The safety and quality of open registers and a new approach for classifying risky ships

Transcription

1 Transportation Research Part E 35 (1999) 135±143 The safety and quality of open registers and a new approach for classifying risky ships K.X. Li 1,2 University of Wales, Cardi, UK Received 2 February 1998; received in revised form 20 December 1998; accepted 20 January 1999 Abstract By examining 20 years data, the paper attempts to outline the safety and quality records of open registers, and to establish a relationship of accidental total loss rate with ship quality factors, e.g., certi cate, crew performance and equipment indicated by detention rates. It con rms that loss rate, detention rate and age of ships are highly correlated with each other. It shows that the loss rates and detention rates of open registers are above the world average, but with the exception of Liberia, Marshall Islands and Barbados. A new approach is suggested for assessment of ships safety score. Ó 1999 Elsevier Science Ltd. All rights reserved. 1. Introduction In June 1998, 150 delegates from seamen's unions, shipowners/operators, shippers, bankers, insurers and port and ag states took part in a conference debating on the ` ag of convenience' (foc) regime, an old topic of continuing interest to the shipping industry. 3 In the conference, it was held from the union side that a foc ship could be described as `a oating death-trap', whilst from ag states, e.g., Bahamas, Liberia and Cyprus, there was a defence of their records, especially from Liberia which claimed a detention rate at 2.1% better than all the European Union 1 PhD Candidate at University of Wales, Cardi, UK, Prof. J. Wonham. Thanks go to Mr. R.W.J. Schiferli, the Secretary of Paris MOU on PSC for supplying their publications, and to the Editor and three reviewers for their valuable comments on an earlier version. The views expressed are solely those of the author, who is responsible for any errors and omissions. 2 LEEKX@cardi.ac.uk 3 Flag of Convenience (foc) has other aliases, in present study ``open register'', ``open-registry'' and ``foc'' are used as interchangeable terms. The foc Conference was held in Oslo in June A full report please see Telegraph Special report, /99/$ ± see front matter Ó 1999 Elsevier Science Ltd. All rights reserved. PII: S (99)

2 136 K.X. Li / Transportation Research Part E 35 (1999) 135±143 (EU) ags. In fact the argument of safety and quality of foc eets is not new. It was observed earlier that foc countries have `neither the power nor the e ective administrative machinery to impose any governmental or international regulations' (Rochdale Report, 1970) and their eets `have poor safety records in comparison to the world average' (Doganis and Metacxas, 1976a). These arguments can however be supported by numerous samples, but `it would require a much more detailed investigation... to e ectively explain the poor safety record of the convenience eets' (Doganies and Metacxas, 1976b). The failure in uni cation of the requirements of ship registration under international law allows the existence of the foc regime (For more detailed examination on legality of foc regime and ship registration, see Li and Wonham, 1999a). Consequently, registration is the only test of the nationality of a merchant vessel, 4 and the `legality of open registers should not be challenged' (OECD, 1993) under the present international law. The ``second registry'' regime has been recognised by traditional maritime nations as a tool to attract back their eets, and has become a strong competitor to open registry. 5 The port state control (PSC) is another practical way of controlling the further development of sub-standard eets, and to ensure the peace and order of sea transport. Since the rst establishment of the Paris Memorandum of Understanding (MOU) on PSC in 1982, there are now ve such programmes, involving 77 port states. 6 The implementation of PSC worldwide and the growing data on inspection records provide a possibility to monitor the quality of ags, shipping operators, classi cation societies and insurers relating to the ships in question. The paper endeavours to test the common understandings of foc eets by quantifying their growth rates and accidental total loss rates compared with the world average. With data covering inspections in 37 nations, the paper uses the detention rate 7 as an indicator to describe the quality of a ship: e.g., its certi cate, crew performance, living and working conditions, accident prevention, and equipment as required and inspected by PSC, to gain an overall quality picture of foc ags. Based on data examination, it then tries to identify the relationship between safety records and quality records; suggestions are made to improve safety management and port-state inspection as to assessment and classi cation of risky ships. 2. Scope and sources of data Open-registry ags in the study include 18 eets, i.e., Antigua and Barbuda, Bahamas, Barbados, Belize, Bermuda, Cyprus, Gibraltar, Honduras, Lebanon, Liberia, Malta, Marshall 4 The International Court of Justice (ICJ) held that the registered tonnage was the only test in determining the eight ``largest'' ship owning nations referred to Art. 28 (a), IMO Convention. The Court rejected the test of the nationality of the bene cial owners of the ships, which was found ``no basis in international practice''. ICJ Rep., Second registry or International register is established by traditional maritime nations to attract back their national eet currently ying foc. Currently Denmark, France, Germany, Italy, Netherlands Norway, Portugal Spain, and UK have their second registries, and Brazil, Finland, Korea, and Turkey set to have theirs as well. Li and Wonham, 1999b. 6 Including Paris MOU, Latin American MOU, Tokyo MOU, Caribbean MOU and US Coast Guard PSC. IMO News, No.1, When a ship has been found de ciency according to safety regulations by Port State Authority, it will not allow to sail from the port until such de ciency being recti ed. The detention rate of a certain ag is de ned as the number of ships detained by the number of ships inspected annually.

