9 Harrenhal

Can Fire Melt Stone? Take Down a Wall?

And the greatest of them, Balerion, the Black Dread, could have swallowed an aurochs whole, or even one of the hairy mammoths said to roam the cold wastes beyond the Port of Ibben.

—Tyrion Lannister, A Game of Thrones

Balerion the Black Dread, along with his kin Vhagar and Meraxes, were responsible for conquering Westeros. Balerion was said to be so large a horse could have ridden down his throat. He could fly (see chapter 7), but most importantly he could breathe fire. Harrenhal was built to be an impenetrable fortress. Unfortunately, no one saw an aerial attack coming. The dragons swooped in, and old King Harren the Black didn’t stand a chance. Harrenhal is described as having melted—not burned down, not crumbled, but melted. Now, stone seems pretty solid. We hear of 1,000-year-old churches and stone buildings from the times of gladiators; stone is not supposed to melt. But maybe it could. And if it could, was Balerion capable of that? There are a few questions to be answered here. First, could fire ever burn hot enough to melt stone? If so, would Balerion’s white-hot flames be enough? We all know Dany wants to ride her three (now two, sniff) dragons to King’s Landing and stake her claim to the Iron Throne, which was forged with Balerion’s fire. But are Drogon and Rhaegal up to the task? Balerion is described as being huge. Does size affect a dragon’s ability to produce fire? Does dragon flame develop in the same way as dragon size?

The most important question, however, is whether or not a dragon can produce fire in the first place. There are a whole lot of explanations on the internet fan forums about how a dragon might produce fire. There are also lots of arguments debunking those theories. What’s missing, however, is some math based on real-world numbers. I’m hoping here to confirm or refute as many of the dragon fire explanations as I can. I will admit from the outset that I have probably missed something. I am sure there is an obscure scientific explanation for how it could happen. However, nature and evolution are really, really good at their jobs. They gave animals giant brains, gills, prehensile tails, and cute faces that make us want to take care of them. Breathing fire is right up there with the opposable thumbs on the usefulness scale. Sure, nature also gave us the appendix, but if there were a way to have given us fire, I’m guessing nature would have figured it out.

What Is Fire?

Before understanding dragon fire, how it might be made, and how hot it can get, it’s useful to start by understanding out what fire is. This is a rather complicated question. You may remember that in 2012 the actor Alan Alda challenged scientists to come up with a good answer for it.1 When he was a child his teacher wasn’t able to explain flame, so as an adult he wanted to fix that. Fire and flame are quite hard to understand and even harder to explain, as many of the contest’s entrants can tell you. The first question many ask on the road to understanding is whether flame is a solid, a liquid, or a gas. The ancient Greeks were adamant that it was an element all its own, and they weren’t far off. Fire isn’t any of those things. It’s energy being released as heat and light during a series of chemical reactions. What you are seeing is the energy that’s released when bonds break and recombine into other things. The color of the flame depend on what atoms and molecules are involved in the reaction and the amount of energy released.

Even though we say the wood (or gasoline, or paper) is burning, it isn’t. The chemical reactions that we see as burning can’t take place when something is in solid or liquid form. The first step in burning something like wood is getting it hot enough to produce vapor that can then burn. The flashpoint of a material is the lowest temperature at which enough vapor is produced to burn. In spite of its colloquial use, the flashpoint is not the point at which something ignites—it is the point at which it has the potential to ignite. The temperature at which it actually ignites is called, fittingly, the autoignition temperature. Take gasoline, for example. Its flashpoint is very low—very, very low, in fact; around −45°F (−42.8ºC). This is why it’s still possible to light gasoline in the winter and why it’s so dangerous to store gasoline. The ignition point, however, is around 536°F (280ºC). The terms combustible and flammable are often used interchangeably, but the first means capable of burning and the second means it can burn when exposed to a flame. Whatever dragons do, it’s got to be flammable, not just combustible.2

When a fuel reaches the flashpoint and is then either brought to its autoignition temperature or comes in contact with a hot enough spark, a combustion reaction starts. A combustion reaction is an exothermic reaction, or a reaction that releases heat. Because lots of heat is being released, the temperature stays above the ignition point and the reaction continues. This is why fire is so dangerous—it’s a self-sustaining reaction. The specific reactions are different depending on the fuel being burned, but in general, the burning material is being oxidized, meaning it’s being combined with oxygen. Air contains roughly 20% oxygen in the form of two oxygen atoms bound together (O₂). During a combustion reaction, the bonds between the oxygen molecules break and the oxygen combines with atoms in the fuel, usually carbon and hydrogen. When wood is exposed to flame, for example, the oxygen molecules break apart and recombine with the carbon and hydrogen atoms present in the wood to produce water and carbon dioxide. Energy can be neither created nor destroyed, so in a chemical reaction, the system has to contain the same amount of energy when it finishes as it did when it started. There is a specific amount of chemical potential energy in the original molecules that make up wood, but there’s much less in carbon dioxide and water. So, when the combustion reaction occurs, that excess energy has to go somewhere. The fire you see is that extra energy being released in the form of heat and light. The color of the flame indicates how hot it is as well as which chemical compounds are releasing that energy. I’ll talk more about that in a bit.

