Nuclear weapons represent an existential risk. I’ll let 80,000 Hours speak for me for a minute:

A survey of academics at the Global Catastrophic Risk Conference by Oxford University estimated a 1% chance of human extinction from nuclear wars over the 21st Century.
…
Luke Oman estimates the probability “for the global human population of zero resulting from the 150 Tg of black carbon scenario in our 2007 paper [delving into the effects of a single nuclear exchange] would be in the range of 1 in 10,000 to 1 in 100,000.” This being said, we think this estimate is too low, as it doesn’t account for the potential for weaknesses in their model or the risk of a societal collapse causing a permanent reduction in humanity’s ability to reach its potential (which is nonetheless an existential risk even if people remain).

If you’re interested in reducing the existential risk that nuclear weapons pose, I’ve identified a few areas where you may be able to make a difference.

8.1 Tactical Weapons

Countries have begun to reduce stockpiles of tactical weapons and put those that remain under better centralized control. No one ever wanted a fresh lieutenant in charge of the nuclear weapon that could eventually set off World War III – it just took everyone a while to realize this.

Still, it seems like this has primarily been possible because of the collapse of the Soviet Union. When the USSR seemed poised to overrun Europe, killing the commies was given priority over keeping humanity alive. Increasing regional tensions between Russia and NATO may result in a resurgence of tactical weapons.

Treaties that ban weapons under a certain yield, or require all nuclear warheads to have locks that can only be released by the civilian leadership of a country would be an excellent way to reduce the risk of conventional warfare leading to a nuclear exchange.

8.2 Arms Reduction Treaties

Not all arms reduction treaties are created equal. The Strategic Offensive Reductions Treaty (SORT) expired on the same day it came into full effect and set non-specific limits; while it may have reduced the total number of nuclear weapons deployed, it probably did this by causing the early retirement of already obsolescent systems. In addition, SORT had no verification provisions. We literally have no way of knowing if it actually had an effect.

On the other hand, the New Strategic Arms Reduction Treaty (New START) has a robust verification mechanism, including demonstrations that technology has been fully decommissioned and eighteen inspector visits each year. New START sets specific limits on ICBMs, SLBMs, nuclear armed bombers, and total deployed warheads. It will be in full force for at least three years, but might be extended longer. It comes up for review in 2019, so convincing the US and Russia to renew it will be very important.

8.3 Anti-ballistic missiles

The US ABM system represents a real threat to global peace. If it is demonstrated to be effective, we could see China rapidly increase its nuclear arsenal. If it’s expanded to the East Coast of the US, or Europe, we could see Russia do the same.

If you live in America, pressuring your congressional representative or senator to vote against any measures increasing funding for the ABM system could be very important. You can call it a waste of taxpayer money, demand it not be built in your backyard, etc.

If you live near one of the current ABM sites, or are near one of the sites for potential expansion, you can engage in direct action.

In addition to organizing protests (it should be easy to get people uneasy about nuclear weapons near them), you can attempt to bog down any expansion or new construction in and endless morass of red tape. If a system is being built near you, you should attend any community meeting you can, be as obstinate as possible, and jump on any zoning violation, rushed environmental assessment, or other bureaucratic mistake like a rabid pit bull. This won’t be very effective if new ABM sites are built entirely within existing military bases, but if even a single support strut has to go up for municipal approval, there’s potential to make an impact.

Current ABM sites are Fort Greely in Alaska and Vandenberg Air Force Base, California. Proposed eastern sites are SERE Remote Training Site in Maine, Fort Drum in New York, Camp Ravenna Joint Military Training Center in Ohio, and Fort Custer Training Center in Michigan.

8.4 Donations

Both 80,000 Hours and OpenPhil have done their own preliminary assessment of nuclear weapon risk.

Neither piece offers firm suggestions for the most effective charity and I lack the expertise to do my own evaluation. Both OpenPhil and 80,000 Hours suggest that there may not be much room for more funding, although OpenPhil suggests that effective anti-nuclear advocacy may be underfunded.

For what it’s worth, I’m donating to the Ploughshares Fund. They seem to have the correct focuses, from preserving the Iran deal, to removing tactical weapons from Europe, to opposing new ABM systems. I don’t think they have that much more room for funding, so I’d welcome a more thorough effectiveness evaluation that would allow people concerned with nuclear risk to confidently donate their money.

Having covered the practicalities of nuclear physics, nuclear weapon design, and nuclear weapon effects, we may now turn our attention to the strategies that have grown out of these physical realities.

7.1 Tactical and Strategic Weapons

Broadly speaking, there are two kinds of nuclear weapons – tactical and strategic. This post has been focused primarily on strategic nuclear weapons, high yield weapons capable of destroying cities and hardened targets. Tactical nuclear weapons have smaller yields, allowing them to be hypothetically used on a battlefield that contains friendlies.

The line between the two gets somewhat blurred with the highest yield tactical weapons. Is a 5kt bomb tactical or strategic? No one really has a clear answer. These already crystal clear waters get muddied further when you add in “dial-a-yield” weapons, which can yield anywhere from <1kt to ~100kt. On the low end, they’re definitely tactical. But at the high end, they’re clearly strategic.

Most of the treaties that deal with nuclear weapons don’t cover tactical weapons. This is a bit of a problem, because tactical weapons are perhaps the easiest way that a conventional conflict could escalate into a nuclear conflict. It goes like this: my army is losing, so I have a fire-team use a tactical nuke launched from a recoilless rifle on your densest concertation of tanks. The 100t weapon totally destroys the formation, swinging the battle in my favour.

The nuke I used on you, the opposing general, is not a strategic weapon, so I don’t need codes from higher up or another person to agree with my decision.

You are now losing because those tanks occupied a key position. You reply with a 1kt tactical nuke of your own, fired from an 8″ howitzer 20km behind your own lines. It takes out 3,000 of my infantry. Satisfied, you go back to conventional war.

But I’m pissed off, so I dial up one of my short ranged tactical missiles to a 10kt yield. There’s a plume of smoke from the rocket launch, a bright flash, and a crater where your army used to be. I’m satisfied with a job well done, but your side is pretty enraged. So they call in a 50kt strategic nuke on an intermediate range ballistic missile that wipes out my forces and camp.

A few more levels of escalation and the ICBMs and SLBMs begin to fly. Once you believe nuclear war is inevitable, the only thing to do is try for a successful first strike and pray for the best.

7.2 First Strike, Second Strike, Counterforce, Countervalue

A first strike is an attempt to destroy an opponent’s nuclear arsenal before they can launch it at you. SLBMs or on station stealth bombers are the only real way to pull this off. The flight time for ICBMs between probable belligerents is much too long for the missiles to reach targets before those targets have a chance to respond.

Targets include airbases where nuclear bombers are known to be based, the known location of any mobile ICBMs, missile silos, docks where nuclear submarines may be resupplying, command and control systems, and key nuclear weapons decision makers. Separately, any nuclear submarines that have be detected will be attacked by the hunter-killer submarines shadowing them.

In all likely nuclear exchanges between larger nuclear powers (NATO/Russia, NATO/China, China/Russia, China/Pakistan, China/India, India/Pakistan, Pakistan/Israel and neglecting North Korea due to the primitive nature of its nuclear program), this won’t be enough. Some of the nuclear weapons will survive.

Baring a truly unlikely display of self-sacrifice and forgiveness, these remaining weapons will immediately be fired at the aggressor in a second strike – a retaliatory attack.

First strikes are predicted to be mostly counterforce, that is to say, aimed at an enemy’s military forces in general and their nuclear forces in particular. There will be civilian casualties, because there always are, especially with weapons as indiscriminate as nukes, but civilian casualties aren’t the goal of a first strike.

A second strike, on the other hand, would be primarily countervalue, that is to say, aimed at the most valuable targets an enemy possesses. Major cities, knowledge centres, and industrial centres are the primary targets in a countervalue strike. Civilian casualties are kind of the point and consequently will be rather high.

7.3 Mutually Assured Destruction

Nuclear policy has for decades been based on the idea of mutually assured destruction (MAD). Mutually assured destruction requires all nuclear armed powers to be committed to a massive countervalue retaliatory strike against any country which deploys nuclear weapons against them. Furthermore, this threat must be credible – enemy decision makers must believe it is real and can be carried out if necessary.

The goal of the mutually assured destruction doctrine is to prevent nuclear war by making it clear that all nuclear wars will be unwinnable, so that no power believes they can gain anything from them. Wargame succinctly summed up the desired end-state of the MAD doctrine with the famous line “the only winning move is not to play”.

For MAD to ensure stability, two things must be true:

1. No actor can destroy the entire nuclear arsenal of another in a first strike.
2. No actor can defend against a second strike well enough that they will escape unscathed.

While the current MAD equilibrium is relatively stable, it has come under threat from both the US and the USSR in the past. The closest we’ve come to nuclear self-annihilation has been those times when MAD had begun to break down.

The Cuban Missile Crisis was ultimately about first strike capability. The US had based ballistic missiles in Italy and Turkey that gave it the easy ability to target the USSR – and Soviet missiles (of which there were only 20 that could hit the continental United States from Soviet territory). With these forward deployed missiles and timely reconnaissance from the U2 spy plane, it had become perhaps possible for the US to wipe out the Soviet Union and face minimal retaliation.

When the Soviets began to build missile sites in Cuba, the problem became mutual. Suddenly each side had first strike capability. If you want to look for the hand of God in human affairs, I would suggest centring your search here. Conditions were riper for a first shot than a Mos Eisley cantina. It’s a genuine miracle that no one launched missiles.

The negotiated resolution to the Cuban Missile Crisis was publically all about Cuba. Kennedy promised to never invade again and Khrushchev promised to remove the missiles. But in secret, Kennedy promised to remove all of the US missiles based in Italy and Turkey. First strike capability was removed and equilibrium restored.

The second threat to MAD came from anti-ballistic missiles (ABMs). By the 1960s, both the US and USSR were working on ABM systems. The Soviets were building a network of ABMs around their capital Moscow, while the Americans were building a similar system around their missile silos in North Dakota.

These systems were imperfect and could be overwhelmed by MIRVs. But both sides were worried about the future; what if their enemy figured out perfect missile defense before they did? Both the US and USSR knew that If one side could gain a critical advantage in missile defense, they would be able to launch a first strike with impunity. The bitter irony was that systems designed to protect against nuclear attacks were making global nuclear war more likely.

This crisis was also defused through diplomacy. Both sides understood the risks and decided they weren’t worth it. The first Strategic Arms Limitation Treaty (SALT I) limited both the USSR and US to two ABM sites. Hot on the heels of this was the Anti-Ballistic Missile Treaty which revised the limit downwards to one site and limited the number of ABMs at each site to 100.

Unfortunately, the US unilaterally withdrew from this treaty in 2002 to begin work on a new missile defense system. If this system ever becomes operational to the point that the US can expect to survive a second strike from Russia or China, nuclear war will become imminent.

7.4 The Nuclear Triad

The delicate balance of MAD is maintained by the nuclear triad: ICBMs, SLMBs, and conventional bombers. Once a nation possesses all three legs of the triad it becomes almost impossible to remove their nuclear capability in a first strike.

Each leg of the triad has a purpose. ICBMs are static, but are relatively cheap and can often be launched quickly. SLBMs are more expensive and slightly harder to launch (they require accurate positioning for targeting) but are very hard to destroy. And conventional bombers have the greatest flexibility in avoiding attack, plus a very long range courtesy of in-air refueling.