3 K.X. Li / Transportation Research Part E 35 (1999) 135± Islands, Mauritius, Panama, Saint Vincent, Sri Lanka, Tuvalu and Vanuatu, which are covered by UNCTAD and ITF. 8 A new register, the Cambodian Register, was born during recent years. As on February 1997, it has more than 120 vessels with an aggregate of over 300,333 g.t. on its book, mainly from Asian shipowners. Any type of vessel, under any ownership and with crew of any nationality, may be registered. There is no tax whatsoever on pro ts, capital gains or anything else, very low registration fee, just US$ 0.05/n.t. (net tonnage), with a tonnage tax of US$ 0.10/ n.t., and a radio tonnage tax of a at of US$ 50. By de nition, the Cambodian Register is a perfect open register, but it is so new that, as on December 1997, there are no data recorded yet in the publication of Lloyd's Register of Shipping. Thus, it is not included in the present study. Regarding ship type, the study focuses on propelled sea-going merchant ships, including cargo and passenger ships that are no less than 100 gross tonnage (g.t.). Preliminary data is collected from Statistical Tables (1978±1991), World Fleet Statistics (1992±1997), Casualty Return (Annually, 1978±1993), and World Casualty Statistics (1994±1997), published by Lloyd's Register of Shipping. The data on ship quality and detention records is pooled from four main PSC, i.e., Paris MOU on European PSC, Tokyo MOU on Asia-Paci c PSC, Australian Maritime Safety Authority and US Coast Guard PSC, covering inspections carried out in 37 states in the period from 1995 to Methodology 3.1. Development of open-registry eets For the purpose of this study, a growth rate (R g ) of a eet is the average annual growth rate from 1977 to 1996, which can be illustrated by formula (1) as: R g ˆ N 96 =N 77 1=19 1 % : 1 For example, in 1977 the world had 67,945 ships (393,678,369 g.t.). In 1996, it rose to 84,264 (507,873,011 g.t.). Thus, the world average growth rate of the past 20 years should be: R g ˆ 84; 264=67; 945 1=19 1 ˆ 1:14% in terms number of ships ; 2 R g ˆ 507; 873; 011=393; 678; 369 1=19 1 ˆ 1:35% in terms of g:t: : 3 Accordingly, we can have the average growth rates of the foc group and of each foc eet. The foc has kept an average growth rate of 4.41%, about 4 times the world average growth rate of 1.14%, and reached the peak of 10% in Since 1994, it has stabilised at around 6% annually. The contribution of open-registry tonnage to the world total has gone up from 27% in 1977 to 47% 1996 in terms of g.t. and from 11% to 20% in terms of number of ships. Only the growth rates of Bermuda, Lebanon and Liberia are below the world average. The study also reveals that the 8 Bahamas, Bermuda, Cyprus, Liberia, Malta, Panama and Vanuatu are listed as major open-registry countries by UNCTAD, see UNCTAD, Antigua and Barbuda, Barbados, Belize, Gibraltar, Honduras, Marshall Islands, Mauritius, Saint Vincent, Sri Lanka, and Tuvalu are listed as foc eets by ITF. See ITF Internet home page,

