by Vaclav Smil
And China will not be the only nation where shale gas development is unlikely to replicate the post-2007 American experience. While blanket bans on hydraulic fracturing (à la France) make little sense, many expectations will be disappointed and many opportunities that might become profitable after periods of trials, errors, and adjustments will be missed because of early failures. In 2013, Talisman Energy of Calgary and Marathon Oil of Houston withdrew from shale gas exploration in Poland, a country that embraced the development of this new resource put by Kuuskraa et al. (2011) at 5.2 Tm3 of recoverable gas in 2011 but reduced by the Polish Geological Institute to just 800 Gm3 a year later. Development of Mexico’s large shale gas resources has been off to a very slow start, with fewer than 25 wells drilled by mid-2014. And the extraction of Argentina’s giant Vaca Muerta shale gas deposit is complicated by the country’s notoriously precarious fiscal situation and uncertainty about its domestic energy policy.
6.2 CBM AND TIGHT GAS
America’s success in shale gas extraction has diverted attention from other kinds of nonconventional source of natural gas, and although these resources may not be developed with a similar speed and at a comparable extent, they will become increasingly more important contributors to gas production. Kuuskraa (2004) emphasized their importance by pointing out that 8 out of 12 of the largest US gas fields are nonconventional: the largest two, Blanco and Basin in New Mexico, are a combination of tight gas and CBM; numbers five, six, eight, and nine (in Texas, Colorado, and Wyoming) contain gas in tight formations; and number four (Wyodak in Wyoming) is a giant coal bed.
A great deal of methane gets generated during coal formation, and its content in coal seams is generally higher in deeper and older layers: more pressurized deposits hold more gas, with typical content of just 0.02 m3/t of coal at depth of 100 m, 1 m3/t at 500 m, and about 3.7 m3/t at 1,000 m (IEA [International Energy Agency], 2005). But the gas is very tightly bound within the fuel’s complex structure that has virtually no permeability. But the solid fuel has many cleats (natural opening-mode fractures) that account for most of coal’s permeability and most of the porosity of coal-bed gas reservoirs (Laubach et al., 1998). Because of its large internal surface area, coal stores six to seven times more gas per unit volume than gas-bearing rocks in conventional hydrocarbon reservoirs—but even the latest and relatively most effective drilling and extraction methods are able to recover only a small fraction of the resource in place.
Methane associated with coal makes the news when an accidental explosion of the gas in improperly ventilated underground mines causes deaths of miners. To use a proper technical term, the gas responsible for these explosions is coal mine methane (CMM) that has been released from coal seams or from collapses of surrounding rock following underground fuel extraction. Proper treatment will dilute the gas below the explosive range (preferably to <1% by volume) and it is removed by large ventilation systems. The gas can be also recovered in order to be converted to CO2 by oxidation or to be used in lean-burn gas turbines, and more than 200 of these processes in 14 countries help to avoid annually nearly 4 Gm3 of CH4 emissions or prevent wasteful flaring (WCA [World Coal Association], 2014).
In contrast, CBM is the gas recovered from unmined coal seams, no matter if they will or will not be eventually extracted. Gas recovery begins with removing coal seam water in order to reduce pressure and release the adsorbed gas; afterward, vertical and horizontal wells drilled through unmined seams are used to recover the gas, and hydraulic fracturing may be used in some seams to release more of the gas from coal seams. Production rate of CBM wells thus has three distinct phases: the first one dominated by water, the second one by rising gas flow and declining water flow, and the third one with both gas and water flows in decline (Garbutt, 2004). The gas has usually a high concentration of CH4 (>93%), and hence it could be used directly as a replacement for conventional natural gas and transported in existing pipelines.
Large-scale commercial extraction of CBM began in the United States during the 1980s; in 2012, it contributed 5% of gross natural gas withdrawals; and its volume was equal to about 15% of shale gas extraction in that year (USEIA, 2014k). In absolute terms, that was nearly 25% below the peak reached in 2008, a consequence of rising extraction of shale gas. Similarly, large CBM resources in Russia remain undeveloped because of the abundance of Siberian natural gas. Obviously, all nations with extensive coal deposits—the United States, Canada, Australia, Russia, China, India, and South Africa—have significant resources of CBM, but so far, only Australia began their large-scale recovery: Queensland’s high-quality bituminous coal seams now supply about a third of all gas in the eastern part of the country, adding to its huge resources of conventional offshore gas (Geoscience Australia, 2012).
