Sunday, March 06, 2005

How Big Is the Threat of LNG Tankers and Terrorists?

Let me make this easy for those of you who like to cut to the bottom line: after reviewing a considerable amount of information, my conclusion is that LNG tankers do not present particularly good terrorist targets.

Let me also say that I was somewhat surprised to reach this conclusion. When I was a young officer of the deck, heading into Tokyo Bay, I was impressed with a couple of things. First, after a long independent ocean transit, the sudden appearance of a huge amount of shipping (the radar screen went white with contacts) and the high speed of the LPG tankers flying through the relatively narrow bay entrance. Knowing nothing about them, my guess was that an LPG tanker is like a bomb ready to go off. In recent years my guess seemed to be confirmed by articles like this: Study: LNG tankers make spectacular targets for terrorists

And, more recently, my concerns were heightened by dramatic reporting from sources like the Washington Post:
Terrorist attacks on tankers carrying liquefied natural gas into a U.S. port could trigger a fire that could burn the skin of people a mile away and cause "major injuries and significant structural damage" within about a third of a mile, according to a government study released yesterday.

The report, prepared by an Energy Department laboratory, said that terrorists could use rocket-propelled grenades, missiles, planes or boats to break open the tankers.

However, I have never heard of an LNG tanker explosion destroying a town (unlike, say the Navy's Port Chicago explosion near San Francisco during WWII). Where were these spectacular blasts? How big a threat is an LNG tanker in the wrong hands or under attack?

To learn more about LNG, I went to the FERC's website, FERC which answered the question -
Is LNG explosive?
In its liquid state, LNG is not explosive. When LNG is heated and becomes a gas, the gas is not explosive if it is unconfined. Natural gas is only flammable within a narrow range of concentrations in the air (5%-to-15%). Less air does not contain enough oxygen to sustain a flame, while more air dilutes the gas too much for it to ignite.
In fact, the FERC site sent me off to a 167-page report (referenced in the Washington Post article) prepared by Sandia National Lab which contains a lot of science and in the Executive Summary the following:
3. Risks from accidental LNG spills, such as from collisions and groundings, are small and manageable with current safety policies and practices.
4. Risks from intentional events, such as terrorist acts, can be significantly reduced with appropriate security, planning, prevention, and mitigation.
5. This report includes a general analysis for a range of intentional attacks. The consequences from an intentional breach can be more severe than those from accidental breaches. Multiple techniques exist to enhance LNG spill safety and security management and to reduce the potential of a large LNG spill due to intentional threats. If effectively implemented, these techniques could significantly reduce the potential for an intentional LNG spill.

Not much in the way of news there.

Historically, the U.S. has had one LNG incident 61 years ago, as reported by the Sacramento Bee:
Questions about the fuel's safety date back to 1944, when a Cleveland, Ohio, tank built with inferior steel ruptured and spilled LNG into the street and storm-sewer system. The gas ignited, triggering an explosion and fire that killed 128 people.

The industry uses better steel today and has reported no further tank ruptures in the United States.
Any accident is bad, but 61 years is a long time.

On the other hand, worst case scenarios are easy to find:
Opponents say terrorists or earthquakes could turn the terminal's vast stores of super-cooled gas into a mile-high fireball, destroying the port and scorching people and buildings a mile away in downtown Long Beach.

"If you are close when that happens and you are out in the open, you are doomed," said James Fay, a Massachusetts Institute of Technology professor emeritus and liquefied natural gas specialist. "No one is going to (be able to) come in and rescue those people."

Again, LNG has an impressive safety record:
LNG has been delivered across the oceans for about 45 years without major accidents or safety problems, either in port or on the high seas. In that time, there have been more than 33,000 LNG carrier trips, covering more than 60 million miles. Today, more than 150 LNG ocean tankers safely transport more than 110 million metric tons of LNG annually to ports around the world. This is more than all American homes consume each year. In 2000, one LNG cargo entered Tokyo Bay every 20 hours, and one entered Boston harbor every week. Japan relies exclusively on imported LNG for its natural gas. According to the U.S. Department of Energy, over the life of the industry, eight marine incidents worldwide have resulted involving accidental spillage of LNG. In these cases only minor hull damage occurred, and there were no cargo fires. Seven additional marine-related incidents have occurred with no significant cargo loss. No explosions or fatalities have ever occurred.
Source Center for Liquified Natural Gas

LNG is transported in double-hulled ships designed to prevent leakage or rupture in an accident. LNG is stored in either double membrane containment systems made of special materials and located within the ship's inner hull or special three-quarter inch thick spherical tanks. For membrane containment systems, a secondary containment system surrounds the primary container. The insulation space between the two has sensing equipment able to detect even the smallest presence of methane (the main component of natural gas), possibly indicating a leak of LNG.

