comparison to flange-to-flange bolted connections. It will be seen from Table 3 that bolted flanges do not feature as a safety hazard, however, QCDCs rank as a significant risk.

However, the fitting of QCDC devices at the end of hard-arms can sometimes be a safety measure, as described in Loading Arms (Hard-Arms) – Specifications, Operation, and Maintenance“Hard-Arm Handling and Operation”. Accordingly, as the use of this equipment has contributed to a number of serious accidents, where fitted, its safe operation should always be ensured.

Emergency Release Couplings

When ERCs were first introduced the lack of inter-linkage between the flanking valves and the dry-break coupling led to a number of serious spillages. This problem has now been largely overcome.

Some problems continue but of a lesser nature. The issues still needing to be addressed include the training of operations and maintenance personnel.

In order to mitigate any leakage after mooring failure and ship break-out, the Emergency Release Coupling ts considered to be a vital second line of defence. Its fitting is therefore strongly recommended.

Pipelines and ESD Valves

An issue not yet fully appreciated, especially in the LPG trade, is the need for properly controlled ship/shore ESD linkage.

Better study of pipeline systems and ESD valve closing times seems essential and, to highlight these issues, this should be included in design criteria for new and existing terminals and in training programmes.

The above table lists all known significant accidents relating to the ship/shore transfer connection having occurred since about 1982. Those accidents considered by SIGTTO to be of major note are highlighted in the “Report Number” column by background shading: (e) – indicates SIGTTO’s estimate in cases where data Is unavailable.

Appendix 1: Known Significant Accidents and Near-Miss Data (c. 1982—1994)

Collisions at Jetties

Report 1.1. Ship collision alongside – fractured hose – fire – injuries.

This major accident occurred in 1984 when an LPG barge was struck by a tug at her berth in America. The barge was discharging butane at the time. The impact tore the barge from her mooring severing the cargo connection. A large gas cloud developed and this ignited in a flash fire. A number of people were burnt.

Report 1.2. Ship collision alongside – near-miss spill of ammonia.

In 1993 a small gas carrier discharging ammonia was struck by a ferry at her berth in Europe. This was a gas carrier of about 80 metres in length having a cargo cubic capacity of 2 000 m³. Fortunately there was no release of cargo as about 100 people on the ferry were at risk.

Mooring Break-out and Related Near-miss Accidents

This summary of ship break-outs includes those events where a direct risk to the hard-arm was perceived. Two events where a ship collided with a gas carrier working cargo alongside are covered in 1 above. This listing excludes a series of quite numerous reports of individual mooring line breakages, sometimes due to stress of weather, but which were not considered critical as far as the hard-arm or hose was concerned. It also excludes a number of reports involving ship movement while mooring lines were being adjusted.

Read also: Emergency Shut-Down and Emergency Release

It is probable that the greater proportion of reports logged in the 1990s is due to better reporting and the better recording of events.

Report 2.1. Ship break-out – damage to ship and dock – no ERS.

This major accident occurred in 1982 when an LNG carrier broke out from her berth in the Far East. This was a ship of about 285 metres in length having a cargo cubic capacity of 125 000 m³. As the hard-arms were not protected by ERCs, there was considerable damage to the jetty, hard-arms and ship.

Report 2.2. Ship break-out.

In 1986 an LPG carrier broke out from her berth in Europe. This was a ship of about 220 metres in length having a cargo cubic capacity of 55 000 m³. Cargo operations had not started at the time.

Report 2.3. Ship listing — possible risk of break-out.

In 1988 an LNG carrier listed heavily at her berth in the Far East. This was a ship of about 290 metres in length having a cargo cubic capacity of 125 000 m³. Owing to human error on board the ship the wrong action was taken and before cargo was stopped the ship was heeled to an angle of 10 degrees.

Report 2.4. Ship break-out – spillage and ship taken out of service – no ERC.

This major accident happened in 1989 when an LNG carrier broke out from her berth in North Africa. This was a ship of about 200 metres in length having a cargo cubic capacity of 40 000 m³. There was considerable damage to the jetty, hard arms and ship including cold-cracking. This accident involved considerable spillage.

Report 2.5. Ship break-out – but ship at a lay-up berth at the time.

In 1989 an LNG carrier broke out from her lay-up berth in Africa. This was a ship of about 280 metres in length having a cargo cubic capacity of 130 000 m³. Subsequently repairs to the ship were necessary. This accident was not cargo related.

Report 2.6. Ship break-out – ERC works but hard-arm snags on ship’s structure.

In 1990 an LPG carrier broke out from her berth in Europe. This was a ship of about 220 metres in length having a cargo cubic capacity of 55 000 m³. The ERC worked well but, as the outer arm did not properly clear the ship and as the ship slid along the jetty, the arm snagged on some superstructure causing severe operational complications.