This type of reaction also occurs in processes like rusting, another oxidation reaction, but in combustion it happens very quickly. To form rust, oxygen from the air breaks its weak double bond and recombines with iron. When this happens, a little bit of heat is released, but it happens so slowly that it usually goes unnoticed because there’s not enough heat to reach the flashpoint. If you soak untreated steel wool in lemon juice to get rid of its protective coating and then throw it in a ziplock bag, you can see it rust very quickly and feel the temperature of the ziplock increase. The reaction still isn’t happening fast enough to be called fire. If you touch a 9-volt battery to very fine, untreated steel wool (#00, available at any hardware store), however, the reaction happens very quickly and causes a fire. This is a great way to start a fire while camping and a terrible situation if you store steel wool in the same drawer as batteries. Different types of fuel, such as gasoline, allow combustion to happen much, much faster. The faster it can happen, the more explosive the fuel is. Controlling the combustion reaction has been integral to modern life. Gas lamps and candles burn slower than jet fuel. By changing the amount and type of fuel and the amount of oxygen or heat provided, it’s possible to control how fast something burns. To put out a fire, one of the three things—heat, oxygen, or fuel—must be removed. In a wood fire, pouring water on it cools it off very quickly and stops the reaction from being self-propagating. This doesn’t work with a grease fire, however, because oil floats on water. Throwing water on it causes the flaming oil to splatter making the fire more dangerous. Dumping baking soda on it starves it of oxygen and puts it out. It’s not a great idea to use flour, though, as I’ll explain shortly.

Figure 9.1 illustrates how the fire isn’t burning the fuel but rather floating on top and igniting the vapors. If you are interested in making a cocktail that tastes like jet fuel and a hangover, this Game of Thrones–inspired “wildfire” cocktail was a combination of melon liqueur, orange juice, and watermelon vodka with a grain alcohol floater to make it light.

How Might Dragons Make Fire?

For a dragon to breathe fire, it needs three things: heat or sparks, fuel, and oxygen. The oxygen is easy to take care of if we assume the dragon has very large lungs. The sparks and fuel are a bit more difficult. Many have made suggestions, but here I’ll address my favorites, or at least the ones I see as the most plausible.

Dragon fuel would need to be a substance that can be kept inside the body in large quantities and easily expelled, and it can’t be too dense, seeing as dragons have to fly. One huge advantage for dragons is that many good fuel sources are organic. The most likely way a dragon could breathe fire is by igniting one of several organic fuels with a spark. The trick is finding a fuel that is both easily made and easily stored internally. Then, there needs to be some sort of ignition. I’ve read many theories on how dragons breathe fire, but the ignition mechanism always seems to be the one that is glossed over. Another rarely discussed issue is fire protection for the dragon. How is a dragon’s mouth not as scorched as Dickon Tarly?

As far as a fuel source goes, there are a few reasonable options. The first one I’ll talk about is based on my favorite demo, the birthday candle torch. (If you are daring, this is something you could try at home, but please don’t sue me or my publisher if something goes wrong. I take no responsibility.) Normally, if you dropped a match on cornstarch, the match would just go out. That’s because one of the three things needed to sustain a fire wasn’t available: oxygen. Sure, there was some at the edges, but not near enough to keep a combustion reaction going. Cornstarch is an interesting substance in that its molecules are spherical. If you rub it between your hands, it feels soft and slippery. That’s because it’s a bunch of really tiny ball bearings. I could talk endlessly about the amazing properties of cornstarch (it’s worth a quick Google, if you have a moment), but here I’ll focus on its flammability. Like all starches, cornstarch is designed to be oxidized, one way or another. Starches store energy to be used by the body later. If you are an endurance athlete, you’ve probably “carb-loaded” to store up energy for a big race. Cornstarch is designed to store energy until it’s needed. When it’s around a flame, or some way of starting a combustion reaction, it releases that energy. If you blow air between the starch molecules and then expose them to a flame, it makes an amazing torch. Because the molecules are so small and spherical, each one has a lot of surface area exposed to the air, so one spark will make it go up in a really nice flame. This is why you occasionally hear reports of grain silos exploding—if there are a lot of grain particles in the air and they come into contact with a spark or other flame, it can cause an explosion. If you want to see this effect firsthand, you can fill a squeeze bottle with cornstarch, light a birthday candle, and blow some cornstarch over the flame. This takes a little practice to do without blowing out the candle and makes one heck of an interesting mess, but it also shoots fire the same way a dragon might. Magicians have also been known to do this trick, but they typically use something called lycopodium powder, which is slightly more flammable and much more expensive. Maybe a better idea would be to watch some nice, safe videos of someone else doing it on YouTube.