A terrorist attack on an LNG facility or LNG carrier would most likely involve an attempt to spill a large quantity of LNG in the hope of generating a flammable vapor cloud. As discussed earlier, this probably requires causing a massive failure of a storage tank, either on the ship or on shore. The force required to breach the tank would almost certainly result in a fire at the tank location thereby limiting the damage potential as a result of the action. Given that other targets that could cause more widespread damage with more certainty of outcome are available, terrorism targeting an LNG facility is thought to be unlikely. Notwithstanding that, law enforcement agencies and LNG operators have developed procedures to thwart any such attempts.

What? No explosion with thermonuclear- like force? Why not? Part of the answer lies in the nature of LNG as explained here:
Like other products that are considered flammable liquids, LNG must first be vaporized, then mixed with air, and then exposed to an ignition source before it will ignite. Natural gas in its liquid form, LNG, cannot ignite. Only the natural gas vapor, which forms when LNG’s temperature rises, can be ignited but its flammability depends primarily on its air content. The flammable range lies between 5% and 15% in air, by volume. For ignition to occur when LNG vapor contacts a hot surface, the temperature of that surface must exceed 1004ºF...When LNG is confined within a tank, ignition of the natural gas vapors cannot occur due to the lack of oxygen. If LNG leaks out and begins evaporating in an open area, the natural gas vapors are often quickly dispersed by wind, making ignition unlikely. If ignition of the natural gas vapors occurs, the gas does not burn rapidly like gasoline, but forms a slow burning flame that burns back to the source of the natural gas vapor, until the fire is extinguished or the fuel is exhausted.

Some members of the public have raised concerns about the possibility of LNG explosions. LNG is not stored under pressure. If the tank is ruptured, there is no massive release of energy and thus no explosion. If an object of force such as a missile punctured an LNG tank a fire could result but the tank would not explode.

For an explosion to occur, LNG must first return to its gaseous state and then the natural gas vapors must accumulate in a confined space in a perfect mixture, of 5% to 15% of gas in air, and encounter an ignition source.

Yeah, but that's an industry source, what does Dr. Fay say? Here's his premise:
The events of September 11, 2001 have raised concerns about the potential for terrorists attacks on the energy system infrastructure of the United States. In particular, the possibility of the use of a boat bomb, such as was used against the USS Cole in 2000 and the oil tanker Limburg in 2002, to attack a marine liquid fuel tanker in a U.S. harbor, was publicly discussed in Massachusetts, where both LNG (liquefied natural gas) and oil product tankers land cargoes in Boston harbor. The consequences of such an incident could be severe, and present a potential problem of great magnitude for public safety officials.

The safety concerns for the public stem from the effects of the burning of the tanker's combustible liquid cargo, which would certainly escape from cargo holds punctured by the force of an explosion. The ensuing fire can spread on the sea surface toward nearby shorelines, and its thermal radiation could produce bodily harm to exposed individuals on shore and possibly set fire to shoreside buildings.

The fire that would ensue from a boat bomb attack on a tanker would be of unprecedented size and intensity. Like the attack on the World Trade Center in New York City, there exists no relevant industrial experience with fires of this scale from which to project measures for securing public safety. Lacking such experience, we must rely on scientific understanding to predict their characteristics, based upon laboratory and field experiments of much smaller fires.
The liquid fuel carried in sea-going LNG tankers is stored in separate holds, each of which may be as large as 25,000 cubic meters holding 10,500 tons of cargo. A powerful explosion close along side the tanker can puncture at least one hold and allow the cargo to drain out upon the surrounding sea surface. The upper part of the cargo fluid that is higher than the sea surface level will first leak out, but additional cargo may also be ejected. Given an explosively formed hole of sufficient size, such cargoes can be disgorged within minutes.

LNG is lighter than sea water. Once spilled, it floats, unmixed, on the sea surface. Most importantly, it speedily spread sideways, exposing the fuel to the air above. Once ignited, as is very likely when the spill is initiated by a chemical explosion, the floating LNG pool will burn vigorously. The time to burn spills of the size mentioned above can be less than five minutes.

Fires that burn thousands of tons of fuel in a few minutes are extraordinarily large, lying well outside the range of domestic firefighting experience. Such fires can be damaging to people and can set afire combustible buildings.
To illustrate the characteristics of such spills in Fall River harbor, we consider a typical spill of LNG. (The relevant spill parameters are listed in Table 1.) The LNG spill volume is 14,300 cubic meters or 3.8 million gallons. Provided the vessel hole area is greater than ten square meters (Eagle1 Note: ~30 sq. yards), the maximum pool fire area is 180,000 square meters (44 acres) and radius is 340 meters (1115 feet), while the fire duration is 3.3 minutes.

