by Vaclav Smil
If a portion of a resource is to be classed in the category of economic reserves, it is not enough just to illuminate an object and ascertain its location and size (now done first through remote sensing and then in detail with on-site geophysical assessment). It is also necessary to probe the resource in sufficient detail (through exploratory drilling) in order to determine what techniques will be used for its recovery (e.g., horizontal vs. vertical drilling) and what cost its extraction and delivery to the market will entail. Consequently, reserves are not a given natural category but a quantity that is derived from resources through scientific, technical, and managerial means: human ingenuity keeps translating a growing share of naturally available resources into economically exploitable reserves.
This process is easily illustrated by following the progression of reserve-to-production ratios. The ratio is expressed in years, and it is simply the quotient of a specific reserve in a given year and output of a commodity during that year. Obviously, there is no economic incentive to invest into exploratory activities that would result in very large (>25 years) R/P ratios, and inversely, there can be no commitment to actual extraction unless the R/P is kept above a certain minimum (at least 10 years). In any case (and contrary to a common misperception), low or falling R/P ratios are not an inevitable sign of any imminent resource exhaustion: long-term trends of R/P ratios for the US natural gas demonstrate this reality. Between the late 1920s and the late 1940s, the ratio rose (with fluctuations) from about 20 to about 40 years, then a steep decline (until 1970, to 15 years) and a more moderate decrease brought it to just 10 years by the early 1990s, and a subsequent slow rise lifted it to 12.5 years by 2012.
This means that the ratio has been fairly steady for more than 40 years, while the US natural gas production first declined by about 20% (between 1970 and 1987) and then kept rising so that by 2011 it was nearly 10% above the 1970 mark. The ratio thus informs us primarily about the economic and technical capacities of transforming a resource to a reserve, not about any imminent peak of extraction or approaching resource exhaustion: it can be falling when low prices do not justify aggressive additions to reserves, it can be rising when new capabilities for tapping nonconventional resources expand the resource base and add rapidly to new reserves. A retrospective look shows the following natural gas R/P ratios for some of the major producing countries in the year 2000 and 2012 (with all numbers rounded to the nearest year): the United States, 9 and 13; Canada, 11 and 13; Norway, 23 and 18; Russia, 83 and 55; Algeria, 55 and 55; and Australia, 41 and 77—and the world 62 and 56 years (BP, 2014a).
In the early decades of natural gas industry, the US reserves dominated the global total: they were put at 424 Gm3 in 1919, and by 1950, they were an order of magnitude higher at roughly 5.5 Tm3. After reaching a peak of about 8.2 Tm3 in 1970, they declined to less than 4.6 Tm3 by 1994 and grew slowly for the rest of the decade, but boosted by shale gas development, they reached 5.8 Tm3 by 2005 and 8.5 Tm3 in 2012. In 1950, the US reserves were more than half of the global total of about 10 Tm3; by 1960, North America was still the leading region with more than 40% of the world’s gas reserves, but during the 1960s, it was surpassed by steadily rising Iranian total, and its reserves became a fraction of the Soviet finds.
These discoveries were made in a huge (about 2.2 million km2) West Siberian Basin, mostly east of the Ob River just before it flows into the Arctic Ocean (Gazprom, 2014). They began with the Taz in 1962 and 4 years later came Urengoy field, at the time the world’s largest, with the estimated ultimate recovery of 9.5 Tm3 (Grace and Hart, 1991). The basin’s second largest field, Yamburg (originally estimated at 4.7 Tm3), was discovered in 1969, and three other extraordinary fields, Zapolyarnoye (2.3 Tm3, 1965), Bovanenkovo (almost 5 Tm3, 1971), and Medvezhye (2.4 Tm3, 1967), helped to make the area the world’s largest concentration of natural gas outside of the Persian Gulf (see Figure 2.2). By 1990, a year before its dissolution, the Soviet natural gas reserves reached 30% of the global total; their official total rose to about 45 Tm3 in 1991. By the year 2012, the countries of the former Soviet Union held at least 55 Tm3 of gas reserves or 29% of the world total, but Russia’s total was, again, surpassed by Iran whose gas reserves grew by 30% during the first 12 years of the twenty-first century.
