Fire Ranks Second in Maritime Casualties

by S. Mendiola & J.J. Achútegui, and M.A. De la Rosa

A survey of total loss accidents in merchant shipping over a period of 25 years shows that these can be arranged in the following order: stranding, fire, water-leaks, gales and collision; other accidents are also taken into consideration. The analysis considers ships over 500grt of different flags, plying any route of navigation and trade. Initially, a sample of 500 merchant ships - of different types and tonnage - and under 15 different flags is analysed to determine age and type of ship, and the causes of loss. On a second analysis, the same 15 flags are considered, but now over a wider range on a sample totaling 1,500 merchant ships. The results of both analyses are compared. It is shown that fire together with explosion amounts to 25% of maritime casualty returns -in the total loss lists- while stranding and fire take more than 50% of the toll.

1. Introduction

Maritime accidents fall into one of the following groups due to several circumstances: those caused by weather conditions, such as gales, reduced visibility [1], ice, etc; or those due to pilot navigation error, narrow [2] and/or congested [3] waters, collision with unknown objects, ship lying at anchor or moored at buoys with strong currents, manoevering at close quarters or with limited space and adverse conditions in port. Cargo related accidents occur through the carriage of dangerous goods, cargo on deck, heavy cargo, or cases relevant to the ship's seaworthiness. Failure in the steering system [4], main engine, different devices, war, terrorism, piracy, collision and misinterpretation in communications at sea, etc [5, 6] can all lead to accident.

Accidents by collision [7] have decreased significantly where a maritime traffic management service or, at least, a Traffic Separation Scheme (TSS) has been implemented. Currently a worldwide maritime traffic management system is being contemplated [8].

The SOLAS (Safety of life at sea) convention rules the safety of navigation in sea trade [9] in shipbuilding and fire-resistant bulkheads, life-saving appliances and facilities, radio communications, grain in bulk and dangerous goods transportation. These international provisions make it compulsory for sea-time training on board merchant vessels, and for fire and abandon-ship drills.

Fire aboard merchant ships is serious, sometimes leading to total loss of the ship and/or her cargo, to gross damage, and to loss of life. In the past, when merchant vessels were built out of wood and propulsion was achieved by wind action on the sails, lighting was achieved by means of oil or paraffin lanterns; tragic fires happened far too often, due mainly to the ship's rolling and the subsequent falling and breaking of the lanterns. In this day and age, flame lights are not allowed on board, or are prohibited by their inefficiency and danger. Nevertheless, fire still poses a high risk for several other reasons.

The stranding of merchant vessels [10] can result in fire and explosions [11] particularly when large tankers engaged in the crude oil trade are involved. Such was the case with the 'Torrey Canyon' in 1967 [12], when a series of explosions and fires after her stranding in the Scilly Isles (Seven Stones, Polard Rock) caused an all-time record in sea pollution. Probably petroleum products, shipped in bulk, present the highest risk, when errors occur [13], but we have other substances such as coal, a number of ores [14], feeding stuff, fertilizers, fish meal [15], etc. which are apparently harmless - when one is not acquainted with case histories - but which are liable to produce a spontaneous combustion.

Accident investigations [16] show that fire leads to serious consequences not only in carriage of dangerous goods, but also poses a risk to other goods which otherwise would not be dangerous and would not create a hazard during sea passage - such as sugar, walnuts, cotton, and the like which can readily be stowed with no apparent fire risk. This kind of cargo burns easily and canbecome a risk if neighbouring hot work or a faulty mains line causes fire to break out in the cargo hold. Extinguishing this fire will prove difficult once it has gained a hold and it will spread quickly if there is sufficient oxygen.

This work analysed various maritime accidents during a 25-year period and, with samples of 500 ships or more, it was found that stranding and fire aboard taken together, amounted to 50 percent of the constructive total losses.

2. Method and Results

The method followed in this research on maritime casualties has consisted of analysing total losses of merchant ships, under different flags, which gross tonnage of 500grt or over, throughout 25 years. Data for accidents were taken from "Modern Shipping Disasters" [17], which lists disasters alphabetically by vessel's name. To obtain useful or reliable results for a given flag, it is necessary to consider a group of 100 ships for each flag. To quantify, causes, ages, and class-type under different flags, a sample 500 ships [18] is needed. Firstly a total of 500 merchant ship losses under 15 flags was analysed, to establish ship's age when lost and the trade she was on (class or type of goods transported). The number of ships, and total gross tonnage per flag was recorded, and this data is presented in Table 1. The reasons for these accidents have been analysed and quantified in Table 2 both as numbers and as a percentage of the total.

Whilst a sample of 500 accidents is sufficient to establish causes, it is not a large enough sample to discriminate behaviour between different flags. In the next two tables, Tables 3 & 4, the same particulars have been analysed, but the number of ships has been increased to a total of 1,500.

