Would You Like to Play a Game? - By A. Nonymous

Would You Like to Play a Game?
A Primer on National Nuclear Strategy, and Individual Survival
By A. Nonymous

Few subjects have been less understood in the last century than global thermonuclear war.  Pop culture and journalism have depicted it as ending human civilization at the least, and all life on Earth at the worst.  In doing so, they have propagated myths and inaccuracies that have actually reduced the chances of individuals surviving.

Nuclear war is a Bad Thing(tm).  That much should be obvious.  But, the details matter.  What would happen to you, as an individual, and your friends and family, would depend on a variety of factors, many of which you actually have considerable control over.

Part One - A Strange Game; The Only Winning Move Is Not to Play

First, let's address one of the biggest elephants in the room:  MAD.  Mutual Assured Destruction has never been an official policy of the United States (or in any records recovered after the fall of the Soviet Union).  Early on in the Cold War, WWII thinking regarding strategic bombardment still held considerable sway, with factories and raw resources considered valid targets.  However, as the number of warheads on each side grew into the tens of thousands, it began to make less sense to target the enemy's cities, because there was now something much more important to shoot at--the enemy's warheads themselves.

Let's play a game of Global Thermonuclear War.  It can be set at almost any time period, but let's just say 1988, as the Soviet Union was beginning its (mostly unseen and unpredicted by Western sources) collapse.  As its chances of world domination began to falter, the temptation was surely there to use its sole trump card to wrest out a victory while it still had the chance.  There was quite a fear in the West of exactly this scenario coming to pass.

You can play as the US, and I'll play as the USSR.  Now, while there are no hard and fast rules to this game, there are loose "rules" based upon each side's desired outcomes.  Let's list the first few priorities:

1.  Keep my country "alive".  My nation must continue to exist as a going concern after the shooting stops.  Otherwise, what's the point?  This means that I must make choices that give my people, my economy, and at least a basic level of government the best chances to survive.

2. Remove the threat from the enemy.  This actually derives from the first point; the enemy has the power to "kill" my country, and the most certain way to ensure that they cannot is to remove that power from their hands altogether.

3. Maximize my country's ability not just to survive, but to thrive.  This includes minimizing damage received from the enemy, protecting the economy and society, and any number of "softer" goals.

The first priority is why deterrence works; the best way to guarantee my nation's continued survival is to not play the game in the first place.  "Not to play" is not the only winning move, but it is the most certain.

The second priority, combined with the first, is why shooting at your cities is just about the last thing that I would ever want to do.  You see, your cities do not directly threaten my country's existence--but your warheads do.

Given this, we can derive a few of those loose "rules":

1.  Whoever runs out of warheads first (or loses the ability to issue timely commands to the warheads they have left) loses completely.  The "winner" (in quotes, because "winning" in this game is somewhat relative) gets to use their remaining warheads to threaten the loser's population and economy with utter devastation.  The loser is obligated to  accept terrible peace terms, or face annihilation.

2. Therefore, if one side believes that it is likely to lose, it immediately comes under increasing pressure to negotiate a peace at something less than a complete loss.  In 1988, this likely would have included removing the loser's troops from the immediate theater of conflict (roughly, from France to Poland), and might have included painful restrictions on the loser's ability to reconstitute a powerful nuclear force (together with a painful inspection regime to ensure no cheating on the loser's side).

3.  The side that believes it is likely to "win" (that is, not lose completely) also faces pressure to negotiate a peace--because the losing side, as long as it has not lost completely, retains the option to flip over the table, invoke MAD, and take the "winner" down with them.  Therefore, unless the losing side plays the game so poorly that it loses completely, the winner is not going to get to fly their flag over the other's capitol, much less demand unconditional surrender; not as long as their first priority is to keep their country "alive".

Thus, even a large nuclear exchange is unlikely to result in the wholesale destruction of cities, because doing so a) reduces the warheads that each side has left to take out the other side's warheads, and b) makes it all the more likely that the other side will flip over the table.  A brutal calculus, but one into which both sides invested massive amounts of brainpower during the Cold War.