Iran’s reserves are dominated by the country’s share of the world’s largest natural gas accumulation which contains almost 20% of all discovered gas. Qatari part (North Dome, al-Idd al-Sharqi) was discovered in 1971 but Iran explored the field’s northern extension, South Pars in its territorial waters, only in 1990 (Figure 2.5). Although South Pars covers only about 40% of the field’s area (total of 9700 km2), it contains just over 70% of its proved reserves, about 25.5 Tm3 of the total of 35.7 Tm3 (Esrafili-Dizaji et al., 2013). With about 51 Tm3 of gas reserves in place, this implies a recovery rate of 70%. The field also contains some 19 Gb (about 2.6 Gt) of recoverable condensate, and at a depth of about 1 km, there are three oil-bearing layers.
Figure 2.5 North Dome/South Pars gas field.
Field sizes also have a highly skewed distribution: nearly 20,000 of hydrocarbon fields have been discovered since the 1850s, but 1,087 giant fields found by 2012 hold about 72% of proved and probable oil and gas reserves (Bai and Xu, 2014). Giant fields are defined as those containing at least 500 million barrels of oil equivalent (Halbouty, 2001), be it just crude oil (that would be about 70 Mm3), mixture of oil, natural gas liquids and natural gas, or nonassociated gas (with 7.2 barrels/t and 42 GJ/t that equals roughly 3 EJ or 85 Gm3 of natural gas). Mann, Gahagan, and Gordon (2001) provide details of tectonic setting for 592 of these formations. As for the conventional reservoirs of natural gas, 10 largest fields contain more than one-third of the world’s 2013 conventional gas reserves, and the share is more than 40% for the top 20 fields.
The only two regions that have not seen steadily rising natural gas reserves during half a century between 1960 and 2010 were North America and Europe. The North American record volume reached in 1983 was not surpassed until 2011, and EU reserves have been in decline since the beginning of the twenty-first century, but in 50 years preceding 2010 the Latin American reserves had quintupled, Middle Eastern reserves grew more than 15-fold, and the reserves of the countries of the former USSR grew more than 25 times. As a result, global reserves of natural gas rose from about 19 Tm3 in 1960 to 72 Tm3 in 1980, to at least 140 Tm3 in the year 2000, and to more than 177 Tm3 in 2010, more than a ninefold rise in 50 years. As with most energy resources, natural gas endowment has a highly skewed spatial distribution, with just four countries (Iran with roughly 34 Tm3, Russia with 33 Tm3, Qatar with 25 Tm3, and Turkmenistan with about 18 Tm3) harboring nearly 60% of global reserves in 2012, a slightly more highly concentrated distribution than that for crude oil: when nonconventional proved oil reserves are included, the top four countries, Venezuela, Saudi Arabia, Canada, and Iran, account for about 56% of the world total (BP, 2014a).
To know (within an acceptable error, say, no more than ±25%) the total volume of natural gas that could be eventually produced would be much more valuable than to know the total volume of oil originally in place. Of course, we know that the resource base (the room, gas originally in place) is finite, but to ascertain its ultimate magnitude is a notoriously difficult challenge, and even if we were to know its size with a high degree of accuracy (after our searchlights will have penetrated the furthermost recesses of the room), we still could not immediately translate this knowledge into clear assessment of ultimate production, the sum of already produced gas, remaining reserves, conventional reserve growth (the overall volume for a given location tends to increase with time as more exploration is done), and undiscovered conventional gas.
Because reserves are determined by our technical abilities and economic possibilities, what appears unattainable today can become a matter of routine recovery in a few years or decades: recent rapid expansion of natural gas pro
duction from American shales is a perfect example of this reality as a combination of newly affordable horizontal drilling and improved hydraulic fracturing transformed a significant share of a previously untapped resource into highly economical reserve of gaseous fuel. But as these nonconventional resources have not been a part of the past global assessments, I will leave them in the following comparisons and do their appraisal in Chapter 6.