4 138 K.X. Li / Transportation Research Part E 35 (1999) 135±143 average size of foc ships is 15,180 g.t., i.e., almost 3 times larger than the world average of 5615 g.t. The average age of foc ships is 18 years, slightly younger than the world average of 19 years Accidental total loss rate For the purpose of study, a total loss rate (R loss ) is de ned as the annual total loss number out of one thousand ships in a certain eet, which can be described as: R loss ˆ N loss =N ship ½ : 4 For example, Bahamas in 1995 had 1176 ships and had ve ships incurring accidental total loss. Thus, the annual total loss rate is R loss82 ˆ 5=1176 ˆ 4:25½: Average total loss rate To gain a general idea of the safety record over the period between 1977 and 1996, the average annual total loss rate can be introduced. For each eet, sum up all the ships in each year (N allship ) and all total loss numbers in 20 years (N allloss ). Thus, an average total loss rate (R) can be calculated by using the formula: R av ˆ N allloss =N allship ½ : 6 Again in the case of Bahamas, the sum of ships over all years is 10,877 ships, and 43 total loss ships over the 20 years period. Thus we get an aggregated annual total loss rate: R ˆ 43=10877 ˆ 3:95½: Improved rate We can compare each year's total loss rate (R) with that of the previous year (R p ) and an improved rate can be yielded by the formula: R imp ˆ R R p =R p % : 8 In 1977, the foc total loss rate was 12.79&, in 1978 it rose to 14.77&, thus the improved rate in 1986 was then: R imp ˆ 12:79 14:77 =14:77 ˆ 13:41%; 9 which indicates that the foc annual total loss number increased by 13.41% in When it is positive, it means that the total loss number has decreased. Accordingly, we can have the aggregated improved rate of foc at 8.76%, the world improved rate at 6.28% Detention rate and quality of ships The quality of a ship can be re ected in its safety management in general, the completeness of certi cates, the e ciency of life saving, re ghting appliances, radio, navigation, anti-pollution, propulsion and auxiliary machinery, loading lines compliance, competency of crew, and quality of food and accommodation. Any de ciency in these areas will cause the ship to be detained. Thus,

5 K.X. Li / Transportation Research Part E 35 (1999) 135± the detention rate can be used as a general indicator re ecting the quality of a eet, which is de ned in the following formula as: Detention rate ˆ ships detained=ships inspected % : Safety and quality analysis 4.1. Safety records of open-registry eets in general According to the methodology discussed in Li and Wonham (1999c), the total loss rate and detention rate can be produced for each individual eet, as well as the world average and foc average (Table 1). It is noted that the annual total loss rate of the open-registry eet (7.31&) is about 2 times higher than the world average. However, it is encouraging to note that its safety record in general has been much improved since The accidental total loss rate is 4.30& in Table 1 Performance of open registers (1977±1996) Flag In year 1996 a Growth Detention Average total Improvement rate (%) b loss rate a (&) rate (%) Ships g.t. (`000) av. g.t. age rate (%) a Column 1 Column 2 Column 3 Column 4 Column 5 Column 6 Column 7 Column 8 Bermuda , Marshall Isl , Liberia ,989 35, Bahamas ,409 20, Antigua & B Vanuatu Sri Lanka Panama ,131 13, Tuvalu Malta ,479 15, Barbados St. Vincent Cyprus ,799 14, Gibraltar , Mauritius , Belize Lebanon Honduras Group 10, ,150 15, World 76, , Score c a Lloyd's Register of Shipping (1997). b Paris MOU (1998), Australian Maritime Safety Authority, Tokyo MOU, US Coastal Guard. c The scores are based on the detention rates pooled from four PSC programmes: score 0 for detention rates <2%, score 1 for P 2±5%, score 2 for P 5±10%, score 3 for P 10±15%, score 4 for P 15±20%, score 5 for P 20%.

