by Vaclav Smil
A Key to Understanding Energy Sources and Uses
Vaclav Smil
The MIT Press Cambridge, Massachusetts London, England
Preface
Acknowledgments
1 How Power Density Matters
2 Quantitative Keys to Understanding Energy
Power of Rates
Power Density: Sorting Out the Rates
Typology and Caveats
3 Renewable Energy Flows
Solar Radiation and Its Conversions to Heat and Electricity
Wind and Wind-Generated Electricity
Water Power and Hydro Generation
Phytomass for Traditional and Modern Uses
Geothermal Energy
4 Fossil Fuels
Coal Extraction, Preparation, and Transport
Crude Oils, Refining, and Long-Distance Deliveries
Natural Gas, Pipelines, and Liquefied Natural Gas
5 Thermal Electricity Generation
Coal-Fired Power Plants
Hydrocarbons in Electricity Generation
Nuclear Generation
Transmission
6 Energy Uses
A Brief Historical Perspective
Hierarchy of Modern Energy Uses
The Challenges of Heat Rejection
7 Making Sense of Power Densities
Recapitulations
Comparisons
Aggregate Land Claims
8 Energy Transitions
Biofuels
Wind and Solar Electricity
What It Would Take
Appendix
References
Index
Modern societies have many concerns about their energy supply. Above all, it should be affordable, reliable, and convenient: affordable in order to drive economic development and improvements in quality of life, reliable so as to be available on demand in its various forms, most of all as uninterruptible electricity, and convenient to give consumers virtually effortless access to preferred household, industrial, and transport energies. During the closing decades of the twentieth century two other concerns became prominent: energy supplies should also be environmentally benign and, preferably, renewable. Since the 1970s-that is, once energy matters began to receive, belatedly, an unprecedented amount of professional and public attention-all of these concerns have been addressed and analyzed in hundreds of books and thousands of papers, to say nothing of the instant expertise proffered by the mass media.
Any modestly informed person knows that energy prices matter (we need only think of OPEC's crude-oil price fixing), as do a reliable electricity supply (to avoid blackouts) and the finiteness of fossil fuel resources (underscored by claims about an imminent peak of oil extraction), and that the environmental impacts of energy use are not only local but worldwide (global warming). But does power density matter? Power is simply energy flow per unit of time (in scientific units, joules per second, which equals watts, or J/s = W), spatial density is the quotient of a variable and area, and hence power density is W/m2, that is, joules per second per square meter. Why should we care about how much power is produced, or used, per unit of the Earth's surface? Some say explicitly that we should not bother at all.
Amory Lovins spent a lifetime making exceedingly optimistic forecasts about the speed with which renewable energy conversions (and other energy-related innovations) would be adopted by modern energy systems (for these claims, see Lovins 1977 and 2011a; for their critique, see Smil 2010a). In 2011 he dismissed any need to consider power densities, concluding that "land footprint seems an odd criterion for choosing energy systems: the amounts of land at issue are not large, because global renewable energy flows are so vast that only a tiny fraction of them need be captured" (Lovins 2011b, 40). In contrast, nearly two generations ago Wolfgang Hafele and Wolfgang Sassin, two leaders of energy research at the International Institute for Systems Analysis, wrote that "the density of energy operations is one of the most crucial parameters that predetermine the structure of the energy systems" (Hafele and Sassin 1977, 18).
Even so, most of the modern energy literature has simply ignored the subject. Energy economists have been worrying about prices, their elasticities, oligopolies, taxation, and links between energy use and economic growth. As interest in energy matters began to rise, Malcolm Slesser offered an explanation of this omission:
Land has always had a firm place in classical economics, where attention focused upon the agrarian economy, and land as a factor of production. Ricardo and Malthus founded their ideas around the land factor, in marked contrast to more recent economic thought in which land virtually dropped out of the scene, and production was viewed essentially as a synergy of labour and capital. (Slesser 1978, 86)
Similarly, modern engineering analyses of energy systems look at fuel qualities, mass throughputs, the rated power of energy converters, and annual and peak production. Of course, the design of specific energy extraction or conversion facilities must, inevitably, consider land requirements and land qualities in requisite detail, but space is not a common analytical denominator used to assess their performance. Still relatively uncommon interdisciplinary inquiries into the nature and linkages of modern energy systems look at per capita energy use, energy's role in economic performance, and the impacts of energy conversions on environmental quality, but only very few recent publications have looked at power densities.
