01:00 AM EDT on Tuesday, May 24, 2005

IN OUR HASTE to secure our critical infrastructure after Sept. 11, 2001, we repaired the most glaring chinks in our collective armor. We filled holes in structures (e.g., airport baggage and passenger screening) and strengthened areas where we were exposed (e.g., visa enforcement).

But despite our efforts, we will always have weaknesses, and attractive terrorist targets cannot be entirely eliminated. The best we can do to protect our citizens is to creatively address our deficiencies and reduce our vulnerabilities.

In this regard, KeySpan's proposal to expand its liquefied-natural-gas facility in Providence would be counterproductive to our homeland-security effort: It would create a vulnerability where none previously existed, and endanger the lives of thousands of Rhode Islanders.

It is tempting to apply traditional risk-management methodologies to the new threat environment. Risk is traditionally thought of by security specialists as the potential for some unwanted event to occur, and is quantified as the combination of the threat, vulnerability, and consequence of a particular action. But such an approach should now be viewed as insufficient to combat the threat posed by modern jihadist terror networks.

Threat is determined by the intent and capability of terrorists to wage a specific type of attack, and by their track record of successful such attacks. This formula would have concluded that the probability of terrorists' using hijacked commercial aircraft to destroy the World Trade Center was zero. We now know that the inability to quantify a threat does not mean that the threat does not exist.

KeySpan would like Rhode Islanders to believe that the expansion of its Providence LNG facility and the weekly voyages of 900-foot-long tankers carrying 38 million gallons of fuel up a 29-mile route through Narragansett Bay pose little threat to the area's people and property. The Federal Energy Regulatory Commission has said that the risks are "manageable" and "acceptable." It points to the commendable record of LNG-facility and tanker safety over the past four decades.

But this rationale is based on the notion that because a terrorist attack against an LNG asset in the United States has never happened, it can never happen. That is dangerously illogical. History is nothing more than a list of events that never previously happened.

We have devised a five-pronged security-risk-management methodology to assess the dangers of KeySpan's proposal:

The first step is to determine whether terrorist groups have shown an intent to attack LNG tankers and/or facilities. Al-Qaida, other jihadist groups, and even domestic terrorist organizations have articulated goals that include killing large numbers of Americans inside the United States, launching "spectacular" attacks, damaging the U.S. economy, and damaging the oil-and-gas infrastructure. Al-Qaida has shown an intent to attack ships and the shipping industry by its assaults on the USS Cole and the French oil tanker Limburg.

The second element we examined was whether terrorists have the capability of launching such an attack. Even after 9/11, al-Qaida has shown an ability to operate in the United States; a recent report indicates that the FBI has over 1,200 full field investigations under way against al-Qaida alone. The weapons needed to attack an LNG facility or tanker are readily available in the United States. Small boats and scuba gear -- in which terrorists have shown an interest, for underwater attacks -- are easily obtainable. General-aviation aircraft can be rented or stolen, and explosives -- both fertilizer-based and commercial -- can be easily manufactured or bought on the black market. Assault weapons, including rocket-propelled grenades and light anti-tank weapons, can be bought illicitly on the international arms market.

The third factor we considered was the vulnerability of the potential target. Narragansett Bay's 29-mile inland-waterway transit provides ample opportunity for attack aircraft, small boats, divers, and small arms fired from shore or the Pell Bridge. If a small boat or aircraft was deemed a threat, the military and law-enforcement officials escorting the tanker would not be authorized by their rules of engagement to use deadly force against what could be simply a wayward civilian craft.

The LNG facility is equally vulnerable. Its security, even with the proposed upgrades, would be unlikely to prevent the two-wave style of attack used by al-Qaida in Saudi Arabia, Iraq, Kenya, Pakistan, and elsewhere: An initial explosion or small-arms attack is aimed at security gates and guards, followed by an attacking vehicle, carrying a large explosive directed at the primary target.

The consequences of an LNG attack were the fourth component we assessed. There is a high risk that such an attack -- especially on a tanker docked at the facility during unloading -- could be catastrophic. Our analysis of the KeySpan proposal shows that 3,000 people could be killed and 10,000 severely burned in just the first minutes of a large-scale LNG release and fire at Fields Point. Thousands of others could be killed from sympathetic fires, which could also generate the release of noxious materials, should storage tanks in nearby chemical facilities detonate. Emergency responders and hospitals' trauma and burn units would be totally overwhelmed.

Furthermore, since LNG fires cannot be extinguished by conventional means, firefighters would have to adopt a "let it burn" response -- forcing many victims to fend for themselves.

The final aspect of our assessment was the recovery costs after an attack: the financial cost of compensating victims and rebuilding damaged or destroyed facilities, as well as the opportunity cost to the local economy. These costs would probably exceed any insurance carried by the owners and operators of the LNG facility and tanker.

There is also a hidden cost. In the absence of adequate insurance, the LNG company would transfer the financial cost of the risk it was creating either to the victims or to governments, or to both.

There are viable alternatives to KeySpan's proposal that would remove the risk associated with LNG facilities and tankers in populated areas. Siting the facility in a non-urban environment or offshore -- thereby eliminating the inland-waterway tanker transit -- would greatly reduce both the attractiveness of the target to terrorists and the consequent management and recovery burdens, should an attack occur.

Although there is no adequate way to determine the probability of a terrorist attack on the proposed KeySpan facility or tanker, a cost trade-off can be precisely measured.

If the Federal Energy Regulatory Commission approves the KeySpan proposal, the federal government will be deciding that avoiding the additional financial cost to KeySpan of a more secure location is more important than avoiding the additional risk to Rhode Islanders of a catastrophic attack inherent in urban LNG facilities.

Richard A. Clarke was the national coordinator for security and counterterrorism for Presidents Bill Clinton and George W. Bush; Evan Pressman and Daniel Dolgin are master's-degree candidates at the Fletcher School of Law and Diplomacy, at Tufts University.