5

This chapter’s title is a deliberately loaded question with multiple answers. The most obvious answer – if the question is taken literally and purely in terms of physical presence – is as long as the planet Earth. Even the best conceivable enhanced recovery techniques will still leave behind a substantial amount of the oil originally present in reservoir rocks. More importantly, the excessive cost of discovering and producing liquid oil that is stored in countless small, marginal or virtually inaccessible formations, and unappealing returns on extracting liquid hydrocarbons from most of the known oil sands, oil shales and tar deposits, will ensure that a significant amount of the oil originally in place in the Earth’s crust will never be brought to the surface.

If the question means simply how long there will be some commercial production of crude oil then the answer is also easy: definitely throughout the entire twenty-first century. Even those who have argued that the peak of global oil extraction is imminent must concede that we have yet to discover many conventional oilfields and that the new ways of extracting oil from nonconventional sources mean that some of these accumulations will remain in commercial production during the closing decades of the twenty-first century. And while accelerated decarbonization can displace a number of specific fossil fuel uses with non-fossil alternatives, other oil uses will prove much more resistant to change. But I will not try to answer the question of how long crude oil will remain the single most important fossil fuel in the global primary energy supply; nor will I predict when global oil extraction will reach its peak and begin to decline. Answers to these questions are contingent on the unknown magnitudes and trends of many variables, and any new predictions of the peak production year would only extend an already long list of failed forecasts of this kind.

What I will do instead is to provide a proper historical perspective on oil’s role in the global energy supply. Then I will survey the fallacies and facts concerning the currently fashionable catastrophic prognoses of an imminent end of the oil era – the sentiment embodied in publications whose titles claim that we have already reached the production peak, that the party’s over, or, in the most extreme fashion, in Richard C. Duncan’s Olduvai theory, that the decline of oil extraction will plunge humanity back to a life comparable to that experienced by the first primitive hominins who inhabited the famous Tanzanian gorge some two million years ago. Finally, I will look beyond oil, outlining briefly some major means of providing liquid fuels from sources other than conventional crude oil.

If oil’s importance were to be judged by the frequency with which the words oil, crude oil or petroleum are mentioned in the media or by politicians, the inevitable conclusion would be that no other source of energy is more important for the survival of civilization. This would be an incorrect and indefensibly exaggerated notion. Undoubtedly, liquid fuels have had an enormous impact on the modern way, and quality, of life, but outside of North America they became very important only during the last two generations, after 1960 in Europe and Japan, and since the 1980s in the populous countries of Asia. Global dependence on oil is thus a relatively new phenomenon, and this reality should forcefully remind us that we should not exaggerate the fuel’s indispensability: we had ingenious industrialized societies capable of delivering a decent quality of life long before oil consumption rose to its current level – and there will be prosperous societies supporting a good quality of life long after liquid hydrocarbons have become minor constituents of the global energy supply. A telling indication of this transition process is that in 2017, a Google search returns nearly as many hits for natural gas as for petroleum and crude oil combined!

ENERGY TRANSITIONS

Pre-hydrocarbon industrializing societies were energized by coal and also by the generation of hydroelectricity. Coal, the quintessential fuel of the nineteenth century’s industrialization, continued to dominate the global supply of commercial energy during the first half of the twentieth century: its share declined slowly, from about 95% of the modern energy supply (excluding traditional biofuels) in 1900 to about 80% of the total in 1930 and to just over 60% by 1950. Meanwhile, crude oil provided just 4% of the world’s commercial primary energy in 1900, 16% by 1930 and 27% by 1950 when natural gas supplied about 10% of the total (see figure 25). In many leading economies coal was much more dominant: even by 1960 its share in primary energy supply was nearly 60% in Japan, 61% in France, 77% in the UK and 80% in Germany, at that time fueling the country’s impressive economic expansion.

Global transition from coal to hydrocarbons accelerated during the 1960s: 1962 was the first year when coal provided less than half of the world’s primary energy. Later, because of oil’s numerous advantages, even the two rounds of OPEC’s oil price increases could not stimulate coal’s comeback. The most important factors that explained coal’s continuing retreat were the relatively high cost of its underground extraction, its inflexibility of use and the considerable environmental impact of its production and combustion (the latter including acid precipitation and high carbon emissions). But there is no doubt that the post-1972 OPEC-driven oil price rises have definitely helped to slow down coal’s global retreat, from about 33% of the total supply of primary commercial energy in 1970 to 28% by the year 2000. But resurgent China disregarded all these drawbacks: it needed enormous amounts of energy fast in order to fuel its rapid rates of economic growth. As a result (and, secondarily, thanks to India’s rising demand), coal’s share in global primary energy supply rose to 33% during the first decade of the twenty-first century and it was almost that high in 2015.

Figure 25 Global shares of primary energies, 1900–2015

But outside China and India coal is now in retreat as its most important use, for generation of electricity, is being replaced by cleaner and more efficiently converted natural gas (particularly by combined cycle gas turbines). Most notably, the US dependence on coal-generated electricity declined from 45% in 2010 to 30% in 2016.

The best global aggregate count indicates that during the twentieth century coal and crude oil supplied roughly the same amount of primary energy, each approximately an equivalent of 125Gt of oil, but during the century’s second half crude oil’s energy surpassed that of coal roughly by a third. Crude oil became the world’s leading source of primary energy during the mid-1960s and its share of the global energy supply peaked during the late 1970s at about 44%; by 1990 it was down to 37% and it remained there a decade later, but by 2010 it had slipped to 32%, the lowest it has been since the late 1950s. This long-term view puts the importance of crude oil in global energy supply into a proper historical perspective. By 2015 oil had been the largest component of the global primary energy supply for only fifty years and during that period its consumption rose from about 1.6 to 4.3Gt a year, nearly a 2.7-fold increase.