So, what does this all mean in practical terms?  As the USSR, my primary targets are not your cities, your refineries, your power plants (particularly risky targets, in terms of table-flipping potential), or really anything except your nuclear forces.

What are your nuclear forces?  Your active warheads, ready to be used.  Your delivery systems (primarily ICBMs, SLBMs, and nuclear-capable bombers).  The communication nodes that you use to transmit orders from NCA (National Command Authority, i.e., the President and senior generals who can in certain circumstances issue launch orders).  And, the NCA itself.

Those are my targets.  Nothing else matters, or rather, everything else pales in importance.  Every warhead that I waste on a target other than your nuclear forces slightly increases the risk that I will lose, and given the stakes of this game--my nation's very survival--I cannot afford to take that risk.

Part Two - The Cold (War) Equations

So, I simply aim one of my warheads at each of your warheads, and they all cancel out, right?  Wrong.  I may need to use several--even a dozen or more--warheads on each target that I select.  Why?  Murphy's Law and Vegas odds.

There is a military concept, used in everything from tanks to ICBMs, called P(k), or Probability of a Kill.  This is a best-guess estimate of the ability of one thing to eliminate the threat posed by another thing.  For example, let's take a Minuteman silo in Montana.  It contains a missile that could destroy one or more of my warheads (or cut a hole a few miles in diameter out of one of my cities, if you were to flip the table).  I need to destroy it, but warheads are not magic wands that you can wave at the enemy and make things disappear.  I need a lot of steps to go exactly right in order to destroy the missile sitting in your silo.

Let's work backwards from the target.  My warhead has to explode correctly in order to produce the blast overpressure (a nuke's primary means of destroying something, especially a hardened target) needed to punch through the reinforced concrete cap on top of the silo and then crush the missile inside.  Any given warhead should in theory go off correctly every time, but there's a small chance it won't.  Maybe the technicians who assembled that particular warhead made a mistake; perhaps one of the components has failed due to age (the nuclear material in warheads produces a tiny amount of radiation that can damage the circuit boards over the course of many years) and got missed by the last inspection; maybe the design of the warhead itself was flawed (this actually happened to the US).  There are a number of things that could go wrong, and the official estimates for how likely they are to happen is, obviously, classified.  For this game, let's give it a 95% chance of producing sufficient overpressure to do the job.

But, then, the warhead has to detonate close enough to the target that sufficient overpressure is actually delivered to the target.  If it misses by even half a mile, it might not be able to crack that concrete lid over the silo.  Hitting within half a mile of a target after flying halfway around the world is a lot harder than it sounds, and the remarkable thing is that it's possible at all.  Let's give that another 95% chance of working, since the USSR partially made up for their historically lower accuracy by using larger warheads (which works to an extent, but is a less efficient solution).

But, there's a chance that the warhead will fail to separate from the bus (the unpowered stage that the warheads ride into space) correctly.  Another 95%.

There's also a chance that the missile will go off course, sending all of the warheads it carries astray.  95%.

Or the warhead, or the bus, or even the missile could be intercepted by a defensive missile.  95%, although in 1988, it was 100%, because the US hadn't developed modern missile defenses (Russia has had defensive missiles around Moscow for decades; they make up for lower accuracy by using nuclear warheads  While it would be somewhat inaccurate to say that they plan to "carpet-bomb" their own sky, that does convey the general idea).

Or the missile could fail to launch (this turned out to actually be a big deal with Soviet missiles, despite their considerable experience in building rockets).  95%.

Or (and here's the big question mark), the missile could get destroyed by one of your warheads before I can make the decision to use it, issue the launch order, have the order transmitted to the people who physically turn the keys to launch the missile, and for them to launch it.  95%, but in truth, this one is the least predictable variables, because we're both aiming for each other's delivery systems at the same time, and the time needed to make a decision, transmit it to the delivery system, and launch, is about the same as the flight time of most missiles (and just reaching that rough equivalency required a massive technological and organizational effort, and still forced both sides to pare the available decision time down to the bone).

To calculate the final probability of success, or P(k), you have to multiply the odds of each step working correctly by each other.  So, .95*.95*.95*.95*.95*.95*.95=69.8%.  A 70% chance for each and every one of your ICBMs that I absolute, positively must destroy in order to protect my nation.  That's not going to cut it.