The one issue with this, though, is that if a dragon’s “fuel bladder” were, say, punctured by a White Walker’s Olympic-level javelin throw, it wouldn’t explode. When Viserion’s neck was punctured, there was a clear explosion out of the side of his neck. As much as I would love dragons to use some sort of powder as a fuel, that probably isn’t the case. To cause an explosion like that, the fuel bladder would have to be full of some sort of pressurized gas. Methane is an example of one type of flammable gas. Cows are a big producer of methane. Microbes called methanogens in a cow’s gut break up the cellulose in the grass it eats to release its nutrients, and methane is a byproduct of that process. If we assume that the average dragon is many, many times larger than a cow, with a correspondingly larger fuel bladder, this means they can produce a lot more methane. Methanogens can’t function in an environment with oxygen. They are anaerobic. This is great news if methane is a dragon’s fuel source. There’s no way methane could be produced in an environment in which it could explode. The two processes, combustion and methane production, are mutually exclusive, so this seems like a great method for producing the fuel. The problem though, is that dragons don’t eat grass—they eat meat. This may be great for the drama, but it’s not so great for methane production. Because the anaerobic bacteria that create methane feast on the cellulose of plant matter, a meat-based diet is not going to produce the needed fuel. It very well may be that Balerion ate grass on the side but, seeing as it wasn’t very “dragonly,” no one ever mentioned it.

The question then becomes: Is it possible for a dragon to create enough methane to endanger all of Westeros just by eating a little grass on the side? A good way to estimate that potential is to look at cows’ methane production. According to a study done on dairy cows, which produce about twice as much methane as beef cows, they produce about 20 g of methane per kilogram of dry food consumed. That’s about 330 g of methane a day, or roughly 1 L of methane a day.3 This is only about 0.04 cubic feet. Methane produces about 1,000 BTUs, or British thermal units, of heat per square foot when ignited.4 A BTU is the amount of heat needed to raise the temperature of one pound of water by 1ºF. In a single day, the methane produced from a single cow could raise the temperature of a pound of water by 40°F. That wouldn’t even make room-temperature water boil. So, what if Drogon was the size of 100 cows with a similar digestive system? That would mean about 4 cubic feet of methane, 4,000 BTUs, and a few pounds of boiling water. This level of fire could definitely be dangerous; indeed, a cigarette can cause a forest fire, but it doesn’t lead to the kind of destruction wrought by the Khaleesi’s children.

For a cornstarch-like material or methane, a spark would be needed to start the flame. This is most likely the easiest part of creating dragon fire. There are a few different options. One would be flint. Pterosaurs, which are potentially related to dragons, had gizzards, as do many birds. Gizzards help grind up food with the help of rocks. The bird (or pterosaur) swallows some rocks and gravel and they are stored in the gizzard. The muscular walls of the gizzard break down tough food with the help of stones. It’s not too far of a stretch to assume dragons could have a “fire gizzard” that stores flint and iron pyrite. Usually steel is used with flint, but it would be hard for dragons to easily obtain steel. When flint is struck against iron pyrite, the reaction is very similar to the one created by touching steel wool to a 9V battery. When a tiny piece of iron is struck off with the hard flint, oxidation occurs quickly and creates a spark. A dragon could easily blow the contents of their fuel bladder over the spark to create a blowtorch.

The spark could also be caused by something like sodium, potassium, or another alkali metal. When alkali metals come into contact with water, electrons from the metal quickly transfer to the water. This creates ions that very strongly repel each other, causing what’s called a Coulomb explosion.5 In a Coulomb explosion, the ions repel each other so violently that they tear apart nearby molecules, releasing enough energy to cause an explosion (a demonstration favored by many chemistry teachers). It is entirely possible that dragons could house some sort of alkali metal, potentially in a different type of fire gizzard, and judiciously introduce water to cause a spark. Although this could work, it would be very dangerous. The alkali metal would have to be stored completely dry. This would be difficult to manage in an animal whose body most likely contains a large percentage of water.

