The Effectiveness of Chemical Weapons: The Banal Truth
The on-going use of chemical warfare agents in Syria and the use of a so-called “Novichok” agent for an attempted assassination the UK have brought non-stop commentary - some from the Kremlin designed to mislead, some simply ill-informed - on the nature of chemical weapons. Chemical warfare agents (the actual chemical compounds used for chemical warfare) and chemical weapons (the systems which employ them) are subjects that are not well understood except by specialists, so some degree of uninformed commentary is to be expected.
Half-truths and misconceptions are both accidentally and deliberately exploited for purposes of disinformation and hybrid warfare. Indeed, it seems that there is a tendency to ascribe properties to chemical weapons and chemical warfare agents which they do not actually have.
The overall point is that chemical warfare agents and their associated weapon systems are generally less lethal than people make them out to be. One dirty little secret of chemical warfare is that it actually has a poor track record when compared to conventional ways of killing people. There are several reasons for this, and I will attempt to unpack them.
Some chemical weapons not designed to kill quickly
One point that is rarely understood is that killing people, and killing them quickly (the so-called “sudden death” or “instant lethality” occasionally referred to in lurid commentary, is generally not the primary design criteria for a chemical weapon. Only some chemical warfare agents are designed to quickly kill people as the direct outcome of their use. Believe it or not, many chemical warfare agents do not kill quickly. Some do not kill very well at all, such as mustard gas, which killed only a low percentage of its victims in the First World War according to a comprehensive post-war study. Many chemical warfare agents and chemical weapon systems are not even designed with killing as the ultimate objective. As a practical matter, a soldier who is too sick to fight, but who might get better with medical care, is more of a logistical burden to many armies than a soldier who dies on the spot. Someone who takes hours or days to die is also quite a blow to morale. The idea that chemical weapons are all instantly lethal or meant to be instantly lethal is generally not supported by the facts.
Nerve agents are almost entirely liquids at normal temperatures and they are broadly divided into two categories, based on how fast they evaporate.
The first is non-persistent weapons, which are designed to cause immediate casualties but not linger about on the battlefield. The casualties may be lethal or not, the objective is to take someone out of the battle and render them unable to fight. Nerve agents can be incapacitating at much lower doses (and, by extension, greater distances downwind) than at levels where they are lethal. Sarin is an excellent examples of a non-persistent chemical warfare agent. It lasts only minutes or only a few hours, unless in very cold weather because it quickly evaporates into a vapour state from liquid.
The second is persistent weapons, such as VX and the various so-called “Novichoks”, which are designed for use in interdiction and area denial. They tend to be liquids that can take weeks or months to degrade or evaporate. While some of them might cause some immediate casualties, their intended use is to be a major nuisance to military operations. They can make land, especially key terrain like bridges, roads, ports, and runways contaminated, so that troop passing through them have to operate in cumbersome protective clothing and vehicles, including tanks.
Actually killing people is a secondary consideration with persistent weapons. Indeed, the first major persistent agent developed, Mustard, is not very lethal at all. Nor is it fast acting, taking many hours before signs and symptoms emerge. But if you want to make a car park full of NATO tanks filthy so that they must be laboriously decontaminated before use, force an invading mechanized infantry regiment to go left instead of right, or make the enemy wear clothing that causes soldiers to pass out from heat injuries, then persistent weapons are a consideration.
Therefore, the idea that someone in the vicinity of a chemical weapon or exposed to a chemical warfare agent is going to die right away is an assumption that may or may not be supported by the actual data. However, given modern concepts of manoeuvre warfare, making a lot of terrain dirty constrains both sides in a conflict, so one can see how modern military doctrine would only find limited use for this kind of weapon.
An important aspect that is often not realized by the online commentariat is that persistent chemical warfare agents are liquids. They evaporate very slowly, or else they would not be persistent contaminants. They aren’t gases. Without a means to make them into aerosols, they are often little or no vapour hazards. This is why one does not need a full hazmat suit to handle them and why, in many circumstances, one needs only careful technique and some gloves, rather than a gas mask, to handle them. A simple pair of gloves is what you need to handle small quantities of VX or a persistent Novichok, not some gas tight suit and air tanks. Of course, UK specialists were not sure what they were dealing with initially, so did use full hazmat suits initially.
