Preface.  I was writing about how deadly and brutal wars are today, nuclear bombs, and all the mafia totalitarian and gangster states like Russia and Mexico, Ecuador, various African states that make the lives of their people miserable and getting pretty depressed.

Then I remembered this book by Vaclav Smil. Energy and Civilization A History (2017) MIT Press. In the future societies can’t get as bad — they won’t have the fossil energy to gain as much territory, power, and control.  Energy determines what societies are capable of.  The ability to make nuclear weapons, jet fighters, AI killing machines and more will no longer be possible.

Eventually societies will go back to wood for fuel, which limits the size of cities, which need at least 30 times as much woodlands surrounding them for energy to cook and heat with. Since it takes a lot of energy to get to a forest, cut it down, and take it back to a city, I think deforestation is more likely from forest fires than being cut down as the shift from fossils back to biomass is made.

Now if we can just avoid nuclear warfare, at some point in the future the earth and much smaller societies forced to live in balance with nature may arise. There will still be wars, but the endless need to buy more stuff will likely be replaced by relationships with fellow townspeople and family, religious and guild festivals.  When war is on the horizon, there will be many more societies and land to escape to than now. Read all about it in Scott’s 2017 “Against the Grain. A Deep History of the Earliest States. Yale University Press” (or read my review at energyskeptic).

Below are my Kindle notes from Smil’s Energy and Civilization.

***

Vaclav Smil. 2017. Energy and Civilization A History. MIT Press. 

Energy is the only universal currency: one of its many forms must be transformed to get anything done.

 

Howard Odum offered a variation on Ostwald’s key theme: “The availability of power sources determines the amount of work activity that can exist, and control of these power flows determines the power in man’s affairs and in his relative influence on nature”

 

One does not have to be a scientist to make the link between energy supply and social advances. This is Eric Blair (George Orwell, 1903–1950), writing in 1937 in the second chapter of The Road to Wigan Pier, after his visit to an underground coal mine: Our civilization, said Chesterton, is founded on coal, more completely than one realizes until one stops to think about it. The machines that keep us alive, and the machines that make machines, are all directly or indirectly dependent upon coal. In the metabolism of the Western world the coal-miner is second in importance only to the man who ploughs the soil. He is a sort of caryatid upon whose shoulders nearly everything that is not grimy is supported. For this reason the actual process by which coal is extracted is well worth watching, if you get the chance and are willing to take the trouble. (Orwell 1937, 18)

The fundamental nature of the concept is not in doubt. As Robert Lindsay put it, If we can find a single word to represent an idea which applies to every element in our existence in a way that makes us feel we have a genuine grasp of it, we have achieved something economical and powerful. This is what has happened with the idea expressed by the word energy. No other concept has so unified our understanding of experience.

Earth is in the goldilocks zone: not too hot, not too cold

Kopparapu and co-workers (2014) concluded that, given its atmospheric composition and pressure, Earth orbits at the inner edge of the habitable zone, just outside the radius where the runaway greenhouse effect would bring intolerably high temperatures. About two billion years ago enough carbon dioxide (CO2) was sequestered by the ocean and by archaea and algae to prevent this effect on Earth, but if the planet had been a mere 1% farther from the Sun, virtually all of its water would have been locked in glaciers. And even with temperatures within an optimum band the planet could not support highly diversified life without its unique atmosphere, dominated by nitrogen, enriched by oxygen from photosynthesis, and containing a number of important trace gases regulating surface temperature—but this thin gaseous envelope could not persist without the planet being sufficiently large to exert enough gravity to hold the atmosphere in place.

Specific energy densities for individual foodstuffs

Foodstuffs                                          Energy density (MJ/kg)

Very low              Vegetables, fruits             0.8–2.5

Low                        Tubers, milk                        2.5–5.0

Medium               Meats                                   5.0–12.0

High                       Cereal and legume grains 12.0–15.0

Very high             Oils, animal fats                 25.0–35.0

The difference fossil fuels made to civilizations was huge

Cities had to draw on nearby areas at least 30 times their size for fuel supply. This reality restricted their growth even where other resources, such as food and water, were adequate. Large preindustrial cities located in a northern temperate climate and relying heavily on charcoal (China’s Xi’an or Beijing would be good examples) would have required a wooded area at least 100 times their size to ensure a continuous supply of that fuel.

The technical advances of the past two centuries have brought many substantial declines in energy intensities. Perhaps most notably, the coke-fueled smelting of pig iron in large blast furnaces now requires less than 10% of energy per unit mass of hot metal than did the preindustrial charcoal-based production of pig iron (Smil 2016).

Box 1.8 Energy intensities of common materials

Energy cost (MJ/kg)

175-200                Aluminum metal from bauxite

1 –  2                   Bricks fired from clay

2 –  5                   Cement from raw materials

90-100                Copper from ore

10-  70                Explosives from raw materials

4 –   10                Glass from raw materials

 

Two kinds of coal may have an identical energy density, but one may burn very cleanly and leave behind only a small amount of ash, while the other may smoke heavily, emit a great deal of sulfur dioxide, and leave a large incombustible residue. An abundance of high-energy-density coal ideal for fueling steam engines was clearly a major factor contributing to the British dominance of 19th-century maritime transport, as neither France nor Germany had large coal resources of comparable quality.

There are other important qualities hidden by abstract measures. Access to energy stores is obviously a critical matter. Tree stem wood and branch wood have the same energy densities, but without good axes and saws, people in many preindustrial societies could only gather the latter fuel. That is still the norm in the poorest parts of Africa or Asia, where children and women gather woody phytomass; and its form, and hence its transportability, also matters because they have to carry wood (branch) loads home on their heads, often for considerable distances.

During the 1890s a dozen powerful American horses needed some 18 tons of oats and corn per year, about 80 times the total of food grain eaten by their master.

Only a few land-rich countries could provide so much feed. Feeding 12 horses would have required about 15 ha of farmland. An average U.S. farm had almost 60 ha of land in 1900, but only one-third of it was cropland. Clearly, even in the United States, only large grain growers could afford to keep a dozen or more working animals; the 1900 average was only three horses per farm (USBC 1975).

there were clear limits to population densities achievable with traditional farming. Moreover, average food supply was rarely much above the existential minimum, and seasonal hunger and recurrent famines weakened even the societies with rather low population densities, good soils, and relatively good farming techniques.

Even in land-rich agricultures with extraordinary feed production capacities the substitution trend could not have continued much beyond the American achievements of the late 19th century. Heavy gang plows and combines took animal-drawn cultivation to its practical limit. Besides the burden of feeding large numbers of animals used for relatively short periods of field work, much labor had to go into stabling, cleaning, and shoeing the animals. Harnessing and guiding large horse teams were also logistical challenges. There was a clear need for a much more powerful prime mover—and it was soon introduced in the form of internal combustion engines.

 

Regardless of the historical period, environmental setting, or prevailing mode of cropping and intensification, no traditional agriculture could consistently produce enough food to eliminate extensive malnutrition. All of them were vulnerable to major famines, and even the societies practicing the most intensive cultivation were not immune to recurrent catastrophes, with droughts and floods being the most common natural triggers. In China in the 1920s peasants recalled an average of three crop failures within their lifetime that were serious enough to cause famines (Buck 1937). These famines lasted on average about ten months, and forced a quarter of the affected population to eat bark and grasses. Nearly one-seventh of all people left their villages in search of food. A similar pattern would be found in most Asian and African societies. Some famines were so exceptionally devastating that they remained in collective memory for generations and led to major social, economic, and agronomic changes.

Why could not preindustrial societies insulate themselves better from the recurrence of such drastic food shortages? They could—either by extending cultivated land, or by intensifying their cropping, or by doing both—and they were constantly trying to do so. But in an overwhelming majority of cases these moves were undertaken reluctantly, and they were commonly postponed for so long that natural disasters were repeatedly translated into major famines. There is a clear energetic reason for this tardiness. Both the extended and the intensified cultivation required a higher investment of energy. Even in societies that could afford a larger number of draft animals, most of those additional energy inputs had to come from longer hours and harder exertion of human labor.

The villages of Carolingian Europe were overpopulated and their grain supplies constantly insufficient—but except in parts of Germany and Flanders there was little effort to create new fields beyond the most easily cultivable lands (Duby 1968). Later European history is replete with waves of German migrations from densely populated western regions. Armed with superior moldboard plows, they opened up farmlands in areas considered inferior by local peasants in Bohemia, Poland, Romania, and Russia—and set the stage for nationalist conflicts for centuries to come.

Seavoy (1986) labeled this behavior—acquiring minimum levels of food safety and material welfare with the least expenditure of physical labor—a subsistence compromise. He also saw high birth rates as the other key strategy to reduce per capita labor exertion. The energy cost of pregnancy and of bringing up another child is negligible compared to its labor contribution, which can start at a very early age. According to Seavoy (1986, 20), “Having many children (an average of four to six) and transferring labor to them at the earliest possible age is highly rational behavior in peasant societies, where the good life is equated with minimal labor expenditures, not with the possession of abundant material goods.

Such sweeping generalizations ignore the influence of many other critical factors. Environmental conditions—soil quality, the amount and reliability of precipitation, the per capita availability of land, fertilizer, and food, the capability to support draft animals—have always made much difference. So did socioeconomic peculiarities (land tenure, corvée, taxation, tenancy, ownership of animals, and access to capital) and technical innovation (better agronomic methods, animal breeds, plows, and cultivation and harvesting implements).

 

A much less contentious generalization concerning the physical work is that it was spread as much as possible. In practice this meant transferring much of it to women and children, generally persons of lower status in peasant societies. Women were responsible for a high share of field and household tasks in virtually every traditional society. And because even pregnancy and lactation did not have to be much of a burden in terms of additional food, and because the children often started working as early as four to five years of age, large families were the least energy-intensive way to minimize adult labor and to secure food in old age when infirmities set in.

In those traditional agricultures that were powered overwhelmingly by human labor it was clearly rational to minimize individual workloads by having large families. At the same time, this strategy made it much more difficult to increase average per capita food availability, and to avoid recurrent famines.

Only the inputs of fossil energies—directly as fuels and electricity, and indirectly in agricultural chemicals and machinery—could sustain both an expanding population and a higher per capita supply of food.

Fossil subsidies began to make a major difference only after 1910 with the diffusion of tractors, trucks, and synthetic nitrogenous fertilizers.

Wind energy and power

Average wind speed increases roughly with one-seventh the power of height. This means, for example, that it will be about 22% higher 20 m above the ground than at a height of 5 m.

As the wind power goes up with the cube of the average speed, doubling the speed increases the available power eightfold. Early (relatively heavy and poorly geared) windmills also needed winds of at least 25 km/h (7 m/s) in order to start milling or pumping; at lower speeds they just turned slowly, but sails had to be trimmed at wind speeds above 10 m/s (and furled at speeds above 12 m/second), providing only a narrow window (5–7 hours of daily rotation) for useful work (Denny 2007).

These realities obviously favored locations with sustained brisk winds. Later, more efficient, smoothly geared, and properly lubricated designs would work well with winds above 4 m/s, delivering 10–12 hours of useful operation a day. Preindustrial societies could capture only wind flowing near the ground, with the spans of most of the windmills being less than 10 m. Wind flows also have large temporal and spatial variation. Even in windy places the annual wind speed means fluctuate by up to 30%, and shifting a machine’s location by just 30–50 m may cut or increase the average speed easily by half.