2.1 Age estimation
In the analysis of '15 flags - 500 ships' (Table 1), the losses have been separated into four periods of a ship's active service - first 0-5 years, second 6-12 years, third 13-20 years, and fourth ages overrunning 21 years. In Table 1, the sum of the losses in the first two periods, 38+72, amount to 110 ships, meaning 22% of the 500 ships analysed. The losses in the third and fourth periods, 185+205=390 ships, make 78% of the total analysed. The first particular to consider is that losses in ships over 13 years do outnumber the others, and are 3.5 times more frequent than in new and middle-aged ships; but it is also true, that the over-ageing of the world shipping, in the 25 years under survey, is a trend to consider. The 205 vessels in the table, aged over 21 years, amount to 41% of the 500 ships and 15 flags of the sample, and only Japan looms as a younger fleet.

2.2 Trade
Regarding ships classified by trade, those carrying general cargo (break-bulk) are the majority, making a total of 333 units, meaning 66.6% of the total; dry-bulkers come to 51 ships with 10.2%; tankers number 55 with 11%; and the rest of ships make a total 61 units, coming to 12.2% of the 500 analysed. Total gross tonnage amounts to 3,941,360 which divided by the 500 ships, comes to a mean of 7,883grt per ship.

2.3 Cause
Table 2 shows 15 flags and 500 merchant ships, entering the nine most frequent circumstances in maritime accidents, resulting in total losses confirmed by ships' classification societies. The results of the table places "stranding" as the leading maritime casualty - in fine weather, reason unspecified - followed by stranding in heavy weather, a total of 146 being entered with 29.2% of the 500 ships involved in accidents. Ranking second is "fire"; fire in the engine room is the most common cause, with 61 cases forming 55% of the total caused by fire. Total accidents by fire amount to 111 ships and represent 22.2% of the total. Third cause of total loss is attached to "water-leaks" with 72 cases and 14.4% of the total. Fourth are "gales" affecting 70 ships and a share of 14%. Fifth place is held by "collisions" with 48 cases and a 9.6% share.

The remainder of the total losses having a lesser frequency impact in this table stand in the following order: explosions, faults in cargo, war, and striking unknown floating objects.

2.4 Ship losses
In Table 3 the same 15 flags are analysed, increasing the number of merchant ships - in total loss casualties - , to 1,500 of 500grt and over, the results being entered by age, class of ship and gross tonnage. In the first nine flags with 100 or more ships, ages are analysed separately, the type of ships involved in the casualties, the mean per flag and the total mean. For the analysis and quantifying of ships the same four periods of Table 1 are maintained.

In considering the first two periods - new and middle-aged ships on one hand, and the two second periods - too long in service and old ships - on the other, if the ratio of the former over the latter is greater than unity, the flag of that merchant fleet can be assumed to be as of new construction. For a valid comparison, there needs to be 100 or more ships per flag in making the estimation.

• Japan has a ratio greater than 2 [(39+50) / (31+7) = 2.342]
• Greece has the lowest ratio, [16/134 = 0.119], indicating the oldest fleet;
• Cyprus [19/131 = 0.145] turns out to be the second oldest flag; following this line Panama with 0.154; Philippines 0.174; Italy 0.244; UK 0.250; Liberia 0.363; and Spain 0.887.

These results do not mean that ships had a casualty for being very old, but rather that the world fleet in the 25 years under survey is quite old, neither doesn't mean that Japan's casualties occur mainly in her new ships, but rather that her fleet is new.

As in Table 1, two letters have been entered at the head of the columns, for readily identifying class of ship, GC = general cargo, BC = bulkcarrier (dry-bulkers), etc. Total contribution of GC ships is 1,034 units, a share of 69% of the total. Next column is for dry-bulkers with 121 ships with a 8.1% share, followed by tankers on trades of crude oil or oil products, with 193 tankers and 12.9%. The rest of ships (all columns to the right) entered in the same line (G total) amount to 152 ships, with 10% of the total. The mean of the gross tonnage of the first four columns with 150 ships per flag is highest for Liberia with 3,524,820 grt/150 = 23,499grt. This figure, representing less GC ships and several in BC and TA ships for a same number of ships, means a higher tonnage average. On the contrary, Panama, with a higher number of GC ships and few BC and TA ships, has the lowest tonnage average of the four analysed with 874,780/150=5,832grt. Total mean (15 flags) amount to 12,472,710 / 1,500 = 8,315grt.

Comparing Table 1 and 3, the total tonnage average per ship is only of 8,315-7,883=432grt. Ages in each table keep a similar ratio, and only column 13-20 years shows a difference as high as 4.8%. Regarding types of ships, the highest difference between tables does not surpass 2.4%.

In Table 4 we have arranged the accidents of the 15 flags in 10 columns, for allocating 1,500 ships per flag and their casualties under their corresponding entries and headings. "Stranding" is still the first reason for accidents totalling 455 cases and representing 30.3% of the total.