What do I do?  Well, the logical answer is to use a second warhead.  But, wait!  It can't be from the same missile, or any of the failures that affect the bus or the missile will cause the second warhead to fail, as well.  And even with a second warhead from a second missile, I still face a P(k) that is just over 90%.  Can I gamble with my people's lives on a 1 in 10 chance of failure?  Would you?

So, now I'm using 3 (and possibly more) warheads, each from a different missile, against a single one of your missile silos (nuclear planning is HARD!).  I'm going to run low on warheads in a hurry, just shooting at your missiles.  I definitely can't afford to hit every military base, power plant, dam, or city--because my top priority is to protect mine from your warheads!

Part Three - It's the End of the Cold War as We Knew It (and I Feel Fine)

Now, we've been playing in 1988, where both sides had tens of thousands of warheads.  What about today?  The current treaty is New START, which limits each side to ~1500 strategic warheads (it also counts a nuclear-capable bomber as a strategic warhead, while not restricting the number of tactical warheads, so it's a little more complicated than it sounds).  The US keeps just over 400 land-based Minuteman III ICBMs, each with a single warhead, and slightly more in the way of tactical bombs for use by bombers.  Over half of all US warheads are carried by 14 submarines, of which several are at sea at any given moment.  Russia, on the other hand, focuses most of its warheads on ICBMs, split roughly evenly between silos and very, very large custom-built trucks, with a smaller number of submarines and bombers.

The US decision to keep as many Minutemen as possible, and only carry one warhead apiece, has been derided by journalists and cost-cutters at times; there have been calls to decommission the ICBMs, which are not considered to be as survivable in a shooting war as submarines.  However, as shown above, the Minuteman fleet serves as an incredible warhead magnet, forcing opposing nuclear planners to expend hundreds, perhaps even a thousand or more, warheads on fewer than 500 targets that are for the most part located in rural areas, a thousand miles away from the densely-packed cities east of the Mississippi.  Since those warheads could not be used against submarines (just the sub bases, one on each coast, which would certainly be hit hard, and repeatedly, just to be sure), the Minuteman fleet in effect protects 500 or more lesser targets, many of which would coincidentally be located in or near major cities.  It is entirely valid to state that they serve as decoys, which cannot be ignored, and must be struck, as part of any full-scale attack.  And they serve this role whether they are launched in time or caught with their missiles still in the silos.  Consider them a form of ablative armor, cheap at the price.

Part Four - A Question of Priorities

Now, to this point, we've just focused on hitting each other's warheads.  But, crippling the other side's ability to make decisions and transmit valid launch orders to the delivery systems a) buys more time to catch warheads on the ground, and b) increases the chance that the other side will begin to lose so badly, and so suddenly, that they become afraid of losing completely, and agree to negotiate at a substantial disadvantage.  Unfortunately, most of these targets, at least in the US, are located in or near cities.

If you happen to live in the Greater D.C. Area, you are in considerable danger.  Between the White House, Pentagon, Andrews AFB (where Air Force One usually sits), and several other nearby targets, you will almost certainly get hit, and probably get hit harder than any other major metropolitan area.  There's only so much that you can do; if one of the dozens of warheads aimed at the region happens to miss its target by a few miles, and lands too close to you, there's not much that you can do about it; at the very least, you would need a proper bomb shelter (not just a fallout shelter; more on those later), and enough warning to get inside in time.

Beyond D.C., strategic nuclear targets are not always what you might think they are.  Texas is an excellent example: home to 2 of the Army's 10 divisions (and the heaviest, most powerful ones, at that), a bomber base, two nuclear power plants, major factories for the V-22 and F-35, and the primary training base for F-16 pilots, plus the largest collection of refineries and chemical plants in the country, and the only factory that assembles and disassembles nuclear warheads (it doesn't make any of the parts, it just puts them together). While target lists are a major secret on both sides, civilian nuclear experts (there are a few out there, and most work for think tanks) suggest that there are exactly two strategic nuclear targets in the entire state.