These aren’t the only three methods that can round out a triad though. Really, the important thing is having three credible and disparate systems, such that it is impossible to remove your ability to make a countervalue second strike. Space based weapons (forbidden by treaties), air launched missiles, carrier based nuclear bombers, or nuclear torpedoes could all be used in place of one of the “standard” legs of the triad.

Only India, China, the United States, and Russia have confirmed nuclear triad capability. Israel is suspected of having a nuclear triad, but refuses to confirm or deny this assertion (Israel does wink suggestively when asked, which has led basically all experts to assume that they do in fact possess a nuclear triad).

7.5 Current Nuclear Strategy

Every country charts its own course on nuclear weapons. From public statements and acknowledged procurements, it’s possible to get some idea of each country’s strategy, but you have to understand that they really don’t broadcast these things. I mean, they broadcast them, but we shouldn’t take those statements at face value. There are a host of reasons – diplomatic, strategic, domestic, that prevent leaders from being entirely candid with their nuclear plans.

When reading into strategies, I focus on questions like: what are the known capabilities of deployed weapons? How many nuclear weapons can a country deploy? What delivery methods does a country possess? Where are its weapons based? What advances in technology are politicians highlighting in public speeches? Where have they faced technical difficulties? What countries are they friendly with? Unfriendly with?

I would also recommend avoiding the common pitfall of obsessing over the total number of warheads a country possesses. This number is much less important than the count of operational or deployed warheads. In any significant nuclear exchange, it is unlikely that any warheads but those immediately at hand for deployment will be used.

This section represents my best guesses at the current nuclear strategies of various countries. Please treat these as speculation, not as fact.

7.5.1 Russia

Russia wants to keep its status as a major nuclear power, but it needs to do it on a tighter budget than the USSR had. This means a focus on land based ICBMs, no truly stealthy fighters, and limited resources for its SLBM program.

Current sanctions on Russia have disrupted Russian supply chains and applied pressure on Putin to slash military spending – right as he becomes more confrontational with the rest of the world. Russia can’t afford to fall behind in the nuclear arms race. Its second strike capability is the only thing stopping the US from giving it much harsher ultimatums. Significant budget cuts would put this second strike capability under threat. But on the other hand, Russia can only afford its current military budget for so long.

This statement falsified if: real stealth bombers enter service with a sticker price of >$500,000,000 per unit, Russia manages a string of successful SLBM launches in 2017, the international sanction regime against Russia collapses with Trump as US president.

7.5.2 China

China’s nuclear arsenal is less advanced than that of Russia or the USA. MIRV-ready missiles have been rolled out only in the last decade and many of their missiles still lack MIRV capability (but they do incorporate some decoys and countermeasures). In addition, Chinese missiles are kept unfueled and without warheads in place, which drastically increases their second strike response time. They make up for this with a massive network of decoy silos, real silos, and tunnels built into the mountains of Central China. China only has a handful of nuclear missile submarines and its conventional bomber force is fairly obsolete.

China’s nuclear policy is explicitly second strike only. Based on all the facts above and what I know about China, I’m inclined to believe this. Historically, China has never cared much about what happens outside of China (broadly defined, of course). Since China already enjoys massive conventional supremacy over its neighbours, it has no need of nuclear weapons to intimidate them.

This statement falsified if: China renounces no first use, China threatens a non-nuclear state with nuclear attack, China has >10 active ballistic missile submarines by 2020, China develops a new dedicated heavy bomber.

7.5.3 India and Pakistan

Neither of these countries have sprung for forces really capable of mutually assured destruction and only India maintains a full nuclear triad. Instead of adopting MAD, they instead both aim to have forces just big enough to make a nuclear attack by the other not worth the risk.

Since these countries really only need to be able to deter each other (and possibly China), they’re freed from the need to spend to keep up with the US or Russia. Both India and Pakistan lack the ability to launch a truly significant countervalue strike in response to a first strike from the US or Russia. Given how unlikely the US or Russia launching a first strike on India or Pakistan is, this is a sensible approach.

This statement falsified if: Either of these countries test Mt range thermonuclear weapons, either of these countries develops an ICBM capable of targeting the continental United States (range >11,000km), either of these countries increases nuclear funding by >200%.

7.5.4 UK and France

These countries keep nuclear weapons because they’re members of the UN Security Council and it comes with the territory. Neither has a particularly robust nuclear force (the UK only has American made Trident SLBMs, France has indigenous SLBMs and nuclear capable bombers). It’s largely a prestige thing though. Neither country has been particularly enthusiastic about the nuclear rigmarole (and its price tag) since the end of the Cold War.

Nuclear policy in the UK and France is closely tied to the nuclear policy of NATO, although both countries do maintain some ability to conduct nuclear war on their own terms. Neither country has ruled out using nuclear against non-nuclear states in response to attacks with conventional forces and France has specifically mentioned that they would be willing to use nuclear weapons against countries that sponsor terrorism against them.

All this being said, it is unlikely that France or the UK would be the ones to start a global nuclear war.

This statement falsified if: Either the UK or France increases their supply of operational warheads, either the UK or France develops a full nuclear triad.

7.5.5 North Korea

North Korea’s nuclear weapons program is especially difficult to assess. In addition to the normal challenges when trying to understand a classified program, there’s the bluster of Kim Jong-un to sort through. Up until recently, experts thought ICBM technology was beyond North Korea. They were thought to be struggling with the re-entry heat shield, struggling with miniaturization, struggling with the whole endeavour.

Now, people are less sure. Has North Korea gained full ICBM capability? Or is this more bluster and staged photographs? There’s probably a dozen men and women in the Pentagon (and in many other places) who would love to know the answer for sure. My personal guess, based on the evidence (and the bluster) is that North Korea has a missile design that in theory could target the US, but they’ll need a year or three to get it working reliably. I don’t think they’ll pull off a successful test this year, but I won’t be surprised if they pull one off in 2018 or 2019. That also seems to be the view of the Ploughshares Fund, an anti-proliferation non-profit.

Even if North Korea can’t attack the US, there’s still Russia, China, Japan and South Korea at risk from its shorter range missiles. This represents a considerable threat to a large number of people. It’s tempting to laugh off these threats in light of the inflated numbers that North Korea likes to give for the yields of their weapons (e.g. claiming a 6kt detonation was a hydrogen bomb that could wipe out the whole US). You could look at the disconnect as evidence of a fizzle, but personally I see it as evidence that North Korea likes to exaggerate. In September, they tested a bomb with a yield of 10-20kt. Fizzle or not, if delivered to Seoul, it would kill over 100,000 people.

Under Kim Jong-il, the prevailing belief was that the nuclear program was a bargaining chip in order to get free food or other concessions from other countries. Its purpose under Kim Jong-un is less clear. Despite punishing sanctions, Jong-un has maintained and expanded the nuclear weapon and missile programs started by his father. Whether he is willing to trade them in for concessions or wishes to use them in an attempt at unification is unknown.

This statement falsified if: North Korea successfully tests a true ICBM (range >5,500km) with successfully atmospheric re-entry by the end of 2017 and analysts believe it has enough additional payload for a miniaturized bomb.

7.5.6 Israel

Israel’s nuclear weapons remain unacknowledged, because Israel has pledged to not be the country that “introduces” nuclear weapons to the Middle East. Israel’s statement should be given all of the skepticism one would give to Bill Clinton using the word “is”. Israel may intend “introduce” to mean “acknowledge” or “deploy”, but we’re all pretty sure they don’t intend “introduce” to mean what it literally does.

Israel wants to have nuclear weapons as the ultimate hedge against military aggression by its neighbours. In addition, it wants to ensure that none of its neighbours possess them. Given the clear support for genocide that some of its neighbours have expressed, this position is understandable. If it appears likely that one of its regional foes will develop nuclear weapons, Israel is likely to launch a conventional attack to stop their development. If a conventional attack fails, a nuclear one is not out of the question.

Many sources talk about how Israel holds nuclear weapons as a “Samson option” and is prepared to use them to utterly annihilate an enemy if it looks like they are in a position to destroy Israel. This is actually how I’d expect most countries to behave, so I think the obsession with the Samson Option in Western reports on Israel’s nuclear program has more to do with the story it makes than a real difference between, say, France and Israel.

Israel probably possesses a full nuclear triad. Since it does not confirm or deny its nuclear program, there are no publically available official details that would allow us to be sure of this. It does make sense though. Israel has the technical know-how to pull off the tricky parts of a triad, like SLBMs.

In the future, we can expect Israel to continue to hold onto its arsenal and continue to neither confirm nor deny its existence. I don’t think Israel is a particularly likely candidate to touch off a nuclear war, as it is unlikely to use nuclear weapons against another nuclear armed state. That said, Israeli use of nuclear weapons would almost assuredly result in many civilian casualties and is still an eventuality that basically everyone would like to avoid.

This statement falsified if: Israel joins the NPT and allows inspectors into its nuclear facilities, Israel publically acknowledges its nuclear arsenal, Israel does not launch an airstrike against any nuclear program started by another Middle Eastern country.

7.5.7 Iran

Iran does not possess nuclear weapons, but as recently as 14 years ago was probably working on them. Most analysts believe that this work mostly stopped with the US invasion of Iraq. Iran had no particular desire to become (more of) an international pariah for developing nuclear weapons, but couldn’t accept the risk of Iraq developing them without an Iranian counter. Iran remembered the nerve agent attacks that Iraq unleashed (with the assistance of the US) during the Iran-Iraq war and felt that any developments in Iraqi weapons of mass destruction had to matched.

Iran has been prevented from coming clean about its past development of nuclear weapons by the belief (rightly or wrongly) than any admission will result in sanctions or attempts at regime change from Western actors.

Iran does have a well-developed civilian nuclear program. Despite the wailing and gnashing of teeth among hawks, the current nuclear deal should prevent any breakout towards nuclear weapons. The deal includes a robust enforcement and inspection regime that has most global powers convinced that Iran won’t be able to restart nuclear weapons work secretly.

Donald Trump’s talk of reversing this deal is just that: talk. It isn’t a bilateral deal between Iran and the USA, it’s a multilateral deal between Iran, the Permanent members of the UN Security Council, and Germany. The US could unilaterally re-impose sanctions, but there would be a significant diplomatic cost for basically no gain; it took a network of international sanctions to bring Iran to the negotiating table the first time around, so it is doubtful that US sanctions alone would change anything. The Iranian economy isn’t very integrated with the US economy, further diminishing the sting of any unilateral sanctions. Honestly, the US would probably suffer just as much from a fresh round of sanctions as Iran would.

This statement falsified if: Iran renounces the nuclear deal, Iran leaves the NPT, Iran refuses to allow inspectors access to a site they wish to visit, inspections turn up clear evidence of nuclear weapons development done since 2005.

7.5.8 The United States of America

When Vladimir Putin goes to his magic mirror and asks “mirror, mirror on the wall, who’s the greatest nuclear power of all?”, the answer is invariably “the United States”. In every nuclear metric that matters (so, discounting the total number of warheads), the United States reigns supreme. It has the best stealth bombers, the most accurate missiles, and the biggest fleet of nuclear submarines. As the world’s one remaining superpower, the United States is the only country that is able to mount even a half-way decent first strike – although it probably can’t launch a successful first strike against any triad state.

I don’t know what US nuclear policy will look like going forward. If Donald Trump maintains good relations with Putin, then a nuclear exchange with Russia will be (even more) unlikely. I do think a nuclear exchange with China has become slightly more likely as a result of Donald Trump’s election, but I hope the risk is relatively low.