 

Nearly all traditional societies could produce heat and light only by burning biomass fuels. Woody phytomass, the charcoal derived from it, crop residues, and dried dung provided all the energy needed for household heating, cooking, and lighting and for small-scale artisanal manufactures; later, in larger proto-industrial enterprises, those fuels were used in firing relatively large quantities of bricks and ceramics, making glass, and smelting and shaping metals.

The provision of biomass fuels could be as easy as making an everyday short trip to a nearby forest or bush or mountainside to collect fallen branches and break off dry twigs, or to gather dry grasses, or to gather some dry straw after the grain harvest and store it under house eaves. More often, however, it could entail long walks, mostly by women and children, to gather combustible biomass; laborious tree cutting; exhausting charcoal making; and the long-distance transport of the fuel in ox-drawn carts or in camel caravans to cities in the middle of long-deforested plains or in desert regions. The abundance or scarcity of fuel affected house design, as well as dressing and cooking practices. Provision of these energies was one of the principal reasons for traditional deforestation.

in France, coal began to provide more than half of all fuel energy by the mid-1870s, and in the United States, coal and oil (and a small volume of natural gas) surpassed the energy content of fuelwood by 1884 (Smil 2010a). But elsewhere the dependence on phytomass fuels continued well into the twentieth century: in the most populous nations of Asia it remained dominant until the 1960s or 1970s, and in sub-Saharan Africa it remains the single largest source of primary energy.

For most people in traditional societies, energy needs have always amounted just to cooking two or three meals a day, in cold climates heating at least one room, and in some regions also preparing feed for animals and drying food.

Wood was used in any available form: as fallen, broken, or lopped-off branches, twigs, bark, and roots—but chopped stem wood became available only where good cutting tools—adzes, axes, later saws—were common. Wood variety made surprisingly little difference. There are thousands of woody plants, and though their physical differences are substantial—the specific density of some oaks is almost twice as high as that of some poplars—their chemical composition is remarkably uniform (Smil 2013a). Wood is about two-fifths cellulose, roughly one-third hemicelluloses, and the rest is lignin;

The traditional production of charcoal, an excellent fuel, was very wasteful. Typical charcoal yields in such kilns were only between 15% and 25% those of air-dried wood. This means that about 60% of the original energy was lost in making charcoal. The payoff was in the fuel’s quality: its combustion could produce temperature of 900°C, and with a supplementary air supply, achieved most efficiently by using bellows, that could be raised to nearly 2000°C, more than enough to melt even iron ores.

 

Requirements of less than 500 kg/capita in the poorest villages of tropical regions. Up to five times as much biomass was used in latitudes with pronounced winters and with a substantial wood-based production of bricks, glass, tiles, and metals and evaporation of brines. In Germany up to 2 t of wood (almost all of it burned to obtain potassium rather than to produce heat) were needed for to make 1 kg of glass, while evaporation of brines in large wood-heated iron pans consumed up to 40 kg of wood per 1 kg of salt (Sieferle 2001).

I estimated that average annual energy requirements in the Roman Empire around 200 CE added up to 650 kg/capita, that is, roughly 10 GJ, or about 1.8 kg/day (box 4.10). The best available reconstruction of firewood demand in medieval London (around 1300) resulted in an annual mean of about 1.75 t of wood, or roughly 30 GJ/capita (Galloway, Keene, and Murphy 1996). Estimates for Western Europe and North America just before their switch to coal show even higher average needs.

Storing enough straw to last through the winter could never be as easy as stacking chopped wood. The low density of crop residues also meant that open fires and simple stoves had to be stoked almost constantly. Because of a number of competitive nonenergy uses, crop residues were often in short supply. Legume residues were an excellent high-protein feed and fertilizer. Cereal straw makes a good ruminant food and animal bedding; many societies (including England and Japan) used it for thatching house roofs, and it was also a raw material for manufacturing simple tools and domestic articles.

Every bit of combustible phytomass was often gathered for household use. Throughout the Middle East thorny shrubs were often burned, and date kernels were used to make charcoal. On the North China Plain women and children with rakes, sickles, baskets, and bags collected fallen twigs, leaves, and dry grasses. And in the interior of Asia, as well as throughout the Indian subcontinent, parts of the Middle East, Africa, and both Americas, dried dung was the most important source of heat for cooking.

A little-appreciated reality is dung’s essential contribution to America’s westward expansion (Welsch 1980). Wild buffalo and cattle dung made possible the early continental crossings and the subsequent colonization of the Great Plains during the nineteenth century. Travelers on the Oregon and Mormon Trails collected “buffalo wood,” and the early settlers stacked winter supplies in igloo shapes or against house walls. Known as cow wood or Nebraska oak, the fuel burned evenly and with little smoke and odor, but its rapid combustion required almost continuous stoking.

In South America llama dung was the principal fuel on the altiplano of the Andes, the core of the Inca Empire in southern Peru, eastern Bolivia, and northern Chile and Argentina. Cattle and camel dung was used in the Sahelian region of Africa, as well as in Egyptian villages. Cattle dung was gathered in largest quantities in both arid and monsoonal Asia, and Tibetans always relied on yak dung. Only sheep dung has generally been avoided because its burning produces acrid smoke.

 

In India, where dung use is still common in many rural areas, both cow and water buffalo droppings are gathered regularly, mostly by harijan (untouchable) children and women, both for their own household use and for sale. Dung was (and is) collected either as dry chips or as a fresh biomass. Fresh dung is mixed with straw or chaff, hand-molded into patties and cakes, and sun-dried in rows, plastered on house walls, or stacked in piles. A recent survey of rural energy use in South Asia found that 75% of Indian, 50% of Nepali, and 47% of Bangladeshi households are still using dung for cooking .

In traditional agricultural societies where grains supplied most of the food energy, their cooking (by steaming, boiling, or baking) was necessary to make the hard seeds edible. But before the grains (stored in baskets, jars, or bins) could be cooked, they had to be processed, and grain milling has been a virtually universal, and also historically almost always the first, processing step; the extraction of oils by pressing a variety of seeds, fruits, and nuts came later. Tubers were processed to remove anti-nutritive factors or to enable their long-term storage, and sugar cane was crushed to express its sweet juice. In all of these tasks, human energy was only gradually augmented by animal labor.

Cooking required relatively little heat energy in East Asian stir-frying and steaming. In contrast, considerable fuel inputs were needed for baking bread, the staple throughout the rest of the Old World, and for the roasting done commonly in the Middle East, Europe, and Africa. In some societies fuel was also required to prepare feed for domestic animals, above all for pigs. Seasonal heating was necessary in the mid-latitudes, but (save for sub-Arctic locations) preindustrial houses were usually heated only for short periods and to relatively low temperatures.

A donkey-powered mill (energy input at the rate of 300 W) produced from less than 10 kg/h to 25 kg/h, while millstones driven by a small waterwheel (1.5 kW) would grind flour at rates between 80 and 100 kg/h. Flour would have been used to bake bread supplying at least half of all food energy in average dietary intake (but bread’s share was often more than 70%).

These massive stones had to be precisely balanced. If they rubbed against each other they could be badly damaged, and they could also spark a fire. If they were too far apart they produced rough meal rather than fine flour. Tolerance requirements were no more than the thickness of a heavy brown paper between the stones at the eye—and that of a tissue paper at the edge.

 

For centuries, milling with water or wind still required a great deal of heavy labor. Grain had to be unloaded and hoisted with pulleys to hoppers; the freshly milled flour had to be cooled by raking, sorted by sieving, and bagged.

 

In heating, the much-needed transition from wasteful, unregulated open fires to more efficient arrangements was very slow. Merely moving an open fire into a three-sided fireplace brought only a marginal efficiency gain. Well-stoked fireplaces could keep an unattended fire overnight, but their heating efficiencies were poor. The best rates were close to 10%, but more typical performances were just around 5%. And often a working fireplace, warming its immediate vicinity with radiated heat but drawing the warm inside air outside, was actually causing an overall heat loss in the room. When this draft was impeded the combustion could produce dangerous, even lethal, levels of carbon monoxide.

The British House of Commons was heated by large charcoal pots until 1791.

Relay postal systems preferred an average speed of 13–16 km/h and a daily distance of 18–25 km/animal to minimize the risk of damage to horses,

The simplest way of transporting loads is to carry them. Where roads were absent people could often do better than animals: their weaker performance was often more than compensated for by flexibility in loading, unloading, moving on narrow paths, and scrambling uphill. Similarly, donkeys and mules with panniers were often preferred to horses: steadier on narrow paths, with harder hooves and lower water needs they were more resilient. The most efficient method of carrying is to place the load’s center of gravity above the carrier’s own center of gravity—but balancing a load is not always practical.

In relative terms, people were better carriers than animals. Typical loads were only about 30% of an animal’s weight (that is, mostly just 50–120 kg) on the level and 25% in the hills. Men aided by a wheel could move loads far surpassing their body weight. Recorded peaks are more than 150 kg in Chinese barrows where the load was centered right above the wheel’s axle.

Massed applications of human labor, aided by simple mechanical devices, could accomplish some astonishingly demanding tasks. Undoubtedly the most taxing transport tasks in traditional societies were the deliveries of large-sized building stones or finished components to construction sites. Large stones were quarried, moved, and emplaced by every old high culture

1880 BCE (Osirisnet 2015). The scene portrays 166 men dragging a colossus on a sledge whose path is lubricated by a worker pouring liquid from a vessel (fig. 4.17). With lubrication cutting the friction by about half, their massed labor, reaching a peak power of over 30 kW, could move a 50 t load.

 

Pulling a 140 t stone, the heaviest block at Ollantaytambo in southern Peru, up the ramp required the coordinated force of about 2,400 men. But we know nothing of the logistics of such an enterprise. How were more than 2,000 men harnessed to pull in concert?

The superiority of horses could be realized only with a combination of horseshoes and an efficient harness. Performance in land transport also depended on success in reducing friction and allowing higher speeds. The state of roads and the design of vehicles were thus two decisive factors. The differences in energy requirements between moving a load on a smooth, hard, dry road and on a loose, gravelly surface are enormous. In the first case a force of only about 30 kg is needed to wheel a 1 t load, the second instance would call for five times as much draft, and on sandy or muddy roads the multiple can be seven to ten times higher. Axle lubricants (tallow and plant oils) were used at least since the second millennium BCE.

The roads in ancient societies were mostly just soft tracks that seasonally turned into muddy ruts, or dusty trails. The Romans, starting with the Via Appia (Rome to Capua) in 312 BCE, invested a great deal of labor and organization in an extensive network of hard-top roads. Well-built Roman viae were topped with gravel concrete, cobblestones, or slabs set in mortar. By Diocletian’s reign (285–305) the Roman system of trunk roads, the cursus publicus, had grown to some 85,000 km. The overall energy cost of this enterprise was equivalent to at least one billion labor-days. This large total prorates to easily manageable requirements over the centuries of ongoing construction (box 4.12) In Western Europe the Roman achievements in road building were surpassed only during the 19th century.