"Fire" ranks second involving 304 ships with 20.3% of the losses; fire in the engine room has the highest rate in this category, with 165 accidents. Third is "water-leaks", with 202 total losses, 13.4% of the total. The accident in the fourth place is "gales" with 157 ships and 10.5%. In fifth place comes "collisions" with 149 ships and 9.9%. Finally, the remaining five case reasons come to 233 total losses with 15.6%.

Analysing individually, the first nine flags in this table with 100 or over ships in casualty per flag, Greece and Spain come into the highest rate in their total losses because of "fire aboard". On the contrary, Japan is the country with the lowest rate of losses by fire, with only five cases out of the 127 total losses, the most important accidents on record being "collisions".

3. Conclusions

On analysing total loss accidents for 15 flags with sample sizes of 500 and 1,500 ships of over 500grt, over a period of 25 years, the first leading circumstances of maritime casualties in the merchant fleet for both sample sizes were in the following order: stranding; fire; water-leaks; gales; and collisions. Other five accident causes were entered, but had little impact. In the reckoning of ships in both models, fire was the second most frequent circumstance in the accidents and, together with stranding, represent more than the 50% of maritime casualty returns and, if we include explosions in the column of fire, these latter items (explosion + fire) would add up to 25% of casualties. On considering flags one by one with over 100 ships, Greece and Spain are the flags where the highest number of accidents by fire is to be found, while Japan is the lowest. In this latter flag "collision" is the leading accident eventually ending in a casualty.

4. References

1. Marí, R., Prevención de los abordajes en la mar. Ed Universitat Politécnica de Catalunya, Barcelona, Spain, 1994.
2. Zhao, J., Price, G., Wilson, P.A. & So, M. The uncertainty and un coordination of mariners' behaviour in collision avoidance at sea. J Navigation, 48, No 3, (1995) 425-435.
3. Wennink, C.J., Collision and grounding risk analysis for ships navigating in confined water. J Navigation, 45, No 1, (1992) 80-90.
4. Achútegui, J.J., Mendiola, S., Sánchez, F.J. & Díaz, A. The "Amoco Cadiz" stranding could have been avoided. First Intern Conference on Environmental Problems in Coastal Region, Rio de Janeiro, Brazil, Ed A.J. Ferrante & C.A. Brebbia, UK, (1996) 105-114.
5. Cockcroft, A.N., The circumstances of sea collisions. Proc IV International Symposium on Vessels Traffic Service. Bremen, 1981.
6. Mendiola, S., ColReg 1972, Rule 13: Overtaking. Seaways, J N Institute. June 1992, 18-20.
7. Holdert, H., & Buzek, F.J. Collision cases. Judgements and diagrams. Lloyd's of London Press Ltd, 2nd Ed London, 1990.
8. Corbet, A.G., Towards Global Marine Traffic -the need, the technical feasibility and the social and political impediments. J Navigation, 45, No 3, (1992) 425-439.
9. Motte, R., Safe navigation, ships loss and applications of technology. J. Navigation, 49, No 1, (1996) 36-44.
10. Cahill, R.A. Strandings and their causes. Fairplay Public. Ltd, London, 1995.
11. Achútegui, J.J., Mendiola, S., & Díaz, A. Protection from oil pollution caused by the grounding of the tanker "Aegean Sea" in Corunna, Spain. Proc VI Congress, IMAM' 93, Ed P.A. Bogdanov, Varna, Bulgaria, 1993, Vol 1, 211-216.
12. Mendiola, S., Achútegui, J.J., Trueba, A., Sánchez, F.J. & Díaz, A. Maritime pollution caused by accidents of oil tankers. Proc VII Congress, IMAM´ 95, Ed V. Androcec & J. Lovrié, Dubrovnik, Croatia, 1995, Vol 1, 434-439.
13. De la Rosa, M.A., Análisis de la incidencia de los principales errores técnicos que condujeron a la pérdida de grandes buques. Doctoral Thesis, Univ of La Laguna, Tenerife, Spain, 1995.
14. Achútegui, J.J., Mendiola, S., Trueba, A., Sánchez, F.J. & Díaz, A. Transporte marítimo de mercancías peligrosas, hierro obtenido por reducción directa a granel, Project No 640-C109, Univ of Cantabria/Caja Cantabria, Santander, Spain, 1996, 1-69.
15. San José, M.J., Achútegui, J.J., Sánchez, F.J. & Mendiola, S. Spontaneus combustion of fish meal: Factors affecting its safe maritime transport. Grasas Aceites, 46, No 1, (1995) 39-44.
16. Steen, G., Maritime accidents investigation. Board of Accident Information. Malmö, 1985.
17. Hooke, N., Modern shipping disasters. Lloyd's of London Press Ltd, UK, 1989.
18. Mendiola, S., Achútegui, J.J., Trueba, A. & Sánchez, F.J. Cicunstancias y motivos de los accidentes marítimos, Proc VI Congress, Marina Civil, Santa Cruz de Tenerife, Spain, Ed COMME, Madrid, 1994, 301-313.