No matter how you cut it, the risk of a nuclear exchange is – and always has been – low. But no one truly knows how low “low” is. Is it 10% over the Trump presidency? 1%? Whatever it is, I wish it was lower.

I also wish Trump was less of a loose cannon. I can’t really make predictions about America’s nuclear policy over the next few years because the information I have is too heavily weighted towards hyperbole and bluster.

All the nukes in the world are useless unless you have a way to get them to their targets. Aside from outlandish and potentially suicidal methods like suitcase nukes or nuclear artillery, there are three main ways of doing this: bombers, intercontinental ballistic missiles (ICBMs), and submarine launched ballistic missiles (SLBMs).

6.1 Bombers

The only nuclear weapons ever used in anger were delivered by B-29 bombers, the Enola Gay and the Bockscar. Because the Allies had attained near total air-superiority over Japan at the time of the bombings, it was possible for these bombers to go in without any real escort. They were accompanied only by weather reconnaissance and observation planes.

In a modern nuclear exchange, total air superiority would probably be required for a country to be able to openly deliver a bomb. If a nuclear bombing is attempted with anything less than total superiority, the attacker can expect to have all remaining anti-air defenses thrown at them.

Absent total air superiority, the only reasonable way to deliver a nuclear weapon via bomber is stealth. Conventional bombers can make a play at this by flying below radar, but this is very risky and doesn’t guarantee success. In states of heightened alert, the bomber would have to avoid both radar and settlements where residents might alert the target to the coming attack.

Stealth aircraft can make a better go at delivering bombs, especially when the target is on high alert. All new combat aircraft are billed as having some stealth capabilities, but the only true stealth bomber is the US B-2 Spirit. Its virtually invisible to most means of detection (it shows up about as well on radar as a large bird or a sphere a foot in diameter would) and can carry 18 tonnes of bombs 11,000km (this gives it a range of about 5,000km, neglecting in-air refueling).

The main disadvantage of bombers then, is the difficulty that they all (except the B-2) face in arriving at their target both intact and at the correct altitude to drop a bomb. There are a couple of significant advantages to bombers though.

They’re controlled by people. Really well-trained, smart people who know what they’re doing. Tell a pilot that it’s up to her to prevent a retaliatory strike on her home country and she’s going to be very motivated to carry out her mission. Maybe she’ll come up with a new trick and make it to her target. If she doesn’t, maybe one of her squadron mates will. Send in a dozen bombers and you force your enemy into a problem of resource allocation. They may end up spread too thin to stop them all. And if just one gets through…

Having pilots also makes bombers fairly accurate. The steep fall-off in weapon intensity with distance from the target (as discussed earlier in §5.1) demands a fair amount of accuracy and pilots can be significantly more accurate than some ICBMs.

Bombers are also the delivery method that’s hardest to stop at the source. During periods of heightened alert, countries will always keep nuclear armed bombers in the air. Any sign of a nuclear attack and they’ll be released for bombing runs on their prearranged targets. Unlike ICBs, there is no silo you can take out to stop them.

Additional Reading: B-2 Spirit, Nuclear Bomber, Stealth Aircraft

6.2 ICBMs

Intercontinental ballistic missiles are missiles capable of delivering nuclear warheads at least 5,500km from their launch site. Intercontinental ballistic missiles are distinguished from cruise missiles by their use of rocket engines and their suborbital trajectories. They leave the atmosphere during their boost phase and come back down on top of their target, using a guidance system and fins for limited maneuvering and final targeting once back inside the Earth’s atmosphere.

US and Russian ICBMs use multiple independent targetable re-entry vehicles (MIRVs). A single ICBM can carry many MIRVs, which allow it to simultaneously strike several targets. MIRVs were invented in response to anti-ballistic missiles, nuclear tipped missiles that detonate near incoming enemy missiles, causing them to fizzle or frying their electronics. The advent of MIRVs made missile defense mostly pointless, because it is much easier to build bigger ICBMs with more MIRV capability than it is to expand missile defense systems to deal with more incoming MIRVs.

MIRVs were one of the main factors pushing down bomb yields. The scaling factors we went over earlier mean that smaller weapons do damage much more efficiently than larger ones. It was MIRVs that made the delivery of these smaller weapons economical. Without MIRVs, launching 8 smaller warheads would require launching 8 smaller missiles, which would end up much more expensive than launching one really huge missile. With a MIRV, you get the best of both worlds – one missile, many warheads.

Improvements in US and Russian missile accuracy over the last 50 years have also driven the decrease in the warhead yields commonly fielded by those countries. Early ICBMs were accurate to within about half a kilometer, which necessitated big warheads if you wanted to be sure of annihilating your target (this is especially important when the target is hardened, like your enemy’s nuclear arsenal is liable to be). Modern US and Russian missiles are now accurate to something like 200m, which (using cubic scaling to approximate destruction) allows for the same change of destroying the target with about 1/15 the yield.

China lags behind on ICBM technology, with Chinese missiles only accurate to within 800m of their target. This results in China using much larger warheads (4-5 Mt) on its ICBMs than the 475kt US MIRV capable W88 or the 200-300kt warheads the Russians favour. It is counter-intuitive yet true that routinely using larger warheads is the sign of a less capable nuclear power.

ICBMs are stored on mobile launching units, or in missile silos away from large cities and built into mountains or underground. Both strategies are designed to protect missiles from enemy attack. Keeping ICBMs mobile makes them very hard to target with other missiles – where they are when the missile is launched may not be where they are when it arrives. Add this to the problem of finding the damn things in the first place and you have relatively good protection.

Missile silos were most useful when missile accuracy was much worse than it currently is. Missile silos are supposed to be impervious to everything but a direct hit from another nuclear weapon. Once a difficult task, this is now fairly possible. If nothing else, missile silos are useful because they will divert some enemy attacks away from populated areas. Still, the expense and relative uselessness of nuclear silos mean that no one is clamouring to build more of them. The Russians are pivoting towards mobile ground based launchers, while the USA is focusing its efforts on SLBMs.

ICBMs have several advantages over bombers. Missiles can endure g-forces, altitudes, and heat that would kill humans many times over. Taken together, this lets missiles launch with very rapid acceleration, travel at very high speeds, and use very efficient trajectories; basically they arrive on target much more quickly than a bomber can. They’re also much harder to intercept. Modern ICBMs come with a variety of countermeasures, from dummy MIRVs, to aluminum balloons, all of which make countering them extremely difficult.

The Iron Dome suggests that >90% missile interception might be technically possible, but intercepting ICBMs remains much more difficult than intercepting short range rockets, so don’t expect much progress anytime soon.

One disadvantage of ICBMS are their highly visible launches and re-entries. A successful stealth bombing provides no warning at all and a successful SLBM attack might provide only 2-10 minutes of warning. ICBMs take something like 30 minutes to cover intercontinental distances (say, Russia to the USA, or China to the USA). During this time targets can be evacuated and retaliatory strikes prepared.

Additional Reading: ICBM, MIRV, Minuteman III, Dongfen 5, RS-26, United States National Missile Defense

 6.3 Submarine Launched Ballistic Missiles

At one point, it made sense to have a firm distinction between SLBMs (which had fairly small ranges) and ICBMs. For modern weapon systems, the ranges are effectively the same. Gone are the days where submarines had to sneak right up to the coast to attack their targets. If you were to look at a globe centred on a modern US or Russian ballistic missile submarine, you’d see all the targets it can hit.

Modern nuclear armed submarines are also very stealthy. As long as they remain submerged they cannot be found by satellites. SONAR and thermal imaging remain about the only things that can find them. Active sonar will smoke a sub out pretty quick – at the cost of letting the submarine know it’s been found and giving away the position of the intercepting vessel. Passive SONAR can occasionally find them, but the ocean is big and nuclear submarine captains know a whole bunch of tricks for masking their SONAR signature.

Thermal imaging can find nuclear powered submarines by the wake of warm water they leave behind (warmed after cooling their nuclear reactors), but savvy captains (and all of them are) can minimize their changes of detection by cruising along boundaries between zones of different temperatures. These zones are so irresistible to subs that a French and British submarine managed to collide as they both (probably) followed one. This was obviously pretty embarrassing to both countries, but should stand as a testament to how damn stealthy these things are. Had either captain realized what was going on, he would have changed course.

Submarines powered by nuclear reactors can run for decades without any refueling. The only practical limit on how long a nuclear sub can stay submerged is the food supply. Water is collected from the ocean and air is recycled much as it would be on the ISS.

SLBMs have all the advantages of ICBMs, with the added advantage of stealth. The US has embraced them wholeheartedly. Its SLBMs are more accurate and more heavily MIRVed than its land based missiles; the Ohio class nuclear submarine carries 24 Trident II SLBMs with up to 8 W88 475kt warheads, each with an effective range of 11,000km and an average error less than 90m. The Russians, Chinese, and others have been less enthusiastic with their nuclear subs. Russia in particular has been having a lot of trouble getting new SLBM systems into active use.

This highlights the one real disadvantage of SLMBs. They’re expensive and technologically complicated. Just getting a missile to successfully launch from beneath the ocean is challenging enough. Add in stealth requirements and a nuclear reactor and it’s no wonder that the sea floor is littered with sunken nuclear submarines, primarily of Soviet origin (one unlucky Soviet nuclear sub actually sunk twice).

You could reduce the technical complexity by using a more traditional diesel-electric submarine, but you’d sacrifice much of the submarine’s endurance and ability to stay submerged. Batteries can only power a submarine for a few days (or hours, if it’s moving as fast as it possibly can) and diesel engines will quickly suffocate the crew if run underwater. Add to this the limited amount of diesel any sub can carry and nuclear submarines become the only real way to do long duration submerged deterrence patrols.

Additional Reading: Nuclear subs collide in Atlantic, Submarine launched ballistic missile, Ohio Class Submarine, Trident II, Borey Class Submarine, List of sunken nuclear submarines, Nuclear submarine accidents, Bulava missile troubles

To understand the effects of nuclear weapons, you first need to understand how those effects scale with weapon yield.

5.1 Scaling

Modern bombs are much smaller than the Tsar Bomba. The standard US nuclear warhead, the W88, is a “mere” 475kt, a yield that is 100x less than that of the Tsar Bomba. On the other hand, the W88 weighs in at 360kg, 75x lighter.

This may seem like a poor trade, but it’s actually a very good one, due to the fundamental properties of explosive scaling. Scaling factors are very important to weapons. They determine the stable equilibriums that designs fall into. For example: we have tanks instead of mechs because strength scaling and mass scaling together make tall vehicles very vulnerable to weapons.

Scaling factors for all nuclear weapon effects (the fireball, the shock wave, and electromagnetic radiation) are different, but we can use the scaling factor of electromagnetic radiation as an example.

A nuclear weapon emits a set amount of energy as photons – gamma rays and X-rays, as well as IR, UV, and visible light. These photons radiate out approximately equally in all directions. The energy that the photons carry is fixed, but the area that the energy covers isn’t. We call the amount of energy per square meter “intensity”.

Intensity is incredibly important. It’s what drives the differences in climate between Mercury (surface temperature: 427ºC) and Pluto (surface temperature: -218ºC). In nuclear weapons, the intensity of radiated photons determines if buildings are set afire from the heat or people die of radiation sickness.

