Twenty-five years ago next Wednesday, a ship called the Torrey Canyon, owned by a Bermudan company, on charter to a British one, flying the flag of Liberia and crewed by Italians, ran aground on a notorious reef off the southwest corner of England. The wreck was unusually dramatic because the Torrey Canyon was an unusually large ship and its cargo was crude oil. Despite the efforts of salvage teams and attempts to set it alight by dropping bombs, around 120 000 tonnes of oil leaked out, creating havoc on holiday beaches and killing thousands of sea birds. The age of the supertanker spill had arrived.
At the time, the Torrey Canyon was the world’s biggest oil spill. However, the years since then have seen four worse accidents. In 1978, the Amoco Cadiz became the worst offender when 220 000 tonnes of oil from its tanks washed onto the beaches of Britanny, just over the English Channel, in what the local people called the maree noire, the black tide. And just over a year later, on 19 July 1979, a collision between the Atlantic Empress and the Aegean Captain off the coast of Trinidad in the Caribbean spilled 270 000 tonnes of oil into the sea. Although this environmental crisis remains the biggest ever, it barely warranted a mention in European and American newspapers at the time.
It took a much smaller tanker accident, the holing of the Exxon Valdez off Alaska in 1988, to stir a serious response. In the league table of spills, this was relatively minor – the 36 000 tonnes of oil that leaked out after the 215 000-tonne tanker ran aground was the 28th largest – but it happened in an area of special environmental sensitivity and, crucially, under the scrutiny of the North American media.
Advertisement
The outcry following the Exxon Valdez accident led to the most important attempt yet to force tankers to be leak proof, the US Oil Pollution Act of 1990, which requires tankers operating in American waters to have double hulls. Last week in London, the International Maritime Organization, the UN agency responsible for regulating world shipping, decided that collection of new – and not so new technologies – would prevent catastrophic oil spills from supertankers, known as VLCCs (very large crude carriers). The meeting, of the Marine Environment Protection Committee, said that any new supertanker must have either a double hull, a type of design known as mid-deck (with a horizontal division of oil tanks) or any other design which is as effective at reducing oil spills. These rules will apply to all new supertankers built after July 1996.
Even relatively small improvements can make a big difference. A design that cuts by half the volume of oil escaping after an accident reduces the length of coastline damaged by the subsequent slick by 10 to 15 times. But questions still remain about whether the new technologies and operating practices will work.
The movement of oil by sea is vital to the world’s economy and one of its biggest businesses, accounting for 40 per cent of all seaborne trade. More than 400 VLCCs are afloat today. A large number fly so-called ‘flags of convenience’: they are registered in states that impose less stringent crewing requirements than the major developed countries. Nearly one in five of all oil tankers is registered in Liberia. Panama is the second most popular flag of convenience with about a tenth of the world’s tanker tonnage. Other countries with sizable tanker fleets, in order of total tonnage, are Norway, Greece, the Bahamas, the US and Japan, which has 7 million tonnes. Even landlocked Czechoslovakia has a tanker fleet of almost 6 million tonnes.
Oil tankers are the largest self-propelled machines that human beings have ever created, and the growth in their size has been sudden and spectacular. At the end of the Second World War, few tankers had a deadweight (a measure of a ship’s cargo capacity) of more than 16 500 tonnes. Thirty-five years later, the largest tanker was more than half a million tonnes. The explosive growth resulted from several new technologies, especially the replacement of riveting by welding as the standard way of building ships. The process began during the Second World War, with the Allies’ need to build thousands of cargo ships for Atlantic convoys. After the war, Japanese shipyards quickly adopted the technique to build ships more quickly and cheaply than anyone else.
Meanwhile, from the 1960s, the introduction of computers turned ship design from a rule-of-thumb skill into a more precise science. Naval architects were better able to quantify the loads acting on different parts of a ship and predict how the structure should behave. Tankers jumped in size from 200 000 to 300 000 and even 400 000 tonnes, earning the title ULCCs, for ultra-large crude carriers.
The process culminated in the construction of the Seawise Giant, fabricated in 1980 by adding a new centre section to an existing VLCC to create a deadweight of 565 000 tonnes. The Giant, which survived a bombing by Iraqi jets in the Iran-Iraq war, will probably remain the largest ship ever built. In the end, it was the restricted depth of the sea, rather than limits to engineering, that stopped the relentless growth of VLCCs. Engineers at the Japanese company NKK, which extended the Giant, say they could go bigger if so commissioned, but they do not expect ever to get such an order. Today, most VLCCs are between 250 000 and 280 000 tonnes.