The Incas, consolidating their empire between the 13th and 15nth centuries, built an impressive road network by corvée labor. Its length totaled about 40,000 km, including 25,000 km of all-weather roads crossing culverts and bridges and equipped with distance markers. Of the two main royal roads, the one winding through the Andes was stone-surfaced. Its width ranged from up to 6 m on river terraces to just 1.5 m when cut through solid rock. The unsurfaced coastal link was about 5 m wide. Neither road had to support any wheeled vehicles, just caravans of people and pack llamas carrying 30–50 kg per animal and covering less than 20 km/day.

At its beginning it was cheaper to import many goods into England by sea from Europe rather than to carry them by pack animals from the country’s interior. Travelers described the state of English roads as barbarous, execrable, abominable, and infernal

 

Roads in continental Europe were in similarly bad shape, and coach horses harnessed in teams of four to six animals lasted on average less than three years.

 

The steam engine had actually expanded the deployment of horses (Greene 2008). Most railway shipments had to be collected and distributed by horse-drawn vans, wagons, and carts. These vehicles also delivered food and raw materials from the nearby countryside.

The stabling of the animals in mews and the provision and storage of hay and straw made an enormous demand on urban space. At the end of Queen Victoria’s reign, London had some 300,000 horses. City planners in New York were thinking about setting aside a belt of suburban pastures to accommodate large herds of horses between the peak demands of rush-hour transport. The direct and indirect energy costs of urban horse-drawn transport—the growing of grain and hay, feeding and stabling the animals, grooming, shoeing, harnessing, driving, and removal of wastes to periurban market gardens—were among the largest items on the energy balance of the late 19th-century cities

This equine dominance ended rather abruptly. Electricity and internal combustion engines were becoming practicable just as the numbers of urban horses rose to record totals during the 1890s. In less than a generation, horse-drawn city traffic was largely displaced by electric streetcars, automobiles, and buses.

Triple-tiered trieres (Roman triremes), the best-performing warships of the classical era, were powered by 170 rowers. Strong oarsmen could propel them with more than 20 kW of power, enough to produce maximum speeds of close to 20 km/h. Even when moving at more common top speeds of 10–15 km/h, the highly maneuverable triremes were powerful fighting machines.

The 126 m long ship was to carry more than 4,000 oarsmen and nearly 3,000 troops, and could theoretically be propelled with over 5 MW of power. But its weight, including the heavy catapults, made it virtually immovable, a costly shipbuilding miscalculation.

The Pantheon’s remarkable design is often cited as an ingenious use of concrete, but the often repeated claim that the Romans were the first builders to use this material is inaccurate. Concrete is a mixture of cement, aggregates (sand, pebbles), and water, and cement is produced by high-temperature processing of a carefully formulated and finely ground mixture of lime, clay, and metallic oxides in an inclined rotating kiln—and there was no cement in the Roman opus caementicium used to build the Pantheon or in any other building until the 1820s

 

The Roman opus caementicium was a mixture of aggregates (sand, gravel, stones, often also broken bricks or tiles) and water, but its bonding agent was not cement (as it is in concrete) but lime mortar (Adam 1994). The mixture was prepared on a building site, and the unique combination of slaked lime and volcanic sand—excavated near Puteoli (modern-day Pozzuoli, just a few kilometers west of Mount Vesuvius) and known as pulvere puteolano (later as pozzolana)—made a sturdy material that could harden even under water. Although inferior in comparison with modern concrete, the pozzolanic aggregate and a high-quality lime produced material that was strong enough not only for massive and durable walls but also for large vaults and domes.

 

Although some cathedrals were completed quickly (Chartres took only 27 years, the original Notre-Dame de Paris took 37), construction was commonly interrupted by epidemics, labor disputes, regime changes, money shortages, and intra- and international conflicts. As a result, the building of a cathedral usually lasted several generations.

 

Armed conflicts have always had a formative role in history: they require the mobilization of energy resources (often on an extraordinary scale, be it by marshalling masses of foot soldiers equipped with simple weapons or by producing highly destructive explosives and machines and laying down supplies for prolonged wars), and they have repeatedly resulted in the most concentrated and the most devastating releases of destructive power. Moreover, the basic energy supplies, whether of food or fuel, of populations exposed to armed conflicts are affected not only during the conflict’s duration (through the acquisition of food for roaming armies, the destruction of crops, or the interruption of normal economic activities with the mobilization of young males and the damage inflicted on settlements and infrastructures) but often for years after its end.

 

All prehistoric land warfare and all conflicts of ancient and early medieval eras were powered solely by human and animal muscles. Warriors wielded daggers, axes, and swords in close combat, on foot or on horseback. They used spears and lances, and they drew bows and much more powerful crossbows (both the Chinese and the Greeks used them since the fourth century BCE) to shoot arrows whose impact would injure and kill unprotected enemies as far as 100–200

 

Handheld weapons could cause grievous injuries, and well-aimed cuts or thrusts could kill instantly, but they required a commingling of fighting forces, and their power was obviously limited by the capabilities of a warrior’s muscles. Bows and arrows allowed the separation of fighting forces, and master archers achieved admirable accuracy over relatively long distances, but archery battles wasted too many arrows through inaccurate targeting and the relatively low kinetic energy of light arrows, and the time needed to reset the arrows between successive discharges limited the magnitude and frequency of injuries that could be inflicted by these weapons. The limits of human performance also determined the daily range of advancing armies, and even if well-rested and well-fed men were able to move swiftly, the army’s progress was often limited by the speed of its supply train, made up of slow-moving animals.

 

The earliest, smaller trebuchets were operated by men pulling ropes attached to the short arm; later large machines had massive counterweights and were able to throw objects weighing hundreds of kilograms (with record weights around or even surpassing 1 t) farther than the range of early medieval artillery. They were also used in defense against sieges, with trebuchets positioned on the high ramparts of castles or city walls, ready to hurl massive stones at any siege construction within their reach.

 

The importance of horses, both in cavalry units and harnessed to heavy wagons and field artillery, persisted in all major Western conflicts of the early modern era (1500–1800), as well in the epoch-defining Napoleonic Wars. Large armies projected far from their home base had to rely on animals to move their supplies: pack animals (donkeys, mules, horses, camels, llamas) were used in difficult terrain;

 

Prussian supplies and animals for the Russian invasion

 

Also put in muscle power

Opening the road to Russia to Napoleon: that is how Philip Paul, comte de Ségur (1780–1873), one of Napoleon’s young generals and perhaps the most famous chronicler of the disastrous Russian invasion, described the Prussian contribution. By this treaty, Prussia agreed to furnish many goods: 22,046 tons of rye, 264 tons of rice, two million bottles of beer, 44,092 tons of wheat, 71,650 tons of straw, 38,581 tons of hay, six million bushels of oats, 44,000 oxen, 15,000 horses, 300,600 wagons with harness and drivers, each carrying a load of 1700 pounds; and finally, hospitals provided with everything necessary for 20,000 sick.  

 

Attackers able to destroy sturdy stone structures from far beyond the range of archers put an end to the defensive value of traditionally built castles and walled cities, and the medieval practice of building relatively compact fortresses with thick stone walls was superseded by new designs of low spreading star-shaped polygons with massive earthen embankments and huge water ditches.

 

Only a small fraction of initially sedimented biomass carbon was transformed into fossil fuels (Dukes 2003). During the formation of coal up to 15% of plant carbon ended up as peat, up to 90% of that could be preserved in coal, and in open-cast mines up to 95% of coal in place can be extracted from thick seams. As a result, up to 13% of the original ancient plant carbon can be extracted as coal; inversely, this means that roughly eight units of ancient carbon ended up in marketed coal (the range is mostly 5–20 units). In contrast, the overall carbon recovery factor is much lower for crude oil and natural gas. These fuels arose from organisms buried in marine and lacustrine sediments, and the production of fossil hydrocarbon recovers at best close to 1% but commonly just 0.01% of the carbon that was initially present in the ancient biomass whose transformation yielded oil and gas. The recovery rate of 0.01% means that 10,000 units of ancient carbon were required to produce one unit of carbon sold as crude oil or natural gas.

 

Every transition to a new form of energy supply has to be powered by the intensive deployment of existing energies and prime movers: the transition from wood to coal had to be energized by human muscles, coal combustion powered the development of oil, and, as I stress in the last chapter, today’s solar photovoltaic cells and wind turbines are embodiments of fossil energies required to smelt the requisite metals, synthesize the needed plastics, and process other materials requiring high energy inputs.

 

BRITAIN: Choosing a precise date for the tipping point between wood and coal would be arbitrary, but his reconstructions show that the most likely date when coal surpassed biomass as the source of heat was around 1620, or maybe even a little earlier. By 1650 coal’s share was up to 65%, by 1700 about 75%, by 1800 about 90%, and by the 1850s more than 98% (the last two shares are for the UK). British coal supremacy lasted for another century: in 1950 coal supplied 91% of the country’s primary energy and still 77% by 1960. As a result, coal dominated (more than 75%) the country’s energy use for 250 years, much longer than in any other country.

 

The Panama Canal could not have been completed so fast (1904–1914) without the deployment of about 100 large Bucyrus and Marion steam shovels, and steam engines found their way even into the American fields with cable plows. But steam engines became victims of their own success: as their typical efficiencies rose and their highest capacities reached unprecedented levels (orders of magnitude above any traditional prime mover), they began to encounter their inherent limitations as more was demanded of them than they could deliver (Smil 2005). Even after more than a century of improvements the most commonly used steam engines remained highly inefficient: by 1900 the typical steam locomotive wasted 92% of coal fed into its boiler. And they remained heavy, limiting mobile uses beyond water and rails where it was easier to support their mass (box 5.5)

 

The first steam engines of the eighteenth century were exceedingly massive, with ratios (600–700 g/W) nearly as high as those of people and draft animals. By 1800 the ratio had declined to about 500 g/W, and by 1900 the best locomotive steam engines weighed just 60 g/W. But that was still too heavy for two very different but fundamental applications: to power land vehicles and to rotate large dynamos in new electricity-generating stations.

 

 In 1894 a new Daimler-Maybach gasoline engine installed in a car that won the Paris-Bordeaux race rated less than 30 g/W, leaving no place for steam engines in road transportation. And even the first commercial design of Charles Parsons’s small steam turbines—a 100 kW unit built in 1891—rated only 40 g/W, and before the beginning of World War I the ratio fell below 10 g/W and their efficiencies surpassed 25%, far above the 11–17% for the best steam engines (Smil 2005). Consequently, 16 large Westinghouse-Corliss steam engines installed in New York’s Edison plant in 1902 were already outdated,

 

When steam engines reached their highest efficiency, greatest power, and the lowest mass/power ratio, these were not achievements paving the way for further dominance: despite its impressive improvements and its recently achieved ubiquity in industry, on railroads, and in waterborne transportation, the dominant prime mover of the 19th century could not continue to be the leading prime mover of the 20th century. That role was split, as steam turbines had rapidly filled the demand for the most powerful prime movers in electricity generation, and internal combustion engines (first the gasoline-fueled machines, starting in the 1880s, then diesel engines) finally provided a prime mover that was light, powerful, and affordable to energize road transport. The ascendance of internal combustion engines was made possible by the availability of inexpensive liquid fuels refined from crude oil: they had higher energy density than coal, were cleaner to burn, and were easier to move and store, a combination that still makes them the best fuel choice for transportation.