Since energy is constant, intensity will depend only on the area the energy is spread out over. Luckily, it’s pretty easy for us to calculate this area as a function of distance from the bomb. Dividing the total energy by the surface area of a sphere of a certain radius, we get the following equation for intensity:

Where E0 is the initial energy of the bomb and d is the distance from the epicentre of the explosion.

Ten meters out from a 10kt (in the SI unit, joules, this is 41.8 TJ) bomb, the energy per square metre 33,300 MJ. Ten times further away, the energy per square metre is 333 MJ, a 100-fold decrease.

To get the desired destructive power, a certain intensity is necessary. An intensity of 372kJ per square meter is necessary to give third degree burns, for example. Ignoring loss from the air, you get this effect from a 10kt bomb out to about 3km. Increase the bomb size 100-fold to 1Mt, and the radius expands only ten-fold, to 29.9km.

These examples illustrate one of the fundamental scaling factors of nuclear weapons. The photon effect radius of a nuclear weapon is proportional to – at best – the square root of the power of the weapon. This means that there are intense diminishing returns to increasing the power of a weapon. Increase its power 25 times and the radius of destruction becomes five times bigger. Increase the power 50 times and the radius only increases sevenfold.

Blast radius and fireball size scale differently than radiation, but in all cases the scaling is square root or worse. In fact, you’ll soon see that radiation scales better than nearly every other destructive effect.

Taking into account this scaling factor, the W88 is (by one measure) 10x less destructive than the Tsar Bomba, but 75x lighter – a trade-off that makes much more sense. Keep this relationship in mind whenever nuclear weapons are described in terms of “the power of the atomic bomb that destroyed Hiroshima”. The policy implications of this relationship will also become clear later. For now, it’s important that you simply understand that relative destructive power and yield don’t correspond 1:1.

Additional Reading: Inverse Square Law

5.2 Direct Effects

Thus far, we’ve restricted our discussion of nuclear weapon detonation to the technical. There’s a cascade of neutrons and a bunch of energy is released as a by-product – that much is clear. But what does all that energy do? How do we get from neutrons to levelled cities and mushroom clouds?

There isn’t as much clear, well-curated information on this topic as I’d like, but perhaps that’s only to be expected. As near as I can tell, here’s what happens:

1. A lot of heat (and some photons and neutrons) get generated by fission or fusion
   1. The fission of an atom releases a huge amount of energy. Only 7% of this in the form of fission neutrons and gamma rays; most is in the kinetic energy of the atom fragments, which will be travelling at about 4% of the speed of light.
   2. Fusion reactions output mainly high energy neutrons. These are captured by the dense tamper around the fusion core, allowing them to cause fission, or to impart their kinetic energy to another atom. Most fusion neutrons that don’t collide with the tamper will escape.
2. About one microsecond after detonation, the core and tamper of a bomb will be a cloud of plasma a couple metres in diameter, with temperatures exceeding 10,000,000ºC.
3. Via blackbody radiation, the plasma emits x-rays. These x-rays heat the surrounding air to a similar temperature.
4. The very air becomes incandescent, releasing a bright pulse of light.
   1. Depending on yield, distance, and atmospheric conditions the pulse can be permanently or temporarily blinding for anyone who looks at it.
   2. As well as visible light, IR and UV light is emitted. The combined radiance of all of this light can set buildings on fire or give people severe burns.
5. The superheated air in the very centre of the explosion pushes against the surrounding air, acting like a giant piston and producing a massive shock front.
6. Compression in the shock front causes it to briefly become a dense plasma, temporarily blocking the light from the incandescent air closer to the core.
7. The compressed air of the shock front begins to expand rapidly, dropping in temperature and becoming clear. The fireball in the centre can once again be seen.
8. This shock wave travels at approximately the speed of sound. Whenever it hits something (a house, a vehicle, a tree), it causes damage via two mechanisms.
   1. Static overpressure: the direct hammer-blow of all that dense air striking at once.
   2. Dynamic overpressure: the drag associated with the wind of the shock-front’s passage, which can tear, tumble and drag objects.
9. Back in the centre of the blast, there’s still a bunch of hot air. It will quickly cool to the point where it no longer glows, but it will still be much brighter than the surroundings.
10. This hot air begins to rise, pulling up debris from the ground (at this point, there will be plenty) into the vacuum its rising creates. The hot air also expands and the outside edges cool, which causes rotating internal air currents that help to draw in more air from below. This forms the characteristic mushroom cloud of a nuclear explosion.
11. Between the vacuum caused by the rising mushroom cloud and the vacuum caused by the initial shockwave, a steady wind will blow back towards ground zero until equilibrium is reached. This wind will further toss around debris and can help to fan the flames of any fires that have started.

[Image Credit: Wikipedia/Mushroom Cloud]

 5.2.1 Fireball

The nearby detonation of an atomic bomb is utterly devastating to infrastructure. Everything within the central fireball will be annihilated. The size of the fireball varies: 150m for a 10kt blast (slightly smaller than Hiroshima), 720m for the modern 475kt W-88 favoured by America, and 4.62km for the monstrous 50Mt Tsar Bomba. Ideally, fireball effects would scale as the square root of the yield. In actuality, they scale a bit worse than this, as the fifth root of the square of the radius (i.e. the 2.5^th root). I’m not sure exactly what causes this, possibly some weird property of plasma, or just general fluid dynamics oddness.

[Data Source: Nukemap 2.42] Trend line has equation .

5.2.2 Shockwave

As deadly as the fireball is, most of the damage from a nuclear explosion comes from the shockwave. Weapon designers carefully model the shockwave and use these insights to come up with an optimal detonation height for various effects – like levelling reinforced buildings or destroying residential areas. Models take into account the “Mach stem”, the angle at which some of the wave will be reflected off the ground and constructively interfere with the rest of it, leading to increased destruction.

To get an idea of the damage a shockwave causes, we can look at the radius within which the maximum pressure increase is 20 psi and the radius within which the maximum pressure increase is 5 psi. 5 psi is the blast pressure required to level residential buildings, while 20psi will demolish even heavy concrete buildings.

Unlike radiation effects, blast pressure increases with cube root of the yield. This scaling – like the shockwave itself – is driven by heat. The pressure air exerts is directly proportional to its temperature (this relationship is linear and given by the ideal gas law: PV = nRT). To get a 5 psi overpressure at a certain radius, the bomb must ultimately heat all the air within the radius to a temperature that will cause a 5 psi increase in pressure. The energy taken to heat a material is proportional to the mass of material to be heated. The mass of air is proportional to its volume. And volume of a sphere is really easy to calculate: . Chaining all these observations together, we’re left with radius increasing as the cube root of explosive power.

This means that for every 1000-fold increase in bomb power, you’ll see a 10-fold increase in destructive radius. In terms of overpressure effects, the W88 is about 4.7 times less destructive than the Tsar Bomba, while being 75 times lighter. This trade-off comes up even better than for radiation.

[Data Source: Nukemap 2.42] 5 psi trend line has equation , 20 psi trend line has equation .

5.2.3 Direct Radiation

Nuclear weapons produce two types of damaging radiation (they’re both photons, just at different energy levels): thermal and ionizing. Ionization radiation (mainly gamma rays) is responsible for causing acute radiation poisoning and later cancers. Thermal radiation (light, from IR to UV) is responsible for the horrendous burns nuclear weapons survivors often bear.

In a vacuum, both of these would exhibit scaling with the square root of the yield, as discussed earlier. In real world conditions though, there’s another key factor in play: the atmosphere. Air is actually fairly good at absorbing gamma radiation. For every 150m that a group of gamma rays travels through the air at sea level, about half of them are absorbed. This is cold comfort if you ever find yourself at ground zero of a nuclear explosion, but it means anyone more than a few kilometres away has fairly little to fear from direct radiation. Once atmospheric absorption is taken into account, the radius where gamma radiation is lethal increases with approximately the 6^th root of the yield. This means that even very large bombs scarcely irradiate more people than their smaller counterparts.

As blasts become larger, fewer and fewer people are affected by direct radiation. The radius of complete destruction (20 psi overpressure) is smaller on 10kt bombs than the radius at which radiation fatalities are common. But the differences in scaling means that by the time a bomb is 575kt the radii are equal. Beyond this size, the radius of complete destruction will be larger than the radius of radiation effects and most casualties will come from the shockwave, not radiation.

IR, visible, and UV light are much less affected by the atmosphere. If the atmosphere absorbed light at the same rate it absorbs gamma rays, the sun would appear 717 billion billion times brighter on an airplane at cruising altitude as it does at sea level. This obviously isn’t the case – the sun is approximately as bright in both cases. This difference in absorption means that the radius at which nuclear weapons cause burns scales with close to the square root of the yield (the small amount of energy absorbed by the air does mean that it scales slightly slower in atmosphere than in vacuum).

[Data Source: Nukemap 2.42] 50% Deaths trend line has equation , 3^rd Degree burns trend line has equation .

Additional Reading: Effects of Nuclear Explosions, Nuclear Weapon Design: Fission, Nuclear Weapon Design: Fusion, Mushroom Cloud, Nukemap and The Nuclear Double Flash.

5.3 Indirect Effects

Part of the horror of nuclear weapons comes from fallout, the lingering radiation left behind after a nuclear detonation. Fallout doesn’t last forever (and remember, the more unstable and radioactive an element is, the quicker it breaks down into something stable), but while it is around it can sicken and kill. Fallout can also cause illness and death far from the site of the initial blast, increasing the human toll of any nuclear weapon detonation.

Fallout is composed of unstable, radioactive isotopes that are scattered after a nuclear bomb is detonated. Fallout is mainly comprised of fission by-products (some of these are stable, but many are themselves radioactive and must conduct additional decay before they become stable) and left over fuel (remember, fission efficiency is almost never higher than 40%, leaving 60% of the fuel weight in radioactive waste). Depending on bomb detonation height, there may also be neutron activated fallout. Neutron activated fallout arises when stable elements are bombarded with neutrons and either capture the neutron, or fission from the energy of the collision. Since nuclear weapons tend to be detonated high in the air (with the hope of maximizing the shock wave), neutron activated fallout is a small portion of the total fallout. Either way, all fallout behaves similarly.

Some of the fallout will be blown up into the stratosphere by the force of the blast, or carried up in the wake of the rapidly rising fireball. These fallout products end up distributed fairly evenly across the whole globe. Nuclear testing before the partial and complete test bans resulted in a dramatic rise in strontium-90 and caesium-137 levels in humans all over the globe. These isotopes can cause cancer and other problems if ingested in significant amounts and the threat posed by steadily increasing global levels was enough to prompt the test ban treaties.

Stratospheric fallout is the minority; most of the fallout remains in the area of the blast. It’s still dispersed in the atmosphere, but its subject to local winds, not the global jet stream. The largest particles begin to fall immediately. Not all of them will be radioactive – some of the falling matter will be simple vaporized dirt or water – but it will never be safe to assume that any particles falling in the wake of a nuclear explosion are benign.

A high altitude burst doesn’t guarantee that the particles that rain down will be fallout free. Even if no neutron activation occurs, some of the material that rains down will be contaminated with radioisotopes produced by the blast. After the Castle Bravo nuclear test, coral contaminated with radioisotopes fell like snow over a large area, causing severe radiation burns whenever it touched human skin.