The ruthless application of new technologies had a cost. In a study published last year entitled Tanker Spills: Prevention by Design, the National Research Council of the US observed: ‘Traditional margins – or ‘safety factors’ – were pared back and ship design became relatively lighter, or less robust.’ One example of this tendency was the growing use of high-tensile steels, rather than the heavier but more forgiving mild steels. A tanker built of high-tensile steel is about 15 per cent lighter than one made of mild steel, and this can cut the cost of shipbuilding. Last year Lloyd’s Register, which carries out structural surveys on many of the world’s ships, warned that tankers built with high-tensile steel were more vulnerable to fatigue, the failure of metal subjected to repeated low stresses. As a rule of thumb, there is an inverse-cube relationship between stress and resistance to fatigue: increasing the stress on a component by half reduces the component’s design life by about 70 per cent. As the research council’s study noted with engineering understatement: ‘Initially small, innocuous cracks can grow to significant proportions, and unexpected problems can evolve.’
VLCCs have acquired a reputation among seafarers for metal fatigue and mechanical breakdown. Japanese oil tankers constructed according to a modern design with high-tensile steel have been recalled for emergency repairs to cracks. The tankers, all with a deadweight of around 250 000 tonnes, were made in the late 1980s to a design in which stronger high-tensile steel replaced mild steel.
At sea, VLCCs are notorious for their difficulty in stopping or undergoing any manoeuvre apart from motoring at 15 knots in a straight line. They are so unwieldy that their crews often abandon traditional seafaring practices. As one skipper with more than 20 years of experience in tankers put it: ‘A VLCC at sea is not going to be keeping a lookout, it’s as simple as that. Yes, there’ll be a mate on the bridge, even a qualified one if you’re lucky. But he’ll be too busy with manifests and other paperwork to look where he’s going.’ Some of the most spectacular oil spills, such as those from the Torrey Canyon and the Exxon Valdez, have been the result of errors in navigation.
Today, shipyards in Japan, South Korea and Denmark (the only country outside East Asia capable of building VLCCs) are gearing up in the expectation of a mass of orders for ships to replace much of the world’s fleet of tankers. There are two reasons for the likely surge in demand for new ships. Besides the need for tankers to meet changed regulations governing their design, there is the age of the existing fleet. According to the Shell oil company, the average age of the VLCCs now in service is 13.6 years, and their design life is only 20 years.
The Torrey Canyon set the pattern for disasters. Its captain, who was suffering from tuberculosis, decided to take a short cut to the oil terminal at Milford Haven in Wales. By the time he realised he was on course for the Seven Stones reef it was too late to stop or steer clear. Eleven years later, another combination of human and mechanical frailty led to the wreck of another Liberian-registered tanker, the Amoco Cadiz. The chain of events began with a relatively minor failure of the ship’s steering gear – the Amoco Cadiz, like most VLCCs, had only one propeller, so it could not steer with its engines alone. A salvage tug was on hand, but its master and the Amoco Cadiz’s captain could not agree on salvage terms until the ship was perilously close to shore. The weather worsened. When the tug tried to pull the tanker clear, the tow rope parted. As a last resort, the Amoco Cadiz dropped anchor. But, against all traditions of seafaring, VLCCs’ anchors are useless unless the ship is already stopped. The anchor’s flukes tore off and the Amoco Cadiz slammed into the Britanny shore, breaking in two and spilling 220 000 tonnes of oil.
In the waters of the US, grounding is the most common type of incident for tankers of over 10 000 tonnes that leads to spillage. The most famous of these involved the Exxon Valdez, which lost 36 000 tonnes or 20 per cent of its cargo. The Valdez drama, played out live on television and later in an American courtroom, was the prime reason for the only earnest attempt so far to use the law to prevent accidental pollution. One of the ironies of the carrier industry is that the coast of the US has otherwise suffered little from major spills.
In August 1990, President Bush signed the Oil Pollution Act, designed to prevent repetitions of the Valdez disaster. Unlike most maritime and environmental laws, the act does not lay down standards that industry must meet. Instead it stipulates a specific design – a double hull, in fixed proportion to a ship’s beam that, on VLCCs, creates a gap of about 3 metres between the outside hull and the oil tanks inside. The requirement for a double hull applies to all new tank vessels ordered after 30 June 1990 operating in the US’s territorial waters; existing single-hulled ships may operate under a series of time limits, which expire in 2010. On top of this legislation, tanker owners operating in the US are liable to unlimited claims for compensation for damage that oil spills cause.