1.  The AT&T network interchange in downtown Dallas
2.  The FEMA Regional Headquarters in nearby Denton

That's it.  The bomber base?  It's home to half the B-1 fleet, and the B-1s have been denuclearized (and Russian inspectors got to verify it) for over a decade (the introduction of the nuclear-capable B-21 Raider will put it back onto the target list).  Without that capability, they fall below everything still nuclear-related on the priority list.  The power plants?  They can't reliably produce bomb fuel (fuel rods in power plants do produce small amounts of Pu-239, but it's contaminated with Pu-240 and other things that make it a serious hassle.  Reactors used to make bomb fuel are designed differently from power plants; that's how we know, for example, that Iran and North Korea are blantantly lying about wanting to "make power"--the facilities they built were terrible designs for making electricity, but excellent for making bomb fuel).  The refineries?  If MAD was the goal, then sure, but if MAD is in play, then both sides have already lost, and neither wants to lose.  Even the Army bases aren't strategic targets; it would take weeks for those divisions to be shipped across the oceans, and a nuclear war is typically measured somewhere between hours and at most a few days.  Those tanks are irrelevant in the current situation.  FInally, the warhead factory might be a low-priority target to improve negotiating positions, but it cannot build or deliver new warheads soon enough to make any difference in a shooting war.

But, why those two targets?  What do AT&T and FEMA have to do with nuclear war?

Quite a bit, actually.  Both double as strategic communications nodes, which are used to transmit valid launch orders from NCA to the delivery systems.  You see, the President can't tweet launch orders, or call the guys sitting at the keys in a missile base up on his cell phone.  The launch command process is built on a complex series of very deliberate methods, in order to prevent a saboteur (or accident) from causing missiles to be launched when they shouldn't be--or to be held back when they are supposed to launch immediately.  The safes, one-time authentication cards, all of those steps that you see in some of the old movies?  Those are actually fairly accurate, albeit just a small part of the command chain that has been hammered out over the course of decades to ensure that missiles are launched when and where they are ordered, and never, ever, at any other time.

So, there are certain "pathways" that launch orders can take, and none of them involve the civilian side of the Internet.  The downside, of course, is that it is therefore possible for enough key communications nodes to be taken out early in a nuclear exchange to delay or disrupt the transmission of those launch orders.  So, Dallas would take it on the chin.  Places like Offut AFB in Nebraska (home of STRATCOM) would get it even worse.  Cheyenne Mountain would probably get hit just in case parts of it are still in use at the time; nearby Colorado Springs would get pasted almost as badly as Offut.

But, most US cities would not be targeted at all; to do so would be a waste of irreplaceable warheads, at a time when making the best use of them is of the utmost importance.

Part Five - Surviving the Explosion

A nuclear detonation produces three lethal components: heat, radiation, and pressure.  Of these, pressure causes almost all of the damage in almost every circumstance.  The heat and radiation attenuate rapidly over distance, while the impulse wave from the pressure (which is created by adding a ridiculous amount of heat into a relatively small area, and letting Earth's atmosphere do the work--just like an old depth charge used the water around it as a weapon) propagates over a far larger distance.  This can be shown on the helpful NukeMap website: as warhead size increases, the lethal blast radius increases as well; however, the lethal ranges of the heat and radiation increase at a far, far slower rate.  An explosion equivalent to 10,000 tons of TNT produces dangerous ranges for each component that are close together, while a 100-kiloton explosion produces far more blast damage than heat or radiation.

This leads to a serious misunderstanding.  The first warheads ever built produced explosions equivalent to 22,000 tons of TNT, 15,000, and 21,000.  These were crude devices by modern standards, and far "smaller" than any strategic warhead today.  One result of the smaller explosions was that a number of Japanese soldiers and civilians successfully survived the blast, only to be overcome in the following days and weeks by the radiation doses that they had received in the same instant.  However, with a modern weapon, if you are close enough to receive a dangerous dose of radiation from the explosion, you are unlikely to survive long enough to even notice it.  Outside of edge cases--if you happen to be near a blast, and are inside a structure that can withstand the immense, building-flattening pressure, but the same structure is not thick or dense enough to withstand the radiation--if you are far enough away to survive the pressure wave, you are far enough away to survive the heat and radiation.