But in global terms, crude oil’s share of primary energy supply could never reach coal’s massive dominance that prevailed during the first half of the twentieth century and, despite the continued absolute growth of its consumption, the fuel has been in relative retreat as coal’s Asian expansion and greater worldwide reliance on natural gas combined to supply nearly 60% of all commercial energy in 2016. Oil’s relative retreat is also illustrated by the declining oil intensity (units of oil needed to produce a unit of economic product) of all major economies as well as on the global level. When expressing the US GDP in constant ($2011) monies, oil intensity of the US economy fell from about 140kg/$1,000 in 1964 to 55kg/$1,000, a 60% decline in fifty years. During the same period the global decline of average global oil intensity was even slightly greater, amounting to about 66%.

These impressive declines have been achieved through a combination of the higher efficiencies with which we now convert refined oil products and their replacement by gaseous hydrocarbons and non-fossil energies. For example, in 1975, tens of millions of American households (mostly in the Northeast) used fuel oil for relatively inefficient (about 50%) home heating; its replacement by natural gas not only eliminated a previously large market segment but it also lowered the atmospheric carbon burden: even so-called mid-efficiency natural gas furnaces had efficiencies around 70% during the 1980s and well above 80% a generation later, and today’s best high-efficiency furnaces convert 97–98% of the chemical energy in methane into heat.

And with the retreat of liquid fuels from electricity generation (less than 7% of all refined fuels were burned in power plants in 2015) and residential uses such as heating and cooking (now also less than 7% of global demand for liquids), the consumption of refined oil products has become even more concentrated in the transportation sector: all major forms of moving goods and people – be it shipping, railroads, trucking, automobiles and flying – rely overwhelmingly on refined fuels. But because this sector has become such a critical component of all affluent economies and lifestyles, it is no exaggeration to conclude that at the beginning of the twenty-first century modern civilization is defined in many important ways by its use of liquid fuels and hence it will go to great lengths to ensure their continued supply.

And it must be repeated that this importance goes beyond the reliance on high-performance fuels in all forms of transportation: oil-derived lubricants are indispensable for countless industrial tasks; modern transportation infrastructures are unthinkable without oil-derived paving materials; and syntheses of scores of plastics begin with oil-derived feedstocks. All of these benefits derive from the extraction of a resource that is not renewable on a civilizational timescale, and since the mid-1990s the questions about its durability have been receiving increasingly worrisome answers from some oil geologists whose arguments have been given prominent coverage by the media. Nearly all of those who have argued that the peak of global oil production is imminent have not foreseen any subsequent comfortable plateau but a rapid decline, and have repeatedly assured us that its inevitable consequence will be the end of modern prosperity and an intensifying fight over the dwindling oil resources. I will deconstruct these scares and show why such bleak scenarios are not likely to prevail.

The basic assumptions tirelessly repeated by the proponents of the theory of an imminent peak of global oil production are as follows. When we compare the total volume of the estimated ultimate recovery (EUR) of oil with the worldwide cumulative production we see that (depending on the somewhat uncertain size of EUR) we have already extracted half of the EUR or are about to do so in a matter of years. Once the global oil production reaches its peak its fairly steep descent will start fairly promptly because the complete extraction cycle must follow a function described by a normal (bell-shaped, Gaussian) curve (with the area beneath the curve equal to EUR). Given the importance of oil for modern civilization this inevitable decrease in annual oil production will have enormous consequences, and some of the leading peak oil theorists went as far as writing obituaries of modern civilization.

L. F. Ivanhoe, an American geologist, believed that an early end of the oil era will bring ‘the inevitable doomsday’ followed by an ‘economic implosion’ that will make ‘many of the world’s developed societies look more like today’s Russia than the US.’ Richard C. Duncan, originally an electrical engineer, saw the peak as ‘a turning point in human history’ leading to massive unemployment, breadlines, homelessness and a catastrophic end of industrial civilization. Indeed, his ‘Olduvai theory’ had humanity returning soon to the condition in which our hominin ancestors lived nearly two million years ago. But all of the leading proponents of the theory of the imminent peak of global oil extraction (Colin Campbell, Jean Laherrère, L. F. Ivanhoe, Kenneth Deffeyes) have resorted to more or less alarmist arguments. Their writings and speeches have mixed incontestable facts and sensible arguments with indefensible assumptions and caricatures of complex processes as they ignore those realities that do not fit their preconceived conclusions.

Their conclusions are based on simplistic interpretations. Values of EUR are not at all certain, and tend to rise with better understanding of petroleum geology, with frontier exploration and with enhanced recovery techniques. Moreover, the proponents of an imminent peak of global oil extraction disregard the role of prices, they ignore historical perspectives, and they presuppose the end of human inventiveness and adaptability. But it is precisely their bias and their catastrophic message that have attracted the mass media (ever eager to spread new bad news) and impressed a scientifically illiterate public. My critique rests on three fundamental realities. First, these recent peak oil worries are only the latest instalment in a long history of failed peak forecasts. Second, the claim of peak oil advocates that this time the circumstances are really different and hence their forecasts will not fail mixes correct observations with untenable assumptions. Third, and perhaps most importantly, when contemplating a world with little or no oil, a gradual decline of global oil production does not have to translate into any economic and social catastrophes.

Public concerns about running out of fossil fuel resources date to 1865 when William Stanley Jevons, a leading economist of the Victorian era, published a book in which he concluded that it is ‘of course... useless to think of substituting any other kind of fuel for coal’ and that the falling coal output must spell the end of Britain’s national greatness. In reality, the UK’s coal extraction continued to expand until the second decade of the twentieth century and its subsequent decline had nothing to do with the exhaustion of resources, and everything to do with the arrival of new fuels. In 2015, when the last remaining British underground coal mine was shut down, the UK still had more than 200Mt of bituminous coal reserves, but extracting the fuel had already become an unappealing proposition decades ago when compared to producing crude oil and natural gas from the North Sea, or to importing much cheaper foreign coal. In 2017 the world’s proved reserves of coal were about 1.1Tt (implying R/P ratio of more than 150 years) and that total could certainly be multiplied with more exploration and with advanced mining techniques. Clearly, post-1950 switching from coal to hydrocarbons (and to primary electricity) has had little to do with ‘running out’ of actual mass stored in the Earth’s crust.

But with oil the argument has been different, as the proponents of imminent peak production anticipated major declines in the fuel’s availability.