6 140 K.X. Li / Transportation Research Part E 35 (1999) 135± , four times less than in 1979 (16.79&), which, however, was still higher than the world general of 2.14& at the same year. Its average annual improved rate was 8.76%, better than the world average of 6.28%. The PSC (port-state-control) data show that the detention rate of foc eet in general is 19.8%, i.e., almost 4 times more than the world average of 5.71% Safety records of individual open-registry eets The average total loss rates of most of the open-registry eets are higher than the world average. According to their average total loss rates, they can be listed in the descending order as Lebanon, Belize, Honduras, Gibraltar, Cyprus, Saint Vincent, Malta, Mauritius, Panama, Sri Lanka, Antigua and Barbuda, Bermuda, and Bahamas. It is noted that both accidental total loss rates and detention rates of Bahamas, Liberia, and Marshall Islands are lower than the world average, and constitute about 40% of total foc tonnage. The safety record of the foc eet has been much improved during the last 20 years, particularly, the eets of Bahamas, Cyprus, and Liberia, while Honduras, Lebanon, Gibraltar, and Antigua and Barbuda eets have not improved in terms of total loss rate Age with quality and safety The correlation test result (Table 2) shows that the ages of ships have a high correlation with both the detention and total loss rate. The correlation coe cient (R 1 ) between ship age and total loss rate is R 1 ˆ The correlation coe cient (R 2 ) between ship age and detention rate is R 2 ˆ 0.81, which shows that the quality of a ship goes down directly with its age. Thus, it is reasonable to suggest that in PSC the ship age should be taken into account for a priority inspection. The average ages of six high-risk eets (total loss rate above 10&) are all more than 20 years old with the only exemption of the Cyprus ag, which has an average age of 16 years. This rea rms the conclusion that `the contention that old ships are extremely vulnerable' made by Lloyd's Register of Shipping after examination of losses between 1967 and 1975 (Cashman, 1977). It suggests that those old-age (e.g., above the world average of 19 year-old) ships, whatever the type, should take an extra boarding priority for inspection (Table 3). The high correlation between total loss rate and detention rate also suggests that the PSC has succeeded in identifying the Table 2 Correlation matrix of Table 1 Column 1 Column 2 Column 3 Column 4 Column 5 Column 6 Column 7 Column 8 Column 1 1 Column Column Column Column Column Column Column The test result of Table 3 by the use of correlation function in Microsoft Excel.

7 K.X. Li / Transportation Research Part E 35 (1999) 135± Table 3 Risk score based on ship age Age group <5 yr 5±9 yr 10±14 yr 15±19 yr 20±24 yr >25 yr Risk score The scores are based on the present study of world merchant ship losses between 1977 and 1996, and the study by Lloyd's Register of Shipping on the analysis of world merchant ship losses 1967±1975, see Cashman, 1977, Table XIII, p. 27. riskiest ships. There is no direct link between the average ship size of a eet with the total loss rate (correlation coe cient R 3 ˆ 0.24) or with the detention rate (correlation coe cient R 4 ˆ 0.32). 5. A new score system on ship quality It is recognised that there are some quality ships among open registers. Even under the same ag, ships that are classi ed by di erent classi cation societies, insured by di erent insurers, manned by di erent nationality crews, managed by di erent operators may have di erent safety standards. The ag of a ship cannot re ect its overall quality and potential risk. A systematic and quantitative approach should be developed to facilitate spotting the riskier ships. A proposed new approach would automatically give a risk score to a ship based on its age, ag, insurers, classi- cations and operators. From the data of the US Coast Guard, for example, we can produce the performance of classi cation (Table 4). The ships associated with Ploski Regestr Statkow, Croatian Register of Shipping, Maritime Register of Shipping, Registrul Naval Roman and Hellenic Register of Shipping have high detention rates, and are assigned the highest risk score of 5. Based on the detention rate and total loss rate, we also give scores to di erent ags (Table 1). At present due to lack of data, performance scores for shipping operators and insurers cannot be produced. This can be done at a later stage with su cient data. Considering all these factors, the risk score of a ship can be calculated by: Risk score ˆ age score classification score flag score 11 Logically a ship with low score, meaning that the ship is registered with a quality register, insured by a quality insurer, manned by quality crew and operated by a quality company, has high quality and should not be targeted by inspectors at any port in the world. By this, the good players will gain some competitive advantages to compensate for their inputs on quality and safety. This new approach will help port states spot the most risky ships more e ciently. Above all, the scoring system will urge all parties involved to place greater emphasis on quality ships. The new system, for example, will encourage ag states, classi cations and insurers, to ensure that newly entering ships meet safety standards by pre-inspection, and will motivate them to maintain a standard on their existing ships. To achieve this, they will play a pre-active role in safety education and training, loss prevention measures and technical service. In the long term, shipowners will not seek association with high-score risky service providers. Equally, service providers will not admit substandard ships and operators. The risk (or quality) score of ags, operators, insurers and classi cations can be calculated from inspection records worldwide. Thus, it is suggested that the