Report 2.7. Ship break-out.

In 1992 an LPG carrier broke out from her berth in Europe. This was a ship of about 225 metres in length having a cargo cubic capacity of 80 000 m³.

Report 2.8. Ship break-out – (near-miss).

In 1992 an LPG carrier nearly broke out from her berth in South America. This was a ship of about 220 metres in length having a cargo cubic capacity of 60 000 m³. A ship manoeuvring nearby caused suction effects, the gas carrier surged and a number of mooring lines parted. It was found that snore wire strops were used on the bollards and that these were of poor quality and of inadequate strength.

Report 2.9. Ship break-out – ERC works satisfactorily – minor damage to hard-arms.

This major accident occurred in 1993 when an LPG carrier broke out from her berth in Europe. This was a ship of about 220 metres in length having a cargo cubic capacity of 55 000 m³. A ship moored at a berth nearby was nearly struck by the first ship as she broke-free The event was the subject of an official government inquiry.

Report 2.10. Ship break-out – hard-arms saved – no ERC.

In 1993 an LNG carrier broke out from her berth In America. This was a ship of about 240 metres in length having a cargo cubic capacity of 70 000 m³. The ship broke-out due to pressure from ice floes created by very strong currents. The hard-arms (which were not protected by ERCs) were disconnected before they could fracture.

Report 2.11. Ship break-out – hard-arms damaged.

In 1993 an LPG carrier broke out from her berth in the Persian Gulf. This was a ship of about 225 metres in length having a cargo cubic capacity of 75 000 m³.

Report 2.12. Ship break-out – no ERC.

In 1993 an LPG carrier broke out from her berth in the Middle East. This was a ship of about 170 metres in length having a cargo cubic capacity of 35 000 m³. The hard-arms were damaged.

Report 2.13. Ship break-out – hard-arms saved – no ERC.

In 1994 an LNG carrier broke out from her berth in America. This was a ship of about 240 metres in length having a cargo cubic capacity of 90 000 m³. Ship broke-out due to suspect mooring line quality of the nylon tails. By quick crew reaction the ships bow thruster was used to keep the ship alongside.

Report 2.14. Ship break-out – hard-arms protected and saved.

In 1994 an LNG carrier broke out from her berth in the Far East. This was a ship of about 240 metres in length having a cargo cubic capacity of 90 000 m³. Ship broke-out due to suspect mooring line quality of the nylon tails.

Hoses

Report 3.1. A multi-ship incident – fire – fatalities.

This major accident occurred in 1985 during a ship loading at a terminal in the Gulf of Mexico. One large ship of about 60 000 m³ was loading LPG at the time. Also there were four liquefied gas carriers alongside other adjacent jetties. As the ship in question was nearing completion, the cargo transfer hose burst. The berth was not equipped with emergency shut-down facilities and product spilled for a number of minutes.

This resulted in the formation of a significant vapour cloud which subsequently ignited. The source of ignition was believed to be the air-intake of a diesel engine on a boat passing nearby. In this case a number of personnel lost their lives and considerable damage occurred to two of the gas carriers.

Report 3.2. Hose burst during the loading of LPG – fire – fatality.

This major accident occurred in 1985 in South East Asia. It involved a self-propelled barge, designed for river work, but used also for domestic trading in exposed waters. The barge had cylindrical pressure cargo tanks on deck rated at 17 bar.

The barge was loading a cargo of pressurised propane and butane mixture. As loading was coming to an end, the cargo transfer hose burst, the spilled product ignited and one crew member died. The fire then gutted the accommodation and caused widespread damage. The fire burned for four or five days.

Report 3.3. Hose burst during discharge of ammonia – vapour cloud – fatalities.

This major accident occurred in 1983 and happened in South East Asia. It involved a pressurised gas carrier of about 2 000 m³ capacity. The ship had four cargo tanks. During discharge of ammonia the cargo transfer hose failed. An ammonia cloud developed and engulfed another ship. No ignition occurred but at least five people died and a further 31 were hospitalised with breathing difficulties.

The cargo hose was certified to 27 bar and, at the time of the incident, the ship’s discharge pressure was 13 bar. In the subsequent investigation, the local authorities said that they lacked the necessary technical expertise and facilities for auditing ship and terminal equipment.

Report 3.4. Hose burst during discharge of ammonia – vapour cloud – fatalities.

This major accident occurred in about 1980 and involved a ship of about 2 000 m³ capacity, loaded with an ammonia cargo of about 500 tonnes. The ship was discharging when the transfer hose failed, releasing a large vapour cloud. The cloud was reportedly 100 metres high and, although the shore hydraulic valve was closed, the escape continued for nearly one hour and involved approximately 180 tonnes of cargo.