OIL’S REPEATED END

Published reports about the imminent end of oil production can be traced as far back as the 1870s, and fears about running out of liquid oil were quite strong in the US during the early 1920s. But the most influential argument was made by M. King Hubbert, an American geologist, who postulated that mineral resource extraction follows an exhaustion curve that has the shape of normal (symmetrical, bell-shaped) distribution: its peak is immediately followed by a decline whose course mirrors the production rise. Hubbert used this approach to accurately predict the peak of US oil extraction in 1970, and the symmetrical exhaustion curve thus acquired the status of an infallible forecasting tool: once the recoverable resources are known and the past production is plotted then a symmetrical continuation of the curve shows the peak extraction year, declining production and the timing of eventual resource exhaustion.

Hubbert’s own forecast put the peak of global oil extraction between the years 1993 and 2000. In reality, global output in 2017 was nearly 40% above the 1993 level, a substantial error that clearly invalidates the original forecast. Hubbert’s fame rested more on his supposedly accurate forecast of US peak oil production in 1970. Indeed, the output peaked at 11.3Mbpd in 1970 (and it was down by 10% ten years later, conforming to predicted mirror decline) – but Hubbert’s forecast pegged it 20% lower and based it on EUR of 200Gb, the total that he himself had raised from the 150Gb that he had estimated just a few years earlier. But between 1859 and 2005, before hydraulic fracturing began to make a substantial difference, the US oil industry had already produced 192Gb, it was still the world’s third largest producer of crude oil, and had 32Gb of remaining reserves. And the difference between Hubbert’s forecast and reality has grown much wider with the post-2005 rise of shale oil extraction.

As a result, the post-peak decline of US oil production has not been a mirror image of the incline, and the country’s oil extraction has not followed a symmetrical curve: in fact, in 2017, the US crude oil output was on the way to surpassing the record set in 1970, and to reach a level nearly five times higher than Hubbert’s forecast for 2017 (see figure 26). This could hardly be seen as an admirable forecasting record. Hubbert underestimated the conventional recovery because he had no knowledge of the Prudhoe Bay supergiant oilfield or of coming giant finds in the Gulf of Mexico, and (as everybody else) he did not even consider that nonconventional resources could become major sources of new supply. Obviously, any EUR is just that, an estimate subject to revision, not a fixed total.

Figure 26 Hubbert’s forecast of US peak oil production and the subsequent extraction decline – and the actual 1970–2016 course

Hubbert has had many followers who have been oblivious to the fact that actual production has not been conforming to his forecasts. In 1977 the Workshop on Alternative Energy Strategies set the global oil peak as early as 1990 and most likely between 1994 and 1997. In 1978 Andrew Flower wrote in Scientific American that ‘the supply of oil will fail to meet increasing demand before the year 2000’. In 1979 the CIA believed that the global output must fall within a decade. In 1990 a USGS study put the peak of non-OPEC oil production at just short of 40Mbpd before 1995 – but the actual output was more than 47Mbpd in 2005. Some peak-of-oil proponents had already seen their forecasts fail: Campbell’s first peak was to be in 1989, Ivanhoe’s peak was in 2000, Deffeyes set it first in 2003 and then, with ridiculous accuracy, on Thanksgiving, 2005.

One would think that this record would dissuade more entries, but the true believers could not resist enlarging this list of failures. In the early years of the twenty-first century we were told that the failure of forecasts produced during the 1970s or 1980s should not be used to argue for a high probability (or certainty) of future failures. Their principal post-2000 argument was that, unlike 20–30 years earlier, exploratory drilling had already discovered some 95% of the oil that was originally in place in the Earth’s crust. Consequently, more frantic and more extensive drilling efforts would make no difference: all it would do is discover the small volume of remaining oil faster. And, the peak oil proponents also argued that efficiency improvements (even outlawing SUVs) or a new-found frugality in affluent countries cannot make any fundamental difference: the slower increase in global oil demand (the increase itself being guaranteed by the huge oil needs of modernizing countries) would only slightly postpone the timing of the peak.

To sum up, the entire notion of an imminent peak of global oil production is based on three key claims: that recoverable oil resources are known with a high level of confidence; that they are fixed; and that the history of their recovery is subsumed by a symmetrical production curve. None of these claims is true. EUR cannot be known with a high degree of confidence and it is not fixed as long as large areas of the Earth remained unexplored or only cursorily assessed, and as long as technical advances are able to transform previously unexploitable resources into major sources of new supply. But even before the re-evaluation of global oil reserves (due to the addition of nonconventional deposits whose recovery became possible thanks to technical advance), and before the rapid ascent of America’s oil extraction using hydraulic fracturing, the claims about an imminent peak of global oil extraction followed by a precipitous demise of the oil era ignored several fundamental facts.

To begin with, if the peak of oil extraction is coming soon, should not then the prospective shortage of the precious fuel result in relentlessly rising prices and should not buying a barrel of oil and holding on to it be an unbeatable investment? These conclusions are patently wrong. Oil is not getting either intolerably more expensive to find or to develop, and a barrel of crude, say West Texas intermediate, bought at $17.65/b at the end of 1986 (that is, $39.38 in 2017 monies) and sold in June 2017 at $50/b would have earned (even when assuming no storage costs) an average annual return of 1.1%, a performance vastly inferior to any guaranteed investment certificates and truly a miserable gain when compared with virtually any balanced stock market fund (and a 1980 barrel sold in 2017 would have resulted in a 50% loss!).

Then there is the basic property of a bell-shaped curve. Even if a peak oil promoter had believed in 2005 that the global oil extraction would start declining in a matter of years, the Hubbertian reality would demand that half of all oil was yet to be extracted after that date, and the decline close to a normal curve shape would mean that we would still have more than a century of oil production ahead of us. This means, for example, that a symmetrical curve with the peak at 83Mbpd in 2010 would indicate global extraction of roughly 65Mbpd in 2030 and nearly 50Mbpd by 2050. Clearly, the oil era would not be nearly over even if we had already reached the peak in oil production in the recent past.

