Schneier on Security
A blog covering security and security technology.
« Solving the Underlying Economic Problem of Internet Piracy |
| Securing iPads for Exams »
February 9, 2012
Security Implications of "Lower-Risk Aircraft"
Interesting paper: Paul J. Freitas (2012), "Passenger aviation security, risk management, and simple physics," Journal of Transportation Security.
Abstract: Since the September 11, 2001 suicide hijacking attacks on the United States, preventing similar attacks from recurring has been perhaps the most important goal of aviation security. In addition to other measures, the US government has increased passenger screening requirements to unprecedented levels. This has raised a number of concerns regarding passenger safety from radiation risks associated with airport body scanners, psychological trauma associated with pat-down searches, and general cost/benefit analysis concerns regarding security measures. Screening changes, however, may not be the best way to address the safety and security issues exposed by the September 11 attacks. Here we use simple physics concepts (kinetic energy and chemical potential energy) to evaluate the relative risks from crash damage for various aircraft types. A worst-case jumbo jet crash can result in an energy release comparable to that of a small nuclear weapon, but other aircraft types are considerably less dangerous. Understanding these risks suggests that aircraft with lower fuel capacities, speeds, and weights pose substantially reduced risk over other aircraft types. Lower-risk aircraft may not warrant invasive screening as they pose less risk than other risks commonly accepted in American society, like tanker truck accidents. Allowing passengers to avoid invasive screening for lower-risk aircraft would introduce competition into passenger aviation that might lead to better overall improvements in security and general safety than passenger screening alone is capable of achieving.
The full paper is behind a paywall, but here is a preprint.
Posted on February 9, 2012 at 6:10 AM
• 43 Comments
To receive these entries once a month by e-mail, sign up for the Crypto-Gram Newsletter.
Isn't the real question the perception of the hijackers/population and not the energy payout structure?
If the hijackers find all of the airplains equally appealing smaller aircrafts don't pose a smaller risk because they both have the same psychological impact on the society when they crash in the city...
I've always wondered why they don't just completely seal off the cockpit from the cabin. Put a restroom up there, remove the door to the cabin, and add an external door so they can come and go at the airport.
Granted, it doesn't deal with people trying to bring down the plane, but that is a separate risk.
You'd also have to redesign airports and have a lower-security terminal for smaller planes and a higher-security terminal for bigger ones.
I don't understand why authorities are concentrating only on improving air travel security when terrorists can detonate a shrapnel bomb in a shopping center by simply waking in and throwing it up in the air a few meters, above the crowd. Compared to airports, it's much cheaper and the likelihood of detection is way smaller. I think open communities, like the countries of the western world, cannot effectively defend themselves against terrorism. On the other hand, even when living under a totalitarian regime during Nazi occupation, we found ways to blow up their supply trains. It's all just for show.
Matic: And for that matter, why not target the huge crowd of people in the airport waiting to go through security? One bomb there would be devastating, and almost definitionally unpreventable.
The relative risk of a hijacked plane is only partly determined by the physical and chemical energy potential of that plane. A significant part of the calculation would involve the potential deadliness of the target. Probably a worse case situation would be a single passenger plane (say, a Cessna) deliberately crashed into a highly toxic (to life) storage tank. The oil and chemical industry has many such storage tanks. This risk is not new. The point is that smaller, less energetic planes are not necessarily less risky than a jumbo jet.
@Matic - they did use to bomb shopping centres. although oddly the people who did it aren't regarded as terrorists by many in the US
@David - but if you wanted to crash a Cessna into a chemical plant you wouldn't need to hijack a commercial flight - you simply rent/steal a Cessna from a small airfield and fly it yourself.
This would remove the ridiculous case of subjecting the pilot of a single seat aircraft to X-ray checks in case he is carrying a weapon which he could use to hijack himself!
Rather than just dismiss "Tanker Truck Incidents", consider why it is that society has chosen to accept the regular death of vulnerable road users to death-by-truck and death-by-SUV-with-DUI-driver.
We accept that when other forms of death that have far less impact on the probability of reaching adulthood (e.g. firearms).
Aircraft safety appears the worst of both worlds: lots of money spent, no real benefit. Pedestrian safety: little money spent, hence no benefit at all.
That's a mighty long paper, just to tabulate a few high school physics functions. The paper's adjacent political/market editorializing could have gone into a blog post instead.