The pool fire, initiated at the time of the explosion, grows in area in proportion to the time since initiation, reaching maximum extent at the end of the burning process. Maximum pool size for an LNG spill located at the proposed LNG terminal: the outer edge of pool fire extends to both east and west shores of the Taunton River. For a spill anywhere along the path of an LNG tanker approaching the terminal, the pool fire would reach Fall River shore. It is most certain that combustible buildings long the waterfront would be ignited by contact with the pool fire.

The extent of the pool fires, which spread to distances greater than the ship length in a short time, would make it impossible to move the stricken vessel away from the waterfront areas. The potential for retarding the pool spread is nonexistent.

Pool Fire Thermal Radiation:

Burning LNG emits thermal radiation that, if intense enough, can cause skin burns on humans exposed to the radiation and can ignite combustible materials on buildings. The more intense the radiation, the shorter is the exposure time needed to cause a skin burn or combustible material ignition.

For human skin exposure to flame thermal radiation, a thermal flux of 5 kilowatts per square meter will result in unbearable pain after an exposure of 13 seconds and second degree burns after an exposure of 40 seconds. Exposure to twice that level, 10 kilowatts per square meter, for 40 seconds is the threshold for fatalities (K.S.Mudan, Thermal radiation hazards from hydrocarbon pool fires, Progress in Energy Combustion Science, 10, 59-80, 1984). Wood can be ignited after 40 seconds exposure at a thermal flux of 5 kilowatts per square meter.

We have chosen a thermal flux of 5 kilowatts per square meter a a criterion for the limit for significant damage to humans and combustible materials and have calculated the distance from the spill site at which that flux would be experienced (These distances are based upon an analysis contained in Fay, Model of large pool fires, submitted to the Journal of Hazardous Materials). As listed in Table 1, this distance is 1100 meters (3600 feet or 0.68 mile) for an LNG spill.

For an LNG spill, the thermal radiation damage zone encloses 940 acres, including about 400 acres of land area in Fall River. Within this zone, extending 3600 feet from a spill site in the main channel of the Taunton River, skin burns to humans exposed for only a fraction of a minute will occur, and building fires can be induced. Beyond the shorefront, at 1600 feet from the spill site, where the thermal radiation flux is 10 kilowatts per square meter, fatalities can ensue.

One cannot exaggerate the thermal intensity of the LNG pool fire. It's average heat release rate is about twice the average thermal power consumption of all U.S. fossil fuel electric power plants.
In short, a short duration, high intensity incident under the worst of all conditions - in short if the terrorists accomplished all of their goals under ideal (for them) conditions. But how realistic is it to assume a perfectly windless day coupled with all the other events that would have to happen in exactly the right sequence? In my view, not very. In my view, you have a greater probability of having the gasoline in the tank of your car cause you a problem. It is not enough for these scientists to merely demonstrate possible worst case situations without providing information on the likelihood of such a event coming to pass. Let me put it this way. It might be possible that a large commercial airliner could crash into my house. If one did, it would cause a great deal of damage in a fairly large area. However, the odds against such a crash occurring, while not completely zero, are so small that it is not worth me worrying about it or taking any steps to ban commercial aircraft from flying within 50 miles of my home. More likely events I take reasonable precautions against. Just as the Coast Guard and others take reasonable precautions against even the remote chance of someone attacking an LNG tanker.

Suffolk Law School Environmental Law Society summarized the industry position:
Industry studies predict that a LNG fire aboard a tanker who be less destructive than Fay hypothesizes, and point to the fact that the hull of an LNG tanker has many layers and incorporates protective structures than single-hulled battleships like the USS Cole do not possess. These studies indicate that while LNG mixed with air is flammable, LNG by itself is not, and that while the gas would rapidly expand from its compressed state, the size of the hole created by such an explosive would alleviate the pressure needed to create a "fireball"� type explosion.
That all sounds bad. But how likely is it to occur? After all, it seems Dr. Fay has postulated an absolute "worst case" scenerio.

One of those experts, Jerry Havens, developed the computer software used by federal agencies to
determine safety buffer zones around facilities handling hazardous materials such as LNG. Havens, a distinguished professor of chemical engineering at the University of Arkansas, has consulted with numerous government agencies, including the Army, the Department of Energy, the Environmental Protection Agency, as well as with Exxon, British Petroleum and other oil
companies. He served as an officer in the Army's chemical weapons division. The other expert, James Fay, is a retired Massachusetts Institute of Technology professor whose work on liquefied natural gas is cited in congressional reports and forms the basis for many of the existing scientific predictions regarding the behavior and dangers of LNG fires. In one of the worst-case fire scenarios calculated by these scientists, people in the neighborhoods closest to an LNG fire would be "in unbearable pain after an exposure of 13 seconds." They said that trees and wooden buildings within a half mile would ignite within 40 seconds due to superheated air created by the burning natural gas. Wooden buildings and vegetation over a broader area would catch fire soon after, they said."If you are standing 1,000 feet away from an LNG pool fire you are simply going to die," Fay said.