And, quite inexplicably, those who forecast an imminent peak of global oil production and a rapid end of the oil era completely ignore fundamental, and proven, economic realities and assume that future demand is immune to any external factors. This is patently false: an indisputable peak followed by precipitous decline in production would not trigger an unchecked bidding for the remaining oil but would rather accelerate an ongoing shift to other energy sources. OPEC learned this lesson in the early 1980s when record high prices were followed not only by the decline of oil’s share in global energy supply but also by an absolute decline in global oil demand and a drastic fall in price (by about 60% for an average OPEC barrel between 1981 and 1986).

Consequently, even if resources were rather constrained and we found ourselves on the declining right-hand side of the production curve, we would still have many decades of oil era left and oil prices would not reach an astronomical level. In reality, we are still on a gently ascending slope and we also have plenty of evidence that the world has more undiscovered oil than the most pessimistic past estimates would indicate. Most notably, the most comprehensive assessment of the world’s undiscovered resources of conventional oil, published by the US Geological Survey in 2000, offered the following division among the three key categories: by the late 1990s roughly 710Gb of oil had been produced worldwide, leaving about 890Gb of remaining known reserves; nearly 690Gb of oil were to come in the future from reserve additions in currently known fields, and roughly 730Gb were yet to be discovered, giving an EUR of about 3.020Tb. In 2012 the USGS updated its worldwide estimate and put the total of undiscovered conventional resources that would be technically recoverable at 565.3Gb of crude oil and 166.7Gb of natural gas liquids, or a total of 731Gb. And even the 95% confidence limit (near certainty) of the original estimate of the undiscovered reserves of conventional oil was 400Gb, that is, nearly three times as much as a typical claim by those who saw an imminent peak of global oil extraction.

According to the USGS about 60% of all undiscovered reserves are almost equally split among three regions, Latin America and the Caribbean, sub-Saharan Africa, and the Middle East, with the following six basins having the largest discovery potential: the Mesopotamian Foredeep Basin, the West Siberian Basin, the as yet completely unexplored East Greenland Rift Basin, the Zagros Fold Belt, the Niger Delta and the Rub’ al-Khālī Basin of eastern Saudi Arabia. In North America, the best prospects for major new oil discoveries are in northern Alaska, in the Canadian Arctic and in the Gulf of Mexico. In Latin America, large reserve additions will come in Venezuela and in Brazil’s offshore waters, perhaps most importantly in Foz do Amazonas, in the delta of the river. Most of Africa’s untapped oil resources are in waters off the Congo and Niger, but significant potential remains in Algeria and Libya. In the Middle East, both of the two leading producers, Saudi Arabia and Iran, will see substantial new discoveries, as will Iraq.

Given all of these uncertainties, a large number of future production curves can be drawn on the basis of different estimates of ultimate conventional oil recovery (see figure 27). For example, the EUR of about 3Tb would (depending on the future rates of consumption) imply a peak of conventional oil extraction sometime after 2020 and it would mean that global production during the 2040s could still be as high as it was in the early 1980s. But the noted assessments offer an inappropriately narrow resource perspective as they account only for conventional resources. As Jean Laherrère (one of the most vocal proponents of an early peak oil production) conceded, with the addition of the median reserve estimates of natural gas liquids (200Gb) and nonconventional oil (700Gb) there would still be some 1.9Tb of oil to be produced, double the amount of his estimate for liquid crude oil alone. Since that time, we have gained a better understanding of nonconventional resources, and the success of US shale oil extraction proved how technical advances can convert a significant share of such resources into economic reserves.

At the beginning of 2010, BP’s annual survey put the world’s conventional crude oil reserves at 1.476Tb after adding reserves in Canada’s oil sands, in Venezuela’s Orinoco heavy oil belt and in the US oil shales. The total stood at 1.707Tb in mid-2017, more than a 15% gain in just seven years. Extraction of North America’s nonconventional oil is already changing the shape of both US and Canadian EUR curves, and more reserves will be added in the future, even in the regions that were previously explored for conventional resources. The latest example of such a gain is the USGS announcement in November 2016 that credits Wolfcamp shale in West Texas with 20Gb of crude oil recoverable with current practices (and 16 Tcf natural gas): at almost three times the total in North Dakota’s Bakken shale this would make it the country’s largest crude oil deposit. Many other countries have extensive shale deposits, and in 2015 the US Energy Information Administration put the worldwide total in unproved technically recoverable reserves of shale (tight) oil at nearly 420Gb: outside North America the largest potential is in Russia, Argentina, China, the United Arab Emirates, Libya and Kazakhstan.

And the coming shape of exhaustion curves will also be greatly influenced by adjustments in demand, a phenomenon clearly demonstrated by the decline and stagnation of global oil consumption between 1979 and 1994 and, again, by the post-2008 slow-down in demand. Even greater discontinuities are possible if deliberate management (such as more aggressive efficiency goals for road vehicles) were to shape the profile of future oil demand. If a looming physical shortage of oil were to become a matter of humanity’s survival, then clear priorities could ensure an extended period of adequate supply by allocating the refined fuel according to a firm hierarchy of priority uses. Fuel for agricultural machinery, indispensable aviation and feedstock for essential petrochemical syntheses would be in the first category; fuels for long-distance transport of perishable goods in the second (but all railway traffic should be electrified); and gasoline for passenger cars would get the lowest ranking while that transportation sector undergoes gradual conversion to electric drive.

All of this means that we do not know when the global extraction will peak, at what level it will, and if it will be followed by decline mirroring the historic run-up, by a more gradual prolonged retreat or by a substantial drop followed by decades-long fluctuating plateaux. We will not get close to fairly accurate answers until we have explored all of the world’s sedimentary basins in great enough detail to offer a narrowly constrained estimate of ultimate reserves. But, as Morris Adelman put it succinctly: ‘To know ultimate reserves, we must first have ultimate knowledge. Nobody knows this, and nobody should pretend to know.’ But even if we had a perfect knowledge of the world’s ultimately recoverable oil resources, the global oil production curve could not be drawn without also knowing the future oil demand.