If you live within a mile of a strategic target, your only chance of survival (other than not being there when it gets hit) is a full-fledged bomb shelter that protects against all three components of a detonation.  If you live within twenty miles, you still run the risk that a warhead will happen to miss its target--in your direction.  It's not much of a risk these days, but it's still a roll of the dice.  If, however, you, like the vast majority of the population, live more than twenty miles away from the nearest strategic target (again, thanks in no small part to the Minuteman silos hogging most of the attention), then you are not going to die from the direct effects of a nuclear war.

Part Six - Fallout

Fallout has been granted almost magical powers by the press and pop culture; it can kill you hundreds of years later, it poisons the soil for thousands of years, it can even wipe out all life on Earth (On the Beach).  None of this is true.

Fallout is radioactive matter left over from a nuclear detonation.  There are only two sources of fallout: atoms that were part of the warhead and blown clear before those particular atoms could explode, and atoms from certain elements that, upon being struck by a neutron released by the blast, become unstable and eventually decay into another element, releasing radiation in the process.  This is called neutron activation.

Imagine, for a moment, that a warhead is filled with raw popcorn.  Each kernel represents an atom of uranium or plutonium (or the small amounts of tritium and other bomb components that are either naturally radioactive or can become radioactive if hit by neutrons).  When the warhead detonates, much of the popcorn "pops" in one sudden burst.  However, many of the kernels are hurled away by the force of the explosion beore they can pop; these kernels are hot enough to pop, but will not pop on their own for a certain period of time (this is where the term "half-life" comes into play; any given kernel will pop at a seemingly random time, but over any measurable amount of kernels, they always average out to an extremely exact rate that varies only by the specific isotope involved).

If the detonation occurs high enough that the fireball does not touch the ground, then a) very few new kernels are created by neutron activation, and b) all unpopped kernels are thrown into the stratosphere and dispersed to such a degree that when they come down again, they are likely to have already popped, and be in such low densities as to not be a serious hazard to health.  Airbursts (used to maximize the range of the pressure wave against targets that are not buried or hardened) are actually considered "self-cleaning" by the military, because they produce so little dangerous fallout.

Ground bursts, on the other hand (used to maximize the heat, radiation, and pressure around a single point, mostly for targets that are buried or heavily hardened, like a missile silo or Cheyenne Mountain), produce significant fallout.  Many of the kernels, whether bomb fuel or created by neutron activation, get swept up with dust and other particulate matter, which keeps most of the kernels from getting all the way up to the stratosphere.  These kernels hang around in the lowest parts of the atmosphere, where weather systems (particularly rain) can cause them to "fall out" of the sky before most of them can pop.

One thing that is often confused with fallout is gamma radiation.  There are three types of ionizing (dangerous) radiation produced by a warhead: alpha, beta, and gamma.  Gamma is the nasty one.  To block enough gamma radiation to stay safe, you need 1" of lead, or 1' of reinforced concrete, or 3' of packed earth, or 8' of water, or a few miles of air.  However, almost all of the gamma radiation produced by a warhead is prompt radiation--created in the first few microseconds of the explosion.  Fallout is mostly alpha and beta radiation.  Alpha can be blocked by your skin, and beta by clothes.