Having a good estimate of ultimately recoverable natural gas would allow major producers to do better long-range planning on local, regional, or national basis, and it would allow us to make better assessments of long-term global energy supply prospects of modern civilization. Here is what we know. The first global assessment of ultimately producible natural gas was done by using a rough multiplier relating gas to ultimately recoverable crude oil. Half a century ago, such crude estimates yielded at least 216 Tm3, and in the mid-1970s, the most likely volume was put at 280 Tm3 (Hubbert, 1978). A decade later, the USGS estimate was nearly 260 Tm3, and then regular global USGS assessments, based on evaluating 171 geologic provinces, raised the volumes to about 295 Tm3 for 1990 and 435 Tm3 for the year 2000 (USGS (United States Geological Survey), 2000).
The year 2000 total was composed of roughly 50 Tm3 of cumulative production (less than 12% of ultimately recoverable reserves, with just below half of that total extracted in the United States), 135 Tm3 of remaining reserves, 104 Tm3 of reserve growth, and 146 Tm3 of undiscovered conventional gas. In 2012, the USGS put its latest estimate of undiscovered conventional natural gas outside the United States about 9% higher, at 159 Tm3 (Schenk, 2012). And the USGS had also estimated that in 2012, reserve growth, that is, potential additions to conventional gas resources of the world (again outside the United States) in discovered giant gas fields (with reported gas in place of at least 85 Gm3), was at least 40 Tm3 of gas and 16 Gb (about 2.2 Gt) of gas liquids (Klett et al., 2012).
But this ascending progression of reserves does not mean that eventually all undiscovered but potentially producible resources will shift into reserve category—and it would be quite misleading to think that the story of exploitation will end with physical exhaustion. Significant shares of such resources will always remain uneconomical (in too small concentrations, too scattered, of exceedingly poor quality), and in the case of hydrocarbons embedded in reservoir rocks, only the actual mining of those (more or less) porous formations (as is done with Alberta with oil shales) could recover virtually all fuels originally present in place. This means that in virtually all cases, conventional extraction of underground reservoirs leaves behind significant shares (as much as 70–75%) of liquids and gases inside the porous rocks.
A more practical indicator of extraction prospects is the cost of marginal production, a changing sum that is affected both by technical advances and by the ability of economies to pay the price. Technical advances may lower the cost of existing extraction or they may increase it but make available previously unexploitable resources and hence improve the security of supply that may be worth of higher costs. Affordability varies among economies, but there is no simple positive correlation between the level of per capita GDP and the ability to pay more (still relatively poor China has huge foreign reserves and the country has been a rising importer of all fossil fuels), and some countries may have little choice paying exorbitant prices (the case of Japan’s LNG imports after Fukushima: for more, see Chapter 6).
Once the marginal costs become unsupportable, a particular extraction locality (and eventually all, or virtually all, production of a specific resource in a country) is gradually abandoned (British underground mining is an excellent example of this reality)—and economies respond by a mixture of reduced uses and resource substitutes. Hence, there are no output peaks followed by precipitous collapses, rather (slower or faster) shifts toward new use and supply patterns. The cited assessments indicate that global natural gas extraction can keep rising to new record levels for decades to come. Assuming that the recovery of additional 400 Tm3 of conventional gas will follow the normal (Gaussian) function roughly fitted to the past progression, the peak of global extraction would come around 2050, and after gradually declining, the annual output at the beginning of the twenty-second century would be still about as large as it was in the year 2000.
But production curves of natural resources are perfectly symmetrical only in theoretical studies; in the real world, they show all kinds of deviations, and their progression is dependent on inherently inaccurate estimates of ultimately recoverable fuel totals: hence, a less steep progression of global gas extraction would bring a production peak perhaps only around the year 2070—and various assumptions regarding the recovery rates of nonconventional natural gas resources (see Chapter 6) could raise the output peak as well as lengthen the duration of natural gas era. Here is perhaps the best illustration why such long-range forecasts are little more than amusing exercises with little relation to reality.