This is impossible because demand will be driven, as in the past, both by predictable forces (including growing populations and higher disposable income) and by unpredictable political and socioeconomic changes and, above all, by new technical advances. Four prominent historic examples illustrate the repeated extent of our ignorance, with unpredictable outcomes having consequences in either direction, that is, both boosting and depressing future price and future oil supply. In 1930 nobody could have predicted the introduction of commercial jet aircraft by 1960, an innovation that created an entirely new economic sector with a large demand for kerosene. In 1960 nobody could have predicted oil prices rising by an order of magnitude as a result of OPEC’s actions, a political shift that, for the first time since the 1860s, led to a notable decline in global oil demand. In the early 1980s, as oil prices set new records, nobody could have predicted that a quarter century later half of the passenger-carrying vehicles in the US would be gasoline-guzzling SUVs, pick-up trucks and vans. And in 2005, as media reveled in reports of the imminent peak of global oil extraction, nobody foresaw that a decade later oil prices would be sliding toward new lows as the world worried about large surpluses of oil supply!

As a result, linear assumptions based on the past rates (be they longer lasting or short-term) are risible. US oil demand rose 50% between 1965 and 1973 but less than 2.5% during the two decades between 1979 and 1999; US oil extraction decreased by 20% between 1995 and 2006, but it increased by about 85% during the next ten years. Which one of these values should we use for a truly long-range, say at least half a century, forecast of demand and supply? And there is yet another enormous uncertainty: as yet we have no idea to what extent the rising concerns about global warming will affect the future extraction of fossil fuels. Study of energy transitions and realities of inertial, embedded energy uses and infrastructure preclude any rapid abandonment of fossil fuels in general and crude oil in particular.

Even with remarkable technical advances there is no doubt that energy transitions present enormous problems for the providers of energies that are being replaced (OPEC members certainly do not look forward to any early end of the oil era), that they necessitate scrapping or reorganization of many old infrastructures (think of all the oil tankers, pipelines and refineries), and that they require the introduction of entirely new links, procedures and practices (no matter if the dominant new resources are solar or nuclear). Resulting sectoral and regional socioeconomic dislocations are thus inevitable and can be deep and long-lasting (think of the economically depressed former major coal-mining regions), the necessary infrastructural transformations will be costly (valued in trillions of dollars) and inevitably protracted (requiring decades rather than years to put in place) and their diffusion will be uneven (they always have been: even in the US many rural areas were electrified only during the 1950s and about 1.5 billion people worldwide still have no electric lights!).

At the same time, we cannot discount the possibility that concerted global action could accelerate the transition to non-carbon energies, especially as far as electricity generation is concerned – but such a shift would have only a limited impact on the global demand for refined oil products whose principal market is, and will remain, in transportation. A greater impact could result from determined efforts to limit the use of fossil carbon in order to prevent excessive rise of atmospheric temperature caused by the emissions of anthropogenic greenhouse gases. In 2017 about one-third of all CO2 emissions from commercial energy use originated in combustion of refined oil products (in 1950 it was about one-quarter), a share too large to achieve meaningful reductions of future emissions by concentrating only on other fossil fuels. So far, our efforts have been quite inadequate. The goal of the Paris (COP 21) meeting of November 2015 was to work toward limiting the rise of average tropospheric temperature to no more than 2oC – yet even if all the national pledges submitted at that time were completely fulfilled, the meeting’s final document noted, with concern, that the estimated emissions in 2025 and 2030 would not fall within the desired scenarios but rather lead to a further increase of annual CO2 generation.

Achieving the desired goal would require actual cuts in the current rates of fossil fuel combustion, including substantial reductions of oil use. The chances of ending the fossil fuel era in a matter of two or three decades appear quite unrealistic: in 2017 the world derived about 85% of its primary commercial energy from the combustion of fossil carbon. The coming years will show how far our efforts will go, but it is most likely that we will not stay below 2oC and will have to do our best to adapt to the resulting warming. At the same time, a long-term outlook is more encouraging: in two generations (2060s) we will have made substantial progress toward decarbonizing the global energy supply and although we will still rely on oil as one of the fundamental energizers of our economies we should be doing so with a greatly reduced environmental impact.

A fundamental general consideration needs to be stressed before I proceed with a brief outline of alternatives to conventional liquid oil. Substitutions that are already technically proven (such as gas-to-liquid conversions) or that appear as highly promising future candidates (for example ethanol production from cellulosic biomass) are nevertheless often seen as unacceptable or impractical simply because they cost (or are projected to cost) more than the conversion they are set to replace. This simplistic cost argument is misleading for three principal reasons. First, it does not acknowledge that the real cost of today’s liquid fuels is higher (often substantially so) than the price directly paid by consumers. Second, it implies that only the resources and conversions secured with the lowest cost are worthy of consideration regardless of the environmental or strategic implications of their use. Third, it ignores the fact that modern societies are already paying far less (as a share of disposable income) for their energy needs than at any time in history and hence even a doubling of such vital expenditures would not be catastrophic.

The last point is true even in the country that is most addicted to excessive driving. Detailed surveys of US consumer spending show that in 2015 an average family spent less on gasoline (3.7% of all expenditures) than it did on entertainment (5%) or food away from home (5.4%). Why then should we be panicked by the prospect of an alternative motor fuel that retails (say, arbitrarily) at twice the price of today’s gasoline? If that new fuel were to be used in vehicles operating with double the efficiency of today’s cars (given the poor average performance of US cars this is an easily achievable goal), then even this low share of expenditure may remain unchanged! And the argument about unacceptably higher costs of alternative fuels is insufficient if they could be produced and converted with lower environmental burdens (including lower emissions of greenhouse gases) or if they will provide important strategic benefits.

Such considerations may drive future efforts to extract even more oil from known reservoirs: after all, even today’s best enhanced recovery methods still leave behind 40–50% of the oil originally in place, and higher prices may justify more expensive recovery techniques. Mining of progressively poorer mineral ores is perhaps the best analogy. And beyond the conventional liquid oil there are vast resources of nonconventional hydrocarbons whose recovery is already contributing to most of the crude oil produced in North America (a combination of American shale oil and Canadian oil extracted from oil sands) and will be gradually adopted in other parts of the world rich in such nonconventional oil resources. This is an apposite place to stress that there are no sharp and obviously apparent divides between the two kinds of oil resources and that the same is true for their extraction. In quality terms, this continuum extends from light oils (API gravity of at least 25o) to medium heavy oils (20–25o API, mobile at reservoir conditions) to heavy oils (10–20o API, still somewhat mobile) to bitumen (less than 10o API, non-mobile) and finally to oil shales with minimal or virtually no permeability.