A third ignition mechanism might be a piezoelectric crystal. Piezoelectric crystals create an electric spark when squeezed. The atoms in these crystals sit in a slightly asymmetrical way. When the crystal is squeezed, the atoms are jammed together, upsetting the usual electrical neutrality and creating an electric charge. In practicality, if a dragon had such a crystal surrounded by a muscle that could squeeze it at the right time, it could produce a spark as the fuel was expelled over it. Luckily for dragons, the collagen in bone has some piezoelectric properties.6 Other naturally occurring materials can produce sparks, including quartz and even sugar. Of the potential issues surrounding dragon fire, it seems that producing a spark is the least problematic.

There’s one way dragons might have evolved to create fire that combines the spark part with the combustible material: hypergolic fluids. Having such fluids as a fuel would still lead to the type of explosion seen in Viserion’s neck, would not require eating grass, and would be self-igniting. They seem like a good candidate. Hypergolic fuels are liquids that ignite when mixed together, no spark needed. They have been used in rocket propulsion, so they can definitely produce the flame needed for a dragon’s blowtorch powers. Just as in other combustion reactions, there is a fuel and an oxidizer, but in this case, they don’t need an external spark to start the self-sustaining reaction. In general, the chemicals used are toxic and corrosive, but this doesn’t necessarily rule them out; hydrochloric acid is pretty darn dangerous, but stomachs don’t seem to mind. Hypergolic fuels were first studied in the 1930s, and BMW developed a hypergolic engine fueled by nitric acid and various other compounds in 1940. The most common hypergolic fuels are hydrazine, monomethylhydrazine, and dimethylhydrazine, and nitrogen tetroxide is typically used as the oxidizer. But are these things a dragon’s body could make and store? There’s no particular reason why these substances couldn’t be made; the bigger question is whether they could be stored in the body. All of these fuels are toxic and highly unstable, so if a dragon were to produce them, it would need something similar to a stomach lining but on a much larger scale. It’s not entirely clear whether that is feasible; however, it’s not out of the realm of possibility. The drawback is the weight. Dragons are already right on the edge of being able to fly—the added weight of that much liquid fuel would most likely render them grounded.

Let’s say the dragon gets everything right and produces a jet of flame. There’s still the question of protecting its throat from the heat. There are several methods that humans have designed to protect themselves from fire, such as Kevlar and fire-retardant gel, but none of those are naturally occurring. There are, however, several organisms that are able to survive in extreme conditions: thermophiles, hydrothermal worms, and the bombardier beetle. Thermophiles, a group of organisms that includes certain types of bacteria and other microorganisms, live up to their name—they love heat. In most organisms, excessive heat breaks down the enzymes required to sustain life, so they cannot survive, but thermophiles can survive temperatures as high as 284°F (140ºC) because their enzymes continue to function at very high temperatures (figure 9.2). That’s great, but the temperatures we’re talking about are much higher than 284°F, as I’ll explain in the next section; at temperatures as hot as dragon fire, the skin of known animals would be burned off completely. Similar to thermophiles, there are hydrothermal worms on the ocean floor that love both high pressure and heat; however, scientists recently discovered that they don’t have enough insulation to survive in temperatures much higher than 140°F (60ºC).7 The bombardier beetle, by contrast, can tolerate high temperatures within its own body—it doesn’t exactly breathe fire, but it can shoot boiling hot toxic chemicals at predators. When the beetle is threatened, it mixes two compounds in a vestibule. The lining of that vestibule acts as a catalyst, causing an exothermic reaction that produces enough heat to bring the chemical mixture to a boil. The liquid is then violently expelled from the creature’s rear end. The pressure from the expulsion mechanism closes off the openings in the vestibule, thereby protecting the beetle’s organs from being melted by the boiling liquid. Dragons aren’t so lucky, however; any fire created within the dragon’s body would have to pass by crucial organs on its way to the dragon’s mouth. And even if a dragon were somehow able to protect its organs, the bombardier beetle has the advantage of a hard, protective exoskeleton that can withstand the heat of the chemical reaction. Dragons have no such exoskeleton. So, although there are creatures that can protect themselves from intense heat, none of them come close to being able to protect themselves from fire hot enough to melt stone.

I wish more than anything that I could write a paragraph easily summarizing how dragons might be able to produce flame. That was my hope when I set out to write this chapter. As seen in the chapter on dragon flight, I am comfortable granting dragons a small, one-time exception to the laws of physics. Despite my best efforts, however, I would have to give dragons no small number of physics exceptions to make fire-breathing possible. Creatures exist that can handle extreme conditions, but nothing near what a dragon’s throat would have to endure. There are ways to make biofuel sources that dragons could use to produce fire, but none of those would be light enough to allow dragons to fly, be produced in high enough quantities, or cause an explosion if the dragon’s neck is punctured. The only thing dragons might be able to do is create a spark. I think this is one mythical occurrence that must stay mythical.

