For some 60 years, the liquefied natural gas industry has been relatively catastrophe free, but how long can this "luck" be expected to last? asks Maria Kielmas

For a business punctuated with high risks throughout the chain from well-head production to end consumer, the liquefied natural gas (LNG) industry has been extremely lucky over the 60-plus years of its existence.

This was underlined in December last year following the devastating Indian Ocean earthquake and tsunami. The ExxonMobil-operated Arun LNG plant and port located at the tip of Aceh, Sumatra, escaped damage because the tsunami wave was smaller there than elsewhere. Wimpi Tjetjep, chief of research and development at Indonesia's mines and energy ministry, said that this was "just luck".

The huge loss of life in the aftermath of the tsunami dominated the media reports, so the lucky escape of the Arun LNG plant was largely ignored.

But this was not the case in January 2004 when an explosion at an LNG plant in Skikda, Algeria killed 30 people and injured a further 74. This accident was the worst of its kind in the industry's history, and in one day destroyed claims that it had a clean safety record (see box).

Now the issue of LNG safety has become a science of its own at a time when the LNG industry stands on the brink of a massive world-wide expansion.

Over the last five years, the global LNG market has expanded by 33%. According to the Paris-based International Energy Agency (IEA) some $3.1trn will be invested in new LNG projects over the next 30 years. The increased demand for gas in North America, Europe and Asia, together with improvements in re-gasification technology that have reduced prices, has boosted the demand. So, as the industry grows and changes its structure from that of dedicated cargoes between supplier and client to one of world-wide trading and a spot market, the safety risks and potential liabilities are changing.


For an industry which had not hitherto experienced a major accident, the consequences of the Algerian fire have had reverberations in the US and have undermined industry arguments about the safety of proposed sites for LNG receiving terminals in New England, the Gulf coast and California.

According to Jerry Havens, professor of chemical engineering at the University of Arkansas, the principal safety question is: If the contents of a single LNG ship tank were released onto water and ignited, how big would the subsequent fire be, and how far from the fire would people need to be to be safe?

Tests by the US Naval Weapons Centre in China Lake, California undertaken in the late 1970s and early 1980s provided some answers. The experiments involved a series of spills of LNG on water that were then ignited. The largest of these was approximately 10,000 gallons, and the subsequent fire exceeded 50 feet in diameter and 250 feet in height. But Havens contends that a typical LNG tanker carries up to five cargo tanks, each holding about six million gallons of LNG. While these ships so far have unblemished safety records, natural disasters, human error, or deliberate sabotage could rupture their hulls. Industry studies examining such an occurrence have provided contradictory results, or have been criticised for underestimating the risks, especially those related to intentional action. So the Department of Energy ordered a study of the consequences of LNG spills over water by its agency, Sandia National Laboratory.


Published in December last year, the Sandia study noted that there is little empirical information on the consequences of breaches or large LNG spills. Existing experimental data on LNG spill dynamics and LNG dispersion over water address spill sizes that are more than a factor of one hundred smaller than the spill sizes currently being postulated for some incidents.

Variations in site conditions, LNG ship design, and environmental conditions further complicate hazard predictions. The Sandia report also noted that recent LNG studies by other parties showed a broad range of results, due to variations in models, approaches and assumptions. The four studies were not consistent and focused only on consequences, rather than on both risks and consequences. While consequence studies are important, the report said, they should be used to support comprehensive, risk-based management and planning approaches for identifying, preventing and mitigating hazards to public safety and property from potential LNG spills.

The Sandia study concluded that LNG cargo tank hole sizes for the most credible threats would range from two to 12 square metres, while the expected sizes for intentional threats are nominally five square metres. The most significant threats to public safety and property exist within approximately 500 metres of a spill, due to thermal hazards from fires, with lower threats at distances beyond approximately 1,600 metres. While large, uncontrolled LNG vapour releases are unlikely, vapour clouds could spread over distances greater than 1,600 metres from a spill. If these clouds were to ignite, the resulting hazard ranges could extend up to 1,700 metres for a nominal accidental spill. For a nominal intentional spill, the hazard range could extend to 2,500 metres.

Professor Havens argues that fires would be likely to spread, and cause a total LNG cargo loss. So, in order to protect the public there should be exclusion zones of between one and two miles around LNG receiving terminal facilities. Such receiving terminals should not be located in an urban environment, as is the case for many proposed terminals around the US coast.


However, the construction of LNG vessels themselves is facing a re-think.

According to the London-based Society of International Gas Tanker & Terminal Operators (SIGTTO), LNG carriers have not been exposed to the same kind of vetting by charterers, buyers, terminal operators and sellers, as have oil tankers and LPG carriers. This is because up to now the LNG trade has been between dedicated seller and buyers. But trading patterns are changing, and the vessels are now trading to ports where they have never called previously. This in turn has prompted concern about standards, especially as plans exist for the construction of ever-larger vessels.

The lack of tanker capacity has become a problem for LNG producers. Alain Gautron, partner and LNG shipping specialist at Law firm Orrick, Harrington & Sutcliffe in Paris, observes that some of them are signing contracts for delivery in 2007, although they will be unable to acquire tankers until 2009. He notes that shortages of trained crew will be an even bigger problem.

An expected, 16 - 20 new ships are to be constructed for the Qatar Liquefied Gas II project, a joint venture between ExxonMobil and state company Qatar Petroleum. These tankers are expected to deliver gas to an LNG terminal under construction at Milford Haven in Britain, and they will have a cargo-carrying capacity at least 40% greater than any gas carrier in service.