Limited volumes of medium and of some extra heavy oils have been produced for many years in Saskatchewan and Venezuela. Alaska’s North Slope also has large deposits of heavy oil (up to 33Gb) and experiments have begun with its commercial production. But most of the world’s 4Tb of heavy oils in place are found in Venezuela (close to 1Tb) and in North America’s most important commercial concentration, in Alberta’s oil sands, with proved reserves of about 166Gb in 2014, making Canada the country with the world’s third largest oil reserves behind Venezuela and Saudi Arabia. Total (mining and in situ) extraction reached 2.6Mbpd in 2017, which means Alberta’s conventional oil contributed less than 15% of the province’s total oil extraction. Much like the future of US oil, Canada’s oil prospects thus depend heavily on the recovery of a nonconventional resource.

OIL FROM SANDS

Extraction of oil from oil sands was commercialized for the first time on a small scale during the late 1960s. Suncor was the first company to produce oil from Alberta oil sands near Fort McMurray in 1967, and the Syncrude consortium, formed in 1965, has been producing in the area since 1978. Both of these pioneering projects operate large oil sand mines, using huge excavators to mine the rock and the world’s largest off-road trucks to transport it to a bitumen extraction plant from which it moves to an upgrading facility to yield light crude oil. Mining of Alberta oil sands, extraction of bitumen and its upgrading to light crude oil returns about six units of energy for every unit invested. Only about a fifth of all recoverable oil in Alberta’s oil sands can be reached by surface mining; the rest will have to be extracted in situ and two techniques (see figure 28) have been commercialized so far, cyclic steam-stimulation (CSS) and steam-assisted gravity drainage (SAGD).

Imperial Oil’s Cold Lake Project was the first CSS recovery that alternates periods of injecting hot pressurized steam (300°C, 11MPa) into well bores with periods of soaking that loosens the bitumen. These cycles last from a few months to three years and the heated bitumen–water mixture is drawn from the same wells that were used for steam injection. On average, this process extracts about a quarter, and with follow-up processes up to 35%, of the bitumen originally present in sands. SAGD was patented by Imperial Oil in 1982 and Cenovus Energy is now its leading practitioner in Alberta oil sands. The process uses two horizontal wells (typically 500–800m long) that are drilled near the bottom of an oil sands formation and are separated by a vertical distance of 4–6m. Steam injected into the top well heats the surrounding bitumen that slowly drains into the bottom well from which the mixture of water and oil is then lifted. Cenovus also injects butane along with steam and adds a horizontal well between a pair of SAGD wells in order to boost recovery to as much as 70%. Separated water is recycled to generate steam, and while the early in situ recovery required (in volume terms) nearly eight units of steam to produce a unit of oil, the latest practices have reduced that (and with it the energy cost to produce steam) to two units or even slightly less. Although most companies still operate with EROEI of 3:1, the best recovery practices have doubled that ratio.

Figure 28 Recovering oil from sands: CSS and SAGD techniques

Petróleos de Venezuela has been exporting heavy boiler fuel (coal or gas substitute), a mixture of 70% natural bitumen, 30% water and a small amount of an additive that stabilizes the emulsion, as liquid Orimulsión burned in electricity generating plants, but any large-scale development of the Orinoco Belt’s enormous nonconventional oil resources will be determined not only by a complex interplay of oil prices and technical advances but also by the political situation in what has become one of the world’s most precarious economies with declining conventional oil extraction (20% down between 2006 and 2015).

Before nonconventional oil became a major addition to the global supply of liquid fuels, there was much interest in gas-to-liquid (GTL) conversions. Their technical feasibility was demonstrated for the first time in 1923 with the production of motor fuels from coal gas by a process invented by Franz Fischer and Hans Tropsch in Germany. During World War II the Fischer-Tropsch synthesis kept the Wehrmacht and Luftwaffe supplied with fuel; the output peaked in 1943 at 124,000bpd before it was reduced by Allied bombing. Inexpensive imported crude oil made further GTL ventures uneconomical, but in 1955 production began at South Africa’s Sasol plant. The improved Sasol process was used abroad for the first time in 2007 when Oryx GTL was completed in Qatar. Shell built its first GTL plant in Bintulu in Malaysia in 1993 and in 2012 completed the world’s largest project, Pearl GLT in Ras Laffan in Qatar, using natural gas from the world’s largest gas field in the Persian Gulf.

But the high cost of these plants and fluctuating, low costs of crude oil have not led to any large-scale adoption of these designs, and in 2017 all operating GTL projects supplied less than half a percent of worldwide liquid fuel production. The ultimate goal of natural gas conversions is to commercialize new methods that would transform methane into affordable energy carriers. George Olah, the Nobel Prize winner in chemistry for 1994, has argued that a methanol would be the best choice. This liquid hydrocarbon (CH3OH) can be prepared by a number of methods – above all, by direct oxidative conversion of natural gas, and also by catalytic reduction of CO2 – and it is safer and less expensive to handle than hydrogen, a gas that has been seen by many as the ultimate energy carrier of future non-carbon economies.

But near-term displacement of oil by gas does not have to go through liquids. Direct replacement of liquids by natural gases has been underway for decades and it will continue for decades to come. As already noted, natural gas replaced fuel oil in heating all but a small share of homes both in North America and in Europe, and it has been substituted for fuel oil in many electricity-generating plants and in industrial enterprises. Natural gas is also highly valued petrochemical feedstock. And natural gas can replace motor gasoline in passenger cars in two ways: directly, and with only minor engine modifications, as compressed natural gas suitable for urban fleet operations and other city driving; and indirectly by using highly efficient combined cycle gas turbines to generate electricity for electric cars. Simply put, with the exception of flying and long-distance land and maritime transport (energy density of natural gas under normal pressure is only 1/1000 that of liquid fuel, making it unsuitable as a portable fuel sufficient for extended travel), everything that is done with liquid fuels can be done with gases.

