It’s the worst of worst-case scenarios: terrorists develop the capacity to unleash a chemical, biological, nuclear or radiological (CBNR) attack on a Western city, killing hundreds of thousands and causing unprecedented destruction.

The nature of this threat – and what role insurance could and should play – was debated at the recent Australian Reinsurance Pool Corporation-OECD Global Terrorism Risk Insurance Conference in Canberra.

Experts agree that while such an attack is possible, it’s highly unlikely.

Terrorists may be capable of deploying chemical or biological weapons in Iraq or Syria, but to do so in developed nations would be an entirely different proposition.

Also, it is important to distinguish between the C, the B, the N and the R. Each threat is unique, and they cannot be considered as one.

Scott Hawkins, GM Non-Life at Munich Re, explained the challenges from a reinsurer’s perspective.

“As we all know, the principle of insurance or reinsurance is we need to know how often something happens, and how much it’s going to cost when it does happen,” he said.

Terrorism presents a very different test to natural catastrophes in this regard, due to its instability, reliance on perpetrators’ states of mind and a lack of previous data.

“We pretty much know from a natural perils point of view how often we expect an earthquake to happen somewhere, how often we expect a cyclone. That may change over time, but generally over a 12-month period it’s fairly stable.

“Clearly when we talk about terrorism, it’s very unstable, and over a one-month period the frequency could change remarkably.

“That’s something that, as an insurance industry, we are not used to.”

It can also be hard to predict the success of any attack, and what may be targeted.

“How do we assess the likelihood of where they will attack? It’s easy to say key targets are things that will create damage, hysteria and publicity.

“But is it actually easy to get those key targets, or is it easier to get a secondary target where you know the security is not there?

“There is a lot of protection in catching these people before they can do it. As a citizen, that is a great comfort, but as someone who needs to run a reinsurance business and have someone tell me how likely something is to happen, that’s very hard to factor in.”

Modelling such attacks has a long way to go – but that doesn’t mean we should give up.

“We need to start off with basic tools or models that give us some insight,” Mr Hawkins said.

“We realise that insight won’t be perfect and we need to improve it over time.

“But the worst thing we can do is wait until we have the perfect model before we implement it. Because, let’s face it, we won’t actually get the perfect model.”

Steve Coates, Chief Underwriting Officer at Pool Re, says the UK reinsurance scheme changed focus to incorporate CBNR following the September 2001 attacks on the US.

“9/11 showed that something completely unprecedented and unpredicted happened,” he said.

“The cover had to change and was broadened out to all risks, thereby [including] CBNR.

“The only things we excluded were war and cyber. The thought was that future attacks might include CBNR.”

He believes terrorists would undoubtedly use CBNR in developed countries if they could, and pools must adapt.

“The ability of terrorists to do stuff in the UK or France or wherever is completely different to the stuff they are doing now in Syria or Iraq,” he said.

“But the million-dollar question is, what happens when some of these guys come back?

“Will they come back with not only weapons training but also some idea about how to sling some radioactive material onto the top of a bomb?

“It’s no good thinking we are going to develop a solution to a problem 20 years ago, and not think that solution might have to change over time, because clearly the risk changes.

“The risk is very different now, it’s much more diffuse, much more complicated, and for us to be relevant to today’s risk we have to make sure we are providing cover against the emerging threats of terrorism in the UK, one of which is clearly CBNR.

“So we’ve dealt with that and in the future we might have to deal with other threats such as cyber.”

Each element of CBNR would produce very different losses, and while all are unlikely, some are more unlikely than others.

Everyone understands how difficult it would be to make or obtain a nuclear bomb, Mr Coates says.

“But from a radiological perspective, it’s a lot easier – easier being a relative word – to get some radiological material, stick it on top of a conventional bomb and make it explode.

“People sometimes say radiological material is very difficult to get hold of. It is, but actually it does sit in some fairly accessible places in all our societies, hospitals being a good example.

“Some radiation and X-ray machines will have caesium and cobalt within them, and if you’ve got somebody who has got a death-wish and they are willing to put their hand in and lift that stuff out and put it in their pocket, then you’ve got a problem.”

Given that CBNR attacks are so unlikely, and so difficult to model, should we even be worrying about them?

Absolutely, according to Mr Hawkins.

“The likelihood of [terrorists] using these sorts of methods going forward is increasing,” he said. “So while it’s easy to say, ‘It’s really difficult and no one’s been able to do it much in the past, we don’t need to worry about it,’ that would be missing the point that this will change into the future. We need to keep one step ahead.”

Chemical weapons are defined as including any toxic chemical or its precursor that can cause death, injury, temporary incapacitation or sensory irritation through its chemical action. Some of the most well known categories are nerve agents such as sarin and VX; blood agents such as hydrogen cyanide; blister agents such as sulphur mustard and other mustard agents; choking agents such as phosgene; and irritants such as tear gas.

Biological weapons, also referred to as bioweapons, are deadly pathogens – bacteria, micro-organisms or viruses – or toxins that can be deliberately released to inflict harm. These weapons can be disseminated through inhalation, ingestion or skin absorption. Some biological agents may be better suited to affecting large numbers of people, such as the highly contagious severe acute respiratory syndrome – or SARS – virus, while others such as Ebola may be less contagious but more deadly.

Nuclear weapons rely on nuclear energy produced by either fission or a combination of fission and fusion of atomic nuclei. Nuclear weapons have not been detonated in armed conflict since 1945, and most concerns today tend to centre on states selling them to terrorists, or the security of highly radioactive nuclear material (uranium and plutonium) that could be stolen for use in an improvised nuclear device.

Radiological weapons disperse radioactive material, often via an improvised explosive device. This is commonly known as a “dirty bomb”. While most immediate fatalities in such an attack would be caused by the explosion rather than the levels of radiation, exposure of people and the environment to radioactive contamination would cause massive disruption and have a severe psychological impact on those affected.

In Australia the ARPC is designed to cover chemical and biological terror attacks, but not nuclear or radiological.

“It was not in the mandate of the scheme,” CEO Chris Wallace told insuranceNEWS.com.au.

“It is a very low risk, but a very high-consequence event, and it would have to be dealt with differently.”

Before 2001 the simultaneous hijacking of four airliners would have been thought to be unlikely or even impossible, had it not been unforeseeable.

The breadth and depth of the terrorism threat is still a perplexing one for the various specialist agencies deployed to anticipate and prevent attacks. For insurers it is in many ways even more difficult.

Insurance works most efficiently when dealing with measurable risks, and terrorism relies largely on the unexpected. It will continue to be a risk that’s difficult to deal with.