Different Colors, Different Sizes

I know I just used science to burst your bubble on fire-breathing dragons, but let’s assume I didn’t—because dragons breathe fire, damn it! Let’s grant the huge myth exception and say dragons can breathe fire. If that’s the case and the laws of physics are being defied, are there laws that they do obey? How does their firepower compare to their size? We know that their necks can explode, and that the potential destruction is directly related to the heat and size of dragon fire. If the fire really is magic in some way and not caused by conventional physical and chemical methods, is it possible to look at the descriptions in the book and the depictions in the TV show to get an idea of how dragons grow both in size and firepower? One of the most dramatic scenes of destruction was Balerion’s melting of Harrenhal. Not destruction—melting. After his attack, the impenetrable castle looked like dripping candles. I’d like to use this section to determine if Balerion could have done that and if Drogon, now that he is fully grown, might be able to follow suit. This requires first answering three questions: How hot were Balerion’s flames? Is Drogon anywhere near his level? And would those flames have been hot enough to melt stone?

Balerion the Black Dread was so named not just for the color of his body but for the color if his flame. It was said that his flame was so hot it was black. From a physics point of view, that can’t really be true. It’s true that flames come in different colors, and I’ll assume it’s true that Balerion was black, but there is no such thing as a black flame. There are some pretty awesome chemistry demos that supposedly produce black “flame,” but they are really just superheated vapors and not a true flame. Nevertheless, black flame remains a staple of the fantasy world. Though there is no such thing as black flame, flame color does have to do with something black: black-body radiation. In this case, “black body” refers not to Balerion but to anything opaque and nonreflective, such as a lightbulb filament or burning wood. Iron is often used as an example of a black body. When black bodies get really hot, they emit electromagnetic radiation. When hot enough, that radiation is in the form of visible light. It’s then possible to determine how hot the flame is just by looking at the color of the light being emitted.

Black-body radiation has been observed for as long as such observations have been made. When forging weapons, blacksmiths noticed that the metal emitted different colors based on its temperature. At the turn of the 20th century, scientists began to try and explain the phenomenon. To understand black-body radiation, it’s important to know two things: what heat does to electrons and how charged particles move. Temperature is a measure of how fast particles are moving around; the faster they move, the higher the temperature. If the particles were to stop moving completely, the temperature would be at absolute zero (0 K), which was discussed back in chapter 1. This is the point where there is no temperature at all. When charged particles such as electrons move, they emit electromagnetic radiation. The faster they move, the higher the energy of the radiation. They have to be moving pretty fast to radiate visible light. Lots of hot things radiate in the infrared spectrum, which is why heat-sensing cameras are able to see things our eyes can’t—they can see energy radiating from objects at much lower levels. As the heat increases, the charged particles move faster and emit light at greater levels of energy. In the visible spectrum (the rainbow, ROY G BIV), the colors appear in order from lowest energy to highest energy. An object radiating yellow light is less hot than something radiating blue light. If something is heated to a high enough degree, it will emit light in all colors and therefore appear to be white. There are multiple wavelengths released at each temperature, as you can see in the graph (figure 9.3), but the peaks indicate the dominant wavelength. You can see that the location of the peak changes based on temperature.

It’s easy to sum up this graph in a way that is actually useful: As soon as the tail of the graph of a certain temperature enters into the visible spectrum (390–700 nm), our eyes can see that color. When all the tails are in the visible spectrum, the glowing object looks white. The peak doesn’t need to be in there, just the tail. The chart in figure 9.4 shows how the color of a fire indicates its temperature. In fact, you may remember from chapters 5 and 6 that smiths needed to know precise temperatures when forging swords. They could judge a blade’s temperature by the color it was glowing. This chart was accurate enough to forge Damascus (or Valyrian, in this case) steel. It is not at all complete because it doesn’t include blue fire, which is fairly common, but for dragon purposes it’s a good start. Blue fire is hotter even than white. When the tail of the graph in the blue section of visible light is sufficiently higher than the other colors, the flame goes from looking white to looking blue. It is possible that when Westerosi oral historians referred to Balerion’s fire as “black,” they really meant “blue.”