Environmentalists criticised the Oil Pollution Act as too little, too late, but the measure shook the oil and shipping industry. Even though American law does not directly affect fleets operating from Europe or Japan, any company wanting to build a new tanker must choose between meeting the requirements of the act or never letting their vessel visit American waters. Japanese ship operators fear that the American law may also set a precedent for international controls.
The Intergovernmental Maritime Consultative Organisation, predecessor of the IMO, was responsible for the main measure taken to prevent oil pollution caused by routine operations, as opposed to accidents. The Marine Pollution Convention (MARPOL) of 1973, which was revised in 1978, stopped a major source of pollution by banning ships from carrying ballast water in their oil tanks. Ballast water is routinely taken on by tankers sailing without cargo, and it inevitably collected dregs of oil from tank walls. When pumped out towards the end of a voyage, it produced chronic pollution that was more pervasive than headline-grabbing spills. Separate ballast tanks cause far less pollution when emptied.
At last week’s meeting of the IMO committee, the US, whose shipping companies are already bound by its own laws, pushed for double hulls to be mandatory worldwide. While a double hull is of little use when a ship breaks clean in half, as did the Amoco Cadiz, it is effective in the more common type of accident where a tanker scrapes over a rock. In high-speed collisions and groundings, however, it would not help much. According to research carried out by Germanischer Lloyd, the German equivalent of Lloyd’s Register, a 3-metre gap between the hulls would be punctured at a collision speed of just over 4 knots. A 6-metre gap would be breached at a collision speed of 7.3 knots.
And there are other snags, notably cost. One Japanese yard puts the cost of a double hull at between 20 and 30 per cent of the total, which is already around $125 million. To modify a tanker would cost about $30 million. Some tanker operators, such as the smaller companies that operate around the Japanese coast, say a requirement for double hulls would put them out of business. However, at a time of relatively low oil prices, the extra cost should be far easier to bear than the sharp rises in the price of crude oil that occurred in 1974 and 1979.
There are also legitimate technical arguments against the double hull, one of which is the question of maintenance. Inspecting an ordinary VLCC for corrosion is already a formidable task. It involves repeated climbs up vertical surfaces – a total of more than 10 000 metres, higher than Everest – and the examination of 1200 kilometres of welding. A double hull adds to the difficulties. Not only is the gap between the hulls difficult to enter, but if used as a ballast tank, as current designs envisage, it is peculiarly liable to corrosion. An empty oil tank is protected by the thin film of oil that is left on its surface; an empty ballast tank, in contrast, is covered with water droplets, which makes rusting inevitable.
When an inspector detects faults, rectifying them by applying a protective coat of paint may be impossible. ‘These are closed spaces, so it is very difficult to do the paintwork,’ says Hisayuki Yamada, a naval architect at the Japanese shipbuilder NKK, which is constructing a double-hulled VLCC of 280 000 tonnes. One solution is to fit painting equipment permanently inside the hull gap. However, installing more equipment to simplify inspection and repair involves another trade-off: the more hardware inside a tank, the more there is to go wrong and the more nooks and crannies there are to trap potentially explosive gases given off by the oil.
Most alternatives to double hulling rely on the phenomenon of hydrostatic balance. To see how this works, it is necessary to understand why oil escapes from a damaged tanker. When a collision or grounding pierces a tanker’s hull, two basic physical forces cause oil to leak out. The first, and most important, is the hydrostatic pressure of fluid inside the tanker, which depends on the fluid’s density and the difference between its level and that of the sea outside. If the pressure inside exceeds the pressure of the surrounding sea, oil will flow out. Ironically, the MARPOL convention exacerbated this effect. Ships fitted with separate ballast tanks ride higher when loaded, increasing the hydrostatic pressure of the oil inside.
Once the oil inside a damaged ship establishes equal pressure with the surrounding sea, hydrostatic losses of oil cease and a second force comes into play. This is the tendency of oil to rise to the top of sea water, which is denser. Because this process requires an inflow of seawater to replace the lost oil, it is slower than losses caused by hydrostatic pressure.
In real accidents, of course, other factors complicate the situation. A certain amount of oil usually flows out as a direct result of impact, and even when the fluids inside and out achieve hydrostatic balance some oil is still likely to escape as an emulsion formed with sea water. Also, a change in sea conditions can disturb hydrostatic equilibrium, as happened in the stranded Exxon Valdez when a falling tide restarted the flow of oil.
The simplest way to prevent leaks is to take advantage of hydrostatic balance and stop filling a tanker while the hydraulic pressure of the cargo is less than that of the sea outside. If the hull is holed, water will tend to flow in rather than oil flow out. This approach, known as hydrostatic loading, appeals to the tanker industry because it is cheap. Operators can begin reducing loads immediately on structurally unmodified ships.