In 1956, M. King Hubbert, American geophysicist and advocate of symmetrical production curves for mineral resources that appear to be well suited to forecast output peaks, put America’s ultimately recoverable natural gas resources at 24 Tm3 (Hubbert, 1956), and in 1978, he raised that total to 31.2 Tm3 (Hubbert, 1978). These totals led him to forecast the US peak gas production in 1973, either at about 400 Gm3 (with the 1956 total) or 620 Gm3 (with the 1978 total). But by the end of 2013, the cumulative US gas extraction reached 34 Tm3, nearly 10% above Hubbert’s grand total of all recoverable gas, and the 2013 annual production was about 12% above the previous (1973) record.
And while the biennial assessments of technically recoverable US natural gas endowment were reporting a fairly steady volume between 1990 and 2004 (rising from 28.3 to 31.7 Tm3), their totals rose to 53.7 Tm3 in 2010, and the latest assessments by the Potential Gas Committee (2013) put the remaining potential gas resources (conventional, tight, shale, and coal bed) at 67.5 Tm3 (with shale gas accounting for 45% of that total or more than the conventional resources in the year 2000), which is more than twice as high as Hubbert’s higher (1978) value: so much for orderly cumulative production curves and pinpointed peaks.
Similar uncertainties regarding ultimate recoveries and the most likely production peaks apply to crude oil, and hence, it is only as an interesting aside that I introduce some published comparisons of the two resources. Laherrère (2000) put the ultimately recoverable gas at 1.68 Tb of oil equivalent, the total nearly identical to his estimate of ultimately recoverable crude oil. In the same year, World Petroleum Assessment put the total volume of recoverable conventional oil nearly 60% higher at 2.659 Tb compared to 2.249 Tb of oil equivalent in natural gas. That is only about 18% difference, most likely within the estimate’s error range, and hence it is a fair statement, as the first order of approximation, that our best understanding of recoverable hydrocarbon resources puts crude oil and natural gas at nearly the same total energy level.
But the combination of more focused appraisals of unconventional resources and technical advances in their recovery may eventually make the marketable reserves of natural gas significantly higher than those of liquid fuels. A recent Canadian gas resource assessment illustrates this continuing process. In November 2013, the National Energy Board released the first study of marketable unconventional petroleum resources in the Montney Formation in west-central Alberta and northeastern British Columbia (NEB, 2013). Montney siltstones and sandstones are estimated to hold 12.7 Tm3 of marketable natural gas, 2.3 Tm3 of marketable natural gas liquids, and 1.1 Gb of marketable crude oil. These reserves are equivalent to nearly half of the reserves in the world’s largest gas field in the Persian Gulf, and the Board concluded that the overall volume of natural gas resources in Western Canada is likely to increase as other unconventional formations receive closer attention. I will take another look at nonconventional resources of crude oil and natural gas when I will compare their global endowment in Chapter 6.
3
/> Extraction, Processing, Transportation, and Sales
As long as most of natural gas production was associated with crude oil extraction (i.e., for most of the twentieth century), the term oil and gas industry was a much more apt description of activities involved in finding, producing, and transporting gaseous fuel, and of course, in many hydrocarbon basins, it is still the case. But numerous post-WWII discoveries of nonassociated (dry or wet) gas; development of new giant and supergiant gas fields in North America, Europe, Asia, Africa, and Australia (including, in all regions, important offshore deposits); construction of transcontinental pipelines; and emergence of large-scale liquefied natural gas (LNG) trade have created a distinct, and now truly world-spanning, natural gas industry that now expects further vigorous growth in decades ahead.
This chapter presents a basic how-we-do-it sequence of activities that make up this quintessential (yet largely hidden) modern industry: how we look for new, commercially viable resources; how we extract the gas from reservoirs, coal beds, and shales; how we process it before we send it into pipelines; how we transport it across long distances and then distribute it to individual consumers; how we store it for often prolonged periods of high demand; and how, and how much, we have been selling to consumers. The only segment of modern natural gas industry that is deliberately left out is the intercontinental transportation of LNG, an increasingly important activity that will get a closer look in the fifth chapter dealing with the emergence of the global gas trade. But before I begin a brief systematic review of exploration, extraction, and transportation activities, here is perhaps the most apposite place for a few paragraphs about the early history of natural gas production and the reasons for its relatively slow pre-WWII development.