An older US estimate indicated about 20 Tm3 of coal-bed gas in place, with about 3 Tm3 economically recoverable (Rice, 1997), but Kuuskraa (2004) concluded that technically recoverable reserves (with 2002 capabilities) are only 820 Gm3. IEA (2005) put the global CBM resources at 143 Tm3. Kuuskraa (2004) estimated global resources in place at 85–222 Tm3 (IEA, 2005). With a typical recovery of about 20% of all gas in place, this means recoverable volume of 17–44 Tm3, compared to 185 Tm3 of proved conventional natural gas reserves in 2012 (BP [British Petroleum], 2014a). The largest extraction potential is in the United States, Canada, Russia, and China, the greatest expansion of CBM recovery expected in China where clean fuel alternatives are limited. There is also considerable potential for recovering methane from abandoned coal mines in old mining regions in Europe (the United Kingdom, Germany, Czech Republic), in the United States, and in China. Existing old ventilating shaft and new wells are used to extract the gas.
6.2.1 Tight Gas
Tight gas is methane stored in rocks—mostly in sandstones but also in siltstones, limestones, dolomites, and chalk—with very low in situ matrix permeability, almost always less than 0.1 mD; recall that most conventional natural gas reservoirs have permeability between 0.01 and 0.5 D, that is 10-500 mD. Tightness of these formations is a consequence of their age: while conventional natural gas comes mostly from younger Tertiary basins, tight gas originates mostly from Paleozoic formations with very low permeability due to prolonged compaction, cementing, and recrystallization. China’s Sulige field discovered in the Ordos Basin in 1999 is an excellent example of challenges facing the harnessing of natural gas in tight formations (Total, 2007; Yang et al., 2008; CNPC [China National Petroleum Corporation], 2014; Figure 6.4). Its permeability is between 0.02 and 2 mD; gas-bearing layers created by an extensive braided river system 250 million years ago are just a few meters thick and are 3.2–3.5 km below the surface. Yet Sulige is the world’s largest sandstone trap whose reserves may be as much as 2 Tm3, and its expensive extraction, began in 2006, should reach 5 Gm3/year.
Figure 6.4 Sulige field in China.
© Corbis.
Because the recovery depends on penetrating as much formation volume as possible, tight gas production depends on drilling the largest (economically justifiable) number of directional and horizontal drilling, preferably from a central drill pad. Additional production stimulation is achieved by hydraulic fracturing as well as by introducing acids to dissolve basic formation substances (limestone, dolomite, calcite) and hence to enhance permeability of tight sediments. Removing water from deeper wells also increases their productivity.
Global resources of gas in tight formations may be as large as 510 Tm3, but existing recovery techniques yield only 6–10% of gas in place, reducing the technically recoverable reserves to no more than 50 Tm3. Estimates of the US tight gas in place have been as high as 26 Tm3 in 1980, Kuuskraa (2004) put technically recoverable total at more than 8 Tm3, and USEIA estimates the total at about 8.8 Tm3, considerably larger than even the highest appraisals of CBM and almost a fifth of the country’s total gas reserves. Only a few companies outside the United States have extensive experience with producing gas from t
ight formations; notable commercial developments have been in Algeria (Timimoun), Venezuela (Yucal-Placer), Argentina (Aguada Pichana), and, as already noted, China (Sulige).
American producers have a fairly long history of extracting tight gas; they began to do so during the 1960s in the San Juan Basin in New Mexico where they had pioneered the relevant recovery techniques. US tight formations are widely distributed—in the Appalachian Basin, Denver Basin (mainly in Colorado and Wyoming), West Gulf Coast Basin (in Texas, Louisiana, and Mississippi), Permian Basin (in West Texas), Uinta Basin (in Utah), and Greater Green and Wind River Basins of Wyoming (USEIA, 2010)—and the United States is the only country with relatively large-scale natural gas production from tight sands, recently about 170 Gm3/year or about 25% of all gas produced.
6.3 METHANE HYDRATES
These remarkable energy stores (also known as clathrates or clathrate hydrates) were created as the gas from methanogenic archaea feeding on organic sediments escaped and became trapped inside lattice cages formed by molecules of frozen water (Kvenvolden, 1993; Makogon, Holditch, and Makogon, 2007; Figure 6.5). Fully saturated gas hydrates (at 2.6 MPa) contain one molecule of CH4 for every 5.75 molecules of H2O, and this means than one m3 of methane hydrate holds up to 164 m3 of methane. Because 80% of hydrate volume is taken up by water and 20% by methane, one cubic meter of hydrate contains 164 m3 of gas. One m3 of water can tie up to 207 m3 CH4 as it forms 1.26 m3 of solid hydrate; without gas, freezing 1 m3 of water expands its volume to just 1.09 m3 of ice.
