A new world of energy

VACLAV SMIL

By the 1750s a few small regions of Western Europe (most notably parts of England, Wales, Scotland, Belgium, and Germany) were in the early stages of energy transition from plant fuels to fossil fuels and from animate prime movers to machines powered by combustion.

Great powers of the past that continued to rely on traditional energy sources and on animate prime movers were swiftly left far behind: although in aggregate terms China, with its large population, remained the world's largest economy until the 1880s, in per capita terms its economic product had stagnated for many generations and by 1913 it was less than a tenth of the British rate, which had quadrupled in less than a century. Similarly, the US economy, whose aggregate size was only about a tenth of India's value in 1820, was 2.5 times as large in 1913, and roughly eight times as rich in per capita terms.

In fundamental energetic terms the path of all late modernizers was thus preordained: in all cases the subsequent economic development had to follow the new Euro-American pattern of energy uses. This first became evident in the case of two new economic and military rivals to the Atlantic dominance - imperial Japan (starting in the 1870s) and Soviet Russia (starting in the 1920s). Both of these had to base their quest for influence on copying, adopting, and adapting Western techniques, resting on a mastery of new energy sources, such as coal, oil, and hydroenergy, and new energy con­verters, including internal combustion engines and electric motors and lights.

Historians have not ignored energy as a subject of their studies but it is undeniable that the production, transport, and conversion of fuels and electricity and multitudes of their final uses have been, at least in explicit terms, relatively neglected as a leading topic of modern historical inquiries. Of course, dealing with energy cannot be avoided. After all, as any physicist would point out, every action (from shining stars to marching armies) requires a conversion of one form of energy into another - and a thermo- dynamicist would add that all those conversions entail a loss of entropy, that is a decrease in energy's utility: no energy ever gets lost but every conversion degrades its quality, so say the two most fundamental physical laws.

So all historians are unwitting illustrators of the universal dictates of thermodynamic laws. In addition, although they rarely use the term power in their writings in its correct physical meaning, that is, as energy flow per unit of time (rather than as a common descriptor of might or military capacity), many historians have contributed detailed insights into the evolu­tion of prime movers (humans, draft animals, waterwheels, windmills, steam engines, internal combustion, rockets), energy sources (wood and charcoal, coal, oil and natural gas, thermal and hydroelectricity generation), and their uses. And the minority of historians dealing explicitly with the complex phenomena of energy has been joined by assorted interdisciplinarians con­tributing their knowledge and insight from other fields.

Consequently, we have been fairly well served with many detailed histor­ical studies of particular energy sectors.

Measuring energy advances

The breadth and quality of existing studies allows me not to recapitulate their findings, many of which provide fascinating but highly time- and place­specific details. Instead I will concentrate on just six universal measures, the understanding of which offers convenient keys to unlocking the meaning and import of past energy developments that have created the modern world through an unprecedented (and perhaps unrepeatable?) combination of advances. These are: energy density, power density, the maximum power of prime movers, the efficiency of energy conversions, the per capita con­sumption of useful energy, and, given the role of violent conflicts in history, the maximum energies of weapons.

Before 1750 most of these indicators had either remained (quantitatively and qualitatively) entirely unchanged or had shown only very slow, hesitant, and uneven advances during the millennia of recorded history. As a result, in 1700 Chinese peasants plowing grain fields, Indian builders erecting tall structures, Turkish merchants trading with distant places, Italian messengers carrying urgent news, English laborers smelting metals, and Inca families cooking their meals relied overwhelmingly on the same energy sources (low- energy-density plant fuels) and on the same prime movers (their own muscles or the exertions of often inadequately fed domesticated animals) that their predecessors used one or two millennia earlier.

There were some notable improvements but they took place very slowly: their use was largely restricted to only a few regions and did not amount to a fundamental qualitative change. By 1700 the best-built English waterwheels, whose origins go back to antiquity, became fairly efficient, as did (after nearly a millennium of evolution) the best windmill designs, but in many regions both devices did not greatly differ from their ancient prototypes.