 

In 1800 the inhabitants of Paris, New York, or Tokyo lived in a world whose energetic foundations were no different not only from those of 1700 but also from those of 1300: wood, charcoal, hard labor, and draft animals powered all of those societies. But by 1900 many people in major Western cities lived in societies whose technical parameters were almost entirely different from those that dominated the world in 1800 and that were, in their fundamental features, much closer to our lives in the year 2000.

 

In 1800 the world consumed about 20 EJ of energy (an equivalent of less than 500 Mt of crude oil), of which 98% was phytomass, mostly wood and charcoal; by 1900 the total primary energy supply had more than doubled (to about 43 EJ, equal to just over 1 Gt of crude oil), and half of it came from fossil fuels, mostly from coal. In 1800 the most powerful inanimate prime mover, Watt’s improved steam engine, had a capacity just above 100 kW; in 1900 the largest steam engines rated 3 MW, or 30 times as much. In 1800 steel was a rarity; by 1850, even in the UK, it “was known in commerce in comparatively very limited quantities” (Bell 1884, 435), and only a few hundred thousand tonnes of it were produced worldwide—but by 1900 the global output was 28 Mt (Smil 2016).

 

By 1900 the United States and France were already overwhelmingly fossil-fueled societies, but the world as a whole still derived half of its primary energy from wood, charcoal, and crop residues—and even in the United States the year of the peak draft horse numbers was still 17 years in the future. And though incandescent lights, electric motors, and telephones were making rapid inroads, most of the electricity used by urban families in the United States or Germany powered just a few light bulbs.

 

The post-1900 extraction of fossil fuels has been marked by three notable trends. First, the global expansion of coal mining and hydrocarbon production raised the annual extraction of fossil carbon roughly 20-fold between 1900 and 2015: from 500 Mt in 1900 to 6.7 Gt a century later and to about 9.7 Gt in 2015

 

American productivities, highest in the world, rose from less than 1 t of coal per miner per shift in 1900 to a nationwide hourly average of about 5 t per worker, with specific rates ranging from 2 to 3 t/h in Appalachian underground mines to about 27 t/h in surface mines of the Powder River Basin in Montana and Wyoming (USEIA 2016a).

 

Coal remained the leading transportation fuel during the first half of the 20th century, but the conversion of locomotives and ships to diesel engines (begun, respectively, in the century’s first and third decade) accelerated after World War II.

 

The specific use of metallurgical coke (kg of coke/kg of hot metal) has been declining for decades, but the worldwide expansion of pig iron smelting, from about 30 Mt in 1900 to about 1.2 Gt by the year 2015, pushed the coking coal output to roughly 1.2 Gt (Smil 2016).

 

Until 1983 the USSR was a larger coal producer than the United States, but after the state collapsed, Russian coal extraction declined as natural gas and crude oil filled the need. India is now the third world’s largest producer (in 2014 only one-sixth of the Chinese output), but its coal is of much lower quality than the Chinese and U.S. deposits and productivity of its extraction is still dismal.

 

Hydrocarbons.  At the beginning of the 20th century oil was produced in quantity in just a handful of countries, and the fuel provided just 3% of all energy from fossil fuels; by 1950 that share was about 21%, energy content of crude oil surpassed that of coal by 1964 and it peaked in 1972 at about 46% of all fossil fuels. The two common impressions—that the 20th century was dominated by oil, much as the 19th century was dominated by coal—are both wrong: wood was the most important fuel before 1900 and, taken as a whole, the 20th century was still dominated by coal (Smil 2010a). My best calculations show coal about 15% ahead of crude oil (approximately 5.2 YJ vs. 4 YJ), and even when nonenergy applications of processed oil (in lubricants, paving materials) are included, coal would still be just ahead of liquid hydrocarbons, or, because of the inherent uncertainties when converting extracted masses to common energy equivalents, the 20th century’s cumulative production of the two fuels would be roughly equal.

 

Offshore rigs (mostly semisubmersible designs) work in waters more than 6550 feet deep (2,000 m). Production platforms installed at major offshore fields are among the most massive structures ever built.

 

The transportation of crude oil has been transformed by seamless steel pipes fashioned into large-diameter trunk pipelines eventually capable of spanning continents. These pipelines are the most compact, most reliable, cleanest, and safest mode of bulk transportation on land.

 

Global oil production grew roughly 200-fold during the 20th century; by 2015 it was (at over 4.3 Gt) about 20% higher than in the year 2000, and since 1964, when its energy content surpassed that of coal extraction, it has been the world’s mostly commonly used fuel. Second, oil is now produced on every continent and from offshore wells in every ocean except the high Arctic seas and Antarctica, and from fields as deep as 23,000 feet (7 km) belowground on land, while Brazil’s Tupi deposits are 2.1 km below the surface of the Atlantic and then nearly 5 km below the ocean’s bottom. Third, oil is the single most valuable traded commodity.

 

Although oil extraction is widely distributed, the world’s largest oil fields were discovered on land, in the Persian Gulf region between 1927 (Kirkuk in Iraq) and 1958 (Ahwaz in Iran). Al-Ghawar, the world’s largest oil field in the eastern province of Saudi Arabia, has been producing since 1951, and the second largest field, Kuwaiti al-Burqan, has been in operation since 1946. Nothing can change this fundamental reality: in 2015 almost half of known reserves of conventional (liquid) oil were in the region, which is, unfortunately, also the world’s most prominent source of complex conflicts and chronic political instability (BP 2016).

 

The 20th century saw a 375-fold increase in total energy derived annually from the fossil fuel natural gas.  A small but very important new market was the use of natural gas as both feedstock and fuel for the synthesis of ammonia—the most important nitrogenous fertilizer, now mostly used as a feedstock to produce solid urea—and for the production of plastics.

 

Because of the rapid growth of populations in rural areas of low-income countries (where there has been no, or a very limited, access to modern energies), the world now consumes more fuelwood and charcoal than ever. According to my best estimates, gross energy in traditional biofuels reached about 45 EJ in the year 2000, almost exactly twice as much as in 1900.

 

This means that in the year 2000 biofuels supplied roughly 12% of the world’s primary energy  (whereas in 1900 they provided 50%).

 

Unfortunately, even that harvest, equivalent to about 1 Gt of oil, has not been enough: with hundreds of millions of people in rural areas of low-income countries of Africa, Asia, and Latin America still burning biomass fuels, the demand for fuelwood and charcoal has been a leading cause of deforestation, most acutely in Sahelian Africa, Nepal, India, interior China, and much of Central America.

 

It is wrong to think about wood only in relation to forests because in many low-income countries large shares of woody matter are gathered by families (in most cases by women and children) from small tree groves and bushes, from tree plantations (rubber, coconut), and from roadside and backyard trees. Surveys in Bangladesh, Pakistan, and Sri Lanka showed that such non-forest wood accounted for more than 80% of all combustion (RWEDP 1997). At least a fifth of all crop residues produced in low-income countries is still burned, and dried dung remains important in parts of Asia, but charcoal has become a preferred biofuel.

 

China and India are the world’s largest consumers of traditional biofuels, followed by Brazil and Indonesia, but in relative terms there is no match for sub-Saharan Africa, where at the 20th century’s end some countries derived more than 80% of rural energy from wood and crop residue, compared to 25% in Brazil and less than 10% in China (Smil 2013a).

 

The global production of liquid biofuels reached about 75 Mt of oil equivalent in 2015, accounting for about 1.8% of energy extracted annually from crude oil (BP 2016). Scaling this industry to supply a significant share of the world’s liquid biofuels is, bluntly put, delusionary (Giampietro and Mayumi 2009; Smil 2010a).

 

PV generation has even lower capacity factors than wind (without tracking in cloudier climates just 11–15%, even in Arizona about 25%), and in 2015 the global electricity generation was only about 30% of the total delivered by wind turbines (fig. 5.16). Again, the industry’s growth has not been a gradual, organic process but a promotion driven by government subsidies:

 

2015 fossil fuels still accounted for 86% of the world’s primary energy, just 4% less than a generation ago, in 1990 (BP 2016).

 

This transformation brought enormous advances in agricultural productivity and crop yields;

 

Coal mining grew 100-fold, from 10 Mt to 1 Gt, between 1810 and 1910; it reached 1.53 Gt in 1950, 4.7 Gt in the year 2000, and 8.25 Gt in 2015 before it declined a bit to about 7.9 Gt in 2015 (Smil 2010c; BP 2016). Crude oil extraction rose about 300-fold, from less than 10 Mt in the late 1880s to just over 3 Gt in 1988; it was 3.6 Gt in the year 2000 and almost 4.4 Gt in 2015 (BP 2016). Natural gas production rose 1,000-fold, from less than 2 Gm3 in the late 1880s to 2 Tm3 by 1991; it was 2.4 Tm3 in 2000 and 3.5 Tm3 in 2015. During the twentieth century the global extraction of fossil energies rose 14-fold in aggregate energy terms.

 

Another way to appreciate the aggregate size of modern energy flows is to compare them with traditional uses, in both absolute and relative terms. Best estimates show the worldwide total of biomass fuel consumption rising from around 700 Mt in 1700 to about 2.5 Gt in the year 2000. This would be from about 280 Mt to 1 Gt in terms of oil equivalent, less than quadrupling in three centuries (Smil 2010a). During the same time, the extraction of fossil fuels rose from less than 10 Mt to about 8.1 Gt of oil equivalent, about an 800-fold expansion (fig. 6.2). In gross energy terms, the global supply of biofuels and fossil fuels was about the same in 1900 (both at roughly 22 EJ); by 1950 fossil fuels supplied nearly three times as much energy as wood, crop residues, and dung; and by the year 2000 the difference was nearly eightfold.

 

But when adjusted for actually delivered, useful energy, the difference in the year 2000 was nearly 20-fold.

 

The energy needs of foraging societies were dominated by the provision of food, and their annual consumption averages did not go above 5–7 GJ/capita. Ancient high cultures added slowly rising energy use for better shelters and clothing, for transportation (energized by food, feed, and wind), and for a variety of manufactures (with charcoal prominent). New Kingdom Egypt averaged no more than 10–12 GJ/capita, and my best estimate for the early Roman Empire is about 18 GJ/capita (Smil 2010c). Early industrial societies easily doubled the traditional per capita energy use. Most of that increase went into coal-fueled manufactures and transportation. Malanima (2013b) put the European averages at about 22 GJ/t in 1500, followed by a stagnation at 16.6–18.1 GJ/t until 1800.

 

Kander (2013) puts the mean for England and Wales rising from 60 GJ/capita in 1820 to 153 GJ/capita by 1910, while during the same period the German rate more than quintupled (from 18 to 86 GJ/capita) and the French rate tripled (18 to 54 GJ/capita), but the Italian rates rose by only 20% (from 10 to just 22 GJ/capita). For comparison, the average U.S. rate rose from less than 70 GJ to about 150 GJ between 1820 and 1910. A century later all richer European countries were above 150 GJ/capita, the United States was above 300 GJ/capita, and as the average consumption rates rose, their composition changed

 

When in 1900 a Great Plains farmer held the reins of six large horses while plowing his wheat field, he controlled—with considerable physical exertion, perched on a steel seat, and often enveloped in dust—no more than 5 kW of animate power. A century later his great-grandson, sitting high above the ground in the air-conditioned comfort of his tractor cabin, controlled effortlessly more than 250 kW of diesel engine power.