The real danger from fallout is if you ingest it.  Note that when Russia murdered the defector with Polonium in his coffee, the people around him at the restaurant were not harmed in any way, but there was no way to save him once he had swallowed it.  Alpha and beta particles may not be able to penetrate your clothes, but once inside of you, they can wreak absolute havoc on your system.  Fallout's greatest danger comes from simply breathing it in, followed by eating food or drinking water that contain fallout.  Fortunately, avoiding this is relatively easy: don't eat, drink, or breathe in fallout particles.  FEMA actually recommends keeping transparent plastic (industrial-grade, not Saran Wrap) and duct tape around to seal up your windows (fresh air can still seep in through cracks, but fallout is unlikely to).  Tarps will do in a pinch; anything that prevents the wind from blowing fine dust into your house.
Fallout is also only temporary.  Remember the popcorn analogy: every kernel that pops, cannot pop again (there are decay chains, where a single atom can "pop" into an element that is still unstable, and thus "pop" again, but the point remains that each atom can only do each type of "pop" one time, and can never "pop" in the same way again; eventually, the atom decays into a stable element and becomes inert).  And, remember that while any given atom may seem to "pop" at random, when taken as a whole, the rate is always the same.  This has allowed scientists to develop the 7/10 rule:  for every 7 times as long, fallout radiation decreases to a tenth of what it had been.  So, 7 hours after a detonation, fallout is 10% as potent as it began.  7 times after that (~2 days), it is 1% as potent.  And 7 times after that (just over 2 weeks), it is 0.1% as potent, and it is considered safe to go outside wearing normal clothes.  So, the primary means of survival should be to find a place to hole up for 2 weeks, and make it... if not necessarily airtight (unless you have a pump or other means of replacing the air), then close enough that fine particles will not blow inside, where they could be inhaled or swallowed.

Part Seven - The Real Danger, Hiding In Plain Sight

None of this should be construed to imply that a nuclear war would be easily survivable.  Rather, it should be understood that the threat is not the explosion itself, or even the wildly-fearmongered fallout.  It is entirely possible that over 90% of the US population could survive the first day, even the first two weeks, after a war, only for 90% of the survivors to perish in the following year.  The culprits would not be heat, blast, or even radiation--they would be old-fashioned starvation, deprivation, disease, and violence over resources (the other three of the Four Horsemen, it should be noted).  It is also possible for there to be no additional deaths after the fallout threat ends.  The key question is infrastructure damage; in particular, the power grid.

Only a bare handful of households in the US are self-sufficient, even over a relatively short timeframe (and almost none are self-sufficient over a lifetime; even the Amish depend upon trade with each other, and to an extent the rest of the world).  Modern transportation systems have allowed things like food to be shipped when and where they are needed, reducing a tremendous amount of waste due to things like spoilage.  The downside is that cities and suburbs are utterly dependent upon these shipments continuing indefinitely.  Take them away, and the cities will literally begin to starve to death within a month.  And the shipments are dependent upon every single step in a lengthy process, from seeds being grown and shipped to farmers, to the farmers with their massive machines that multiply their work by thousands of times, to factories (industrial-grade bakeries, slaughterhouses, etc.), to trucking and rail companies, to warehousers and retailers who actually distribute the food to individual people.  Break that chain in such a way that it cannot be repaired quickly enough, and people will starve.

The fastest, surest way to break that chain--and do far worse, and more immediate damage, as well--is to take out the power grids.  As long as the damage is localized, or even contained to a manageable region, emergency services, volunteers, and ultimately the free market can work around the damage until it can be repaired.  There have been a number of multi-state blackouts in recent decades, from several different causes.  All were resolved with minimal loss of life, even in harsh winter conditions, because the economy outside of the affected region was intact and could make the changes necessary to adjust to the situation.

This solution falls apart if there is no unaffected region left to send help from.

There is actually no single power grid that services the entire country.  Instead, there are three: EAST, WEST, and ERCOT (Texas--excluding the El Paso area, which is part of WEST).  There are interconnections between EAST and WEST, and minor ones between each and ERCOT, but each grid is mostly independent.  This is a good thing!  It prevents a single failure from cascading to the point where it blacks out the entire country, and guarantees that under normal circumstances, there will always be an unaffected region to come to the rescue if needed.

A large nuclear exchange, however, could take down all three grids, even if it was not deliberately intended to do so (and indeed, targeting the other side's power grids is a good way to convince them to flip over the table and switch to MAD; it is therefore not a recommended nuclear strategy).  Aside from nuclear-generated EMP, the effects of which are still hotly argued today, simply hitting too many cities at once (for their communications nodes) could bring down an entire grid.