By contrast to the unchanging, stagnating, or very slowly improving measures concerning the quality of energy sources and the performance of prime movers before 1750, the centuries after this brought great changes, introducing not only impressive quantitative gains but also fundamentally superior qualities. When seen from physical thermodynamic perspectives, the magnitude of those stunning, concatenated post-1750 changes can be traced by following the (rising or declining) trajectories of the fundamental measures and by comparing their outcomes at the end of the twentieth century with the performances that prevailed in the middle of the eighteenth century. Doing so requires, as energy studies invariably do, plenty of numbers, but I will deploy them in a user-friendly way, explaining the quantities and key assumptions in order to allow an interested reader to verify and replicate the calculations as a means toward a deeper understand­ing of fundamental energy realities that help to explain much of post-1750 world history.

Energy density

The first of my six universal measures is the fundamental matter of energy density, that is, the quantity of energy in a unit mass of matter: in modern scientific parlance this is joules (J) per gram (g) or (because both of those are such tiny units), megajoules (MJ) per kilogram (kg) or gigajoules (GJ) per tonne (t), and because both of those units go up three orders of magnitude in each step the number itself does not change. At a basic existential level, the importance of this variable is best grasped by comparing the energy densities of different types of food, because acquiring food determined much of the behavior of our species during the first 95 percent of its existence.

Getting that amount from fruits (as do chimpanzees, our closest primate relatives) would require finding 5 kg of them a day, an equivalent of more than thirty apples, sweet but with no fat and with only a low share of the protein needed for growth; not surprisingly, in their quest for protein chimpanzees also hunt colobus monkeys and “fish” for termites with blades of grass. And securing 13 MJ a day by foraging for nutritious underground biomass required digging 5 kg of tubers by hand or with simple digging sticks, and would still lead to deficiencies in protein and fat. And, obviously, fruits and tubers could be had in temperate and boreal regions only seasonally, restricting the extent of human habitation.

In contrast, killing of megaherbivores (mammoth, bison, giant elk) pro­vided hundreds of kilograms of fatty meat and internal organs whose high energy density made it possible to survive in boreal environments. Eating just i.5-2 kg of this food supplied not only all the needed energy but also plenty of the best-quality protein and fat. No wonder our ancestors were willing to take their chances by hunting animals whose body mass was ten or even fifty times greater than their own. These quests left them well fed and with enough free time to create admirable paintings of animals on the cave walls of Altamira, Lascaux, or Chauvet.

Advantages of higher energy density are no less discernible in the quest for fuels. Although some fossil fuels were used millennia ago, such as outcropping coal in Han China and bitumen pools in ancient Mesopotamia, the only fuels used by most societies since antiquity were a few varieties of phytomass: wood (usually burned after a period of air­drying), charcoal derived from it, crop residues (mostly cereal straws), roots, leaves, and grasses.

Between 1750 and 1900 there was an exponential expansion of coal mining in Europe and the United States (as well as the beginning of modern coal industries in Russia, China, and India), and during the last decade of the nineteenth century the energy content of extracted coal had surpassed the aggregate energy content of phytomass fuels burned worldwide to generate household and industrial heat. And by that time coal was not the only fossil fuel that was extracted on a commercial scale: new industries tapping crude oil emerged in the United States, Russia, Burma, and Indonesia and the United States pioneered the use of natural gas. Comparison of energy densities of these fuels makes it clear why modern civilization is so funda­mentally different from its predecessors.

All phytomass fuels are products of photosynthesis, with cellulose and lignin dominating their chemical composition and their energy density. As a result, their energy density is around 15 MJ per kg of air-dry matter. There was an option of using pyrolysis (slow heating in the absence of oxygen) to produce charcoal from wood (and sometimes from other biomass): every Old World civilization used it, from antiquity into the twentieth century. But while charcoal (a nearly pure carbon) has energy density of 29 MJ/kg (nearly twice that of air-dry wood), its traditional production required at least five units of wood per unit of charcoal, resulting in 60 percent loss of initially charged energy and making charcoal much more expensive than wood and beyond the reach of most people.

Anthracite, the best variety of coal, has pretty much the same high energy density as charcoal (29-30 MJ/kg), and that is why its accessible reserves were the first ones to be exhausted: its extraction peaked in Pennsylvania in 1917 (at about 90 megatonnes (Mt)Zyear) and by 1970 it was less than 10 percent of that peak. Most bituminous coals (used to generate electricity or to produce metallurgical coke) have densities 1.5 times that of air-dry wood (22-23 MJ/kg), while a kilogram of refined oil products (be it gasoline, kerosene, or diesel fuel) packs about 42 MJ, nearly three times the energy density of air-dried wood. Moreover, crude oil (petroleum) and liquid fuels produced from it by the refining of crude oil (gasoline, kerosene, diesel fuel, heavy oil) are particularly convenient: easy to transport (by pipelines, ships, barges, trucks), easy to store (underground or in aboveground tanks), flexible to use (for any stationary application and for land, water, and air transport). As a result, economies relying on liquid fuels are far more efficient and flexible than those dependent on coal or wood.