But there are drawbacks too. The first is that hydrostatic balance can never stop pollution entirely, and would fail completely in major accidents. Secondly, it is difficult to police. Unscrupulous operators – the very shipowners most prone to accidents – could gain a competitive edge by filling tanks above the level of hydrostatic balance. Finally, operating ships with partly full tanks creates a new hazard. In rough seas, oil sloshing around inside causes stresses for which the ships were not designed. To make matters worse, it is the older tankers that shipowners want to keep in business through hydrostatic balancing. A corroded hull could break up under such forces.
One refinement to hydrostatic loading is a Swedish idea, assisting balance artificially with vacuum pumps switched on after a collision or grounding. The drawback is that operating pumps in the hours after a collision or stranding could cause an explosion. There is also the question of disposing of the vapour – and probably oil too – that the pumps remove in the process.
Another refinement, which the Japanese government supported at last week’s meeting, is to divide ships’ oil tanks to enhance hydrostatic balance. The giant engineering company Mitsubishi Heavy Industries proposes splitting tanks with a horizontal deck along the waterline, which would reduce the hydrostatic head. The idea is not new. More than 100 years ago, the Royal Institution of Naval Architects was told about the steamer Gluckauf, which had oil tanks divided in the same way. In a case of ‘the ship’s bottom being entirely knocked out . . . the lightness of the oil itself is such that it will remain in the upper part’. Like today’s tanker owners, the institution was chiefly concerned about the insurance consequences of a spillage, although underwriters were liable for only the loss of the cargo, not the damage it caused.
Although the horizontal mid-deck is cheaper to manufacture than a double hull, it is prone to some of the same problems. The mid-deck design creates dark spaces that are difficult to clean and where it is difficult to inspect for faults, such as corrosion.
Philip Embiricos, chairman of the Safety and Technical Committee at Intertanko, the organisation of independent tanker owners who operate 70 per cent of the world’s tankers, favours a concept that takes advantage of both hydrostatic balance and the mid-deck design. In this scheme, known as the rescue tank, a series of valves and pipes connects the tanks of the mid-deck design to ballast tanks. The ballast tanks would be around six metres broad, thus reducing the chance of a collision breaching the inner oil tanks in the first place. But if any oil tanks are breached, hydrostatic pressure forces water into them and oil automatically flows through the pipes and valves into those ballast tanks that remain intact.
In a high-speed collision or grounding, this design causes a small amount of pollution at first as a result of the impact. But tests by Lloyd’s Register at Croydon, near London, showed that the rescue tank concept creates a greater safety margin than the mid-deck design, especially if there is subsequently a very large drop in the level of the sea, as in the case of the Exxon Valdez. The drawback of the design is that it requires a network of pipes and valves to be maintained in working order for an eventuality that is fairly remote.
Perhaps the most exotic technology under discussion is to line the insides of oil tanks with a flexible membrane that would survive a collision without being punctured. This poses a whole set of new difficulties, not least of which is that no known flexible material is up to the job. The need to withstand fire and sharp impacts rules out rubber and any known polymer. Even if engineers developed a suitable membrane, fitting it around the pumping and cleaning machinery scattered around an oil tank would be difficult.
A group of IMO experts studied the rival designs at a series of meetings, the last one of which was in January. The group concluded that the mid-deck design and double hulls gave ‘equivalent protection’ to the environment. It found that in 8 out of 10 groundings, the inner hull on the double hull design would not be ruptured, so this design would not pollute. The mid-deck design would cause some pollution, but most of the cargo would stay on board. However, in high-speed groundings the mid-deck design has the advantage. Collisions severe enough to rupture the inner hull of a double-hulled ship would cause far more pollution than would occur with the mid-deck design.
For all the good intentions of the IMO, new technologies are only part of the answer. Accidents are usually complex and some devices intended to reduce pollution can increase the risk of shipwrecks. Shipowners cite another reason why new technologies might cause more accidents: reducing the amount of oil each ship can carry makes necessary a bigger fleet, increasing the risk of collisions. In the 25 years since the wreck of the Torrey Canyon, human ingenuity has produced some breathtaking technology for extracting and transporting oil. Only now is the same ingenuity being applied in earnest to stopping oil leaking from shipwrecks, and that is a matter of concern – and shame.
Michael Cross is a freelance journalist.
Further reading Tanker Spills: Prevention by Design, a study by the National Research Council, published in 1991 by the National Academy Press, Washington DC. Significant Ships of 1991, published by Royal Institution of Naval Architects, London. Price £15.