This can be seen in many of the large regional blackouts.  All electricity must be used within a matter of seconds after being generated.  If there is too much power on a grid, circuit breakers and fuses can trip, and unprotected circuits can overheat and cause physical damage.  Too little, and electrical devices (including light bulbs) can fail.  Power plants must constantly adjust the electricity they produce as a group to meet the current demand on their grid (there is a small leeway, enough to work with at the level of a grid without causing blackouts when one person flips a light switch).  In one case, a squirrel got into a substation and managed to kill itself and take down the substation in the process.  This caused a blackout for one town, but a power spike for everybody else on the local grid.  The spike was large enough to trip the circuit breakers at nearby substations, blacking out their communities as well.  This left even more excess electricity on the grid, causing other facilities to go down in a snowball effect.  Normally, the grid managers are able to work with the power plants to adjust on the fly and keep the amount of power being generated within the safety margins of the power being demanded (this is actually a quite amazing process), but in this case, the blackout moved faster than they could compensate for it, spreading and spreading until the entire Northeastern US/Canada region had gone down, perhaps a quarter of the entire demand for the EAST regional grid.  It took painful weeks to get all of the circuit breakers reset, the fuses replaced, the lines checked, and every last community back online.  And this was after the fuses and circuit breakers had done their jobs, and protected the nigh-irreplaceable substation transformers from receiving any damage.

A nuclear war would not be so kind.  Entire neighborhoods could go offline in an instant, causing chaos at the grids.  Some power plants and substations might be hit, if they were too close to a strategic target.  Nuclear plants would probably begin controlled shutdowns, and grids most likely have emergency plans for keeping their equipment as safe as possible (an entire region can be brought partially back online in days, and fully in weeks, if the major equipment is saved--but, if it is damaged or destroyed, it can take a year or more for replacements to be built, and some of those items are no longer made in the US, which could cause further delays).  If, after the shooting stops, there is too much physical damage for two (or all three) of the grids to recover within a month, then the affected cities--and everyone living in or near them--are in mortal danger.

No power means food shipments are haphazard at best, and unlikely to be sufficient if the entire regional grid is down.  No fuel shipments, either, which exacerbates the problem, as well, as endangering future food supplies from the farmers (who in the short-run cannot produce enough food for themselves, much less anyone else, without fuel to run their expensive machines).  Worse, in the cities, it means no water or sewage, which means dehydration and disease run rampant within a matter of days.

And all of this causes fear.  Economic disruption and panic add tremendous amounts of additional stress to the system, which is already hard-pressed.  Truly bad things begin to happen at that point, as people become desperate, which makes the recovery process much harder and longer.

Part Eight - The Big Picture

So, what does this all mean?  Well, a full-out nuclear war has a very good chance of killing a massive number of people--not everybody, and it's unlikely to even end modern civilization, but it still has the potential to be horrifically bad.  The cause of most deaths will not be the bombs themselves, or the fallout they produce, but indirect effects to the economy, in particular the power grid and the supply of food, water, medicine, and other necessities to the affected populations.

This does mean that there is a cause for optimism, or rather, a way to manage the problem.  Storage and delivery of necessities is a solvable problem.  It just isn't being well-solved today, in no small part because the popular culture has convinced so much of the populace that there is no point in doing so, because the bombs and the fallout will kill everybody before a lack of necessities becomes a problem.  And this is the real tragedy: we have in our power the ability to somewhat mitigate the worst effects of a nuclear war, but we do nothing to strengthen that power, because we have become convinced that the worst effects are minor, and the lesser, more immediate effects are not only the worst, but unsurvivable.  Civil Defense has been almost entirely dismantled, and the can-do/work-the-problem mentality replaced with an apathy and resignation that will turn into blind, destructive panic if the situation (which can be caused by a number of other things in addition to nuclear war) ever occurs.

One last piece of (good?) news: this can be fixed.  It might not, given the public misunderstanding, but it can.  At the individual or small community level, prepping for power outages--particularly, for long-term ones--is a start.  At higher levels, a return of Civil Defense would go a long way, as would establishing a strategic reserve of key replacement parts like power substations.  And, keeping power generation and consumption off of the unsecured Internet would be a good idea, too; the much-proclaimed Internet Of Things has proven to be dangerously insecure, and nuclear war is not the only way that the regional power grids can be brought down.