In the first day after a bomb is detonated, half of the local fallout will be deposited – unless it rains, in which case even more will fall. Rain is actually quite likely after a nuclear weapon detonation because fine particles dispersed in the atmosphere (like the dirt drawn up in a mushroom cloud) can help start the process of raindrop formation. Rain makes decontamination even harder, as the radioactive ions that travel down with raindrops tend to bond to external surfaces, requiring sandblasting to remove them.

The danger of the fallout varies with local atmospheric conditions, the size of the bomb, the efficiency of the bomb, and where the bomb detonates. Winds can disperse fallout, affecting more people, but giving each person a lesser dose. Bombs with a high efficiency create less fallout than bombs with lower efficiency. Bombs that get more of their power from fusion will generally have less fallout than a bomb of the same yield that gets more energy from fission.

Because of the potential variance in conditions, there are few general rules about fallout. I’ve seen something called the “Seven Ten” rule bandied about. The rule claims that every sevenfold increase in time after the detonation leads to a tenfold decrease in radiation from fallout. So after seven hours there will be ten times less radiation compared to the first hour and after forty-nine hours (approximately 2 days) there will be one hundred times less radiation.

I spent a lot of time confused by this. Remember half-lives from earlier? Instead of a fixed ratio of times (i.e. seven times) producing a fractional drop in radioactivity (i.e. ten times), I thought there should be fixed period of time that produces a fractional drop in radioactivity. If there really is a tenfold drop in radiation after seven hours, then it means that the half-life of the isotopes in the fallout would average out to 2.1 hours. Which should predict there would be a one-hundred hold decrease in fallout after fourteen hours (The equation is  which gives approximately 1/100). This definitely isn’t the case, which was the cause of my confusion.

I think the 7-10 rule is correct because of the mixture of isotopes we see in fallout. The most common isotopes in fallout (and their half-lives are): neptunium-240 (61.9 minutes), neptunium-239 (2.1 days), uranium-237 (6.75 days), iodine-131 (8 days), tritium (12 years), caesium-134 (20 years), strontium-90 (28.8 years), caesium-137 (30 years), technetium-99 (211,000 years), and zirconium-93 (over 1,000,000 years). The widely disparate half-lives lead to a variety of zones, each of which is dominated by a different half-life.

The 7-10 rule is an approximation that more or less fits the wonky data that results from this mixture. But it isn’t a very good approximation – it’s accurate to within 25% for the first two weeks and accurate within 50% for the first six months. After that, fallout is dominated by isotopes with similar half-lives and fits an exponential decay model, rendering the 7-10 rule basically useless.

Iodine-131 is one of the most famous fallout products and is primarily dangerous if consumed. Once in the body, it will be concentrated in the thyroid, where it can cause immediate damage or eventual cancer. This effect can be minimized with iodine tablets, which essentially “fill-up” the thyroid with safe iodine, leaving no room for iodine-131. Recommended dosage of iodine varies with the severity of the radioactive contamination and isn’t recommended for anyone over 40 except in cases where they may face acute thyroid damage. It’s also recommended to avoid drinking milk if you suspect any contamination with iodine-131.

Neptunium-239 and uranium-237 are responsible for most of the gamma radiation in fallout in the early days after a nuclear blast (neptunium-240 dominates for a few hours, but promptly fades). This is actually fairly good. After 20 days, there will be 1000-fold less radiation from neptunium-239 as there was immediately preceding the blast. Uranium-237 is a bit longer lived, requiring about 67 days for a 1000-fold decrease in radioactivity. Whether this represents a safe level is dependent on the initial amount and the relative abundance of neptunium-239/uranium-237 and other elements.

If you know or suspect that your area has been exposed to fallout, minimize your time outdoors. Your dwelling will offer excellent protection against alpha and beta particles and some protection against gamma radiation. If you must venture outside, wear many layers of clothes to protect yourself from beta particles. Don’t drink water without filtering it first. Try to avoid food produced after the fallout started, especially milk. Food produced before the nuclear exchange should be safe, as long as the outside is washed with clean water. Peeling fruit will also suffice. You can relax these precautions as the weeks and months wear on. In the years after fallout occurs, be aware of your surroundings. Hot spots (with fatal levels of radiation) may exist for several years. If you see dead trees or animals with no signs of rot, head in the other direction quickly.

This is just the barest of outlines. There are many guides out there that will go into specifics. Unfortunately, many guides have an implicit agenda (either: “scare people into becoming anti-nuclear activists”, or “keep our population from panicking about nukes”). If you know of a good, scientifically accurate guide, please link it in the comments. As additional reading, I’ve provided the official US Government planning guide, which seems to be accurate. In addition, I’ve linked to a more technical overview of fallout production and decay (Residual Nuclear Radiation and Fallout).

Additional Reading: Iodine Prophylaxis, Nuclear Fallout, Planning Guidance for Response to a Nuclear Detonation, Residual Nuclear Radiation and Fallout

The last section required that you take it on faith that nuclear weapons are hard to design. Now it’s time to get into the nitty-gritty details of weapon design and understand why that is.

Nuclear explosions require a critical mass of the right unstable isotope. But there’s no safe way to store an assembled critical mass. As soon as you get to the critical mass, the chain reaction starts and an explosion will occur without drastic countermeasures.

All nuclear weapon design ultimately starts with this problem of assembling a critical mass in situ (and only ever in situ).

The first atomic bombs used one of two methods: gun assembly or implosion. These methods are still used to this day in fission weapons or in the fission first stage of multiple stage weapons.

4.1 Gun Assembly Design

[Image Credit: Wikipedia/Nuclear Weapon]

The gun assembly method was used in Little Boy bomb dropped on Hiroshima. In this method, sub-critical hemispheres (or a ring and a cylinder) are combined by conventional explosives into a critical mass (inside a neutron reflective tamper, which further decreases the necessary fissile mass).

This method is highly inefficient and suffers from “fizzling”. Basically, the pieces tend to become supercritical before they fully touch. The massive energy released by this reaction will then blow the pieces far apart before the reaction can complete. Only 1% of the uranium in the Hiroshima bomb was used up in the explosion. The rest was scattered and eventually rained down as fallout.

The only possible fuel in gun-assembly bombs is U-235. Plutonium fissions much more quickly, leading to the sub-critical pieces being blown apart before any significant mass of the plutonium can fission.

Due to these problems, this method has mostly fallen out of use, except in applications where only one dimension can be large (such as artillery shells).

Additional Reading: Little Boy, Nuclear Artillery, Fizzle, and Gun-type assembly weapon

4.2 Implosion Design

[Image Credit: Wikipedia/Nuclear Weapon]

Implosion type weapons start with a sub-critical mass and use explosives to compress it until it becomes critical. As the density increases, fewer neutrons escape the mass without triggering further fission events, leading to criticality despite the lower than normally necessary mass. An outer shell directly around the core that reflects neutrons can further decrease the mass necessary for a detonation.

The fizzling problem that plagues the gun-type device is also present here – eventually, the nuclear chain reaction will overwhelm the force imparted by the conventional explosives and blow the fuel apart, stopping the fission progress. Tamper design plays a factor in this; due to its very strong and heavy uranium tamper, the Fat Man plutonium bomb held together for a few hundred additional nanoseconds, allowing it to attain 20% efficiency.

In addition to tampers, there are various strategies to mitigate the breakup, all adding cost or technical complexity. For example, some nuclear weapons add a layer of aluminum-beryllium alloy between the explosives and the tamper to slow down the explosive shock wave and cause the core to be held compressed for slightly longer. This tweak can marginally increase the efficiency of a weapon.

Additional Reading: Implosion-type weapon and Fat Man

4.3 Miniaturizing Weapons

The core of a nuclear weapon is very small and relatively light (compared to conventional munitions). Most of the size and weight of a nuclear weapon comes from the explosives, tamper, and superstructure necessary to hold it all in the right place. The Fat Man bomb was egg shaped, 3.3 metres long and 1.5m in diameter at its thickest. The whole thing weighed 4.67t. the plutonium core was only 6.4kg, 0.14% of the total weight.

There’s no point having a weapon you can’t deliver. Methods that worked to deliver bombs across the 2500km that separated US island airbases and Japan weren’t feasible when dealing with the vast expanse of the Pacific, or the 8500km between the missile silos in the central United States and Moscow. Mass is the principle limit – the lighter its payload, the further a missile or bomber might fly.

Making weapons more powerful and miniaturizing them are two sides of the same coin. Every innovation that makes a bomb blast bigger can also potentially allow a smaller weapon to have the same effect as a larger weapon.

There are many unique challenges to miniaturizing weapons, such as correctly focusing explosions to create the desired pressure pattern. Regardless of the specific features necessary for miniaturization, it’s important to understand why miniaturization is such a big deal. North Korea, for example, has only possibly succeeded in miniaturizing its weapons to the point that its missiles can carry them any significant range. Until it thoroughly conquers this technical hurdle, only its closest neighbours have anything to fear.

4.4 Boosted Weapons

Boosted weapons were the first atomic bombs to make use of both fusion and fission, although they still get almost all of their energy from fission. Boosted weapons are all of the implosion type. Instead of a solid core, a boosted weapon will use a core with a hollow centre. In this hollow will be an equal mixture of tritium and deuterium.

Detonation proceeds as normal in the implosion assembly method, but with one extra wrinkle. The intense heat and pressure of the plutonium explosion compresses the deuterium/tritium mix to the point where these atoms begin to fuse into helium. Deuterium has two neutrons and one proton; tritium has three neutrons and one proton. Helium most commonly has two of each. Which means that every single fusion event also releases a neutron and imparts into that neutron a great amount of kinetic energy – it will be moving. These neutrons are the goal of the fusion stage.

Recall that fission chain reactions occur because neutrons from each fission event create further fission events. Adding a bunch of bonus neutrons to the mix puts the whole process on steroids. It helps that fusion neutrons move even faster than fission neutrons, so when they cause fission to happen, even more neutrons are released.

Using a boosted weapon, efficiency can be doubled. Which means that the same mass can give twice the yield, or the same yield can be attained with half the mass (because of this, boosting is one of the principle mechanisms of miniaturization).

There are a host of other benefits to boosting. Because the efficiency is increased, a smaller, lighter tamper can be used (these days they’re made of beryllium, which reflects neutrons but is not nearly as dense as uranium). Because the tamper is lighter, less explosives are required to compress the whole package. Boosting also gives a higher efficiency, so there’s no need for an additional focusing layer of aluminum/beryllium, leading to further savings in direct mass and mass of explosives necessary to drive the implosion.

All of these, combined with the reduction in plutonium mass necessary to get the same yield means that bombs have shrunk dramatically since the earliest days of nuclear weapons. The Fat Man bomb was 4.67t and 3.3m by 1.5m. It had a yield of 20kt. The first boosted weapon, the US “Swan” had a weight of just 47.6kg and measured 58cm long by 29.5cm in diameter, with a yield of 15kt, there was only a small loss of power for a 100x decrease in mass.

While it’s possible to build bombs in the >100kt range using boosted fission, this is rarely done. Fission remains fairly inefficient, so the amount of fissile material and tritium that must be used to get these yields quickly becomes wasteful. Very high yields remain the sole providence of fusion bombs, which can achieve these yields much more cheaply (recall how difficult and expensive enrichment is – to put a dollar value on it, the US pays $5,000 per gram of weapons grade plutonium-239).