Figure 6.5 Methane hydrate cage.
Specific density of hydrates (depending on composition, pressure, and temperature) ranges from 0.8 to 1.2 g/cm3, and the gas composition is dominated by methane, ranging from as slow as 66% to as high as 99.7% CH4 (Taylor, 2002). Hydrates can be stable under an enormous range of pressures and temperatures (from just 20 nPa to 2 GPa, which is 17 orders of magnitude, and from 70 to 350 K). Hydrates form when water and gas are subjected to low temperature and elevated pressure, an all too common occurrence in natural gas pipelines where hydrate plugs can damage equipment and where precautions must be taken to limit their formation. The solid plugs were first ascribed to the freezing of condensed water, but Hammerschmidt (1934) proved them to be a hydrate of the transported gas. Interest in hydrates as a potential source of commercial energy began in 1963 with the drilling of the Markhinskaya well drilled in Yakutiya (Central Siberia): the presence of a hydrate layer led Makogon (1965) to conclude that gas hydrate accumulations might be found wherever conditions are right to create cooled layers.
In 1965 and 1966, Makogon’s laboratory experiments demonstrated the formation hydrates, and by 1970, the presence of hydrates was confirmed in the West Siberian Messoyakha field. The field began to produce natural gas from a reservoir situated beneath the gas hydrate layer, but its output was soon overshadowed by such supergiant West Siberian fields as Urengoy and Medvezhye. But some geologists concluded that the depressurization of Messoyakha reservoir also led to depressurization and gas dissociation from the overlying hydrate formation, making it the only instance where conventional production methods would yield gas from hydrates. But a reexamination of accumulated evidence by Collett and Ginsburg (1998) suggested that gas hydrates may not have contributed to gas production in the Messoyakha field.
Distribution of hydrates is obviously restricted by temperature of surrounding sediments. The two natural environments favoring the formation of hydrates are sediments in the Arctic (only about 3% of the total resource base) and sediments beneath the ocean floor. Some hydrates appear as small nodules of just 5 cm in diameter, others form continuous pure layers several meters thick. More than 200 gas hydrate deposits have been identified, mostly in waters off the coasts of Americas and East Asia and in the western Pacific (Figure 6.6). Some hydrate deposits in the North American Arctic are only about 100 m below the surface. Among the studied ocean hydrate deposits those off California are just 600 m below the sea bottom and underneath 600 m of water, while those in the deepest parts of Japan’s Nankai Trough are in water depths of around 4,700 m and 4,800 m below the sea bottom, and off Guatemala both depths are still another 1,000 m greater (Makogon, Holditch, and Makogon, 2007).
Figure 6.6 Methane hydrate global deposits.
Not surprisingly, estimates of the total volume of methane held in hydrates are even more uncertain than the shifting assessments of shale gas or coal-bed gas. But these differences are only of secondary importance because even very conservative estimates indicate the enormity of resources in place. Dillon et al. (1992) put the mass of organic carbon in methane hydrates at 10 Tt or roughly twice as much as carbon stored in all conventional fossil fuels. Lowrie and Max (1999) estimated that the volume of gas beneath the seabed of the US coastal waters may be as much as 1,000 times the volume of all US conventional gas reserves.
The range for Canadian resources is between 43.6 and 809 Tm3 (Center for Energy, 2014). Makogon, in many publications, uses a grand global total of 15 Pm3 (which is roughly 15 Pt) of potential reserves, and Kvenvolden (1988) went as high as 20 Pm3, which is almost exactly 100 times the total volume of conventional global gas reserves in 2013. And a recent assessment of methane reservoirs beneath Antarctica showed that some 21 Tt of organic carbon is buried in deep organic sediment underlying the continent and that the sub-Antarctic hydrate inventory of 131–728 Tm3 could be of the same order of magnitude as the latest estimates for the Arctic permafrost (Wadham et al., 2012).