Many limits and possibilities of everyday options and historical developments can be explained by comparing energy densities: the more concentrated sources of energy offer many great advantages in terms of their extraction, transportation, and storage costs, flexibility of use and conversion options. An excellent historical example of economic (and hence social) constraints due to energy density concerns the efficiency of transatlantic crossings. For a single crossing in 1907, the ill-fated Lusitania (launched in that year as the biggest passenger ship ever and torpedoed by a German U-boat in May 1915) displacing 45,000 t, had to carry about 5,500 t of coal taking up 10,000 m³ of bunker space (assuming 23 GJ/t, 0.6 t/m³ of pulverized coal); if it were fueled by wood chips (15 GJ/t; 0.3 t/m³ of wood chips) its fuel load would have been more than 50 percent heavier and it would have taken up three times as much space - while an oil-fueled vessel of the same displacement would have needed only 3,000 t of fuel oil (42 GJ/t; 0.85 t/m³) in tanks taking up only a third of coal storage’s space.

Actual mass and volume differences would be even larger because of higher combustion efficiencies of coal versus wood and, even more so, of oil versus coal; but even this simplified comparison makes it clear why there could never be affordable wood-fueled transatlantic crossings that would have transported more than 40 million European emigrants to North and South America, and why the shipping companies switched from coal to fuel oil as soon as the latter high-energy density liquid became reliably available: as with so many technical innovations, that switch was led by the military, as the British Navy, prodded by Winston Churchill, began converting its ships to oil just before the First World War.

Power density

Many developments can be best explained by examining combinations of energy density and power density. The latter ratio has a number of meanings in different science and engineering disciplines, but I favor its use as a universal measure of energy flow (that is power) per unit of horizontal surface area (W/m²). The main advantage of this measure is its suitability for comparing virtually all natural or anthropogenic energy flows, and it is particularly revealing for highlighting the differences between traditional renewable energy flows (characterized by low to very low power densities) and modern fossil fuel uses (proceeding at medium to very high power densities). I will illustrate this by using two revealing examples. The first one demonstrates why mass-scale urbanization, a key attribute of moderni­zation, could not have been fueled by wood, even less so by charcoal, because those phytomass fuels have inherently low power density. The second one calculates the demand that the smelting of civilization's most important metal would have made on the world's forests in the absence of coke made from metallurgical bituminous coal.

During the first decade of the nineteenth century, London's population surpassed ι million people and its annual energy demand was equivalent to about 4 t of wood per capita used for heating, lighting, and a multitude of small and large manufacturing enterprises. Photosynthesis converts solar radiation into new chemical bonds in wood with inherently low efficiency, and hence all phytomass fuels are produced with very low power densities. Even if London were surrounded by highly productive forests (good beech and oak stands) whose annual wood production averaged 5 t/ha, their cutting would produce fuel with average power density of only about 0.25 W/m² and the city would have needed a reserve of about 800,000 ha (or a forested circle around the city with a radius of just over 50 km) of which it would have to clear-cut, and promptly replant, some 50,000 ha of mature trees (containing 100 t/ha) every year.

Cutting and transporting that much wood would have been costly and inconvenient; burning it within the city would have been very polluting: polycyclic organic matter released by wood combustion is highly carcino­genic. Replacing all that wood by charcoal would have needed at least 4 tonnes of wood (15 GJ /1) to produce a tonne of charcoal (29 GJ/t), that is, a 50 percent loss of energy due to charcoaling resulting in a power density rate of just 0.12 W/m² in order to produce convenient, light, sulfur-free, and nearly smokeless fuel. Such low power densities of energy production would have required very large areas to secure the fuel: with minimized transport, forests would have to encircle the city with a radius of 80 km - and that was for only i million people. Not surprisingly, by the early nineteenth century all but a very small share of London's energy came from coal.