8 142 K.X. Li / Transportation Research Part E 35 (1999) 135±143 Table 4 Performance of Classi cation Society (1995±1997) a Name Abbreviation Inspection Detention Rate (%) Score b Det Norske Veritas DNV American Bureau of Shipping ABS Nippon Kaiji Kyokai NKK Lloyd's Register LR Reigistro Italiano Navale RINA Germanischer Lloyd GL Korean Register of Shipping KRS Bureau Veritas BV Bulgarski Koraben Registar BKR China Classi cation Society CCS China Corporation Register CR Polski Rejestr Statkow PRS Croatian Register of Shipping CRS Maritime Register of Shipping RC Registrul Naval Roman RNR Hellenic Register of Shipping HR Total and average 22, a Source: The US Coast Guard (1998). b The scores are based on the US Coast Guard detention rates: score 0 for detention rates <2%, score 1 for P 2±5%, score 2 for P 5±10%, score 3 for P 10±15%, score 4 for P 15±20%, score 5 for P 20%. PSC should record all relevant information about sub-standard ships as well as their ags. The uni cation of PSC data recording and publication can be arranged under the auspices of the IMO PSC Committee. 6. Final comments The study con rms that open registers as a group are still the fastest growing eets, and still have higher total loss rates and detention rates compared with the world average, while it is noted that their safety records in general have been improved greatly. About 40% of the total foc tonnage are above average quality eets. It reveals that quality and safety records are correlated with each other, and directly linked to ship ages. Thus the age of a ship should be taken into account as a priority in port inspection. The study reveals that PSC has succeeded in spotting the riskier ships. The practice of categorisation of ships only by its ag should be abandoned. Instead, an integrated-assessment system of the quality of a ship, taking into account the performance records of all parties involved, should be built up worldwide. References Cashman, J.P., Analysis of world merchant ship losses 1967±1975, paper no. 2 in Safety At Sea, Proceedings, London, 23±27 May.

9 K.X. Li / Transportation Research Part E 35 (1999) 135± Doganis, R.S., Metacxas, B.N., 1976a. The impact of ags of convenience. Polytechnic of Central London and Ealing Technical College, London, pp Doganies and Metacxas, 1976b. The impact of ags of convenience. Polytechnic of Central London and Ealing Technical College, London, p. 97. ICJ Rep. (1960). pp. 160±170. Li, K.X., Wonham, J., 1999a. New developments in ship registration. The International Journal of Marine and Coastal Law 14 (1), pp. 128±144. Li, K.X., Wonham, J., 1999b. New developments in ship registration. The International Journal of Marine and Coastal Law 14 (1), 142. Li, K.X., Wonham, J., 1999c. Who is safe and who is at risk: a study of 20-year-record on accidental loss in di erent ags. The Maritime Policy and Management, 1999 Forthcoming issue. pp. 1±8. OECD (1993). Maritime transport 1992, p. 20. Rochdale Report, Report of Committee of Inquiry into Shipping. London, HMSO. Telegraph Special report, Flags of convenience. August 1998, pp. 14±15. UNCTAD (1997). Review of Maritime Transport 1997: Report by the UNCTAD Secretariat. Table 17, in p. 30.