Several options might lead to commercial recovery of hydrates: simply exposing their formations to lower pressure created by well; removing water and other gases from wells drilled into hydrate formations; releasing the gas by injecting steam or hot water in order to disassociate the gas hydrate; and injecting CO2 into hydrate reservoir in order to displace a substantial share of methane. The first test of decompression and heating was done in Canada in 2001. Gas hydrates have been encountered for decades in many wells drilled in Canada's Mackenzie Delta, and one of these sites was chosen by an international (Canada, Japan, the United States, Germany, India) project at the Mallik Bay Gas Hydrate Research Well conducted in late 2001 and 2002 (Dallimore et al., 2002).
Three wells were drilled (one for production, two flanking wells for monitoring), and pressure reduction was tested in six zones as a 17 m layer with high gas hydrate saturation was heated by hot water for several days. This showed that commercial production from land-based hydrates might be eventually feasible, and during winter of 2008 a 6-day test (139 hours) at the Mallik site was able to maintain flows of 2,000–4,000 m3/day with cumulative gas production of about 13,000 m3 (Yamamoto and Dallimore, 2008).
Shale gas development in North America had soon displaced all other gas-related news, but hydrates were back in March 2013 with news from Japan where hydrate research program began in 1995. Its first important finding was the discovery of extensive methane hydrate deposits with high CH4 concentrations (gas filling 60–90% of the pore space) in sand reservoirs off the coast of Japan. This was a welcome departure from a common occurrence of hydrates in unconsolidated muds where they fill only 10% of the pore space (Boswell, 2009).
The Japanese test was conducted by Japan Oil, Gas and Metals National Corporation, and research ship was used to drill a 270 m deep well to reach a 60 m thick hydrate reservoir 1 km below the ocean bottom and about 80 km off Atsumi Peninsula (the east coast of Japan) at Daini Atsumi Knoll in the eastern Nankai Trough (JOGMEC, Japan Oil, Gas and Metals Corporation, 2013). Pump was used to reduce the formation’s pressure from 13.5 to 4.5 MPa, letting the gas rise and flow to a platform on the ship where it was flared starting on March 12, 2013. Although the designers tried to prevent the intake of sand by installing two sifting devices, the pump clogged on the sixth day of the trial, but up to that point the gas flow averaged 20,000 m3/day, an order of magnitude more than obtained by the Canadian depressurization test, and the total yield was about 120,
000 m3.
This was an encouraging experiment as significant volumes of gas were produced for nearly a week, but no commercial breakthroughs are imminent. Not surprisingly, North American interest (the US hydrate research program began in 1982) was diverted by shale gas development, but besides Japan, there are relatively small research programs in South Korea (since 1999), China (since 2004), and India (since 1996). Advances of these efforts and other hydrate news can be followed in Fire in the Ice, a periodical publication by the US National Energy Technology Laboratory (NETL, 2014).
7
Natural Gas in Energy Transitions
Development of modern economies has been characterized by a number of universal transitions. They included changes of sectoral contributions (as the dominance of wealth creation shifted from agriculture to industrial production and then to services), labor participation (the end of child labor, entry of women into labor force, postponed age of the first employment due to longer periods of education), and capital intensity (from artisanal manufacturing requiring often no or minimal capital inputs to very expensive megaprojects undertaken by national governments). The most fundamental shift has been the evolution of modern energy supply as the urbanizing and industrializing societies had first moved away from traditional biofuels (wood, charcoal, in rural areas also crop residues) to coal and then to refined liquid fuels and natural gas while concurrently generating more primary (hydro, nuclear, and now also wind and solar) electricity.
Wood, and to a lesser degree crop residues (mainly staple grain straws and stalks), dominated millennia of premodern and the three centuries of early modern (1500–1800) energy supply. Its replacement by coal began first (already before 1600) in England, but in the continental Europe and in the United States, it took place, fairly rapidly, only during the latter half of the nineteenth century. When leaving traditional biofuels aside, coal delivered about 96% of global total primary energy supply (TPES) by 1900, with crude oil at about 3% and natural gas (overwhelmingly due to the US consumption) a mere 1% (Smil, 2010a). Subsequent transition from coal to oil was slowed down by the two world wars and by the global economic crisis of the 1930s, but by 1950, coal’s share of global TPES declined to just below two-thirds, while oil rose to roughly 30% and natural gas was still the distant third at 10%. The next two decades saw the emergence of new, postwar economies based on inexpensive crude oil: the United States remained its largest producer, but the output in Saudi Arabia and the USSR was catching up fast.