Similarly sized contemporaneous Beijing managed with much less coal (extracted in small mines just west of the capital since the tenth century) because its per capita energy use was considerably lower, due to minimal heating in winter (hence cotton-padded clothes worn inside) and to the absence of energy-intensive manufacturing enterprises within the city. Obviously, even in i8i0, energizing a rapidly growing and industrializing London-like city solely with wood or charcoal would have been a most impractical and exceedingly costly proposition, and large cities and conurba­tions of the late nineteenth century and megalopoli of the twentieth century became direct expressions of high-energy-density fossil fuels produced with high power densities: both coal and crude oil could be extracted with power densities commonly i,000 and often i0,000 times higher (>100 - >1,000 W/m²) than the harvests of phytomass fuels.

Similar constraints apply to smelting iron, the metal (now mostly con­verted to a wide variety of steels) that was and remains the signature material of the modern world: its global annual output is larger than that of the next four most important metals (aluminum, copper, zinc, and lead) combined. For nearly three millennia iron was smelted from its ores by using charcoal. By i800, global production of i Mt of iron would have required (with mass ratios of 8/1 for charcoal/metal and 1/5 for charcoal/wood) about 40 Mt of wood, or the cutting down of 4,000 km² of high-density virgin forest a year. If widely spread it would have been a negligible burden at a time when the world still had large areas of intact virgin forests. By the year 2000, efficiencies had improved (charcoal/metal ratio at 0.75/1, charcoal/wood at 1/4) but the global iron output rose to 580 Mt and it would have required about 1.75 gigatonne (Gt)Zyear of wood.

Even if all of it came from high-yielding tropical plantations (power density of roughly 0.5 W/m²) it would still mean cutting down annually at least 1.75 million km² of trees, an area equal to slightly more than all forests in the European Union. In contrast, when all surface structures needed for coal mining and coking are accounted for, the power density of coke production will be at least 1,000 W/m² (compared to 0.012 W/m² for charcoal). Global output of coke thus occupied less than 10,000 km² of land, a difference of four orders of magnitude. This is certainly one of the best examples illustrating the consequences of using energies produced with low power density in modern high-energy civilization. Moreover, besides obvious cost advantages, coke has an added structural superiority as it can support much heavier charges of ore and limestone than the friable and easily crushable charcoal, making larger and cheaper blast furnaces possible.

Maximum power of prime movers

Maximum power of individual prime movers matters because it determines the maximum magnitude of many tasks. Mass deployment of smaller power units is often impractical or outright impossible: we would not use ten low- power steam locomotives to pull a heavy train; there is no way to harness one hundred horses to pull in a single direction under a driver's command (the maximum in California's grain fields of the 1890s was more than thirty draft animals harnessed to the earliest combines) and passenger aircrafts with a range of just 200 km cannot be used in a staggered manner to accomplish transoceanic flights.

The post-1750 increase in the maximum power of prime movers controlled by a single person during the performance of quotidian tasks has been enormous. The greatest exercise of concentrated power performed daily by millions of people in 1750 was leading draft animals as they plowed or harrowed crop fields or pulled wagons and coaches on roads. A peasant walking behind his weak ox harrowing a field commanded less than 200 W of draft power (or less than a third of mechanical horsepower of 745.6 watts). During the last decades of the nineteenth century, many farmers plowing the Great Plains wheat fields guided six powerful horses (at least 4 kW), and with the first grain combines pulled by thirty-two animals more than 20 kW of draft power was under a single command at the beginning of the twentieth century.

Another hundred years later and many Great Plains farmers drive 300-horsepower (224 kW) tractors sitting in air-conditioned cabins, while anybody with a driver's licence can be behind the wheel of an even more powerful SUV (the most powerful one on the market in 2011, the BMW X6, rated 407 hp or 303 kW). This means that since 1750 the maximum power of prime movers commonly controlled by millions of individuals in the course of their daily life rose 1,000-fold (three orders of magnitude), from about 200 W to more than 200 kW.

Gains in prime movers commonly used for public transport on land have been even more stunning. In 1750 a driver of a fast four-horse coach held reins of about 2,600 W (2.6 kW). By 1850 an engineer of a steam locomotive

Table 6.1 Maximum power of prime movers in field work

Table 6.2 Maximum power of prime movers in land transportation

controlled more than 200 kW; by 1900 his grandson in control of a transcon­tinental train traveling 100 km/h had at his disposal about one megawatt (MW) of steam power. Manually stoked locomotives could do about 120 km/h, while trains in Japan (Shinkansen) or France (trains a grande vitesse) have at their command electric motors with power rating approaching or even exceeding 10 MW. And another order of magnitude has to be added for air travel: pilots of the Boeing 747 (first flown commercially in 1969) or Airbus 380 (in commercial service since 2007) cruising 11 km above sea level have computerized control of more than 100 MW developed by four gas turbines mounted under the planes' wings.