Cargo-carrying capacities as high as 250,000 cubic metres are mooted for the ships, to be known as Q-max vessels.

This increase in carrier capacity has prompted further fears about fire safety. In late 2004, top officials at the US Department of Homeland Security and the US Coast Guard acknowledged that LNG carriers are constructed using tons of highly flammable polystyrene insulation. A paper by SITTGO states that this material is so flammable that dozens of shipyard workers have been killed in accidental fires during tanker construction and repair.


The LNG industry should also "repaper" for strict liability, thinks Philip Weems, partner at law firm King & Spalding in Houston, Texas. The International Maritime Organisation's (IMO) International Convention on Liability and Compensation for Damage in Connection with the Carriage of Hazardous and Noxious Substances by Sea, known as the HNS Convention, is the first treaty to impose strict liability on the LNG industry for maritime accidents.

It is a major change from other treaties that actually limit the LNG ship owner's liability when at fault. The Convention is a two-tiered liability system up to limits of about $400m. The liability is shared between ship owners and LNG cargo titleholders, and includes compulsory insurance.

Approved by the IMO in 1996, the treaty was ratified by the requisite 12 states in May last year and will come into force later in 2005. Philip Weems thinks that LNG terminal owners will also face potential liabilities in the event of an accident.


The perception of potential terrorist threats to LNG infrastructure sharpened following the 9/11 attacks in the US, and has provided a platform for an increasing number of anti-LNG advocacy groups. "You see a dynamic which has, and will have, a large impact on how the (LNG) market develops in the US," says Dianne Phillip, partner and LNG terminal specialist at law firm Holland & Knight in Boston. Although over 50 LNG receiving terminals have been proposed around the US coast, Dianne Phillips believes that local opposition, which will be taken on board by the federal authorities, will halt the development of most of them. "The only projects that have succeeded in getting through are on the Gulf of Mexico", she says, adding that further receiving terminals are more likely to be located on the Canadian or Mexican coast rather than in the US.

Nevertheless, the energy industry's enthusiasm for LNG has resulted in the entry of a large number of "LNG wanabees" into the market, she thinks, and this will make risk management in the sector more difficult. Previously, the sector was composed of individuals with 25 years or more experience, whose personal and professional relationships were invaluable in a crisis, such as managing cargo delays and diversions. These can occur as the result of the perception of a terrorist threat or of environmental damage and become dominated by regulatory politics. "What are you going to tell a liquefaction facility when you don't know when the next shipment is due?" she asks.

Reinsurers remain calm. According to a spokesperson for Munich Re, the transport of LNG is seen as a high risk, but the catastrophe exposure involved does not compare with that of the company's large accumulation scenarios, and there has been no large marine loss to date. Munich Re considers that accidents involving the handling of LNG pose a considerably lower hazard in terms of widespread spills than the transport of crude oil and crude oil products. As the LNG business grows, reinsurers must hope that such risk perceptions remain correct.

- Maria Kielmas is a freelance journalist and consultant.


LNG is a natural gas that has been cooled in huge refrigeration units known as "trains" to -161 degrees Celsius (-260 degrees Fahrenheit) and where small quantities of nitrogen, oxygen, carbon dioxide, sulphur compounds and water have been removed. The gas is typically 90% methane and should not be confused with liquefied petroleum gas (LPG) that consists of the heavier hydrocarbon compounds butane and propane. After liquefaction the LNG occupies only one six hundredth (1/600) of its gaseous volume and is stored in insulated tanks as a "boiling cryogen", a very cold liquid at its boiling point for the storage pressure. Such storage is necessary while awaiting shipment as well as on ocean-going carriers.

An LNG tanker typically holds 125,000 to 138,000 cubic metres of LNG.

During the journey LNG that evaporates is used to fuel the engines of the carrier. Once it reaches a receiving terminal the fuel is unloaded, stored and regasified. The regasification process makes the gas suitable for transport by pipeline to eventual end-users.


Over the past 60 years of the LNG industry's existence there have been more than 33,000 LNG carrier voyages without a serious accident or cargo security incident, according to the US government agency, Federal Energy Regulatory Commission (Ferc). There are approximately 40 LNG receiving terminals operational at present, mainly in Japan, South Korea, Western Europe and the US. There have been four serious accidents to date:

1. 19 January 2004 - Skikda, Algeria

(Pictured above) A steam boiler that was part of a LNG production plant exploded. This triggered a second, more massive vapour cloud and fire that took eight hours to extinguish. As well as the fatalities and injuries to the plant workforce, the fire destroyed a part of the LNG plant and caused material damage outside its boundaries. The destroyed LNG units will cost $800m to replace. Thirty people were killed and a further 74


2. 1979 - Cove Point, Maryland

A natural gas leak caused an explosion killing one plant employee and seriously injuring another and causing $3m in damages.

3. 1973 - Staten Island, New York

A fire started during repairs to an empty storage tank. The resulting pressure increase inside was so great that the concrete dome on the tank lifted and then collapsed down inside the tank killing 37 construction workers inside.

4. 1944 - Cleveland, Ohio

A tank failed and spilt its contents into the street and storm sewer system. The resulting explosion killed 128 people. The tank was built from a steel alloy with a low nickel content, a consequence of the diversion of nickel supplies to the World War II effort - which made it brittle in the extreme cold of the LNG.