Is the conclusion not that airport security policies in their current form are not effective to reduce aircraft terrorism?
Given the amount of money expended on these policies, the next question then is: What are airport security policies intended to achieve if it is not a reduction in aircraft terrorism?
I know that we should not "Ascribe to malice what can be explained by Incompetence".
On the flip side smaller aircraft require more landing slots, more pilots, more check in staff and are probably less fuel efficient with it too.
If the economics of going small were compelling then the airlines would be flying small planes on all their routes.
Therefore I don't think there would be much if any competition unless a rival carrier were to discover that passengers would pay a substantial premium to avoid the grief of extra screening. Even if they did, how would passengers avoid screening that other passengers were subjected to?
Of course if one were to play security theatre, I wonder what the impact of a small aircraft hitting an oil refinery, a stadium or some other economically important or prominent structure would be.
For those not living in the few dozen hub cities, commercial aviation already is smaller aircraft with lighter fuel loads. And those are the same airports where the cost of full-up security is so disproportional to passenger traffic. Meanwhile, "regional" flights are already conducted from separate terminals at many major airports, so we're talking about a "simple" matter of rearranging the security perimeter.
This would save billions of dollars in decreased procurement and staffing, not to mention the delay costs. But of course that in itself is a reason it's unlikely to happen.
@wintermute: Shrapnel bombs don't get set off in airport screening lines because anyone trying to bring down (pro-)Western governments know the security measures to be expensive theatre. If earlier terrorist proclamations are to be believed, one of the long-term goals of the 9/11 terrorists was to drive the U.S. into bankruptcy; drawing such intense scrutiny to massive inefficiencies like airport screening would be inconsistent with that goal.
@Adam - they are and the resulting congestion is a big source of complaints from airports,ATC and pilots.
Routes that used to be a major airline 737 are now being flown on Bombardier/Embraer turboprop by small cheaper contract airlines.
Another concern is what the long term safety record will be for a more congested airspace full of small planes flown by overworked and inexperienced pilots operated by companies on very tight margins.
@David: it still has to have enough mass to actually breach the container. I suspect no matter how hard you threw it an empty beer can would not disable a car. A full one launched by a rocket probably could.
Those of us in the aviation community (ie pilots) have been saying since 9/12 (unfortunately to deaf ears in the government) that tiny airplanes are not much of a risk; quite possibly less risky than an SUV because the energy (chemical or kinetic) just isnt there.
That copycat kid who crashed a cessna into a building in FL right after 9/11 didnt even knock the lights out in the office he hit. The picture looked pathetic, a little crumpled plane carcass stuck to the outside of a basically pristine office building like a bug on a truck windshield. Same thing with that guy in Italy the next year, although he at least managed to start a fire.
Back in WWII a B-25 (which is waaaay bigger and faster than a Cessna/Piper/Beech) crashed into the Empire State building. It made a decent hole and started a fire that was out the same day.
Its a logarithmic damage curve with speed across the bottom and weight up the side and light planes (generally considered .lt. 12,500# GTW) are lost in the noise if a 757 or bigger is on the chart at all.
Picture a graph where M1 Garand rifles, M4 Sherman tanks and Tsar Bomba class nuclear weapons are all on the same chart. The tank and the rifle would look like the same dot even though they are vastly different from each other. Light airplanes are the same way. So by extension a small airliner (picturing a 20-30 seat RJ) should need less "security".
I'm reminded that the WTC design anticipated being struck by an aircraft, but the engineers were thinking about a 707 getting lost on final approach - less mass, less velocity, less fuel.
As Tom Clancy once imagined, a 747-400 with fuel for Tokyo could be used to implement regime change.
@SteveL: "Pedestrian safety: little money spent, hence no benefit at all."
We spend money on sidewalks and cross walks and street lights. I actually expect it has a much larger benefit than the TSA.
@bob however a Cessna with 500lb of explosives crashed into say the Superbowl would make something of a splash.
That's (allegedly) the reason we are now restricted on flying model aircraft - apparently a model plane with 2oz of payload is a potential cruise missile!
What a silly premise! 9/11 style attacks are practically unrepeatable: what made them possible, is that the crew and passengers were not aware of the actual danger. Even if aircraft-as-weapon attacks were eliminated as a risk, the paranoia-ocracy would still insist on grotesque screening procedures, lest someone threaten some nearby passengers or the flight itself.