The story doesn’t fully end there, but it’s a good start. Fire color isn’t only based on black-body radiation. When different chemical elements are added to a fire, it’s possible to produce different flame colors based on how the elements behave when they get hot. When certain compounds are heated, the electrons jump up in energy levels. When they drop back down again, they emit light. The color of the light emitted depends on the difference between the high and low energy levels. Because the interval between the high and low energy levels of each element or compound varies, the colors are different. If, for example, you throw sodium on a flame, the flame turns yellow, whereas copper gives the flame a greenish hue. Chemists use the aptly named “flame test” to discover the composition of an unknown substance, and certain birthday candles rely on the same effect to create colorful flames. (If you want to try this at home, you can find instructions online for making pine cones that will add some color to the flames in your fireplace—it makes for a great new holiday tradition.) In all of these cases, the color of the flame changes depending on how the compounds’ electrons are moving between energy levels, not the temperature of the flame. For dragons, however, I think it’s safe to focus on black-body radiation instead of elemental influence on flame color.

All this is well and good, but what does it mean for dragons? It means that by looking carefully at the dragon fire in the show we can tell how hot it is. In season 3, Drogon made his first real kill when he torched Kraznys mo Nakloz with his yellowish-orange flames after the slave trader handed over the Unsullied in exchange for the dragon. That puts his flame at roughly 900°C. When he reappears in the fighting pits of Meereen at the end of season 5 and kills again, his fire is much lighter in color. This indicates a temperature closer to 1200°C, far hotter than before. It would seem that as Drogon grows in size, the temperature of his fire increases. Personally, I always wondered why Dany waited so long to use her dragons in Westeros—they’re flying flamethrowers! That had to be worth something. But it seems her little blowtorches just weren’t hot enough yet. Balerion’s fire was, if you look at fan art, a dazzling white, or if you read the books, black, which I think we can assume is blue. This puts the temperature at 1800°C or higher. Would that have been enough for the Black Dread to have melted Harrenhal? If so, is Drogon now big enough to follow suit and take on the Red Keep?

What Is Melting, and Can It Happen to Stone?

Clearly, dragon fire is pretty hot, but the next step in understanding the destruction it causes is looking at how stone melts and determining the temperature required for that to happen. At this point, some of you are probably thinking it’s impossible to melt stone—only things like ice and metal can melt. But stone can melt, too. If you’ve ever been lucky enough to visit Hawaii, you may have seen melted stone—I’m talking specifically about lava. Chapter 2 focused on ice, from how it’s structured to how it can build a structure. One of the important takeaways from that chapter is how ice melts; in particular, how it melts under pressure. As we all know, though, ice—and everything else, for that matter—melts when it gets hot, too.

The castles of Westeros are made of granite (according to the descriptions in the book, at least). Granite is an igneous rock, meaning it was created by magma solidifying under the heat and pressure of Earth’s mantle. Granite is typically a combination of quartz, feldspar, and other trace compounds, though the percentages of each mineral can vary widely. It comes in several different colors, including the pink of the Red Keep. Granite is formed by solidifying molten rock, so it makes sense that extreme heat can put it right back into a liquid state. Granite is extremely strong as a building material, and that strength comes from the way the crystals form. It’s certainly not the only rock with that particular chemical composition, but because the rock cools slowly deep underground, it has time to form complex crystals. As the rock cools, the different minerals begin to solidify into crystals at different temperatures and different rates. Granite isn’t composed of one single type of crystal the way ice is, but rather from different crystals interlocking. In chapter 2, I talked about how ice cracks along fracture planes and that introducing something like sawdust strengthens the ice by putting little roadblocks in the fracture planes. Granite is strong for the same reason: the crystals interlock, so there aren’t really defined fracture planes. The fracture plane of one crystal has another crystal as a roadblock. You can even see these interlocking crystals with your eye; its speckled, grainy appearance is what gives granite its strength as well as its name.