There’s only one real drawback to boosted weapons: tritium. Tritium has a half-life of just 12 years. Even worse, it breaks down into helium-3, a form of helium with a single neutron that really “wants” to have two neutrons instead. Because of this, helium-3 “poisons” nuclear weapons by capturing neutrons that would otherwise cause fission. This means that boosted weapons must be periodically serviced and the helium-3 removed and replaced with fresh tritium. This is expensive (tritium cost $30,000 per gram in 2003 but is probably much more expensive now; we can’t calculate the cumulative cost because no government is willing to share the amount of tritium that is in each of their nukes, but it’s probably large) and almost certainly tricky (again, I can’t be sure, because no one releases their refueling procedures).

Additional Reading: Swan, Tritium, and Boosted Fission Weapon

4.5 Alarm Clock/Sloika

The Alarm Clock (named because Teller thought it would “wake up” the United States to the possibilities present by fusion weapons) or Sloika (Russian for a type of layered cake dessert) is a single stage boosted fission weapon that gets up to 20% of its yield from fusion (compare the ~1-2% that boosted weapons get from fusion).

The Sloika is made up of three layers. In the centre is a classic fission core, like in the normal implosion detonation design. Surrounding this is lithium-6 deuteride. The outer layer is unenriched uranium-238.

When this whole contraption is detonated (and detonating it takes a lot of explosives due to the three layers), three reactions occur in sequence. First, the core goes critical and fissions, ejecting a bunch of neutrons. Second, these neutrons hit the lithium-6 deuteride, which first cause lithium-6 to fission into tritium, which promptly fuses with the deuterium in the intense heat. The fusion reaction releases a lot of heat energy, a lot of radiation, and a lot of very fast neutrons. Finally, these fast fusion neutrons hit the natural uranium and cause it to fission. Since fusion neutrons are about 14 times more energetic than fission neutrons, they can cause fission even in unenriched uranium-238. The neutrons the uranium-238 releases during fission actually lack the energy to cause a self-sustaining chain reaction within itself, but they are capable of converting lithium-6 into tritium and helping the tritium fuse with deuterium. For a brief moment, this bomb has a stable multi-step chain reaction (fusion à high energy neutrons à fission à lower energy neutrons à lithium converted to tritium à fusion). Then it tears itself apart.

The Sloika derives most of its explosive power from relatively cheap natural uranium, which makes it a more cost effect way of getting yields in the 100s of kilotons compared to pure or boosted fission. It’s also safer, because most of its yield comes from a fuel incapable of spontaneous fission. Despite this it can’t truly compete with fusion weapons for cost effectiveness and lacks the ability to scale to the Mt range (as the amount of explosives necessary to detonate it grows infeasibly large at higher yields).

Because of these inadequacies, the Sloika proved a dead end in weapon design and is not used in any current nuclear weapons.

Additional Reading: Joe 4 and Alarm Clock/Sloika

4.6 Teller-Ulam Design

The Teller-Ulam design is necessary to economically achieve multi-megaton yields with devices small enough to be delivered intercontinentally. Teller-Ulam weapons (also known as thermonuclear bombs or H-bombs) use multiple stages of fission and fusion and incorporate concepts previously seen in boosted, implosion, and Sloika weapons. Here’s what the design looks like:

[Image Credit: Wikipedia/Thermonuclear Weapon]

The primary is a standard boosted implosion bomb, normally with a yield in the range of 10-20kt. The secondary is made up of a “sparkplug” of highly enriched plutonium or uranium, a supply of lithium-6 deuteride (the fusion fuel), and a natural or depleted uranium tamper (made primarily of uranium-238). In many designs, the bomb is also packed tightly with a plastic foam.

Detonation begins with the primary, which works exactly like the boosted implosion design described above (while boosting isn’t strictly necessary, boosting is important for reducing the size of the weapon and so in practice will be used in almost all modern thermonuclear weapons). As the primary detonates, three separate forces begin to exert immense pressure on the secondary.

First is X-ray radiation. The primary produces x-rays (in addition to neutrons and gamma rays), which can be mirrored by the outer casing to focus on the secondary. These x-rays have momentum and therefore exert pressure when they collide with the secondary. The large amount of high energy x-rays exerts a pressure equivalent to tens of millions of atmospheres, beginning the process of compression.

Plasma from the foam also plays a role in compression. X-rays aren’t just hitting the secondary. They’re also being absorbed by the foam. No plastic can withstand such an intense barrage of x-rays – so it almost immediately converts into plasma. As this plasma tries to expand in the confines of the bomb, it exerts even more pressure on the secondary, probably an order of magnitude more than the x-ray radiation alone.

Finally, the x-rays cause the tamper to begin to vaporize. Gaseous uranium boils off the surface under the intense x-ray bombardment. For every action, there must be an equal and opposite reaction and such is the case here. The uranium boiling off the surface pushes the rest of the uranium into the fusion fuel and it pushes it hard. The ablation pressure in a thermonuclear weapon is an appreciable fraction of the pressure at the centre of the sun, equivalent to many billions of Earth atmospheres.

Which one of these ultimately causes the secondary to detonate is intensely classified. What we do know is that one (or more) of these eventually causes the fissile “sparkplug” to go supercritical, which pushes on the fusion fuel from another direction and provides neutrons to activate it. Shortly after the sparkplug ignites, fusion begins.

Fusion provides a bit less than half of the total power of the average thermonuclear bomb. The rest comes from the fission of the uranium tamper with the very fast neutrons produced in the fission process.

Total yield varies from a hundred kilotons to a dozen megatons. It’s actually possible to make one bomb design that can have multiple yields (called “dial-a-yield”), by varying the amount of tritium in the primary (and therefore the pressure under which the secondary is compressed).

The Teller-Ulam design has the same disadvantage as boosted weapons: constant maintenance is required to replace the unstable tritium necessary to boost the first phase. This disadvantage pales beside the massive power of this design. It’s the only economical (in terms of cost, weight and size) way to achieve yields in the megaton range.

While a two-stage design is by far the most common, there is no theoretical cap on the number of stages this weapon can have. Each stage provides the compressive force to ignite the next. In this manner, yields up to a gigaton of TNT can be achieved, although a bomb this large would be very difficult to deliver, even with the general efficiencies of this design. The largest bomb ever detonated, the Soviet Tsar Bomba used three stages, massed 27t, and had a total yield of 50 Mt (although it used lead tampers to minimize fallout; had the tampers been uranium the yield would have been 100Mt – and fallout from the fission would have irradiated large swathes of the USSR).

Additional Reading: Thermonuclear Weapon, History of the Teller-Ulam Design, and Tsar Bomba

There are currently nine countries with acknowledged or suspected nuclear arsenals. Five of them are signatories of the Non-Proliferation Treaty (NPT), the main international treaty aimed at minimizing the number of nuclear armed states. Ideally, no country or group would have nuclear weapons. Unfortunately, we don’t live in an ideal world; the NPT is maybe the next best thing.

The NPT acknowledges the right of the permeant UN Security Council members (UK, USA, France, China, and Russia) to possess nuclear weapons even as it bans anyone else from getting (or trying to get) them. The remainder of the nuclear armed states (Israel, Pakistan, India, and North Korea) haven’t signed on to NPT or signed and later withdrew from it. South Sudan also isn’t a signatory of the NPT – I think they just haven’t gotten around to it – but no one is particularly worried about that (for reasons that will soon become apparent).

In the nuclear sense, proliferation – the thing this treaty is trying to prevent – is the spread of nuclear secrets, nuclear capability, or nuclear weapons to state or non-state actors that do not already possess them. Proliferation is fairly universally regarded as a bad thing. Luckily, attaining nuclear weapons is very difficult.

Earlier, I talked about how much information is classified. Interestingly, it tends to be only the specifics of nuclear weapons that are classified. Schematics and detailed procedures are under lock and key, but general principles are all over the internet. The next section of this post series covers historical and modern nuclear weapon design. This is totally legal. There’s no law in Canada or the US against disseminating any of this information (really mom and dad, I promise, this is okay).

There are a couple reasons for this laissez-faire attitude.

First, the general design of nuclear weapons is fairly easy for any competent physicist to derive from first principles, once they know what they’re looking for. The Teller-Ulam design was probably re-created 2-4 times by separate groups of physicists. Some of these re-creations used only basic knowledge of what byproducts the bomb produced and its approximate yield. Since computers will only get more powerful and physicists will only understand physics better, it stands to reason that this feat is becoming easier and easier to repeat. In any given university physics department, there are probably physicists who’ve guessed at the parts of the Teller-Ulam device that are still classified.

Second, this doesn’t really matter. Nuclear weapons are incredibly hard to actually build. Understanding in abstract how something works doesn’t mean you can go out and build it. Think about cars; even though you might understand the principles behind an internal combustion engine in the abstract, you’re probably incapable of building one. I know I certainly couldn’t. I don’t have the right materials. Even if I had them, I don’t have a blueprint to go by. And even if I were to come up with a design and assemble it, I’d most likely end up creating an underpowered and unreliable engine, not up to the standards of Ford, let alone Ferrari.

This maps well to the problems a state or non-state actor would face if they tried to build a nuclear weapon. First, they’d have difficulty acquiring materials. Then they’ve have more serious difficult coming up with a blueprint that makes good use of those materials. If they persevere and successfully create nuclear weapons, they’ll at first only have weak weapons, not up to the standards of the rest of the world.

In terms of materials, all nuclear weapons require inherently unstable isotopes – otherwise there would be no neutrons with which chain reactions could occur. These unstable isotopes (mainly plutonium-239 and uranium-235) are rare (or in the case of plutonium-239, basically non-existent) in nature. The half-lives of these isotopes see to that. Pu-239’s half-life is many orders of magnitude less than the age of the world. All of the Pu-239 the Earth original had is long since decayed. U-235 is longer lived than Pu-239 (it’s half-life is about 1/5 the age of the Earth), but it’s still rare; uranium deposits are mostly of the more stable (and therefore useless) isotopes.

Any moron can cause a nuclear detonation if given a critical mass of pure plutonium-239 (for the record, this is 12kg). Luckily (and despite what Doc Brown thought about the future), it is almost impossible for anyone (let alone morons) to get their hands on any amount of pure Pu-239. Furthermore, there is a huge gap between setting off an explosion in your lab and building a weapons system that can reliably deliver a nuclear payload to a hostile target.

For the actors that wish to try this, there are two paths to a bomb. The first requires uranium, the second, plutonium.

3.1 Uranium

There are a few advantages of using uranium in a nuclear weapons program. You don’t require research reactors just to get fuel for your bombs and the actual bomb design is much simpler. That said (and fortunately for the world), it is much more difficult to enrich uranium than it is to enrich plutonium.

Enriching the uranium – separating out different isotopes so that you’re left with only the most unstable ones – is a massive technological undertaking. First you have to mine it or import it. Either way every other country with a functioning intelligence service is going to find out about it and take note. But that’s just the first and easiest step.

Once you have your uranium ore, you have to dissolve it in nitric acid. Then you add ammonia. Then hydrogen gas. Then you mix it with hydrofluoric acid (which is incredibly nasty to work with). Then you fluoridate it some more with fluorine gas (this is even worse; it’s poisonous and it turns into hydrofluoric acid in your lungs if you happen to breath it in). This laborious and dangerous process gives you uranium hexafluoride, which is a real joy to work with (in the same way that being on the beach in the middle of a category 5 hurricane is a relaxing tropical vacation). Uranium Hexafluoride (hex for short) is incredibly toxic, explodes on contact with water, and corrodes most metals.