NATURAL GASES: PROPERTIES AND RESERVES

Much like oils, natural gases are mixtures of variable proportions of hydrocarbons, but unlike oils they are primarily mixtures of just three of the simplest alkanes: methane, ethane and propane. Higher homologues (butane, pentane and hexane, are separated as natural gas liquids) and CO2, H2S, N, He, and water vapor, found in many gases, are also separated before the gases are compressed and transported by pipelines. Natural gases are commonly associated with crude oils but they also exist as free (dry) gases without any contact with crude oil in an oil reservoir or in entirely separate gas-bearing formations. Their heat content ranges between 30 and 45MJ/m3 (35.5MJ/m3 for CH4), they are the least polluting fossil fuels, and they generate the least amount of CO2 per unit of energy. As with crude oil, conventional gas reserves have been steadily increasing and by 2017 they had reached 190Tm3, or in energy terms nearly as much as the total 2016 reserves of conventional crude oil.

This increase has not only accommodated the expanding extraction (it rose more than 2.1-fold between 1985 and 2016, to 3.5Tm3) but it maintained the global R/P ratio at more than fifty years. Conventional reserves are concentrated in Iran (about 18% of the total), Russia (about 17%), Qatar (about 13%) and Turkmenistan (nearly 10% of the total). The Middle East claims just over 40% of the global total, much less than its share of crude oil. Gas associated with oil used to be simply flared as an unwanted by-product but this wasteful practice has declined with the rising demand for clean household and industrial fuel. In 1975, gas equal to about 14% of worldwide production was flared, with major sites visible on nighttime satellite images as lights brighter than those of many large cities. Worldwide flaring has been slowly declining since the beginning of the new century but in 2015 it had a slight uptick and it still accounts for slightly more than 4% of global production (more than China’s annual output in that year!), with Russia, Iraq, Iran, the US (due to rapidly expanding shale oil extraction) and Venezuela being the top offenders.

For decades those large natural gas reserves that could not be accessed by a pipeline could not be used, creating large stores of so-called stranded gas. This limitation began to change only during the early 1960s when the first liquefied natural gas (LNG) tankers were used to export Algerian gas to the UK and France, and Indonesian gas to Japan. But the gas liquefaction plants and special tankers were expensive and the export remained limited. LNG occupies only about 1/600 the volume of natural gas but it must be cooled to −162°C and regasified after delivery. As both the liquefaction process and the construction of special tankers with highly insulated containers became more affordable, LNG trade took off during the 1990s and its unfolding expansion has finally changed the natural gas trade into a truly global endeavor with substantially lower prices and a growing number of suppliers and buyers. In 2017 eighteen countries were exporting LNG from more than 70 liquefaction plants (with Qatar, Malaysia, Australia and Nigeria in the lead) and thirty-nine countries were importing the fuel, with all of the largest buyers (Japan, South Korea, China and India) in Asia. Nearly 400 LNG tankers were used in the exports, the largest ones with capacities of more than 250,000m3, and 105 of them carried almost 33% of all traded gas.

Prospects for long-term supply of natural gas look highly promising. In 2000 the US Geological Survey put cumulative production of conventional natural gas at 292 billion barrels of oil equivalent (Gboe) and remaining reserves at 800Gboe, and assumed reserve growth of about 610Gboe and undiscovered reserves at 866Gboe for the total EUR of 2.57Ttoe, only about 15% less than EUR of crude oil. All of these figures refer to conventional resources only. The only nonconventional gas resource that has been exploited for decades is coalbed methane, but since 2005 the US production of shale gas, which resulted in a 50% increase of output by 2015, showed that hydraulic fracturing can transform the industry as much as it has transformed he extraction of crude oil. The US Energy Information Administration put the undiscovered but technically recoverable resources of shale gas at more than 210Tm3, slightly larger than the existing reserves of conventional gas. The countries with the largest shale gas potential are China, Argentina, Algeria, the US, Canada and Mexico.

As large as shale gas resources may be, they are insignificant when compared to methane hydrates (clathrates) that were formed by the gas released from anoxic decomposition of organic sediments by methanogenic bacteria and are now trapped inside rigid lattice cages of frozen water molecules. Fully saturated gas hydrates have one CH4 molecule for every 5.75 water molecules and hence 1m3 of hydrates contains as much as 164m3 of CH4. Two environments favor the formation of hydrates: polar continental sediments (at depths between 100m and 2.5km) and sediments beneath the ocean floor in many latitudes. More than 200 gas hydrate deposits have been identified worldwide, including those in the Russian, American and Canadian Arctic, off California and Guatemala and in Japan’s deep (water depths about 4,700m) Pacific Nankai Trough (hydrates about 4,800m below the sea bottom). The resource base of methane hydrates is so immense that only gross approximations of its size are now possible, and even very conservative estimates indicate the enormity of resources in place. The global total may be as much as 10Tt, or about twice as much as all carbon stored in coals and conventional hydrocarbons, while hydrates below the seabed of American coastal waters may be three orders of magnitude more voluminous than the country’s conventional gas reserves.

The first small-scale production (decompression and heating) trials were done in Canada in 2001 and 2002, but abundant shale gas supplies ended further experiments. In 2013 the Japanese experimented with the recovery of Nankai Trough hydrates and obtained good flow rate before the pumps became clogged by incoming sand, and in 2017 the Chinese conducted successful recovery tests in the South China Sea. No commercial breakthrough will come anytime soon and any large-scale developments will have to take into account the possibility of a sudden catastrophic release of some hydrate deposits into the atmosphere, a most unwelcome contribution to anthropogenic greenhouse gas emissions. But these challenges are hardly a valid argument for eliminating this enormous resource from future consideration as a major source of hydrocarbons (perhaps even by the middle of this century): doing so would be akin to claiming in 1930 that any extraction of oil from offshore fields out of sight of land was impossible, or to concluding in 1950 that no oil or gas would ever be produced from shales.