I chose the measure as a key analytical variable in General Energetics (Smil 1991) and in its second, expanded edition, titled Energy in Nature and Society (Smil 2008). David McKay's Sustainable Energy-Without the Hot Air (2008) also uses the rate as an essential indicator that offers revealing insights into unfolding energy transitions. And Juan Moreno Cruz and M.Scott Taylor offer a rare example of economists who have explicitly incorporated energy density and power density in their new analysis, concluding that "perhaps the most important attribute of an energy source is its density: its ability to deliver substantial power relative to its weight or physical dimensions" (Cruz and Taylor 2012, 2) and that "the very density of the energy resource we seek, fuels our efforts to obtain more. Therefore differences in power density across energy resources create large differences in energy supply" (Cruz and Taylor 2012, 48).
In this book I demonstrate that power density is a key determinant of the nature and dynamics of energy systems. Its careful quantifications, their critical appraisals, and their revealing comparisons bring a deeper understanding of the ways with which we harness, convert, and use energies. A careful assessment of power densities is particularly revealing when contrasting our dominant fossil fuel-based energy system with renewable energy conversions. But before I start laying the foundation for my systematic power density assessments (by introducing the principal variables and reviewing different power density concepts), I tell a story of early modern charcoal-based iron smelting and of its replacement by coke-based production in blast furnaces. This historical appraisal offers a compelling (yet curiously overlooked) example of why power densities matter, why energy's rate of flow per unit of surface area is a key determinant of the structure of energy systems, and how it constrains their development.
All of my books written between 2000 and 2009 had, besides many photographs, numerous original illustrations on whose creation I worked with Doug Fast, a photographer and graphic artist at the University of Manitoba. Doug left the university in 2009, and as a result, all of my post-2010 books had either only rudimentary graphs or no original illustrations (just some photographs evoking the contents of individual chapters), and Making the Modern World, the last book to come out, in 2013, had no images at all.
Fortunately, this has changed, thanks to the assistance of Ian Saunders, who served as the project's creative director, and a team that in
cluded Anu Horsman, Jen Krajicek, Luke Shuman, Carl de Torres, and Wendy Woska. Additionally, Noli Novak created stippled portraits from low-resolution photographs of John Henry Poynting and Nikolai Alekseevich Umov. My thanks to all: as a lifelong graphics aficionado I greatly appreciate the difference simple, elegant illustrations can make.
This brief chapter starts with a 1548 King's Commission, proceeds to a great English invention of the eighteenth century (not James Watt's steam engine!), and ends with the consequences of trying to supply a rising pig iron output in nineteenth-century blast furnaces by metallurgical charcoal.
In November 1548 the King's Commission, given to 20 men in Sussex, was to examine "the hurts done by iron mills and furnaces made for the same" on the Weald, the region of southeastern England between South and North Downs that was the center of English iron-making during the sixteenth century. The commission's most pressing questions to witnesses were these (Straker 1969, 115):
5. If the said iron mills and furnaces be suffered to continue, then whether thereby there shall be great lack and scarcity of timber and wood in parts near the mills?
6. What number of towns are like to decay if the iron mills and furnaces be suffered to continue?
The cause of these hardships was obvious:
The iron mills and furnaces do spend yearly... above 500 loads of coals, allowing to every load of coals at the least three loads of wood; that is every iron mill spendeth at the least yearly 1,500 loads of great wood made into coals.