You’re going to need to find a metal it doesn’t corrode though (I recommend aluminum) because the next step is putting this terrible chemical into a giant centrifuge, adding some heat to turn it into a gas, and swinging it around as fast as you can. Since all of the isotopes have different masses, this will eventually create a distribution, with heavier isotopes (those with more neutrons) near the bottom. It’s the same principle as sand settling to the bottom of water in gravity, except that here gravity alone isn’t enough.

Even the immense force of the centrifuge really isn’t enough to get an appreciable amount of the necessary isotopes. You need to repeat the process with a thousand other centrifuges, all feeding forward, all enriching the uranium just a bit more, until you get a few kilograms of highly enriched uranium (you’ll have started with several dozen tons of uranium ore). The amount of energy, engineering, and time this takes is staggering.

Actually, this laborious and energy intensive process is the “easy” way of enriching uranium. Prior to the invention of centrifuge enrichment, a process known as gaseous diffusion was used. Compared to it, centrifuges require very little energy. This has been great for the energy return on investment of civilian nuclear plants, but less good for proliferation risk.

This process may eventually become even easier, due to technological advances like laser enrichment. But for now, you have to put in this work if you want uranium for your bomb.

Again, all of this is necessary for uranium bombs because naturally occurring uranium is depleted. To make nuclear weapons, you need highly unstable uranium. But all the uranium on earth was created billions of years ago, in the hearts of now long-dead stars. Over time, more and more of the unstable uranium has broken down, leaving mostly the more-stable isotopes.

2 billion years ago, fission could occur in uranium deposits. That’s now impossible.

When people discuss enriched uranium, they frequently talk about enrichment percentage. This is the relative mass of the fissile isotopes compared to the total mass of the material. For most power plants, an enrichment of 3-5% is used. For certain experimental reactors, including those that produce the radioisotopes necessary for medicine, 15-20% is more common. For nuclear weapons, the uranium must be enriched to about 80%, although higher enrichment is better. Low amounts of contaminants are necessary for nuclear weapons to function, whereas contaminants are well tolerated in nuclear reactors.

“Simple” enrichment of uranium to weapons grade costs untold millions of dollars. Of all the countries that do not currently possess nuclear weapons, only Iran, Germany, and Japan could enrich uranium up to weapons grade in a reasonable time frame. Non-governmental actors (like Al-Qaeda or Daesh) would find it essentially impossible to create their own fissile materials. The only way that they could gain access to the fuel for a nuclear weapon is if given it by another party.

3.2 Plutonium

It is much easier to acquire enough weapons-grade plutonium for a bomb than it is to acquire weapons-grade uranium. There are two reasons for this: separation is easier and the mass required to produce a plutonium bomb is lower than the mass required to produce a uranium bomb. Plutonium has a critical mass of 11kg, about five times less than that of uranium. This means that a viable plutonium bomb requires about one fifth the fissile material as a viable uranium bomb.

Weapons-grade plutonium starts with uranium. Like I said before, the plutonium-239 used in bombs no longer exists in nature.

To get plutonium from uranium, you need a nuclear reactor that’s running on uranium. This means that even when a country builds a bomb using plutonium, they must enrich some uranium. This presents all of the challenges outlined above, partially allayed by the fact that the enrichment percentage required is not very high, merely a few percent.

Inside any uranium fuelled nuclear reactor, some U-238 atoms will absorb neutrons, becoming U-239. This is a very unstable isotope (half-life: 23 minutes), which tends to briefly moonlight as Np-239 (half-life: ~2 days) before fulfilling its destiny and becoming Pu-239 (half-life: 24,110 years). Thus in many nuclear reactor designs, this Pu-239 will hang around in the fuel rods, occasionally decaying, fissioning, or absorbing neutrons (to form Pu-240), but largely just sitting there, waiting to be recovered.

If you want to get plutonium for a bomb, recover it you must. There are a variety of ways to do this, but all of them are much easier than centrifugation. The most common one is PUREX, which I’ll use as a representative example. In PUREX, you take the uranium fuel pellets from the reactor and dissolve them in very concentrated nitric acid, discarding anything that doesn’t dissolve. You then run a mixture of tributyl phosphate and kerosene over the acid. Any uranium and plutonium will move into the kerosene phase, while other fission products will remain in the acid. To separate the uranium from the plutonium, you run the kerosene mixture over water with ferrous sulphamate dissolved in it. The plutonium will react with the ferrous sulphamate, pick up a charge, and move into the water. The uranium will stay in the kerosene.

It’s easy to separate uranium from plutonium because they have a different number of protons. This means that they react differently with many chemicals. Extraction schemes take advantage of these different reactions to set up a scenario where the plutonium ends up in one solvent (like water in the PUREX example) and the uranium ends up in another (like kerosene). Complicated centrifuge arrangements are only necessary when dealing with different isotopes of the same element, where you can’t use tricks like this.

Weapons-grade plutonium (largely Pu-239) does have some problems with other plutonium isotopes. Remember how the critical mass for a plutonium bomb is much lower than that of a uranium bomb? This is possible because plutonium-239 is on a hair trigger where fission is concerned. This is doubly true for plutonium-240, in a way that is very problematic for weapons designers. Plutonium-240 is so prone to detonation that it often detonates early, blowing apart the bomb prematurely and leading to what’s termed a “fizzle” (this will be covered more in the next section; for now, you’ll just have to trust me that fizzles lead to uselessly weak bombs).

No one wants to separate Pu-239 from Pu-240; one of the major advantages of plutonium bombs are that you don’t needs to set up a huge centrifuge plant. To avoid having to separate the two isotopes, groups that are preparing plutonium for bombs aim to avoid the production of Pu-240 altogether. This means that any reactor optimized for producing plutonium can only run for about 90 days at a time. After uranium has been in a reactor for 90 days, the Pu-240 concentration is too high to easily build a working bomb. Despite this, the International Atomic Energy Agency runs audits on spent fuel from reactors, ensuring that no plutonium can be diverted from civilian reactors to weapons.

To economically start and stop a reactor every 90 days, you need a special reactor design. Civilian light-water nuclear reactors won’t do the trick. They have a huge pressure vessel that needs to be laboriously disassembled by specially trained divers every time they needs to get at the fuel. This is so inconvenient that civilian reactors are normally only cracked open once a year – by which point the plutonium has far too much Pu-240 to be viable in most weapons.

Reactors optimized for plutonium production allow for the rapid cycling of U-238 pellets through the reactor core. This requires quite a bit of engineering work – which has to be done from scratch unless you can find a country willing to share their schematics with you. It’s also basically impossible to hide the purpose of a reactor like this. Any IAEA inspector who sees your reactor will understand right away what you’re doing with it.

Once you have your reactor and your fuel, you have to decide how you want to run it. The shorter your cycles, the closer to your plutonium to pure Pu-239, but the more you’re going to pay per gram (in reagents, labour, wasted fuel rods, and wasted time as you cycle the reactor). The US uses >97% Pu-239 in the nuclear weapons on its submarines (Pu-240 produces a lot of gamma rays, which would be dangerous for the crews in close quarters), while its silo based weapons only use ~93% pure Pu-239.

To give an idea of the cost difference for increasing purities, the US Government helpfully lists 87% Pu-239 as $5,840 per gram, while 94% Pu-239 is $10,990 per gram. At these prices, the 3-4kg of plutonium in an atomic bomb would cost about $44 million if weapons-grade (and untold millions more if super-grade).

3.3 Next Steps

I asserted above that it is much more difficult to build a bomb out of plutonium than uranium. This is because the simplest type of nuclear weapon, the “gun” based design, does not work with plutonium. I’ll cover this in depth in the next post in the series. Here, I’m going to discuss problems common to all nuclear weapons designs because there’s significant overlap. Even with the simplest gun design, building a reliable, deliverable nuclear weapon is a significant engineering challenge. There’s no off the shelf design to copy, so you’re going to have to come up with your own solutions to problems like:

• How can I get this to detonate every single time?
• How can I make this package small enough that it is easy to deploy?
• How can I ensure this thing detonates at the correct altitude?
• How can I package this such that it doesn’t get messed up by wild swings in temperature and pressure?
• How can I mount this on a missile?
• How can I keep that missile from burning up on reentry? (This is one of the last things preventing North Korea from having a weapon system capable of targeting the USA)

Eventually, you’re going to have to test out your design. This means you’re going to have to set up specialized test chambers underground. You won’t want to detonate a bomb in the atmosphere because a) this is super-duper illegal under international law and b) It’s super-duper obvious to all of the UN Security council members (and any country with a functioning espionage program) that it was you who just test detonated a bomb. A combination of a) and b) means that atmospheric detonations are a one-way ticket to becoming an international pariah and all the sanctions that entails.

Underground tests are only slightly better. Seismometers will detect all but the smallest nuclear test detonations when they’re conducted underground, allowing several countries (most notably the US, with its excellent seismometer arrays) to detect and precisely locate the test detonation and estimate the yield. On the bright side, if you have the good sense to conduct your tests underground, you’ll probably avoid crippling sanctions (just don’t be surprised when you can no longer buy anything that might be used in a missile from any other country).

While it’s probably possible for a sufficiently advanced country to build a very simple uranium-based nuclear weapon and have it function with some degree of reliability without testing, any design that uses plutonium, as well as all of the largest, most complicated weapons systems (like the Teller-Ulam design) require either testing, or advanced simulations packages seeded with top-secret data from past tests. Without these tests, your weapon has a very low chance of actually working in the field.

This doesn’t even get into the colossal amount of effort required to develop and test all of the technology necessary to successfully deliver a warhead to a target. Stealth bombers, advanced missiles with MIRV technology, and silently running nuclear armed submarines are all necessary for a country to have full nuclear capability and all of these are impossible to develop in secret.

In summary, there are 6 things a country or non-state actor most do to develop a full, modern nuclear weapons program:

1. Enrich or acquire a suitable supply of fuel
2. Design a weapon to make use of that fuel
3. Miniaturize the weapon
4. Design robust (capable of surviving an initial nuclear strike) delivery systems for the weapons
5. Design and test boosted weapons and fusion weapons
6. Tests these systems, individually and as a whole

There are currently 4 or 5 countries that have completed these steps: China, Russia, America, and India (Israel keeps so much of its nuclear program secret that no one knows if it’s completed all these steps or not). Pakistan, the UK, France, and North Korea have checked off some, but not all boxes on this list, leaving them vulnerable to pre-emptive nuclear attack.

Even if checking off all the boxes is hard, there is a significant risk whenever a country or non-state group checks off any of them. Enriched materials can be used to make dirty bombs, conventional explosive devices that disperse highly radioactive materials (as an area denial or terror tactic) even if a true nuclear device cannot be made. Potentially city destroying nukes require relatively little fissile material or technical expertise. Untested weapons can still work some of the time. The UK, France, and Pakistan all possess a lot of warheads and their inability to protect themselves against a pre-emptive nuclear strike from Russia or the US doesn’t mean they can’t annihilate less powerful countries if they so choose.

All of these represent significant threats to individuals. A small nuke or a large dirty bomb can do a lot of damage if used on the correct target. But so can a conventional weapon. A MOAB or FOAB used on a sporting event would be at least as devastating as a dirty bomb. It is only large arsenals of modern thermonuclear weapons, possessed by countries in numbers large enough that some can reasonably penetrate any countermeasure that represent a true existential threat to humanity. Furthermore, all groups or countries that haven’t completed every single item on my list are vulnerable to having their whole nuclear program destroyed by an adversary before they can use it at all; this severely limits their ability to make threats.