The real world is not like the laboratory
Another important factor relevant to chemical warfare is the large gap between behaviour and conditions in the laboratory and actual behaviour of chemical warfare agents out in the field. A good example of this is toxicity, the amount of material needed to kill someone or something. Statements along the lines of “enough Novichok to kill 4000 people” or “10 times deadlier than VX” are inherently misleading because they lack essential context. Toxicity is general expressed as a “LD50” – a dose that kills 50% of the exposed population and it is measured in terms of mass, and sometimes mass per body weight. For example, the although references vary, one source gives the LD50 of Sarin in an average adult male soldier as approximately 25 to 50 mg per kilogram of body weight. However, this is Sarin already absorbed into the human body, and takes no regard for how it got there. Was it breathed? Absorbed through the skin? Into the eyes? Eaten? Injected? These all have widely different efficiencies. Nor does it take into account variance in body weight. Is it 25 mg per kg for a svelte 40 kg person, in which case an LD50 would be about a gram. Or is it 50 mg per kg for a 100 kg person, in which case it is five times higher, 5 grams. Also, human LD50 is calculated with a wide margin error because it is based on extrapolations of data from animal studies. Dogs, cats, guinea pigs, goats, pigs, monkeys, and other species have been studied. But an actual human study would be extremely unethical and has never been done with Sarin. It would involve administering Sarin to a large population and seeing the point at which half the exposed population died. Not even the Nazis did this.
A statement like “X is enough Sarin to kill 4,000 people” is based on LD50 figures, but conveniently ignore that the only way it would be enough to kill 4,000 people would be to line people up and inject them. It doesn’t reasonably account for absorbing through skin or inhalation, which are the normal routes of entry. If you wanted to kill 4,000 people through inhalation, you would need a lot more. And even more if you wanted to kill by absorption through skin. When it actually comes to launching shells, dropping bombs, or painting door knobs, there is a wide gulf between laboratory statistics and actual practical field results.
There is a high degree of inefficiency in actually disseminating chemical warfare agents. Let’s look at a typical weapon system – an artillery shell or bomb filled with Sarin, fused for detonation on contact with the ground. (Sarin is almost always done as a surface bursting weapon, whereas persistent agents tend to be done as airburst weapons, detonating at some height above ground.) An artillery shell has a casing that is thick enough to safely store the contents and keep them inside until the shell hits its target. A charge of high explosives is used to burst the shell. It needs to be big enough to crack open the shell, but not so big that it destroys the Sarin or divides it too finely. The ideal would be to have the highest possible percentage of the Sarin dispersed as fine droplets in an aerosol. But in reality, a lot of chemical agent is wasted or dispersed in a way that it will not create casualties. Some of it sticks to the fragments of the shell, some of it is destroyed by the explosion, some of it may be in globules too big to be aerosolized (deadly if one were to hit you, but not good for a wide area vapour hazard), and some will stay in liquid form in the shell crater. Likewise, if you are using a persistent agent, you want to create larger droplets over a reasonable target area. It is far better to make a lot of tanks a bit dirty rather than making one filthy and leaving the others clean.
In practical terms, however, a huge number of conditions in the field environment make chemical weapon systems highly unpredictable. Temperature (both of air and of any surfaces a drop of agent may alight upon), terrain, wind, humidity, moisture on surfaces, precipitation, foliage, and the presence of temperature inversions in the air are only some of the factors. For a fuller explanation, look at how an old US Army manual describes the myriad ways in which chemical warfare agents interact with the environment. The vast majority of factors serve in ways that decrease rather than increase the efficiency of chemical weapons. Indeed, one can see how field commanders, who might not be able to wait for the optimum conditions to come along, would tend to prefer more highly predictable conventional means of warfare.
What does all of this mean? It means that even with unusual unconventional uses of chemical warfare agents, such as deliberately targeting non-combatants in Syria, or use in assassinations in Wiltshire or Malaysia, actual results in the real world may not match the expectations that their theoretical specifications may imply.
The fact is that if you touch a door handle coated with Novichok with your bare hand, you would definitely become very ill and would die without proper treatment. But quick treatment ensured the Skripals survived (although we may never know the long time health consequences of the attack). Poor Dawn Sturgess sprayed the substance on her bare skin, didn’t get treatment in time, and died.