Above are two comparative ladders of maximum power ratings (all numbers are in watts) for 250 years between 1750 and the year 2000, the first one (Table 6.1) for those prime movers that are commonly used in field work (starting with a value for sustained human labor; peak exertions lasting a few minutes can be an order of magnitude higher), the second one (Table 6.2) for prime movers used in land transportation of passengers (starting with a value for a horse-drawn carriage, the best means of overland travel before the emergence of mechanical prime movers).

Efficiency of energy conversions

In this respect historians have not been nearly as guilty as the modern media, whose idea of dealing with technical advances is to personalize them by focusing on the lives of great inventors (preferably those of a heroic, or at least highly charismatic, cast such as Edison or Tesla) or innovators (Gates, Jobs) and to ignore the process of diffusing those inventions, turning them into acceptable commercial choices, and, most of all, relentlessly improving their often initially poor performance. Without the latter advances most modern energy conversions would play surprisingly marginal roles because they would be too expensive or too inconvenient to use.

If intercontinental flights were still powered by the first generation of commercial gas turbines introduced before 1958 they would consume 50 percent more kerosene. And if the owners of a new house built in California who installed, as is now common, more than fifty lights were to pay their electric bill in pre-1910 electricity prices, this would be at least forty times higher than today, due to the low efficiency of the thermal power plants that provided electric power a century ago. Actually, there is no better example to illustrate the long-term conversion efficiency gains than lighting, as no other common use of energy has seen such gains since 1750. Only a few technical improvements have had such a profound effect on human affairs as extending the day at will (and thus reading, writing, travel, and work in factories), banishing the night and making stairways, houses, and cities safer.

In 1750 candles remained the most convenient source of relatively clean but expensive indoor light. They came in a variety of sizes, qualities, and costs and their production eventually became a major manufacturing endeavor - but their wax-to-light efficiency remained dismal: in 1750 an inexpensive (and smelly) tallow candle converted just 0.01 percent of chemical energy in that animal lipid to light. Banishing the darkness thus remained nearly as ineffi­cient during the time of the first encyclopaedists as it was more than one-and- a-half millennia before their time, when Marcus Aurelius wrote down his meditations among the Quadi on the Granua. The first improvement came with lights that used flammable gas made from coal, commercially produced since 1812, which multiplied those dismal efficiencies, with early lamps reach­ing about 0.04 per cent. In 1881, Edison's earliest carbon filaments converted only 0.15 percent of electricity to light.

By 1900 efficiencies of incandescent lights reached 0.5 percent with metal filaments and by 1950 lights with coiled tungsten filaments were closing on

Table 6.3 Efficiency of lights for indoor illumination

2 percent efficiency. But by that time the fluorescent lights, commercially available since the 1930s, were on the top, with efficiencies as high as 12 percent. This performance was eventually surpassed by low-pressure sodium lights that convert close to 30 percent of electricity to visible radiation (they have been used almost exclusively for outdoor city illumination and they impart characteristic yellowish hue to cityscapes at night) while the best indoor options (linear fluorescents and screw-in metal halide light bulbs) convert electricity to visible radiation with more than 15 percent efficiency (Table 6.3).

Even a small new American house will now have at least thirty lights (or no less than 2.4 kW of installed power) that can be turned on and off by a simple flip or push of a switch. Assuming a mixture of incandescent and fluorescent lights, the overall luminous efficiency will be close to 10 percent and providing the same amount of light (but, obviously, not with the same flexibility and convenience) would require about 10,000 standard paraffin candles. Logistics (and fire hazards) of lighting and extinguishing them aside, it would also be a very costly operation. As for the cost, William Nordhaus calculated that by the end of the twentieth century the true price (in constant monies) of illumination (the cost of the service rather than of the good, in this case a source of light) was four orders of magnitude lower (the actual ratio was about 0.0003) than it was in 1800. Similarly, Roger Fouquct's detailed calculation for the United Kingdom showed that, again in constant monies, light was roughly 6,000 times more expensive in 1750 than it was in 2000.