Because the crystal structure and composition of granite differs from those of ice, it means that granite melts differently, too. In ice, there is one specific melting temperature, and when that temperature is reached and energy is continuously added to the ice, the solid turns to liquid. The extra energy needed to melt something, as I mentioned in chapter 2 in relation to melting ice, is called the latent heat of fusion, and it’s about the same for granite as it is for ice. Although the basic principle is the same—energy must be added to raise the material’s temperature to the melting point, at which point additional energy (the latent heat of fusion) is needed to fully melt it—it works a little differently with granite because granite’s crystals have different melting temperatures. The majority of granite is composed of feldspar interlocked with quartz, the first and second most abundant compounds on Earth, respectively. Quartz has a melting temperature of roughly 1650°C, whereas feldspar’s melting point is 450º lower at 1200°C. That, however, is not the entire story. The two minerals have different melting points, so you’ve probably guessed that granite is very different from ice in that granite can be mostly melted and yet still slightly solid. It’s possible for the feldspar to melt but still have pieces of solid quartz within the flowing molten rock. To make it even more complicated, quartz doesn’t just up and melt; the molecules twist, crack, and then melt. In the same way that ice comes in a number of different crystal structures, quartz and feldspar also come in different structures. Furthermore, quartz is structured in such a way that it doesn’t simply melt at a given temperature; rather, it heats up until its structure changes, cracks a bit in the process, and then decides to melt. Normal quartz has atoms of silicon and oxygen arranged in a slightly twisted pattern. This twisting pattern makes it a chiral molecule, and the direction in which it twists is called the chirality. When quartz hits 1063°C, the chirality abruptly changes directions. This is a pretty jarring process for the crystal and usually leads to some cracking. Feldspar is much less interesting. Add heat and it melts. Putting this all together, here’s what happens when granite gets hot: Not much happens until it gets to 1200°C, at which point the feldspar starts to melt and the quartz starts to crack. The stone will start to flow a little bit, but it’s not fully melted; there will still be chunks of quartz within the flow of molten feldspar, cracking as they go along. As more heat is added, the quartz begins to melt, and after reaching a temperature of at least 1650°C, the stone will be completely molten. For Balerion to have given Harrenhal its characteristic melted candle appearance, his fire would need to have reached at least 1200°C, but getting it all the way to 1650°C would result in a more uniform melted effect. For Drogon to melt the Red Keep, then, he’ll need to blast the granite with a sustained source of heat of at least 1650°C. Would that be possible?

What about Harrenhal and Balerion the Black Dread?

This is the point in the chapter where I go through all the physics I’ve explained so far and tell you how science could make what seems fantastical possible. I’ve done that for most other chapters; in this case, however, there are a lot of obstacles to overcome for this to work. Although many have made vague arguments about how dragons might produce fire, when one looks at the math and science behind it, it just isn’t possible. It might be possible for animals to breathe fire on a small scale—like, beetle-small—but for something the size of a 747 to produce a jet of fire worthy of an industrial flamethrower is beyond the laws of physics. Dragons might be able to fly, but flame is just a burned bridge too far.

Let’s take the completely unreasonable leap in logic anyway and assume that, thanks to a healthy dose of magic, dragons can be the flaming beasts they are described as. The science question now becomes whether or not Balerion—or, more importantly, Drogon—could melt a castle. We’ll start with Balerion. Although there are no official pictures of Balerion, the books’ description puts his fire in the bluish range, or about 1800°C, as I said earlier. This is certainly hot enough to melt the feldspar in granite and just enough to melt the quartz, too. The trick, however, is with the time it would take. Latent heat of fusion makes everything more complicated. Either Balerion or Drogon could crack Harrenhal and leave it in ruins, but it would take time for either of them to make the castle melt. As I said in chapter 2, it would take about 5.5 minutes for the tiny coils of a red-hot space heater to fully melt a kilogram of ice that had reached its melting temperature. A dragon has the benefit of higher flame temperatures as well as the ability to cover a wider area with its flames. Assuming Balerion spouts a square meter of white-hot flame, he could melt that same kilogram of ice in about a minute and a half, which isn’t that much time. But granite, as always, is more complicated. First, there’s the time it would take to raise the blocks of granite to the melting temperature. From there, it’s another minute and a half of direct heat to melt the stone—and that’s just for a kilogram of stone. A kilogram of ice is about 10 cm³, but a kilogram of granite is much, much smaller. I’m guessing the turrets of Harrenhal were built from a whole lot more than a kilogram or two of granite. Much like Balerion, Drogon was big enough by the end of season 6 to produce fire hot enough to melt feldspar and almost hot enough to melt quartz at a sustained flame. Considering his flames grew hotter as he grew bigger, his flame will soon be hot enough to melt stone. If we gloss over the fact that there’s no plausible way for an animal to make fire on that scale, Balerion and Drogon would certainly be capable of producing the kind of castle-melting heat that puts Targaryens on the Iron Throne. The question, really, is whether he has the patience for it. And, of course, just when things start to make sense, GRRM up and kills a character, leading us to both soul-crushing sadness and a whole new branch of science.