To stop the escape the ship’s manifold valve (which was enveloped in the ammonia cloud) had to be closed by emergency service personnel. The master and chief officer of the ship lost their lives. The hose was a new hose, suitable for LPG, but not suitable for ammonia.

Report 3.5. Hose burst during STS transfer of pressurised LPG – vapour cloud.

This incident happened in 1989 and involved a ship to ship transfer operation in South East Asia. Operations were being carried out from a refrigerated carrier of about 55 000 m³ capacity to a pressurised coaster of about 1 500 m³ capacity. During cargo transfer the cargo hose burst and a spill of about one cubic metre of liquid occurred. The resultant vapour cloud drifted away from the operations area into the relative safety of the open sea.

It was found that the hose in use was not properly specified for LPG. It was of too low a pressure rating.

Report 3.6. Hose burst during STS transfer of refrigerated LPG – vapour cloud.

In 1990 during a ship to ship transfer of refrigerated LPG, the ships’ hose was used and this burst during transfer. Transfer was taking place from a 60 000 m³ carrier into a 4 000 m³ carrier. It was found that the ships’ hoses had not been properly tested.

Hard-arms and Quick Connect/disconnect Couplings

Hard-arm related

Report 4.1. Hara-arm fracture and collapse – surge pressure and incorrect materials.

This major accident occurred during a ship discharge of LPG when a quick acting valve in the terminal was shut by mistake. This caused the fracture of the discharge manifolds at two of the ship’s pumps. In addition, the hard-arm fractured and collapsed.

The main cause of failure was found to be excessive surge pressures. Furthermore, the material of construction of the hard-arm was found unsuitable for the minimum operating temperature.

Report 4.2. Faulty hard-arm relief valve.

In 1988 at a terminal in America handling LNG a fault was discovered in the hard-arm relief valve. Due to the vibrations caused during normal cargo discharge it was found that the valve could change its set point. Inadvertent realignment of the valve could allow LNG to escape at normal operating pressures.

QCDC related

Report 4.3. Release of hard-arm – poor maintenance – no injuries but severe damages.

This major accident occurred in 1991 at a terminal in Australia during cargo discharge of propane. The ship was a 20 000 m³ carrier. In this case the hard-arm released from the ship’s manifold. The ship’s ESD system was operated and no damage was sustained to the ship. But the hard-arm was wrecked as it flailed about continuing to spill product and with the shore ESD damaged a large vapour cloud formed. It was found that the hard-arm was inadequately maintained and that training for shore staff on maintenance and operation was lacking.

Report 4.4. Faulty hydraulic system on QCDC.

In 1982 a release of a QCDC occurred due to the use of incorrect equipment in the hydraulic control system. This caused a high back pressure in the system and resulted in the QCDC opening so releasing the hard-arm from the ship’s manifold.

Report 4.5. Release of hard-arm – operator error – injuries.

This major accident occurred in 1994 at a terminal in the Caribbean after completion of loading ammonia. In this case the ship’s manifold valves were left open for line draining purposes. For some reason the jetty operator released the QCDC and product spilled onto the ship’s deck where the crew were working. Two crew members were seriously affected by the fumes and had to be hospitalised.

Emergency Release Couplings

Report 5.1. Lack of training — release of ERC – vapour cloud – fatality and injuries.

This major accident occurred in 1980 in southern Europe. It involved the inadvertent release (due to maloperation or design error) of the ERC during discharge of propylene. The dry-break valves failed to close due to design error. A dock worker was killed and some injuries resulted.

Report 5.2. Poor operational procedure – too many reducers.

This incident occurred in 1979 at a northern European terminal and involved the accidental release of an ERC following connection of a large bore hard-arm to a small diameter manifold. The operation had required the use of a poor arrangement of additional spool pieces to be fitted to the ship’s manifold.

Report 5.3. Serious spillage – ERC releases – poor maintenance.

This major accident happened in Scandinavia in 1978 and involved the inadvertent release of an ERC during discharge of refrigerated LPG. The double ball valves flanking the coupling did not close and substantial spillage took place. Fortunately there was no loss of life. The incidents were attributed to the incorrect fitting of an hydraulic block valve.

Report 5.4. ERC release – poor maintenance.

In 1979, a total of seven ERC releases occurred at a northern European terminal. In no case was cargo being transferred. The cause was attributed to the ingress of water into the ERC hydraulic return line. The water then froze causing overpressure of the ERC actuator and incorrect release.

Report 5.5. ERC release – poor maintenance.

This incident occurred in 1982 in a northern European terminal and involved the release of an ERC during the transfer of pressurised LPG. The cause was attributed to the incorrect fitting of a non-return valve in the hydraulic control system.

Report 5.6. ERC release – poor maintenance.