Greater use of natural gas can lower our reliance on crude oil in many economic sectors. Abundant and affordable natural gas can eliminate all remaining uses of crude oil and refined oil products in electricity generation and space heating; vehicles used in city traffic could easily be converted to run on compressed natural gas; trucks and LNG trucks could be powered by liquefied gas; and many industrial enterprises relying on refined products for process heat or hot water can generate those needs with higher efficiency by switching to gas. Highly efficient combined cycle gas turbines could produce electricity to power electric cars, and their generation costs could be lowered by integrating them with renewable conversions in sunny and windy places. At the same time, it will be a much greater challenge to reduce further, and eventually eliminate, our dependence on refined fuels in shipping and flying.

That is why any long-term assessment of oil’s futures must consider both the supply prospects and the imperatives of demand. As for the supply, we have still formidably large (and still growing) conventional crude oil reserves, at least as large (but most likely considerably larger) technically recoverable reserves of oil and natural gas in shales and improving capabilities of gas-to-liquid conversions: this combination means (even when assuming no early breakthroughs in hydrate recovery) that hydrocarbons should remain the leading source of global commercial energy supply far beyond the middle of this century. On the demand side, many existing uses of refined oil products could be replaced by natural gas or by primary (renewably generated) electricity, but it is most unlikely that by 2050 we will see the entire global car and truck fleet operating without gasoline and diesel, and the prospects for concurrently eliminating diesel fuel and kerosene from, respectively, shipping and air transport are considerably less.

In a speech accepting the Biennial OPEC Award for 2006, the late Peter Odell, one of the most astute, life-long, observers of the global oil scene, concluded that ‘peak-oilers’, much like their numerous predecessors, will soon be proven wrong, that the present contribution of oil and gas to the global energy supply will be only modestly reduced by 2050 and that natural gas will surpass oil as the leading source of primary energy. At that time, I concurred with Peter in my writings, and more than a decade later I still see such conclusions as the most sensible qualitative forecasts of the world’s energy futures (as always, I stay away from any specific quantitative details, as such long-term predictions are beyond our capability).

And as the importance of this mixture of liquid and gaseous and conventional and nonconventional hydrocarbons recedes, first in relative and later also in absolute terms, renewable energy flows will be gaining higher shares of global primary energy supply. Liquids from biomass (ethanol from sugar cane, biodiesel from oil seeds) are already displacing some gasoline and a small share of diesel fuel. But the future should not belong to the currently heavily promoted (but in many ways problematic) conversion of grains (above all corn) to ethanol and oil seeds to biodiesel, but to innovative bioengineering processes converting more abundant cellulosic biomass, including the share of crop residues (mainly straw) that does not have to be recycled to maintain soil quality and high-yielding perennial grasses planted on non-agricultural land.

Wind-powered electricity generation, onshore and offshore, still accounts for only a small part of global generation (about 4% in 2016) but it has already made major strides in a number of European countries and larger turbine sizes (in 2016 up to 8MW) and higher capacity factors (some offshore wind farms can generate electricity for up to 47% of the time) will raise its contributions for many years to come. In 2016, photovoltaic conversion of solar radiation contributed even less to the global electricity generation total (just over 1%) but its numerous advantages (quiet, no moving parts, durable, high power density), falling costs and improving efficiencies guarantee its continued future expansion. Although these two conversions may seem to do little for expanding the supply of liquid fuels they could actually fit perfectly into a system of rechargeable cars, be they hybrids or pure electrics.

The inherent fluctuations and unpredictability of photovoltaic electricity generation in temperate latitudes would be much less of a problem for recharging plug-in hybrid cars than for lighting houses or running machines in a factory where electricity must be available on demand. In contrast, a car, be it in a garage or in a parking lot at a place of work, could be recharged whenever a surge of renewable electricity became available.

Regardless of the actual rate of oil extraction and the eventual date of the highest annual production of oil from any resources, there is no reason to see the transition to the post-oil era as a period of unmanageable difficulties or outright economic and social catastrophes. Historical evidence is clear: energy transitions have always been among the most important stimuli of technical advances (think of new prime movers, new materials and new energy converters), promoting innovation (such as the profound managerial and organizational changes brought by computers), higher efficiency (for example, a gas turbine vs. steam engine) and resource substitution (like the substitution of coke made from coal for charcoal from disappearing forests during the late eighteenth and nineteenth centuries). Their outcomes – coal replacing wood, oil replacing a great deal of coal, now natural gas already replacing a great deal of oil – have shaped modern industrial, and post-industrial, civilization, leaving deep imprints on the structure and productivity of economies as well as on the organization and the quality of life of affected societies.

There is no doubt that the unfolding energy transition (whose eventual outcome would be the replacement of fossil fuels by non-carbon energies) is extraordinarily challenging, mainly due to the scale of our reliance on fossil fuels, cost and inertia of enormous infrastructures required for its reliable functioning, and the continuing concentration of humanity not just in cities but in megacities containing more than 10 million people. As with all long-range perspectives, it is counterproductive to pinpoint any rates of progress, dates or shares for new energy conversions that will be needed to accomplish the epochal shift. This we know for certain: although a small nation can switch from one dominant form of energy to another in a matter of years, energy transitions are normally protracted affairs, extending across decades rather than years. And neither the tempo nor the eventual achievements of these long transitions from first commercial uses to widespread embrace to eventual domination can be judged by the state of affairs during the initial stages of expansion. Precisely because this transition will have to be gradual and protracted, technical means available two or three generations from now may provide effective solutions for many of today’s intractable problems.

Energy transitions – from biomass to coal, from coal to oil, from oil to natural gas, from direct use of fuels to electricity – have stimulated technical advances and driven our inventiveness. Inevitably, they bring enormous challenges for both producers and consumers, necessitate the scrapping or reorganization of extensive infrastructures, are costly and protracted and cause major socioeconomic dislocations. But they have created more productive and richer economies, and modern societies will not collapse just because we face yet another of these grand transformations. The world beyond oil is still several generations away but we should see the path toward that era as one of very challenging, but also immensely rewarding, opportunities as modern civilization eventually severs its dependence on fossil carbon.

5

How long will oil last?