If there were a lot more flights of much smaller aircraft, there would be almost certainly be more accidents, and possibly greater total loss of life due to accidents.
@mcb: The structural engineers of the twin towers didn't consider fire subsequent to aircraft impact, only the stress loads. Even though the attacking planes had much greater energy than was contemplated in the twin towers design, the structures performed well in response to the impact damage. I believe that if one could magically eliminate the fire and its consequences, the structural damage to the towers would likely have been feasible to repair.
"an energy release comparable to that of a small nuclear weapon"
Really? Really? I would love to see the physics justification for that.
Simple. Foreign terrorists don't vote in the USA. Truckers and TSA employees do.
NobodySpecial: Black Sunday. Bruce Dern. Dirigible. Flechetes. 40 year old movie ...
"an energy release comparable to that of a small nuclear weapon"
747 MTOW is 350ton, at mach1 that's 20GJ of kinetic energy, 1 kiloton of TNT is 4000GJ
Even if the 747 was packed full of TNT that's only 0.2 Kilotons
@Joe: Regarding "A worst-case jumbo jet crash can result in an energy release comparable to that of a small nuclear weapon":
gives 590T for an A380, and 440T for a B747. Using 500T as a nice average, we can calculate the kinetic energy embodied in one of those planes going 500mph or about 200m/s:
Ek = m*v^2/2 = 500T * (1000 kg/T) * (200m/s)^2 / 2 = 20e9 kg*m^2 / s^2 = 20GJ
Jet-A has an energy density of 43MJ/kg or 35MJ/L; the A380 can take 320 kL of fuel, while the B747 can take 240 kL. Using a lower-end average of 250 kL, we have:
Ef = 250kL * (1000 L/kL) * 35MJ/L = 8.7e6 MJ = 8.7 TJ
(Granted, that's for complete burn-up, but even a fraction of that completely overwhelms the kinetic energy component. This agrees with the comments upthread about the Twin Towers surviving the initial impact but being undone by the ensuing fire, which included the jet fuel as well as combustibles already in the building.)
For comparison, one ton of TNT is considered to release 4.2GJ, so a kiloton of TNT would release 4.2TJ. That's certainly comparable with the 8TJ that could be liberated from the fuel load. And while 1-2kT is on the very low end of yield for nuclear weapons, the USA certainly did build and deploy such beasts (e.g., Davey Crockett / SADM variable yield 1-10kT).
Reading all the interesting comments here, echoing so much of what we've all said about the TSA in the past, I think perhaps we all have been successfully misdirected. The TSA is not about preventing terrorists from getting on planes. It's about controlling the movements of the civilian population. When the large employers loot the pension plans and then convert them to defined contribution plans, then Wall Street loots the 401K's, when jobs are shipped overseas, unions broken and house foreclosed... a population could get restless, could try to organize.
TSA is about controlling Americans. All of Homeland Security is. It's right there in the name.
Remarkably sensible. In New Zealand, there's no security screening if you're on a turboprop flight, and minimal (no liquid restrictions, but metal detector/xray) screening for domestic jet flights.
Sure, the threat level in NZ is arguably lower than in the US, but it's good to see some granularization of US policy.
I did all the calculations for energy contained in the fuel of an A380, then I discovered it had been done for me (and agreed with my figure) on Wikipedia.
The maximum energy in the fuel of an A380 is 11 x 1012 J. The Fat Man nuclear bomb at 21 kiloton yield is 88 x 1012 J. So it's within an order of magnitude. However, it also ignores the difference between energy released in less than a second and energy released over a longer period.
The max cargo capacity is about 150,000 kg. If that were loaded with pure dynamite, it'd add about 1 x 1012 J. Jet fuel actually has about 6 times greater energy density than dynamite. Jet fuel has just about the highest energy density of any substance that could be realistically used. Liquid hydrogen is about 16 times greater.
On the other hand if a small turboprop was filled with anti-matter AND we assume that all TSA intelligence comes from Dan Brown novels anyway.....
With regards the small nuke -v- full load mega-jumbo...
First of apples are not oranges or even close.
Most nukes even large one's have nothing close to the mass of a full load mega-jumbo, and small nukes are fairly easily man transportable so less than the weight of an adult male...
This effects the way they would be used as weapons because simple energy release is very very inefficient at causing damage.