This incident happened in the Middle East in 1983 It involved the release of hard-arms due to the incorrect setting of a trip switch. This switch was fitted to sense the limits of the operating envelope On this occasion the ball-valves closed satisfactorily and spillage was limited thereby.

Report 5.7. ERC release – poor maintenance.

This incident occurred in 1994 at a terminal in the Far East and involved the unexplained release of three ERCs during the transfer of LNG. On this occasion the flanking valves operated. A surge pressure occurred and a safety relief valve on the ship’s manifold functioned but a manifold flange gasket was damaged. The shut-down was attributed to a loose screw causing a limit switch to function incorrectly.

Pipelines and ESD Valves

Report 6.1. Leakage of LNG from drain line – hydraulic hammer.

This major accident occurred in 1988 during the discharge of LNG from a 125 000 m³ carrier at a terminal in America. A stoppage occurred but on restarting a serious shock load was transmitted along a branch line used for draining purposes. Flanges at the end of the pipeline were damaged and a leakage of about 80 m³ of liquid occurred. This was contained on the jetty in a bunded area – so vaporising gradually.

It was found that some liquid in the branch line had previously vaporised and the shock load was due to a “liquid hammer” effect as the pipeline re-filled with liquid. Subsequent preventative measures included shortening the length of the drain line.

Report 6.2. Leakage of LPG from holed submarine pipeline.

This incident happened in the Mediterranean in 1993. It involved an LPG carrier discharging at an off-shore terminal. During discharge it was discovered that LPG was escaping from the submarine pipeline. This was probably due to poor inspection and maintenance routines.

Report 6.3. ERC release – poor operational procedures.

This incident occurred in 1992 in Australia and involved a sudden closure of a shore ESD valve, during the discharge of a 20 000 m³ carrier. Surge pressure developed and blew the shore pressure relief valve. The transfer hose was damaged but it was found that the shore ESD valve had been adjusted to close in 5 seconds while the ship’s valve closing time was retained at 30 seconds. It was found that the procedures of exchanging valve closing times, as required by the Ship/Shore Safety Check List had not been followed and the SIGTTO recommendation to close the valve nearest to the pump first had not been attended.

Report 6.4. Unexpected ship ESD – serious pressure surge.

In 1992 at a European terminal during the loading of an ethylene carrier of about 5 000 m³ a serious pressure surge occurred with considerable vibration ringing through the system. This happened during a ship black-out when, as can be expected, the ship’s ESD valve closed. As the terminal was not outfitted with a linked ship/shore ESD system the shore valve was not shut first and the pressure surge caused damage to the hard-arm. In this case damage was minimal but the loading arm swivel seals were replaced as a precautionary measure.

The accident was compounded by the fact that the ship’s manifold was fitted with a “Y” piece connected to two liquid lines. The single leg of the “Y” prece was connected to the hard-arm With this arrangement the shore was loading through two of the ship’s ESD valves. As will be found in a reference some care has to be taken when following such a practice. When loading through two or more ESD valves the effective closing time of the combination is less than that presented by one ESD valve, In such circumstances, loading rates should be reduced to suit the circumstances.

Appendix 2: Recommendations for Terminals – Ship Break-out

For ship Suitability assessment (at the time of charter) terminals should have the ability to assess site-specific loads in individual ship’s mooring lines and these loads should be calculated for the design maximum wind and current speeds of the berth:

• Terminals should stop berthing operations when forecasted wind speeds at the berth are expected to exceed a pre-stated maximum.
• For their berths, terminals only ought to allow the deployment of a ship’s mooring system having balanced geometry and equal line-length; for (a) head and stern lines (b) forward breast-lines and aft breast-lines and (c) forward spring lines and aft spring line.
• Terminals should include in Jetty Regulations the actions required by ship and shore during strong winds. These should include wind speed limits for stopping cargo operations and any subsequent disconnection of hard-arms. This Jatter information is often provided by hard-arm manufacturers.
• Terminals should have direct emergency communication with the port authority.
• Terminals ought to consider the need for a (conveniently located) stand-by tug while gas carriers are alongside.
• Terminals should have a pilot on scene if tugs are in use.
• Port authorities should have an effective Emergency Plan for all terminals handling dangerous-substances.
• Terminals should have wind speed and direction instrumentation giving direct read-out in the control room.

Other environmental factors to be considered by terminals. These are listed below:

• Wind speed in gusts;
• Wind direction (off-shore or on-shore);
• Sea conditions (waves or swell) causing unacceptable ship movement;
• Strong currents;
• Currents not in line with the berth axis;
• The suction effect caused by ships passing nearby;
• The use of additional or special shore-based mooring arrangements;
• Tidal height for sufficient water depth in the channels;
• Tropical storms;