Because of the difficulties inherent in proliferation, the greatest gains to be made in protecting humanity from nuclear weapons must necessarily come from convincing existing nuclear powers to decrease their stocks of weapons or delivery vehicles. This is especially true of Russia and the US, which together possess more than 93% of the world’s nuclear weapons. The challenge will be to do this while maintaining a credible deterrent against any rogue nation that wishes to develop a nuclear weapons program of its own.

Additional Reading: NPT (Full Text), Enriched Uranium, “Weapons Grade” Nuclear Material, Plutonium-239, Nuclear Reprocessing, Proliferation/Breakout Capability, Dirty Bomb, and Nuclear Weapons Testing

For this to all make sense, we should start with a brief review of atomic theory.

All matter is made up of atoms. Atoms have an outer shell of negatively charged electrons (more accurate descriptions exist, but I’m not going to delve into them; throughout this section I’m going to use simplified models wherever they’ll do the topic justice) and an inner core containing uncharged neutrons and positively charged protons.

The number of protons in an atom determines which element the atom is. All atoms with two protons are helium, all atoms with six protons are carbon, and so on. Much of the time, elements will have the same number of electrons as they have protons, so that the charges cancel each other out. Forms of elements with differing numbers of electrons are called ions. Ionization is a very common phenomenon. You observe it whenever you see lightning or dissolve salt in water.

Neutrons aren’t as simple; there’s no 1:1 correspondence or linear ratio between the number of neutrons in an atom and the number of protons. Even more confusingly, any sufficiently large sample of most elements will contain atoms that have a different number of neutrons than is most commonly observed.

More concretely, while hydrogen atoms normally have no neutrons, they sometimes have one or two neutrons. Hydrogen with one neutron is commonly called deuterium; hydrogen with two neutrons is commonly called tritium. These unique names bely the fact that regardless of the number of neutrons, the elemental classification of the atom is hydrogen; in almost all cases, the atoms behave identically, regardless of number of neutrons. A deuterium atom, hydrogen atom, and oxygen atom could come together to form what would essentially be water. In fact, you drink water in this configuration every single day!

Different forms of an element separated by the number of neutrons are called isotopes. Normally, chemists assume that isotopes will be in their naturally occurring proportions, which are heavily biased towards stable isotopes. If a specific isotope is being referred to, then it will be referenced with the name of the element and the total number of protons and neutrons; for example, uranium-235.

Isotopes are broken up into two categories: stable isotopes and unstable isotopes. Stable isotopes are at a fundamental resting state. If not acted upon by external forces, they will never change form (to a reasonable first approximation). Unstable isotopes are not at this fundamental ground state and will eventually return to it. The process of returning to the ground state radiates energy. For this reason, unstable isotopes are also sometimes called radioisotopes.

Unstable isotopes also have a characteristic half-life – the amount of time necessary for half of the element to break down into other elements through decay. Elements with a shorter half-life are more unstable, emit more radiation each second, and break down more quickly. Elements with a longer half-life are more stable, emit less radiation each second, but also persist much longer.

The energy released when an unstable isotope reverts to the ground state is commonly termed radiation – the same radiation produced by nuclear weapons, nuclear reactors, and nuclear waste. It mainly comes in four flavours: α-particles, β-particles, γ-rays, and free neutrons.

α-particles are high energy helium nuclei. Because they have a relatively high mass, they tend not to travel far and are unable to penetrate obstacles. A single sheet of paper can stop anα-particle – but their danger shouldn’t be underestimated. An unstable isotope of polonium that decayed and producedα-particles was used in the murder of Alexander Litvinenko.

β-particles are high energy electrons. Since they have less mass and interact less with matter, they can travel much further than α-particles. It takes a few millimetres of aluminum to stop a β-particle.

γ-rays are what many people think of when they think of radiation.γ-rays are photons, the same as radio waves, microwaves, and light. γ-rays have much more energy than more innocuous photons, which causes them to have much smaller wavelengths. Here’s a good rule of thumb about photons: they interact with and can be intercepted by things about the size of their wavelength. TV antennas give you a clue as to the size of radio waves, the mesh or your microwave the size of microwaves, and the tiny rods and cones in your eyes are sized just right for visible light.

γ-rays are much smaller than visible light – they’re sized just right for electrons. This means that they can travel very far, as electrons are very small and any individualγ-ray is unlikely to hit one. Once a γ-ray does hit an electron, it will transfer most of its energy to it, ionizing the atom. This can be very dangerous to humans, killing cells if enough ions are created in them and damaging DNA even when the cells survive. To block out γ-rays, you need to put a lot of electrons between you and them. Four metres of water, two metres of concrete, or 30 centimetres of lead will do the trick.

Breakdown by ejecting free neutrons is comparatively rare. It generally only occurs in significant amounts in the isotopes uranium-235 and plutonium-239.

Any ejected neutron will eventually hit another atom. When it collides, it can cause the atom to immediately fission, releasing more neutrons or be captured by the atom. Whether capture or fission occurs depends on the energy (read: speed) of the neutron. If the neutron is moving at the correct speed, fission will occur. Otherwise the neutron will be captured or bounce off. If a stable isotope captures a neutron, the result is almost always an unstable isotope. Therefore, neutron radiation is the only kind of radiation that can make other substances radioactive.

The observable characteristics of a fission reaction depend on how many neutrons are released (on average) in each collision and subsequent breakup or absorption; none will be immediately released in an absorption, but multiple can be released in a breakup, giving a wide range of possible average values. In nuclear power plants, on average only one neutron is created by each emitted neutron. This causes a slow and steady “burn” of the uranium or plutonium fuel, producing heat that can be harnessed for power. As long as each neutron produces at least one more neutron, the result is a nuclear chain reaction.

In nuclear fission weapons, each emitted neutron generates multiple new neutrons. This quickly leads to a large proportion of the fuel being consumed and turned into absolutely colossal amounts of energy. The smallest tactical nuclear weapons are equivalent to the detonation of dozens of tonnes of TNT, the largest equivalent to the detonation of millions of tonnes. This is where phrases like kilotons (kt – equivalent to 1,000 tonnes of TNT) or megatons (Mt – equivalent to 1,000,000 tonnes of TNT) come from.

The mass at which this out of control reaction takes place is called critical mass. Beyond the critical mass, the reaction is supercritical. Critical mass varies with purity (how great a percentage of the isotopes are the fissionable ones) and shape. A sphere is the shape with the lowest critical mass. This is because in a sphere, the greatest percentage of emitted neutrons are emitted back into the bulk of material – a sphere minimizes the surface area for any given volume. The critical mass of plutonium-239 is 11kg (equivalent to a 10cm sphere), while uranium-235 has a critical mass of 56kg (equivalent to a 17cm sphere).

Criticality can occur with smaller amounts of isotopes than the critical mass in certain cases. Critical mass is tied to density. If you increase the density, you decrease the critical mass by making collisions more likely. You can also lower the critical mass by employing a tamper, a layer around the core that reflects escaping neutrons back towards it, or a moderator (like water) that reduces the speed of the fastest neutrons to one more optimal for sparking further fission.

There is one other type of nuclear weapon. Fusion weapons push atomic nuclei together to form new, heavier nuclei. Fission/fusion weapons aren’t an either/or proposition though. Many nuclear weapon designs incorporate multiple stages. For example, some designs will use the energy from fission to start a fusion reaction and get most of their power from this. Other designs make use of a small amount of fusion to release extra neutrons, which allows more of their fuel to be consumed.

Additional Reading: Proton, Neutron, Radioactive Decay, Nuclear Chain Reaction, Electromagnetic Radiation, and Isotope

With President Trump in possession of the nuclear launch codes, I have a feeling that many people who’ve neglected nuclear weapons as an important cause area may begin to sit up and take notice. This is a good thing. There currently exist basically no checks and balances on a US President’s ability to go to nuclear war. Harold Hering was cashiered from the Air Force in 1973 after asking (on the subject of nuclear weapons launch) “How can I know that an order I receive to launch my missiles came from a sane president?”. Nothing has changed since then.

This post series is meant as a non-exhaustive primer on the (declassified) physical and strategic realities of nuclear weapons. It’s supposed to get you up to the point where you can begin asking the right questions in a relatively short time period. If you want more information, I’ve included relevant links (mainly to Wikipedia) at the end of every section. The quality of Wikipedia does vary when it comes to nuclear weapons, so take what it says with some salt.

One final caution: everything in this post series could be obsolete and we would have no real way of knowing. Nations reserve their highest level of classification for their nuclear capabilities and plans. Some of these plans and capabilities can be inferred from what we know of the physics of nuclear weapons or by looking for equilibriums between decision makers, but a lot is hidden from public view.

Main Series

1.0 Introduction
2.0 Basic Science
3.0 Proliferation
    3.1 Uranium
    3.2 Plutonium
    3.3 Next Steps
4.0 Nuclear Weapon Design
    4.1 Gun Assembly Design
    4.2 Implosion Design
    4.3 Miniaturizing Weapons
    4.4 Boosted Weapons
    4.5 Alarm Clock/Sloika
    4.6 Teller-Ulam Design
5.0 Nuclear Weapon Effects
    5.1 Scaling
    5.2 Direct Effects
        5.2.1 Fireball
        5.2.2 Shockwave
        5.2.3 Direct Radiation
    5.3 Indirect Effects
6.0 Nuclear Delivery Mechanisms
    6.1 Bombers
    6.2 ICMBs
    6.3 Submarine Launched Ballistic Missiles
7.0 Nuclear Weapon Strategy
    7.1 Tactical and Strategic Weapons
    7.2 First Strike, Second Strike, Counterforce, Countervalue
    7.3 Mutually Assured Destruction
    7.4 The Nuclear Triad
    7.5 Current Nuclear Strategy
        7.5.1 Russia
        7.5.2 China
        7.5.3 India and Pakistan
        7.5.4 UK and France
        7.5.5 North Korea
        7.5.6 Israel
        7.5.7 Iran
        7.5.8 The United States of America
8.0 High Value Anti-Nuclear Activism
    8.1 Tactical Weapons
    8.2 Arms Reduction Treaties
    8.3 Anti-Ballistic Missiles
    8.4 Donations
 

Special Topics

Laser Enrichment
    Laser Enrichment – How It Works
    Laser Enrichment – Proliferation
North Korean Nuclear Weapons Program FAQ
The Graph Model of Conflict Resolution As Applied to the Current North Korea Crisis
    The Graph Model of Conflict Resolution Sensitivity Analysis
The (Nuclear) International Monitoring System

I’ve also published an introductory presentation: Nuclear Weapons Explained Backwards

Changelog

As I receive feedback, I intend to update this post series. All major changes (e.g. not copy editing) will be posted here.

February 5, 2017

• In 3.0 Proliferation, clarified that the design of a nuclear weapon is at least as challenging as enrichment, if not more so.
• In 7.5.5 North Korea, clarified my stance on North Korean fizzles and underlined the danger that even relatively small weapons pose.

February 12, 2017

• Added the first “Special Topics” post: Laser Enrichment.
• In 3.0 Proliferation, added a reference and link to laser enrichment.

April 1, 2017

• Significant rewrites to the first part of section 3.0 Proliferation.
• Added a new section to the Proliferation post, covering Weapons Grade Plutonium.

December 19, 2017

• Added link to the North Korea Nuclear Weapons FAQ
• Added links to my two relevant GMCR posts
• Added link to Nuclear Weapons Explained Backwards

April 14, 2019

• Added link to my post on the International Monitoring System