The scientists and the ExxonMobil officials agreed that their very different scenarios may hinge
on calculations of the potential size of a fire. The independent scientists estimated that a much larger and more serious fire could result from a serious accident, which they speculated would most likely result from some kind of terrorist
attack that penetrated a ship's hull. They also warned that the federal regulatory process does not address many of the safety issues generated by an LNG terminal and in no way addresses the hazards posed by fires emanating from LNG tanker ships. The scientists said they were concerned that researchers have never investigated fundamental safety questions regarding the tankers that deliver liquefied natural gas around the world. They also said that all of the existing studies examined the fire risks posed if just one of the five storage containers on an LNG tanker was breached. "No one should assume a tanker is only going to lose one cargo tank. The ship will be engulfed in an enormous and very intense fire," Havens said. "I do not think anyone has studied what will happen to the ship as a result of this fire. It is possible the other tanks could also lose their cargo. If other tanks fail and contribute to the fire, the fire would get much larger. I certainly think this question needs to be investigated before you put these terminals in populated areas."

Of course, LNG tankers are not allowed to willy-nilly enter U.S. ports (and neither are other types of ships under the Homeland Security laws). Every LNG tanker planning to enter U.S. waters will have a Coast Guard inspection, ecorts and other efforts to minimize a terrorist take-over of the ship. In addition, when moored, a LNG tanker will be protected by some form of barrier precluding the use of an "exploding boat" as set out in the premise of Dr. Fay.

What about an "at-sea" take over and the use of an LNG tanker as a shipborne improvised explosive device (SIED)? Again, it wold take a substantial amount of work on the part of terrorists and everything they did would have to work perfectly in a perfect environment. Breach the LNG vessels, spill the LNG, let the LNG reach the right 5-15% air mixture and try and accomplish all this uderway while ramming or approaching other ships at speed without wind. Not likely.

I encourage doubters to look throgh all the literature and make a compelling counter argument (you can email your argument to me and I promise I will post it). In my opinion, there are easier targets for terrorists that could achieve the spectacular results they seek. Don't lose much sleep over this issue.

Update: About BLEVE (Boiling Liquid Expanding Vapor Explosions) and VCE (VaporCloud Explosions) - according to one study BLEVE and VCE are "generally not credible for LNG marine releases."
A BLEVE (Boiling Liquid Expanding Vapor Explosions) is most often associated with fire impingement of a pressurized liquefied gas (e.g. propane or butane) contained in a pressure vessel. While the pressure relief valve can maintain the pressure within the vessel's allowable limit, the fire can weaken the metal shell in the upper vapor space (but not the lower liquid space) so that it can fail. The sudden opening of the pressure vessel releases large quantities of pressurized liquid that immediately flashes to gas at atmospheric pressure. The flashing gas is ignited by the fire and expands rapidly and turbulently, radiating heat as a fireball. No significant pressure wave is generated.

An LNG-related BLEVE would require that a large external fire could exist and cause the pressure inside the LNG tank to rise to a significant level such that upon tank failure due to thermal weakening a large flash of pressurized methane would occur. There is no potential for this to occur either for the membrane or spherical design. The cargo tanks are not designed for significant pressure and pressure relief would limit the pressure rise to a small amount insufficient to cause a BLEVE event.

A VCE (Vapor Cloud Explosion) event was reviewed for its potential. Bull (2004) has summarized the now well advanced current knowledge on vapor cloud explosions and the conditions necessary to cause combustion of a cloud of flammable gas to accelerate sufficiently to generate pressure effects. Methane in the open (as over water) has no mechanism to cause the flame front to accelerate. Work by Mitzner and Eyre (1983) which ignited several LNG flammable clouds found most flame
fronts progressed slowly, often under 10-12m/s. They reported that in one case the flame front could not progress against a wind speed of 5m/s and that several clouds also self-extinguished. LNG vapors are very cold and the cloud will be filled with
condensed fog. These create very poor combustion conditions. DNV believes that LNG VCE events are not credible over water, but that they may need to be considered in specific situations if flammable vapors can drift into highly congested areas where there is sufficient obstruction present to result in an acceleration of the flame front to speeds where overpressures are developed. Typically, such events are considered for releases within process plants where there is a large number of process equipment such as piping, vessels, and equipment supports. (emphasis added)
The point is that the LNG tanks are not pressurized, but are kept at normal atmospheric pressure, thus negating the concern over BLEVE except under extraordinary conditions. Source: here.

1 comment:

  1. This comment has been removed by a blog administrator.