 

In 1900 an engineer operating a coal-fired locomotive pulling a transcontinental train at close to 100 km/h commanded about 1 MW of steam power, the maximum performance permitted by manual stoking of coal. By the year 2000, pilots of a Boeing 747 retracing the transcontinental route 11 km aloft could choose an auto-mode for a large part of the journey as four gas turbines developed up to about 120 MW and the plane flew at 900 km/h (Smil 2000a).

 

Fossil fuels and electricity have become indispensable inputs in modern farming. They are used directly to power machines and indirectly to build them, to extract mineral fertilizers, to synthesize nitrogenous compounds and a still expanding variety of protective agrochemicals (pesticides, fungicides, herbicides), to develop new crop varieties, and most recently to energize the electronics used in many functions that now support precision farming. Fossil fuels and electricity have brought higher and more reliable yields. They have displaced virtually all draft animals in all rich countries and greatly reduced their importance in the poor ones, and the replacement of muscles by internal combustion engines and electric motors sustained the reduction of labor started by preindustrial farming advances.

 

Machinery’s embedded energy cost is a fraction of the energy used directly to run tractors, combines, and other harvesters, to pump water, to dry grain, and to process crops. Because of their inherently high efficiency, diesel engines have come to dominate most of these uses, but gasoline and electricity are also major energy inputs.

 

The mechanization of field work has been the main reason behind the rising labor productivity rise and the reduction of agricultural populations: a strong early 20th-century Western horse worked at a rate equal to the labor of at least six men, but even early tractors had power equivalent to 15–20 heavy horses, and today’s most powerful machines working on Canadian prairies rate up to 575 horsepower (Versatile 2015).

 

Average labor inputs to American wheat farming from about 30 h/t of grain in 1800 to less than 7 h/t in 1900; by the year 2000 the rate was down to about 90 minutes.

 

American draft horses reached their highest number in 1915, at 21.4 million animals, but mule numbers peaked only in 1925 and 1926, at 5.9 million (USBC 1975).

 

The synthesis of nitrogenous fertilizers claimed about 6.3 EJ of energy, or just over 1% of the global energy supply (Smil 2014b).

 

No other energy use offers such a payback as higher crops yields resulting from the use of synthetic nitrogen: by spending roughly 1% of global energy, it is now possible to supply about half of the nutrient used annually by the world’s crops. Because about three quarters of all nitrogen in food proteins come from arable land, almost 40% of the current global food supply depends on the Haber-Bosch ammonia synthesis process. Stated in reverse, without Haber-Bosch synthesis the global population enjoying today’s diets would have to be almost 40% smaller.

 

Synthetic nitrogen provides about 70% of all nitrogen inputs in China. With over 70% of the country’s protein supplied by crops, roughly half of all nitrogen in China’s food comes from synthetic fertilizers. In its absence, average diets would sink to a semi-starvation level—or the currently prevalent per capita food supply could be extended to only half of today’s population.

 

The mining of potash (10 GJ/t K) and phosphates and the formulation of phosphatic fertilizers (altogether 20 GJ/t P) would add another 10% to that total.

 

The global extent of irrigated farmland roughly quintupled during the 20th century, from less than 50 Mha to more than 250 Mha, reaching about 275 Mha by 2015 (FAO 2015a). In relative terms this means that about 18% of the world’s harvested cropland is now irrigated, about half of it with water pumped mostly from wells, with about 70% of the irrigated land being in Asia. Where irrigation draws water from aquifers, the energy cost of pumping (using mostly diesel engines or electric pumps) invariably accounts for the largest share of the overall (direct and indirect) energy cost of crop cultivation. Irrigation still supplies most of the withdrawn water to furrows,

 

During the 20th century, as the world population grew 3.7 times and the harvested area expanded by about 40%, anthropogenic energy subsidies soared from only about 0.1 EJ to almost 13 EJ. As a result, in 2000 an average hectare of cropland received roughly 90 times more energy subsidies than it did in 1900     Howard Odum (1971, 115–116):

 

A whole generation of citizens thought that the carrying capacity of the earth was proportional to the amount of land under cultivation and that higher efficiencies in using the energy of the sun had arrived. This is a sad hoax, for industrial man no longer eats potatoes made from solar energy, now he eats potatoes partly made of oil.

 

In 1900 the gross global crop output (before storage and distribution losses) provided only a small margin above the average human food needs, which meant that a large share of humanity had only barely adequate or inadequate nutrition, and that the share of the harvest that could be used to feed animals was minimal. Greatly increased energy subsidies allowed the new staple cultivars (hybrid corn, introduced during the 1930s, and short-stalked wheat and rice cultivars, first adopted during the 1960s) to reach their full potential, resulting in rising yields of all staples and in an overall six-fold increase in harvested food energy (Smil 2000b, 2008).

 

Once energized by coal and steam power, traditional manufacturers could turn out larger volumes of good-quality products at lower prices. This achievement was a necessary precondition for mass consumption. The availability of an inexpensive and reliable supply of mechanical energy also allowed increasingly sophisticated machining. In turn, this led to more complex designs and greater specialization in the manufacture of parts, tools, and machines. New industries energized by coal, coke, and steam were set up to supply national and international markets with unprecedented speed. The making of high-pressure boilers and pipes started after 1810. The production of rails and railway locomotives and wagons rose rapidly after 1830, as did the making of water turbines and screw propellers after 1840. Iron hulls and submarine telegraph cables found large new markets after 1850, and commercial ways of making inexpensive steel—first in Bessemer converters after 1856, then in open-hearth furnaces during the 1860s— found new large markets for finished products ranging from cutlery to rails and from plows to construction beams.

 

Steelmaking itself is increasingly done using electric arc furnaces, and new lighter but stronger steels have found many uses, particularly in the auto industry (Smil 2016). And before terminating this list, which could run for pages, I must stress that without electricity there could be no large-scale micromachining producing parts with exacting tolerances for such now ubiquitous applications as jet engines or medical diagnostic devices, and, of course, there would be neither accurate electronic controls nor the omnipresent computers and billions of telecommunication devices now in global use.

 

The smelting of iron ore in blast furnaces, followed by steelmaking in basic oxygen furnaces, and the use of recycled steel in electric arc furnaces dominate steel production. The massive growth of steel production would have been impossible without the much larger and more efficient blast furnaces.

 

Few production structures with a medieval pedigree remain as important for the functioning of modern civilization as do blast furnaces. As noted in chapter 5, Bell’s 1840 redesign quintupled their internal volume, bringing it to 250 m3. By 1880 the largest furnace surpassed 500 m3; it reached 1,500 m3 by 1950, and by 2015 the record inner volumes were between 5,500 and 6,000 m3 (Smil 2016). The resulting increases in productivity brought the output of hot metal from 50 t/day in 1840 to more than 400 t/day by 1900. The 1,000 t/day mark was approached before World War II, and today’s largest furnaces produce around 15,000 t/day, with the record rate at POSCO’s Pohang 4 furnace, in South Korea, about 17,000 t/day. The mass and energy flows needed to operate large blast furnaces and associated oxygen furnaces are prodigious (Geerdes, Toxopeus, and Van der Vliet 2009; Smil 2016). A blast furnace producing daily 10,000 t/day of iron and supplying an adjacent basic oxygen furnace will need 5.11 Mt of ore, 2.92 Mt of coal, 1.09 Mt of flux materials, and nearly 0.5 Mt of steel scrap.

 

Modern furnaces now produce hot metal continuously for 15–20 years before their refractory brick interior and their carbon hearth are relined.

 

In 1850, before the beginning of modern steel production, fewer than 100,000 t of the metal were produced annually in artisanal ways, a mere 75 g/year/capita.

 

My calculations show that in 2013 the worldwide production of iron and steel required at least 35 EJ of fuels and electricity, or less than 7% of the total of the world’s primary energy supply, making it the world’s largest energy-consuming industrial sector (Smil 2016). This compares to 23% for all other industries, 27% for transportation, and 36% for residential use and services.

 

Power densities of traditional urban energy supply and use

 

With average per capita food intakes of about 9 MJ/day originating, as preindustrial diets usually did, overwhelmingly (90%) from plant foods, and with typical grain yields of just 750 kg/ha, a traditional city of 500,000 people would have needed about 150,000 ha of cropland. In a colder climate, annual fuel (wood and charcoal) needs would have been about 2 t/capita. If supplied on a sustainable basis from forests or from fuelwood groves with annual yields of 10 t/ha, around 100,000 ha would have been needed to fuel the city. A densely populated city of that size occupied as little as 2,500 ha and had to rely on an area about 100 times its size for its food and fuel. In terms of average power densities, this example implies about 25 W/m2 for total energy consumption and 0.25 W/m2 for the supply. The actual range of power densities was fairly large. Depending on their food intakes, cooking and heating practices, energy requirements for small manufactures, and combustion efficiencies, the total energy consumption of preindustrial cities was between 5 and 30 W/m2 of their area. The sustainable production of fuel from nearby forests and woodlots yielded anywhere between 0.1 and 1 W/m2. Consequently, cities had to rely on cropped and wooded areas 50–150 times larger than their own size—and the absence of powerful and inexpensive prime movers limited the capacity to transport food and fuel from distant regions, putting pressure on the plant resources of the surrounding areas.

 

Common rural environmental burdens include very high concentrations of indoor air pollutants (particularly fine particulate matter) from unvented biomass combustion, inadequate heating in colder climates, unsafe water supplies, poor personal hygiene, dilapidated, overcrowded housing, and minimal or no opportunities to see the children properly educated. Moreover, the drudgery of field labor in the open is seldom preferable even to unskilled industrial work in a factory. In general, typical factory tasks require lower energy expenditures than does common farm work.

 

Caro describes the drudgery, and danger, of ironing with heavy wedges of metal heated on wood stoves, the endless pumping and carrying of water for cooking, washing, and animals, the grinding of feed, and sawing wood. These burdens, which fell largely on women, were much harder than the typical labor requirements in poor countries as the Hill County farmers of the 1930s strove to maintain a much higher standard of life and run much larger farming operations than peasants in Asia or Latin America. For example, the water needs for a family of five came to nearly 300 t/year, and to supply them required an equivalent of more than 60 eight-hour days and walking about 1,500 miles (2,500 km). Not surprisingly, nothing could have been as revolutionary in the life of these people as the extension of transmission lines.

 

The Persian Gulf basin is an unparalleled singularity: it has 12 of the world’s 15 largest oil fields, and in 2015 it contained about 65% of the world’s reserves of liquid oil (BP 2015). These riches explain the lasting interest in the region’s stability. This desire is immensely complicated by the near chronic disarray of the area, which is made up of artificial states separated by arbitrary borders cutting across ancient ethnic group and containing complex religious enmities.

 

Firebombing of Tokyo, March 9–10, 1945.  The raid, the largest of its kind ever, involved 334 B-29 bombers that offloaded their bombs at low (about 600–750 m) altitude. Most of these were large, 230 kg cluster bombs, each releasing 39 M-69 incendiary bombs filled with napalm, a mixture of polystyrene, benzene, and gasoline; simple 45 kg jelled-gasoline and phosphorus bombs were also used. About 1,500 t of incendiary compounds were dropped on the city, and their total energy content (assuming an average density of napalm of 42.8 GJ/t) amounted to about 60 TJ, nearly as much as the power of the Hiroshima bomb. But energy released by burning napalm was only a small fraction of the total released by the city’s incinerated wooden buildings. According to the Tokyo Metropolitan Police Department, the fire destroyed 286,358 buildings and structures (U.S. Strategic Bombing Survey 1947), and conservative assumptions (250,000 wooden buildings, just 4 t of wood per building, 18 GJ/t of dry lumber) result in in some 18 PJ of energy released from the combustion of the city’s wooden housing, two orders of magnitude (300 times) larger than the energy of incendiary bombs. The destroyed area amounted to about 4,100 ha, and at least 100,000 people died. For comparison, the totally destroyed area in Hiroshima was about 800 ha, and the best estimate of immediate deaths was 66,000.