If you look back to WWII "earthquake bombs" although not having more explosive mass than other bombs did tend to do considerably more damage to man made objects designed to be "bomb proof" such as the V1 assembly plant in France and various U-Boat pens.
Part of this was the fact they were designed to not just "impact" the target but to actually "dig into it" and that it had charecteristics not to disimilar from a shaped charge. So the method of deployment is actually of more importance than up and up explosive mass. We see this in modern warfare, weapons are getting a lot lighter in mass especialy in the warhead, to significantly rreduce collateral damage, and to make up for this the deployment system has become very very times more accurate and the energy release more highly focused. So if you look at Afghanistan for instance the level of "wanted destruction" is about the equivalent of that done by all the bombs dropped in WWII, however the actual tonnage of explosives is but a very small fraction, likewise the "unwanted damage".
So I would argue that for a specific fixed man made target, a mega-jet crashed into the target makes a better delivery system than an older style conventional delivery system for a small nuke such as the Davy Crocket.
Thus having delivered the payload we have to look at the effect. An air burst explosion using conventional explosives is usuall a compleate waste of energy as it achieves very little as the energy density drops rapidly from the point of the initial explosion (volumetricaly for those with a burn rate below the speed of sound, as the surface of an expanding sphere for those that burn above the speed of sound).
Thus In a conventional "iron bomb" what does the damage is the energy transfered into projectiles formed from the bomb casing, like bullets these tend to re-release their energy on impact some considerable distance from the original point of explosion. Bombs with low or no case density are usualy deployed against areas where the near blast wave would pick up objects from the environment around the point of explosion. Fuel-Air or FAE / FAX explosives work by using a compression wave around an object inside the fuel burn volume, and like conventional high explosives to objects outside the burn volume. In incendiary devices the idea is to release the energy in a slow controled way so as to encorage as many objects as possible to reach a temprature where combustion is self sustaining, depending on the target the bomb might spray out a burning liquid that has high adhesion properties (any one remember Vietnam footage?). Then there are the "thermabaric" and "bunker busting" or "tunnel busting" bombs, these tend to uses two or more explosives, the first to create a very high impulse preasure wave to rapidly disperse the second explosive which is often actually a fuel only such as flare material (think powdered metals or plastics such as PTFE) this the burns like an FAE creating a masive compression wave, an intense heat and also removing the oxygen from the air, making the chance of survival fairly minimal in a bunker or tunnel.
This emphasises a critical difference between conventional explosives and FAE's, something like 80% of the mass of some conventional explosives is actually not the "fuel" but the "oxidizer". An FAE uses instead atmospheric oxygen, thus giving considarably more "bang per pound" or energy release.
Air burst FAE's are known to create wide area explosions where the compression wave is certainly easily comparable with nuclear devices used in air burst mode.
So it is also fairly essential to know not just how much energy is involved but also, how it is directed, focussed or conveyed to the target, the way the energy released is used to damage the target and thus where the best pllace to place the bomb for any given target.
As with cryptography the name of the game these days is not "brute force", but "know your target's weaknesses" and apply appropriate leverage at the weak spots.
Thus as little as a pound of conventional explosive placed inside a critical control area or a gallon of petro could quite easily bring a city wide or even close to a nationwide power blackout subsiquently followed by lack of other utilities such as gas, water, telecomunications and transportation. And the likely hood of this happening is going up not down due to freemarket preasures reducing spending not just on hardening infrastructure, but actually maintaining and repairing it...
The proposed solution is even worse than the current 'solution'. It would cost a lot more in fuel and labour costs, would be less comfortable, most major airports are already having troubles with numbers of flights, and air traffic control would become harder.
Um, I know, fly in smaller airplanes that won't do as much damage if the bad guys crash 'em.
Great idea! Now write that up into a 10,000 word academic paper and grab yourself a professorship on the way out....
And while 1-2kT is on the very low end of yield for nuclear weapons, the USA certainly did build and deploy such beasts (e.g., Davey Crockett / SADM variable yield 1-10kT).
Yes, I get that. However, the average smallish nuclear weapon in the US arsenal today is in the range of 5 MT, with larger ones around 8-10 MT. So that is three orders of magnitude larger than the plane.
I still call foul, it's very difficult to justify the comparison between planes and nukes.
And while 1-2kT is on the very low end of yield for nuclear weapons, the USA certainly did build and deploy such beasts (e.g., Davey Crockett / SADM variable yield 1-10kT).