Viserion’s Magic Fire

Run-of-the-mill yellowish-white fire was just not enough for the huge world of Game of Thrones. At the end of season 7, there was some pretty impressive blue “fire.” As Neil DeGrasse Tyson correctly stated in a tweet, and as I’ve described elsewhere in this chapter, blue fire is much hotter than orange fire.8 However, there has been much discussion as to whether it is normal blue fire or some sort of freezing anti-fire. Just such a flame was first described in GRRM’s book The Ice Dragon, which was both shockingly short and written for kids. I didn’t think he could write a book that accomplished both those things. Fandom seems split on whether or not Viserion’s fire is very hot or, as GRRM described the breath of an ice dragon in A World of Ice and Fire, “a chill so terrible that it can freeze a man solid in half a heartbeat.” The Wall came down due to the blasts from Viserion, so it seems clear that it must be hot flame. But whether the flame is hot or cold, it’s still not that simple. A process called thermal fracture happens when there are temperature gradients within a material. If you’ve ever seen a glass window fracture when it’s warm inside and cold outside, you’ll know what I’m talking about; it happened to my windshield once. It’s possible that if Viserion’s breath were like an ice dragon’s, he could blow enough cold air to knock the Wall down via thermal fracture. As I’ve already said, however, melting by adding heat isn’t a straightforward process either. It takes heat to raise the temperature to the melting point plus even more heat to actually melt something. Could blue flame even accomplish that? The action seen at the end of season 7 might happen with hot dragon flame or ice fire, just through different physics . . . or maybe it can’t happen at all. Let’s look at both options and see if one, both, or neither makes more sense given how it was presented visually. It would be much easier if GRRM would just get to writing about this bit already so I had some actual text to go off of, but alas, even if a third or fourth edition of Fire, Ice, and Physics is released, I doubt I will have much more source material to work with by then.

It’s been so much fun talking about melting in this chapter, so let’s keep rolling with that. I explained how stone melts and how the latent heat of fusion is what actually makes it melt. The same is true of the Wall. Viserion’s fire would have to be both extremely hot and capable of expending energy quickly enough to melt the ice. Blue fire is hotter than white fire and produces about twice as much energy per square meter. This means it would take about 45 seconds to melt a kilogram of ice. That seems pretty darn quick, but remember, that’s one 10 cm cube of ice out of an extremely thick wall. It would take Viserion about 45 seconds to make even a very small nick, maybe the size of a wildling’s crampon, in the giant Wall. Even though blue fire has about twice the energy per square meter of white fire, it’s still just not enough to bring down a wall, even a questionably built one. It’s even worse when you watch the episode and see him cruising along the length of the wall at a fast clip. I hate to say it because it seemed so likely, but no dragon could take down a wall like that, zombie or otherwise.

Maybe thermal cracking can take this from magic to reality? Before going into the specifics for ice walls, here’s a general overview of why things crack when one part is hot and one part is cold: For something to crack, it needs to encounter a force of some kind. In most solids, the molecules spread apart when heat is applied. I talk a lot about what happens to glasses when they get cold in chapter 8, but that discussion didn’t involve parts of the material getting hot and others staying cold. When the temperature in one part of a material differs from the temperature in another part, it creates stress. The molecules on the hot side speed up and spread out while the ones on the colder side either slow down or condense. This creates stress, and, in many cases, if the temperature difference is extreme enough, the material will crack, sometimes explosively. This is of particular concern to laser scientists who are dealing with crystals that get very hot and could crack if not cooled well (pew pew pew!). When Viserion hits the Wall with a very cold flame, he is cooling down one part of the Wall while the other stays warm. This temperature gradient causes stress, so the question becomes: Is it enough stress to make the Wall crack? Add to that the fact that parts of the Wall might not be fully frozen.

If you’ve read chapter 2, you know that ice is very, very weird. Pretty much anything that can be said about normal solids doesn’t apply to ice. Ice does, however, lose density, or contract, as it gets colder. If it’s hit with an icy blast, there will be stress from the ice contracting as it cools. When it’s cold, it’s also stronger. One group found that ice actually gets significantly stronger the colder it gets. The force needed to crack ice just from whacking on it increases as temperature decreases. If zombie Viserion was blowing icy fire, he might actually have been strengthening the wall.9 But what about the stress caused as ice gets colder? At liquid nitrogen temperature (−196ºC), there is enough stress between the outside and inside of an ice cube to make it crack pretty quickly. This doesn’t exactly scale up well. Sure, if ice flame is the temperature of liquid nitrogen, it will definitely crack the outside of the Wall, as seen in the show. But, again, we run into the pesky problem of the Wall’s thickness. There is simply no way to crack the Wall all the way through with the cold.

As I said in chapter 2, the Wall is most likely reinforced with sawdust and a prayer, so I don’t think it would take much to knock it down, but the power from Viserion, whether hot or cold, would not be enough to cause the wall to crumble. We can now all rest easier knowing that science says the Seven Kingdoms are safe, and Tormund will be just fine. It looks like Martin should have gone with the fan theories of freezing the Shivering Sea at Eastwatch and allowing White Walkers to just walk across. Now that is something Viserion the ice dragon could have pulled off.