The petitioners went to great length to enumerate how wood scarcity would make it impossible to build houses, water mills or windmills, bridges, sluices, ships, boats, wheels, arrows, hogsheads, barrels, buckets, saddletrees, bowls, dishes and to supply timber "for the King's Majesty's towns and pieces on the other side of the sea" (Straker 1969, 118; Boulogne was an English possession between September 1544 and March 1550). As long as the deforestation caused by the rising demand for charcoal was limited to a few counties, local smelting would simply decline or cease, and the production would move elsewhere. Data assembled by King (2005) show that the production of pig iron from Weald's charcoal furnaces (about 4,000 tonnes [t] at the time of the 1548 Sussex petition) peaked by 1590 (at 14,040 t), and that by 1660 the region was making less iron than in 1550, while the metal's output kept on increasing in the rest of England.
English Iron Smelting
By 1620 England was producing more than 26,000 t of pig iron annually, but then production began to fall, and it was only about 18,000 t by the end of the seventeenth century. We can reconstruct fairly reliably what this meant in terms of energy demand and environmental impact. In the early decades of the eighteenth century, charcoal-fueled blast furnace campaigns usually extended from October to May, and during those eight months a typical furnace produced 300-340 t of pig iron (Hyde 1977). The efficiencies of wood conversion to charcoal and the charcoal requirements for pig iron smelting and for subsequent conversion of the metal to bar iron varied considerably (and about a third of the metal was wasted in the conversion), but the best available data (Hammersley 1973) indicate that at the beginning of the eighteenth century, at least 32 t of wood were needed to produce a tonne of bar iron.
This means that a typical furnace and an associated bar forge (sometimes located closer to the market) would have required at least 9,600-10,900 t of wood-and in the early seventeenth century, with less efficient conversions, the total could easily have been twice as much. The preferred wood supply came from coppiced plantings of beech or oak cut in 20-year rotation, which would yield an annual increment of about 5 m3/ha; with an average wood density of 0.75 g/cm3, that would be 3.75 t/ha, and operating a furnace and a forge would have required harvesting about 2,700 ha/year. In 1700 British furnaces produced about 12,000 t of bar iron, and hence they consumed on the order of 400,000 t of charcoaling wood. With an average productivity of 4 t/ha, this amount of wood would have required at least 100,000 ha of coppiced growth, a square with sides of nearly 32 km.
Obviously, the availability of suitable wood was a limiting factor in the further expansion of the English iron industry. Hammersley (1973) estimated that the maximum countrywide harvest would have been on the order of 1 Mt/year. Not surprisingly, during the eighteenth century England became highly dependent on iron imports: Swedish exports rose rapidly after 1650, Russian exports began to dominate after 1750, and by the 1770s England was covering two-thirds of its iron demand with foreign metal (King 2005). An end to the charcoal ceiling on English pig iron production, the elimination of imports, and a massive expansion of iron output (from only about 25,000 t in 1750 to 100 times that much, 2.5 Mt, 100 years later) were possible because of the switch from charcoal to coke. Coke is a lightweight (apparent density of just 0.8-1 g/cm3) but energydense fuel (29 MJ/kg) that is produced from coal by pyrolysis, or heating in the absence of oxygen.
In England, the first use of coke was for drying malt in the 1640s, Shadrach Fox was the first industrialist to use it on a small scale, in a blast furnace, during the 1690s, and, starting in 1709, Abraham Darby became the fuel's best-known promoter (Harris 1988). Coke's initially high production costs limited its rapid adoption, and the two fuels coexisted in England for most of the eighteenth century, and in the United States well into the second half of the nineteenth century. This shift from metallurgical charcoal to coke was part of a much larger energy transition from traditional biomass fuels (woody phytomass, charcoal, and also crop residues, mainly cereal straws) to fossil fuels, first to coals and then to hydrocarbons (Smil 2010b).