Let's put it another way for conversation. If someone managed to pack a 747 full of TNT, and managed to efficiently detonate the entire load, after the entire kinetic force of the load has already been stopped and absorbed, and we compare that with the smallest nuclear weapon ever developed by the US, tuned down to the lowest yield possible to have it still work, they would barely overlap.
Meanwhile, the largest weapon ever built by the USSR was detonated at 50,000 times the energy output, or four orders of magnitude of the plane, but had been tuned down from it's design output of 100,000 times the energy output, or five orders of magnitude larger than the TNT-Plane idea.
I also was skeptical of the "comparable to" claim. When I computed maximum kinetic energy, I got something like 1-2 kT. (I wasn't considering the potential energy of the fuel, which I supposed not to be a consideration of the plane-as-missile attackers -- thanks to the commenter who ran the numbers on that!)
The smallest nuclear explosion I was able to find was a North Korean test, estimated at 0.55 kT. This may well be the smallest that has ever been achieved.
Joe, I think your info about the US arsenal is out of date. As far as I am aware, the current air-dropped bombs and long-range ballistic missile warheads range from 0.1 (on some SLBMS) to 1.2 MT, with the Minuteman ICBM warheads at 0.17 MT. Some of the smaller warheads (e.g., for cruise missiles) can be dialed down to 0.005 MT.
I don't think nuclear weapons are practically usable at all, but to the extent that you consider them as tools to attack an enemy, the only value of giant yields is to compensate for lousy targeting accuracy. US delivery systems are expected to by quite accurate, and the yields are sized accordingly.
Of course, jet fuel isn't a high explosive. So while it may have "comparable" energy, that energy is spread over a much longer time frame. Unless you can aerosol all the fuel and get it all to combust at once.
"...psychological trauma associated with pat-down searches...."
One can't believe we are talking about descendants of those people, who conquered the wild nature of America and transformed it to what it is today.
"A worst-case jumbo jet crash can result in an energy release comparable to that of a small nuclear weapon..."
This sounds like a propaganda garbage for redneck, something in the line of the recent "Iran is preparing an attack on USA, we must act before they destroy us". Who ever made a calculation? How small would the nuclear weapon be? They are not exactly pocket size. A minimal size is required so that the chain reaction occurs.
I thought it was actually the burning office material (paper, etc) that did the real damage; that the jet fuel would have burned up fairly quickly, but those materials kept going long enough to do the structural damage that resulted in collapse.
If I could fly with just pre-911 security screening, plus the current reinforced cockpit doors and passengers wiling to revolt against any attackers, I would sign up now.
If only that option was available. :-(
I thought it was actually the burning office materials (paper, etc) that did the real damage
Eventually, yes, but there is a problem with that, and it is to do with energy release over time.
You say that,
the jet fuel would have burned up fairly quickly, but those materials kept going long enough to do the structural damage that resulted in collapse.
But you have forgoton just how difficult it would have been to ignite the office materials.
If the energy of the combusting fuel had been released very quickly like a conventional high explosive, then the office materials would have been "disrupted" not "ignited" by the blast. If however the fuel had burnt very slowly then the metal and other items with good heat conduction would have conducted the thermal energy away and out the building before they got to levels where structural damage would have occured.
That is buildings don't collapse and paper does not catch fire because you have the heating on high. Nither does paper catch fire if you use conventional explosives as the two large IRA bombs in London showed.
Paper in solid form like a block of A4 paper in a stationary cupboard is extreamly difficult to set alight. As a little experiment take such a block of paper into your garden and put it on the barbecue and try setting it on fire with a paint striping blow torch. It chars a bit but does not catch fire. You need a long sustained temprature above it's dry ignition point for it to start to burn, then it tends to burn more like charcoal brickets because it does not have an internal oxidizer and is a very poor conductor of heat.
Because of it's poor burn qualities and good thermal inssulation properties "fine wood pulp" (which is what paper is) is often used as part of fire retardation materials to protect structural steel members in buildings. In general it is far better than concrete for this purpose as it is considerably lighter for the same fire based insulation qualities and unlike concrete it tends not to have corrosive chemicals in it which attack the steel.
A point to think about for those discussing the energy output of nukes -v- a mega-jumbo with a full load of fuel.