Perhaps the second most impressive illustration of improved conversion efficiencies is the performance of prime movers powered by combustion of fossil fuels. The first commercial models of such a machine, Thomas Newcomen’s atmospheric steam engine, were extraordinarily inefficient, converting no more than half a percent of charged coal into reciprocating motion, and hence suited only for installation at coal mines where there was

Table 6.4 Top efficiency of internal combustion engines

no need for transporting the fuel. In 1750, John Smeaton managed to raise the efficiency above ι percent and James Watt's subsequent famous improve­ments (patent of 1769) were only relatively impressive because the absolute efficiency of his machines remained pitiful, no better than 4 percent during the 1780s. But the gain was sufficient to site them more flexibly away from mines, particularly where the fuel could be brought by vessels. Only high- pressure steam engines did considerably better, surpassing 10 percent by 1850, and by the century's end the most efficient triple-expansion designs con­verted more than 15 percent of coal's thermal energy into useful mechanical energy.

That was a performance superior to new internal combustion engines, but those lighter and much more flexible machines were improved rapidly, particularly as Rudolf Diesel's inherently efficient high-compression engines began to conquer first shipping and then heavy land transport. In 1897 the final prototype of Diesel's new engine had thermal efficiency of almost 35 percent; before the Second World War the best machines surpassed 40 percent and by the year 2000 the massive diesels powering the world's largest container ships and oil tankers were the world's most efficient engines, converting just over 50 percent of energy in heavy fuel oil into propulsion (Table 6.4).

Their closest competitors in terms of efficiency are large gas turbines (jet engines) that have been powering modern long-distance commercial aviation since the late 1950s: in 1940 the first British and German prototypes converted less than 10 percent of kerosene into useful thrust. By the mid- 1960s the first generation of jetliners had engines operating with 25 percent efficiency, and by the century's end gas turbines had thermal efficiencies of about 40 percent. These gains have made mass-scale air travel affordable: in 1950, at the beginning of the last decade of air travel powered by reciprocating engines, the world's airlines logged 30 billion passenger-kilometers; in 2000 the total reached 2.8 trillion.

Per capita consumption of useful energy

Supply of phytomass fuels and mechanical power (human and animal mus­cles, waterwheels and windmills - relatively common in some regions, rare or absent in others) remained low and stagnating for millennia. My recon­struction of fuel supply in Rome during the first two centuries of the Common Era came up with at least 10 GJ and possibly 15 GJ/capita. Estimates for average annual wood consumption in London of 1300 put this at 1.51 or just over 20 GJ/capita, which means that the national average had to be below 20 GJ, not that different from the Roman energy consumption. Peasants on the deforested North China Plain burned annually no more than 10 to 15 GJ/capita of straw, roots, leaves, and grasses even during the early decades of the twentieth century.

Extraction of fossil fuels had multiplied the average supply, often in just one or two generations. Historical comparisons of energy use per capita rates are calculated after converting different fuels and other forms of primary energy (hydro, nuclear, geothermal, wind, and solar electricity) into a common energy denominator. The joule should be the preferred choice, but international statistics used to convert to a tonne of coal equivalent (tce, equal to 29 GJ, because it took the highest quality coal as the standard); but since the 1980s all of the most often cited sources (United Nations, OECD, the annual survey of world energy by British Petroleum) use a tonne of oil equivalent (toe, equal to 42 GJ or 1.45 tce).

British data show the average per capita use of primary energy rising from about 30 GJ in 1750 to just over 150 GJ in the year 2000, a relatively small (five-fold) multiple because the country already had a fairly high per capita use by 1750; the French use, reliably traceable since 1850, went from 20 GJ to 180 GJ in 2000, a nine-fold increase in 150 years. The multiple for Japan, a latecomer to modernization (begun only after the Meiji Restoration of 1868) was nearly eighteen-fold during the twentieth century (from about 10 GJ to just over 170 GJ/capita), and China's belated modernization was energized by going from just 1.5 GJ/capita (fossil fuels and hydroelectricity, excluding phytomass fuels) in 1950 to 35 GJ in 2000, a twenty-three-fold gain in just five decades (Table 6.5).

Impressive as these multiples may be, they obscure the real gains in the consumption of modern energy: because of the just-reviewed gains in typical

Table 6.5 Average annual consumption of primary energy (rounded to the nearest 5 GJ/capita and including all phytomass and fossil fuels and primary electricity)