This transition introduced fuels that had a generally higher energy content: a kilogram of high-quality coal has nearly twice as much energy (29 MJ) as a kilogram of air-dried wood (17 MJ); a kilogram of liquid fuels refined from crude oil had three times as much energy (42 MJ) as straw used for cooking (14 MJ). The higher energy density also made fossil fuels cheaper to transport and easier to store. But the replacement of charcoal by coke did not bring in a fuel with superior energy content, as both charcoal and coke are essentially pure carbon with a virtually identical energy content of about 30 MJ/kg, but the power densities of their production are orders of magnitude apart.
From Charcoal to Coke
Harvesting coppiced beech or oak would yield, at 5 m3 or 3.75 t/ha and an energy density (of dry mass) of 19 GJ/t, an annual phytomass harvest of about 0.22 W/m2. In contrast, a typical late eighteenth-century deep mine-based on detailed data for the Toft Moor Colliery of the 1770s (Hausman 1980)-would have produced about 15,000 t/year, and all of that coal would have been hauled through a single narrow shaft. With the mine's surface structures occupying no more than a hectare (10,000 m2) of land, that extraction would have produced fuel with a power density of nearly 1,200 W/m2, more than 5,000 times higher than the power density of charcoal production.
But more land was needed near the pithead because the fuel was usually sorted at the site to remove associated rocks and to improve the coal's quality. If we assume that the incombustible material (amounting to 10% of the total mass, with a density of 2.5 t/m3) was deposited nearby in a conical heap just 20 m tall, the area claimed after 50 years of a mining operation would have been no larger than 4,500 m2, and the overall power density of coal production would have been no less than 800 W/m2, or 4,000 times higher than the power density associated with harvesting wood.
And we have to make adjustments for conversion to, respectively, charcoal and coke (both calculations disregard the relatively small areas needed for charcoaling or coke ovens). In the early eighteenth century a typical charcoaling ratio (by weight) was five units of wood for a unit of charcoal, which means that in energy terms (with 29 GJ/t of charcoal and 19 GJ/t of wood), it was about 3.3 units of wood for a unit of charcoal, and the power density of charcoal production would have been only about 0.07 W/m2. Early coking in simple beehive ovens was also inefficient, with up to 1.7 units of coal needed to produce a unit of coke. The overall power density of mid-eighteenth-century English coke product
ion was thus roughly 500 W/m2, approximately 7,000 times higher than the power density of charcoal production.
This shift from charcoal to coal and the huge difference in accompanying power densities of the two products had many economic, social, and environmental consequences. England could rapidly reduce, and soon eliminate, its dependence on iron imports from Sweden and Russia. The removal of one of the greatest burdens on the country's forests opened the way to reforestation, and the high power density of coal and coke production made it possible to concentrate their output in a progressively smaller number of facilities (large coal mines and coking plants, often attached to large blast furnaces), from which the products could be distributed not only nationwide but also exported abroad.
And the shift made no smaller difference even in those countries that were initially rich in natural forests because they too were not immune to the charcoal limit on iron-making. In the early nineteenth century, American iron-makers had no problem harvesting the needed wood from majestic Appalachian forests, but 100 years later it had become impossible to acquire that much wood from eastern forests. Although smelting in blast furnaces became much more efficient during the course of the nineteenth century (by 1900 it required just 5 kg of wood for every 1 kg of hot metal), the US pig iron output surpassed 25 Mt by 1906, and maintaining that level of production alone (excluding all charcoal needs for further metal processing) would have required (even when assuming a high average increment of 7 t/ha in natural forests) an annual wood harvest of about 180,000 km2 of forest (Smil 1994).
That is an area the size of the entire states of Missouri or Oklahoma (or one-third of France), equal to a square with a side equivalent to the distance between Philadelphia and Boston, or Paris and Frankfurt-to be harvested annually! Obviously, the power density of producing metallurgical charcoal would have been far too low to enable even forest-rich America to industrialize on that renewable energy basis. And what would charcoal-based ironmaking require today? Could the combination of high-yielding clones of fast-growing trees planted in the tropics, higher efficiencies of the best charcoaling techniques, and the lower specific energy requirements of metal smelting provide a land-sparing renewable solution for a modern iron industry?