As I said earlier you have to understand how the weapon is deployed and how the energy is released against a given target before you can make any meaningful comparisons.
When a nuke explodes by far the majority of it's energy will not be in a usable form for the usual military purpose of either disrupting or transfering kinetic energy to things. It uses the effect of it's mainly ionising radiation on the surounding atmosphere to transport the energy down the electromagnetic spectrum. Thus a very high percentage of it's energy is effectivly lost as it radiates through the atmosphere as light and radio waves etc. And in most cases a considerable percentage of the usefull energy that imparts kinetic or thermal energy onto the target is lost upwards and outwards in the usual "air burst" detonation.
The release of kinetic energy on impact and burning fuel however tends to keep most of it's energy localized and directed at the target for the purpose of destruction. Further unlike a nuke that just produces energy, fuel by the nature of it's combustion process produces expanding gases for the energy to work on.
The difference in how the energy behaves can be seen on a much lesser scale. As those who have played with it know thermite produces a lot of thermal energy, but does not produce expanding gases so no blast wave etc unlike black powder or gun cotton or other more modern explosives used in mining, demolition or military activities. So whilst it might make a good incendiary it has little or no kinetic effects. The energy stays local and thus will melt through most metals, rather than just bending, deforming or disrupting them as a conventional explosive would do.
However again as those who have played with it know conventional high explosives don't on their own do much. That is if you put a block of plastic or stick of dynamite explosive on top of a rock and stood a reasonable distance away and detonated it you would get a loud bang, some smoke and a traveling blast wave that disipated rather rapidly. However drill a hole the right distance into a very large rock or rock face and explode the same size of explosive at the bottom of the back filled hole and you have to be standing one heck of a sight further to avoid the effects of the expanding gases have on disrupting the rock into fragments with a high kinetic energy.
So as I said it's "the how" of the use of "the type" of available energy against the "particular target", not the plain brut force quantity of the energy that counts.
"However, the average smallish nuclear weapon in the US arsenal today is in the range of 5 MT, with larger ones around 8-10 MT. So that is three orders of magnitude larger than the plane."
Not quite true. Modern US warheads are believed to be much smaller yielding than the behemoths of the 1950s. The W88 warhead on the Trident SLBM, for instance, is reported to be about 475 KT. Still a lot bigger than the kinetic energy of an airliner, though as other commenters have noted the physical effects are very different.
In any case, the main point of the article - that small turboprop airliners represent a relatively small risk in terms of being used as kamikazes - is true but not that significant. Small RJs and airliners use far more fuel per passenger-mile than a comparably modern airliner, and airports in major cities are constrained by how many planes they can land. As such, even if security was relaxed for small aircraft, flights on busy routes would still be far cheaper on larger airliners.
@ kinetic energy,
... also why not allow them to deploy some kind of gas that can knock everyone out in the cabin, it is safe of course but u cant smell or taste it and would make the terrorists unconcious
This sort of knockout gas is the "holy grail" of surgery, and obviously currently does not exist.
The problem with gases that render you unconcious is that not only do they render you "lifeless" in the desired way, they also depress "respiratory function". The result is the faster it knocks you out the faster it kills you as you are less capable of breathing without assistance.
Another problem is the gas density that is it is unlikely to be the same density as air, and thus will either rise or sink, in either case it will not be uniformly distributed in a given volume. Also some of these gases are just "hydro carbons" some of which are basicaly fuels and thus the volume may well become a "Fuel Air Explosive" just waiting for an ignition spark...
As far as I'm aware most gases currently in use have a very nonlinear response in humans and thus have a critical level which is not just patient specific but also posture specific as well. Thus the level required to render a physicaly fit terrorist unconcious could well be killing children and grandparents due not just to their sensitivity but also they are likley to be huddled close to the floor and likely to be getting several times the dose of an upright and mobile terrorist.
As far as I'm aware the only use of such gases against terorists was by the Russians in a theater. The result was it killed many of the hostages.
Now that it is known by terrorists that such gases might be used, research on the internet will tell them the most likely candidates and how to deal with them.
In the case of "suicide terrorists" a simple "dead mans switch" would be a suitable defense as when they pass out they fall over and boom the bomb goes off and potentialy the fuel air explosive of the gas...
Not exactly pretty, or politicaly desirable...
Schneier.com is a personal website. Opinions expressed are not necessarily those of Co3 Systems, Inc.