 

The Hiroshima bomb released 63 TJ of energy, about half of it as the blast and 35% as thermal radiation (Malik 1985).  The Nagasaki bomb released about 92 TJ.

 

1961 Soviet tsar bomba tested in 1961 209,000,000,000,000,000

 

Biomass supply in wooden-era societies was not a matter of assured renewability: excessive tree cutting followed by destructive soil erosion on vulnerable slope lands destroyed the conditions for sustainable forest growth over large areas of the Old World, especially around the Mediterranean Sea and in North China.

 

Figure 7.1 The prolonged dominance of human labor, the slow diffusion of water- and wind-driven machines,

 

In contrast to its massed deployment to accomplish remarkable feats of heavy construction, old high cultures, whether based on slave, unpaid labor, or mostly free labor, never took the steps to the really large-scale manufacture of goods.

 

Outside Europe and North America, relatively small-scale, artisanal manufactures remained the norm until the 20th century. The lack of inexpensive land transportation was obviously a major factor militating against mass production. The costs of distribution beyond a relatively small radius would have surpassed any economies of scale gained by centralized manufacturing. And many ancient construction projects did not really require extraordinarily massive labor inputs either. Several hundred to a few thousands of forced unpaid laborers working for only two to five months every year could erect enormous religious structures or defensive walls, dig long irrigation and transportation canals, and build extensive dikes over a period of just 20–50 years.

 

My reconstruction of global energy transitions shows coal (replacing wood) reaching 5% of the global market around 1840, 10% by 1855, 15% by 1865, 20% by 1870, 25% by 1875, 33% by 1885, 40% by 1895, and 50% by 1900. The sequence of years needed to reach these milestones was 15–25–30–35–45–55–60. The intervals for oil replacing coal, with 5% of the global supply reached in 1915, were virtually identical: 15–20–35–40–50–60 (oil will never reach 50%, and its share has been declining). Natural gas reached 5% of the global primary supply by 1930 and 25% of it after 55 years, taking significantly longer to reach that share than coal or oil.

 

The similar progress of three global transitions—it takes two or three generations, or 50–75, years for a new resource to capture a large share of the global energy market—is remarkable because the three fuels require different production, distribution, and conversion techniques and because the scales of substitutions have been so different: going from 10% to 20% for coal required increasing the fuel’s annual output by less than 4 EJ, whereas going from 10% to 20% of natural gas needed roughly an additional 55 EJ/year (Smil 2010a).

 

The two most important factors explaining the similarities in the pace of transitions are the prerequisites for enormous infrastructural investment and the inertia of massively embedded energy systems.

 

Fossil fuels being replaced by new conversions of renewable energy flows), will proceed at a similar pace, the odds are highly in favor of another protracted process. In 2015 the two new renewable ways of electricity generation, solar (at 0.4%) and wind (at 1.4%), were still below 2% of the world’s primary energy supply (BP 2016

 

Even then we would still face the challenges of replacing billions of tonnes of high energy-density liquid fuels in transportation, and producing pig iron, cement, plastics, and ammonia without any fossil carbon.

 

Every new inanimate energy converter became eventually much lighter and more efficient. The steady decline in the weight/power ratio of leading prime movers means that the best internal combustion engines now weigh less than 1/1,000 of what the equally powerful draft animals or early steam engines weighed.

 

Annualized power rose from about 650 GW in 1800 to 12.2 TW in 2000, a nearly 20-fold rise in two centuries, and by 2015 it had increased by more than 40%, to about 17.5 TW. The increase in fossil fuel extraction between 1800 and 2000 was nearly 900-fold, from less than 0.4 EJ to more than 300 EJ.

 

The densities of foraging populations spanned a wide range, but, with the exception of some maritime cultures, they never surpassed one person per square kilometer. Even the least productive shifting farming raised this rate at least ten times. Permanent cropping resulted in yet another tenfold increase. Intensification of traditional farming required higher energy inputs. As long as animate labor remained the sole prime mover of field work, the share of the population engaged in crop cultivation and animal husbandry had to remain very high, more than 80%, commonly over 90%. The net energy returns of intensive farming involving irrigation, terracing, multi-cropping, crop rotation, and fertilization were generally lower than those in extensive agriculture but allowed unprecedented population densities.

 

In contrast to the slow, cumulative transformations of traditional societies, the socioeconomic consequences of fossil fuel-based industrialization were nearly instantaneous. The substitution of fossil fuels for biomass fuels and the later replacement of animate energies by electricity and internal combustion engines created a new world within just a few generations

 

Electricity must be singled out for its many unique roles. Reliance on this most flexible and most convenient form of energy has rapidly developed into an all-encompassing dependence. Without electricity, modern societies could not farm or eat the way they do: electricity powers compressors in both ammonia plants and domestic refrigerators. They could not prevent disease (now controlled with refrigerated vaccines) and take care of the sick (with diagnoses dependent on electricity-powered machines, from venerable x-ray machines to the latest MRI, and with extensive monitoring in intensive care units), control their transportation networks, or handle their enormous volume of information (with data centers becoming some of the largest point consumers of electricity) or urban sewage.

 

As the per capita availability of grain rose, enough cultivable land could be set aside to grow good-quality animal feed: in the United States its supply eventually claimed nearly 25% of all farmland,

 

The administration of distant territories and trading and military ventures were made difficult not only by the slowness but also by the unreliability of both land and sea travel. Sailings from Rome to Egypt, the empire’s largest producer of surplus grain, could take as little as a week—and as long as three months or more (Duncan-Jones 1990).

 

And as late as 1800 English ships had to wait, sometimes for weeks, for the right wind to take them into Plymouth Sound (Chatterton 1926).

 

Even a basic list of the universal infrastructures of a high-energy civilization is a lengthy one: coal mines, oil and gas fields, thermal power stations, hydroelectric dams, pipeline networks, ports, refineries, iron and steel mills, aluminum smelters, fertilizer plants, countless processing, chemical, and manufacturing enterprises, railroads, multilane highways, airports, skyscraper-dominated downtowns, an extensive suburbia.

 

The two most obviously worrisome consequences of the dependence on high energy flows is the restriction of choices (that is, the impossibility of abandoning existing practices without causing numerous massive dislocations) and degradation of the environment. The first phenomenon is perhaps best illustrated by the impossibility of cutting off high-energy subsidies in modern farming without profoundly transforming the entire society.

 

Replacing the existing American field machinery by draft animals would require horse and mule stock at least ten times as large as its record numbers from the early 20th century. Some 300 Mha, or twice the total area of U.S. arable land, would be needed just to feed the animals, and masses of urbanites would have to leave the cities for farms.

 

China’s dependence on fossil energies in food production is even higher.

 

Special interest groups have become certainly the most important macroparasites in modern affluent countries. They range from various professional associations and labor unions with restricted entries to oligopolistic industrial cartels and lobbies. By shaping, and often vetoing or disabling, government policies and fixing prices, these groups work against the optimized use of all resources, and inevitably they have a notable effect on the development of energy resources and on the efficiency of their use. They have been behind the heavy subsidies that for decades have gone to producers of various fossil fuels and nuclear electricity, and now they are behind the new subsidies promoting the construction of solar photovoltaics and wind turbines

 

metallurgical coke (now requiring every year about 1 Gt of coking coal and used not just as the source of energy for reducing iron oxides but for its structural role in supporting charged iron ore and flux in blast furnaces producing annually more than 1 Gt of iron); lubricants (essential for functioning of both stationary and transportation machines); and paving materials (inexpensive asphalt).

 

Georgescu-Roegen’s (1975, 379) was not hopeful: “Perhaps, the destiny of man is to have a short, but fiery, exciting and extravagant life rather than a long, uneventful and vegetative existence. Let other species—the amoebas, for example—which have no spiritual ambitions inherit the earth still bathed in plenty of sunshine.

 

Senancour (1770–1846), Man perisheth. That may be, but let us struggle even though we perish; and if the nothing is to be our portion, let it not come to us as a just reward. (Senancour 1901 [1804], 2:187)

 

Energy and the evolution of homo sapiens

 

Bipedalism—essentially a sequence of arrested falls—is inherently unstable and clumsy: “Human walking is a risky business. Without split-second timing man would fall flat on his face; in fact, with each step he takes, he teeters on the edge of catastrophe” (Napier 1970). Besides making us prone to musculoskeletal injuries, bipedalism also leads to age-related bone loss, osteopenia (lower than normal bone density), and osteoporosis (Latimer 2005). Many answers have been offered to the obvious question of why we do this. It may be that bipedalism offered sufficient biomechanical advantage and energy efficience:  after measuring energy expenditure in walking chimpanzees and adult humans, it was found that human walking costs about 75% less energy than both quadrupedal and bipedal walking in chimpanzees.

 

Bipedalism started a cascade of enormous evolutionary adjustments. Upright walking liberated hominin arms for carrying weapons and for taking food to group sites instead of consuming it on the spot. But bipedalism was necessary to trigger hand dexterity and tool use. Hashimoto and co-workers (2013) concluded that adaptations underlying tool use evolved independently of those required for human bipedalism because in both humans and monkeys, each finger is represented separately in the primary sensorimotor cortex, just as the fingers are physically separated in the hand. This creates the ability to use each digit independently in the complex manipulations required for tool use. But without bipedalism it would be impossible to use the trunk for leverage in accelerating the hand during toolmaking and tool use. Bipedalism also freed the mouth and teeth to develop a more complex call system as the prerequisite of language (Aiello 1996). These developments required larger brains whose energy cost eventually reached three times the level for chimpanzees, accounting for up to one-sixth of the total basal metabolic rate (Foley and Lee 1991; Lewin 2004). The average encephalization quotient (actual/expected brain mass for body weight) is 2–3.5 for primates and early hominins, while for the humans is a bit higher than 6. Three

 

A higher encephalization quotient was critical for the rise of social complexity (which raised the survival odds and set hominins apart from other mammals) and was closely related to changes in the quality of food consumed. The brain’s specific energy need is roughly 16 times that of skeletal muscles, and the human brain claims 20–25% of resting metabolic energy, compared to 8–10% in other primates and just 3–5% in other mammals

 

That expanded the range and quality of available food, and also because its adoption brought many physical changes (including smaller teeth and a less voluminous digestive tract) and behavioral adjustments (such as the need to defend stores of accumulated food, which promoted protective female-male bonds) that led eventually to complex socialization, sedentary lives and “self-domestication.

 

Simplistic theorizing about affluent foragers ignored both the reality of much of the hard and often dangerous work in foraging and the frequency with which environmental stresses and infectious diseases ravaged most foraging societies. Seasonal food shortages forced the eating of unpalatable plant tissues and led to weight loss, and often to devastating famines. They also resulted in high infant mortality (including infanticide) and promoted low fertility rates. And, not surprisingly, a reanalysis of energy expenditure and demographic data collected in the 1960s found that the nutritional status and health of Dobe !Kung “were, at best, precarious and, at worst, indicative of a society in danger of extinction”. As Froment (2001) put it, “Coping with hazards and a heavy burden of diseases, hunter-gatherers do not live––and have never lived––in the Garden of Eden; they are not affluent, but poor, with limited needs and limited satisfaction.

 

Agricultural energy

 

Because the climate was too cold and CO2 levels were too low during the late Paleolithic and because these conditions changed with the subsequent warming, Richerson (2001) have argued that agriculture was impossible during the Pleistocene but mandatory during the Holocene. This argument is strengthened by the fact that between 10,000 and 5,000 years before the present, cropping evolved independently in at least seven locations on three continents (Armelagos 2005).

 

Helland (1980) illustrated low labor requirements in pastoral societies by pointing out the large numbers of major livestock species managed by a single herder in the East African setting: up to 100 camels, 200 cattle, and 400 sheep and goats. Khazanov (1984) lists similarly large figures for Asian pastoralists: two mounted shepherds for 2,000 sheep in Mongolia, an adult shepherd and a boy to tend 400–800 cattle in Turkmenia.

 

The appeal of low labor needs was one of the key reasons for the reluctance of many pastoralists to abandon their peregrinations and to become settled farmers. As a result, many nomadic societies existed for generations as neighbors of sedentary farmers and abandoned their herds only because of devastating drought or substantial loss of available pastures. Minimum per capita counts for pastoral subsistence were 5–6 heads of cattle, 2.5 3 camels, or 25–30 goats or sheep. Much higher cattle ownership among traditional Maasai (13–16 heads/capita) is explained by the minimum requirements for blood harvesting, done by piercing a tightened jugular vein and drawing 2–4 L every 5–6 weeks. During the periods of drought a herd of 80 cattle was needed to provide blood for a family of 5–6 people, or 13–16 animals/capita. In all cases, nomadic population densities were low compared with those of settled farmers, even where cropping replaced natural grasslands (entailing a much smaller loss of stored phytomass), this renewability was no guarantee of sustainability. Poor agronomic practices lowered soil fertility or caused excessive erosion or desertification, resulting in reduced yields and even the abandonment of cultivation. In most regions traditional farming progressed from extensive to intensive cultivation: its prime movers—human and animal muscles—remained unchanged for millennia, but cropping practices, cultivated varieties, and the organization of labor were greatly transformed.

 

The advancing intensification of farming sustained higher population densities, but it also demanded higher energy expenditures, not only for direct farming activities but also for such critical supportive measures as the digging of wells, the building of irrigation canals, roads, and food storage structures, and the terracing of fields. In turn, these improvements required more energy to make a larger variety of better tools and simple machines, powered by domestic animals or by water and wind. More intensive cultivation relied on animal labor at least for plowing, usually by far the most energy-demanding field task.

 

Keeping domestic animals required more intensive cropping to produce feeds. Animals were also used extensively for many other field tasks, as well as in grain threshing and milling, and were indispensable for the land-borne distribution of food. Their stabling, feeding, and breeding and the production of harnesses, shoes, and implements introduced new complexity and skills.

 

Tools and simple machines made field operations easier (thus providing a mechanical advantage) and faster, raised productivity, and enabled fewer people to grow more food, and the resulting energy surplus could be invested in structures and actions: without sickle and plow there would be no cathedrals—or no European voyages of discovery.

 

A great deal of traditional farming required heavy work, but such spells were often followed by extended periods of less demanding activities or seasonal rest, an existential pattern quite different from the nearly constant high mobility of foraging.

 

In all of the Old World’s high cultures the sequence started with plowing.  Plowing prepares the ground for seeding much more thoroughly than hoeing: it breaks up the compacted soil, uproots weeds, and provides loosened, well-aerated ground in which seedlings can germinate and thrive. Simple plows, which merely opened up a shallow furrow for seeds and left cut weeds on the surface,

 

The addition of a moldboard was by far the most important improvement. A moldboard guides the plowed-up soil to one side, turns it (partially or totally) over, buries the cut weeds, and cleans the furrow bottom for the next turn. A moldboard also makes it possible to till a field in one operation rather than by cross-plowing

 

In most soils plowing leaves behind relatively large clods, which must be broken before seeding. Hoeing will work, but it is too slow and too laborious. That is why harrows have been used by all old plow cultures. Their development led from primitive brush harrows to a variety of wooden or metal frames to which were fastened wooden pegs or metal teeth or disks. Inverted harrows or rollers were often used to further smooth the surface. After plowing, harrowing, and leveling, the ground was ready to be seeded.

 

Broadcast seeding by hand—wasteful and resulting in uneven germination—remained common in Europe until the 19th century. Simple drills that dropped seeds through a tube from a bin attached to a plow started to spread, first in northern Italy, during the late 16th century. Before long, many further innovations turned them into complex seeding machines.

 

The inter-cultivation of growing crops was done largely by hoeing. Manures and other organic wastes were brought to fields on carts, in wooden cisterns, or in buckets carried at the end of shoulder beams, a common practice in East Asia. Then the wastes were pitchforked, poured, or ladled onto the field.

 

Cutting with sickles was slow, and scythes, equipped with cradles for grain reaping, were preferred for harvesting larger areas (fig. 3.2). But sickle harvesting caused lower grain losses from ear shattering than broad sweeps with scythes, and the practice was retained in Asia for harvesting easily shattered rice.

 

Mechanical reapers came to American and European grain fields only in the early part of the nineteenth century (Aldrich 2002). Harvests were brought home as sheaves carried on heads, in panniers hung on shoulder beams or the sides of animals, and in wheelbarrows, carts, or wagons pushed or pulled by people or draft animals.

 

A considerable amount of energy went into crop processing. Grain spread on a threshing floor was beaten with sticks or flails; sheaves were hit against grates or pulled across special combs. Animals were used to tread on the spread grain or to pull heavy sleds or rollers over it. Before the adoption of crank-turned fans, winnowing, the separation of chaff and dirt from grain, was done manually with baskets and sieves. Tedious manual labor was also needed for grain milling before animals and water and windmills were used to mechanize the task. Oil was extracted from seeds by manual or animal-operated presses, as was the sweet juice from cane.

 

Tubers are low in protein (typically about one-fifth of that in cereals: some durum wheats have up to 13% protein, white potatoes just 2%). Legumes have twice as much protein as cereals (peas about 20%, beans and lentils between 18% and 26%) and soybeans more than three times as much (35–38%, with some cultivars reaching 40% by weight). But average yields of leguminous crops are a fraction of staple grain harvests: average yields for U.S. cereals were 2.5 t/ha in 1960 and 7.3 t/ha in 2013, with analogical pulse yields at 1.4 t/ha and 2.5 t/ha (FAO 2015a).

 

Dependence on cereal grains is thus a matter of clear energy advantages. Their primacy stems from the combination of fairly high yields, good nutritional value (high in filling carbohydrates, moderately rich in proteins), a relatively high energy density at maturity (roughly five times higher than for tubers), and a low moisture content suitable for long-term storage (in well-ventilated storage they do not spoil when grains contain less than 14.5% water).

 

Notes: Light work (L) consumes less than 20 kJ of food energy per minute for an average adult man. Moderate (M) exertions range up to 30 kJ/min, and heavy (H) ones up to 40 kJ/min. Analogous rates for women are about 30% lower.

 

Farmers had to set aside a portion of every harvest for the next year’s seeding. The combination of low yields and high seed waste in hand broadcasting could mean that as much one-third or even one-half of medieval grain crops had to be set aside. With increasing harvests these shares fell gradually to less than 15%. Some grains are eaten whole, but before actual food preparation (cooking or baking) most cereals are milled first, and in the process lose a significant share of the whole grain’s mass

 

Corn milling removes the tip cap, bran coat, and germ, leaving the endosperm, about 83% of the kernel. Corn flour for making tortillas and tamales, masa harina, is produced by nixtamalization, or the wet milling of kernels soaked in lime solution. This loosens the hulls from the kernels, softens the kernels by dissolving hemicellulose, reduces the presence of mycotoxins, and enhances the bioavailability of niacin (vitamin B3).

 

Storage losses on traditional farms—to fungi and insect infestations and to rodents able to access bins or jars—would commonly reduce the edible grain total by anywhere between a few percent to more than 10%. Improperly stored grain can be consumed by rodents. Even as recently as the mid-18th century a combination of seeding requirements and storage losses could have reduced the gross energy gain of European grains by around 25%.

 

Wheat was able to spread from the Middle East to all continents because it does well in many climates (in semi-deserts as well as in rainy temperate zones, and it is the leading food cultivar in the temperate zone between 30 and 60°N) and elevations (from sea level to as high as 3,000 m above sea level) and on many soils, as long as they are well drained.

 

Before the introduction of reapers and binders, manual grain harvesting was the most time-consuming task, taking three to four times longer than plowing, and it put clear limits on the maximum area manageable by a single family.

 

China’s traditional farming: planting and harvesting (between March and September) in China’s double-cropping area used virtually all (average of 94–98%) available labor. In parts of India the two peak summer months required more than 110% or even 120% of actually available labor, and a similar situation prevailed in other parts of monsoonal Asia. This common energetic bottleneck could be overcome only if all families worked arduously long hours—or by relying on migrant labor.

 

The maximum work periods for South China’s water buffaloes were two months of planting, harrowing, and grading in the early spring, six weeks of summer harvest, and a month of field preparation (again, plowing and harrowing) for a winter crop, altogether about 130–140 days, or less than 40% of the year.  In single-cropping regimes of northern Europe draft horses would do only 60–80 days of strenuous field work during the fall and spring plowing and the summer harvesting, but most of them were used extensively for transport. A typical working day ranged from just five hours for oxen in many African locations to more than ten hours for water buffaloes in Asian rice fields and for horses during European or North American grain harvests.

 

In dryland farming animal labor replaced human labor and sped up considerably many field as well as farmyard tasks, freeing people to pursue other productive activities or to work shorter hours. This prime mover shift did more than make the work quicker and easier; it also improved its quality, whether in plowing, seeding, or threshing. Second, irrigation and fertilization moderated, if not altogether removed, the two key constraints on crop productivity, shortages of water and nutrients. Third, growing a greater variety of crops, either by multi-cropping or in rotations, made traditional cultivation both more resilient and more productive.

 

Draft animals became indispensable for fertilization, both as the source of nutrients in manures and as the prime mover distributing them to crops. In many places they also energized irrigation. More powerful prime movers and better water and nutrient supply also brought more multi-cropping and crop rotations. In turn, these advances could support large numbers of more powerful animals as the three intensification paths were linked by mutually reinforcing feedback loops.

 

Traditional agricultures used animals for a variety of field and farmyard tasks, but plowing was undoubtedly the activity where they made the greatest difference. In general, the tractive force of working animals is roughly proportional to their weight, and other variables determining their actual performance include the animal’s sex, age, health, and experience, the efficiency of the harness, and soil and terrain conditions.

 

A typical draft is 15% of animal’s body weight but for horses it is up to 35% during brief exertions (about 2 kW) and even more during a few seconds of supreme effort (Collins and Caine 1926). The combination of large mass and relatively high speed makes horses the best draft animals, but most horses could not work steadily at the rate of one horsepower (745 W), and usually delivered between 500 and 850 W.

 

Horses are the most powerful draft animals. Unlike cattle, whose body mass is almost equally divided between the front and the rear, horses’ fronts are notably heavier than their rears (ratio of about 3:2), and so the pulling animal can take a better advantage of inertial motion than cattle. Except in heavy, wet soils, horses can work in fields steadily at speeds of around 1 m/s, easily 30–50% faster than oxen.

 

Horses also have better endurance (working 8–10 hours a day compared to 4–6 hours for cattle) and they live longer, and while both oxen and horses started working at 3–4 years of age, oxen lasted usually just for 8–10 years, while horses carried on commonly for 15–20 years.

 

A horse’s leg anatomy gives the animal a unique advantage by virtually eliminating the energy costs of standing. The horse has a very powerful suspensory ligament running down the back of the cannon bone and a pair of tendons (superficial and deep digital flexors) that can “lock” the limb without engaging muscles. This allows the animals to rest, even to doze, while standing, with hardly any metabolic cost, and to spend little energy while grazing. All other mammals need about 10% more energy when standing as compared to lying down.

 

Besides speeding up plowing and harvesting, animal labor also made it possible to lift large volumes of irrigation water from deeper wells. Animals were used to operate such food-processing machines as mills, grinders, and presses at rates far surpassing human capabilities. Relief from long hours of tiresome labor was no less important than the higher output rates, but more animal work required more cultivated land to grow feed crops.

 

In North America and in parts of Europe, the upkeep of horses at times claimed up to one-third of all agricultural area. In China’s traditional farming the cultivation of feed for draft animals claimed only about 5% of the annually harvested area. In India, fodder crops also accounted traditionally for about 5% of all cultivated land but most of this feed went to milking animals.

 

Indian or Chinese draft animals were clear energetic bargains. Many of them did not compete for crop harvests with people at all, while some preempted at most an area of cultivated land, which would grow food for one person a year. But their useful annual labor was the equivalent of three to five peasants working 300 days a year.

 

An average 19th-century European or American horses annual useful labor equal to about six working farmers, and the land used for its feeding (including all the nonworking animals) could have grown food for about six people. Strong, well-fed horses could perform tasks beyond human capacity and endurance.

 

American farmers were advised to feed their working horses 4.5 kg of oats and 4.5 kg of hay a day (Bailey 1908), which translates to about 120 M J/day. With an average power of 500 W, a horse would do about 11 MJ of useful work during six hours, and while an average male human would contribute less than 2 MJ, though he could not maintain steady exertion above 80 W and managed only brief peaks above 150 W, a horse could work steadily at 500 W and have brief peak pulls in excess of 1 kW, an effort that would require the exertions of a dozen men.

 

Horses could drag logs and pull out stumps when humans converted forests to cropland, break up rich prairie soils by deep plowing, or pull heavy machinery. There were additional energy costs of animal labor beyond maintaining a breeding herd and providing adequate feeding for field labor; these additional energy costs appeared above all in the making of harnesses and shoes and the stabling of the animals. But there were also additional benefits derived from the recycled manure and from milk, meat, and leather. Manure recycling has been important in all intensive traditional agricultures as the source of scarce nutrients and organic matter. In largely vegetarian societies, meat (including horsemeat in parts of Europe) and milk were valuable sources of perfect protein. Leather was used in making a large number of tools essential in farming and in traditional manufactures. And, of course, the animals were self-reproducing.

 

Up to 1,500 tonnes of water are needed to grow 1 tonne of wheat.

 

Irrigation was also necessary to cope with seasonal water shortages.

 

In river valleys with minimal stream gradients and on large cultivated plains, it has always been necessary to lift large volumes of surface or underground water. Many lifts were only across a low embankment, but often they had to surmount steep stream banks or come from deep wells. Unavoidable inefficiencies, aggravated by the rough finishing of moving parts and often by the absence of lubricants, prolonged the task. Irrigation powered by human muscles represented a large labor burden even in societies where tedious work was the norm.

 

Much ingenuity went into designing mechanical devices using animal power or water flow to ease that work—as well as to make higher lifts possible.

 

The energy cost of human-powered irrigation was extraordinarily high. A worker could cut a hectare of wheat with a cradle scythe in eight hours, but he would need three months (8 h/day) to lift half of its water requirement just 1 m from an adjoining canal or stream.

 

Winter wheat yield declines by half when a 20% shortfall in annual water supply is concentrated during the critical flowering period. A good late Qing dynasty harvest of 1.5 tonnes/hecare would be thus reduced by about 150 kg on a typical small field of 0.2 ha. Making up the deficit of 10 cm of rain by irrigation would require 200 t of water—but as ridge-and-furrow irrigation was only 50% efficient (owing to soil seepage and evaporation), the actual supply from a canal had to be twice as large. Lifting 400 t of water less than 1 m by a ladder operated by two treading peasants would take about 80 hours and require about 65 MJ of additional food energy, while the increased wheat yield would contain (after subtracting about 10% for seed and storage losses) about 2 GJ of digestible energy. Consequently, water ladder treading would return about 30 times more food energy than its food cost.

 

Cereal straws were a major potential source of nitrogen, but their direct recycling was limited. Unlike modern short-stalked plants, the traditional cultivated varieties yielded much more straw than grain, with straw-to-grain ratios commonly at 2:1. Plowing in so much plant mass could have been a great strain on many animals—but such a situation almost never arose. Only a small fraction of crop residues was returned directly to the soil because they were needed as animal feed and bedding (and only then recycled in manures), as household fuel, and as a raw material for construction and manufacturing.

 

In Chinese cities, high shares of human waste (70–80%) were recycled. Similarly, by the 1650s virtually all of Edo’s (today’s Tokyo) human wastes were recycled (Tanaka 1998). But the usefulness of this practice is limited by the availability of such wastes and their low nutrient content, and the practice entails much repetitive, heavy labor. Even before storage and handling losses, the annual yield of human wastes averaged only about 3.3 kg N/capita (Smil 1983). The collection, storage, and delivery of these wastes from cities to the surrounding countryside created large-scale malodorous industries, which even in Europe persisted for most of the 19th century before canalization was completed. Barles (2007) estimated that by 1869, Paris was generating annually about 4.2 Mt N, about 40% from horse manure and about 25% from human wastes;

 

The recycling of much more copious animal wastes—which involved cleaning of stalls and sties, liquid fermentation or composting of mixed wastes before field applications, and the transfer of wastes to fields—was even more time-consuming. And because most manures have only about 0.5% N, and pre-application and field losses of the nutrient had commonly added up to 60% of the initial content, massive applications of organic wastes were required to produce higher yields.

 

Every conceivable organic waste was used as a fertilizer in traditional farming: pigeon, goat, sheep, cattle, all other dung, composts made of straw, lupines, chaff, bean stalks, husks, and oak leaves.

 

Any theoretical estimates of nitrogen in recycled wastes are far removed from its eventual contribution. This is because of very high losses (mainly through ammonia volatilization and leaching into groundwater) between voiding, collection, composting, application, and eventual nitrogen uptake by crops. These losses, commonly of more than two-thirds of the initial nitrogen, further increased the need to apply enormous quantities of organic wastes. Consequently, in all intensive traditional agricultures, large shares of farm labor had to be devoted to the unappealing and heavy tasks of collecting, fermenting, transporting, and applying organic wastes.

 

Scale of traditional recycling (and hence the energies devoted to gathering, handling and applying the waste biomass) had to be so large because the organic materials applied to field or plowed in (as green manures) had very low nitrogen content: human and animal wastes are largely water, as are green manures; only oil cakes (residues after pressing edible oils) have relatively high nitrogen content. For comparison, urea, the leading modern synthetic fertilizer, contains 46% of nitrogen.

 

In China’s rice-growing areas applications of animal manure and night soil commonly averaged 10 t/ha during the late nineteenth and early twentieth centuries. Huge amounts of organic wastes were collected in cities and towns and moved to the countryside, creating a large waste-handling and transportation industry (box 3.9). This high intensity of Chinese manuring was admired by Western travelers, who were curiously unaware of how closely it was matched by earlier European experience. But no culture surpassed the highest known applications of organic wastes that supported intensive farming in the dike-and-pond region of Guangdong province in South China, between 50 and 270 t of pig and human excrements per hectare.

 

At least 10% of all labor in Chinese traditional farming was devoted to managing fertilizers.

 

The performance of the best twine-binding harvester was soon surpassed with the introduction of the first horse-drawn combines, marketed by California’s Stockton Works during the 1880s. Housers, the company’s standard combines after 1886, cut two-thirds of California’s wheat by 1900, when more than 500 machines were working in the state’s fields. The largest ones needed up to 40 horses and could harvest a hectare of wheat in less than 40 minutes—but they tested the limits of animal-powered machinery because harnessing and guiding up to 40 horses was an enormous challenge.

 

At its beginning, a farmer (80 W) working in a field was aided by about 800 W of draft power (two oxen); by its end, a farmer combining his Californian wheat field had at his disposal 18,000 W (a team of 30 horses).

 

In 1800 New England farmers (seeding by hand, with ox-drawn wooden plows and brush harrows, sickles, and flails) needed 150–170 hours of labor to produce their wheat harvest. By 1900 in California, horse-drawn gang-plowing, spring-tooth harrowing, and combine harvesting could produce the same amount of wheat in less than nine hours. In 1800 New England farmers needed more than seven minutes to produce a kilogram of wheat, but less than half a minute was needed in California’s Central Valley in 1900, roughly a 20-fold labor productivity gain in a century.

 

Naturally, these huge advances were only partially due to much higher efficiencies resulting from better machinery. The other principal reason for the rapidly rising energy returns of human labor was the substitution of horse power for human muscles. American inventors produced a vast range of efficient implements and machines, but they had only limited success in displacing draft animals as farming prime movers.

During the first two decades of the twentieth century the numbers of American horses and mules stayed around 25 million. Growing enough feed for their maintenance and work required about one quarter of America’s cultivated land

In 1910 America had 24.2 million farm horses and mules (and only 1,000 small tractors); in 1918 the draft animal herd peaked at 26.7 million and the number of tractors rose to 85,000. With an average daily need of 4 kg of grain for working animals and 2 kg of concentrate feed for the rest, the annual feed requirements were roughly 30 Mt of oats and corn. With grain yields of about 1.5 t/ha, this would have required planting at least 20 Mha to feed grains. To supply roughage, working horses needed at least 4 kg/day of hay, while the rest could be maintained with about 2.5 kg/day, requiring an annual total of roughly 30 Mt of hay. With average hay yields of about 3 t/ha, at least 10 Mha of hay had to be harvested. Land devoted to horse feed had to be no less than about 30 Mha, compared to around 125 Mha of annually harvested land, which means that America’s farm horse herd (working and nonworking animals) required almost 25% of the country’s cultivated land. The USDA’s (1959) calculation came up with a nearly identical total of 29.1 Mha.

In early 19th-century Europe the typical ratio of human/animal power capacity rose to around 1:15, but on the most productive American farms it was well above 1:100 during the 1890s. Human labor became a negligible source of mechanical energy, and the value of farmers’ work shifted mostly to management and control, tasks of low-power needs but high-output rewards.