Fuels for the future

Abstract

There are unconventional fuels that may serve as near term major replacements for conventional mineral oil and natural gas. These include fuels from oil shale and bitumen, liquid fuels from coal, methane from methane hydrates, biofuels and the secondary fuel hydrogen. Here, these fuels will be reviewed as to their presumable stocks and life cycle wastes, emissions and inputs of natural resources. The unconventional fuels are usually characterized by a relatively poor source-to-burner energy efficiency when compared with current conventional mineral oil and gas. Apart from some varieties of hydrogen and biofuel, their life cycles are characterized by relatively large water inputs, emissions, and wastes. The unconventional fuels shale oil, bituminous oil, coal liquids, and methane from methane hydrates are based on natural resources which are practically finite. This does not hold for biofuels, but the sustainable supply thereof is severely limited when large numbers of people also have to be fed. In view of the problems associated with unconventional fuels, there is a case to consider fuel-less replacements for many fuel applications.

Keywords:

• unconventional fuels
• synfuels
• coal
• bitumen
• oil shale
• methane hydrates
• biofuels
• hydrogen

1. Introduction

Current energy inputs by humankind into the world economy are mainly fuels. Fossil fuels, nuclear fuels, and biofuels supply ∼97% of the energy input, whereas power sources generating usable energy without the intermediate use of fuels such as hydroelectric systems, solar collectors and cells, and windmills supply about 3% (Quadrelli and Peterson 2007). Fossil fuels cover about 80% of worldwide energy demand, with conventional mineral oil (including heavy acid oil) and natural gas (including coal bed methane) together supplying ∼60% of that demand (Salameh 2003, Szklo et al. 2006, Lysen and van Egmond 2008, Shafiee and Topal 2008).

In the near future, there will probably be a large shift in the position of the major fuels powering the world economy. Actual consumption of conventional natural gas and mineral oil exceeds the rate of their geological formation during the last 400 million years by a factor of probably at least 10⁶ (Patzek and Pimentel 2005). Major price rises over the 2004–2008 period suggest that supply has difficulty in keeping up with expanding demand for conventional mineral oil and natural gas. It is moreover expected that production of these fossil fuels will start to fall in the not too distant future, which will necessitate replacement (GAO 2007, Heiman and Solomon 2007, Bentley et al. 2008, Kaufmann and Shiers 2008). There is concern about the geopolitical implications of most of the remaining conventional oil and gas reserves being in a limited number of countries (Heiman and Solomon 2007). Also, the consumption of fossil fuels in general is under pressure because of its share in climate change, with coal most objected to because, in supplying energy, coal is (ceteris paribus) associated with much higher CO₂ emissions than conventional mineral oil and natural gas (Quadrelli and Peterson 2007; Hansen et al. 2008).

Against this background, there are increasing calls for new fuels to replace conventional fossil fuels in coming decades. It has been argued that the calls for new fuels have added urgency as there may be rapidly growing shortfalls in the supply of conventional natural gas and mineral oil, necessitating a rapid and massive phase-in of new energy sources (Salameh 2003, Hirsch et al. 2005, Bockris 2007). Also, a rapid phase-out of conventional fossil fuels as currently used (without CO₂ sequestration), has been advocated in view of the expected impacts of climate change (Glasby 2006, Hansen et al. 2008).

Partly in view of such alleged urgency, fuels which are far from commercial application or so far elusive, such as fuels for nuclear fusion or abiogenic mineral oil and gas, will not be included in this review. Also only new fuels will be considered that may be major fuels in the future. Unconventional fuels that are generally thought to only allow for a modest contribution to future worldwide fuel supply, such as shale gas (Nederlof 1988), will not be discussed.

The new fuels, which are currently considered for near term application and will be discussed here, include unconventional fossil fuels. These are: shale oil, oil from bitumen, or ‘tar sands’, and coal-derived liquids. Other fuels to be discussed here, are methane hydrates, ‘modern biofuels’ (such as ethanol from sugar, starch and lignocellulose and oil from terrestrial oil crops and algae), and H₂ (hydrogen). Together these will be called ‘unconventional fuels’ in the following. It should be noted that designations such as ‘unconventional (fossil) fuels’ and ‘modern biofuels’ are shaped by perceptions in the late 20th and early 21st century. In fact, making oil from oil shale started in Europe during the 17th century and use of such oil, and of liquefied coal, was substantial in the 19th century (Butt 1965). Ethanol from sugar and vegetable oil were used to power early motor cars (Knothe 2001, Dimitri and Effland 2007). Shale gas, currently considered to be unconventional, was actually in 1821 the first commercially produced gas in the United States (Nederlof 1988). And substantial amounts of liquid fuels from coal were produced in Germany during World War II (Vallentin 2008a).

In the following, it will firstly be considered what the perspectives are for unconventional fuels in the coming decades: are they already in commercial use? Can they help in reducing geopolitical problems that beset the current exploitation of conventional mineral oil and natural gas? What are their presumable stocks? and What is their expected life cycle energy efficiency (energy payback) and environmental performance? As to the life cycle, in analogy to terms current regarding mineral oil, a distinction will be made between the ‘upstream’ and ‘downstream’ part thereof. ‘Downstream’ refers to activities such as oil refining and the distribution of refined products. ‘Upstream’ refers to the production of crude oil.

Secondly, in view of the weak points in the characteristics of unconventional fuels, the question will be discussed: Why fuels at all? And finally the matter will be raised: Which fuels can to what extent be used sustainably, that is: without running out of natural resources, decreasing ecosystem services or increasing waste and emission-related problems for future generations? The latter question is important for the long term viability of energy sources, but it may also be an important criterion in selecting unconventional energy supply options for further development.

The questions will be considered from a long range natural science perspective and not in a detailed way. Social and economic aspects of unconventional fuels will not be reviewed here, though the matter of cost will be raised occasionally.

2 Unconventional fossil fuels (also called: synfuels)

Shale oil, coal derived liquids, and oil from bitumen have been selected by the US Task Force on Strategic Unconventional Fuels (2007) as unconventional fuels which will be important for the replacement of conventional mineral oil. Worldwide, oils from coal and bitumen and oil shale are currently produced commercially on a limited scale (Fischer 2005, van Dyk et al. 2006, Patel 2007). The South African company Sasol (operating in Sasolburg and Secunda) is currently the main producer of liquid fuels from coal, using gasification followed by Fischer-Tropsch synthesis (Hoogendoorn 1981; de Klerk 2008). The commercial production of synthetic crude oil from bitumen is practiced in Canada and Venezuela (Patel 2007, Attanasi 2008). Shale oil is currently used on a limited scale in countries such as Estonia, Brazil and China (Fischer 2005, Jiang et al. 2007b, Raudsep 2008). Using coal, oil shale and bitumen to produce liquid fuels for fuelling the economy, expands the natural resources for the production of liquid fuels far beyond conventional mineral oil, as the presumable cumulative resource base is equivalent to several times current conventional mineral oil reserves (Patel 2007, Brecha 2008). On the other hand, their way of production implies that the input of energy to produce a certain amount of synthetic crude oil is higher than in the case of current conventional oil (Patel 2007, Task Force on Strategic Unconventional Fuels 2007, Attanasi 2008, Gagnon 2008, Parks 2008). For instance, for making crude oil from coal or oil shale, about 50% of the energy present in coal or oil shale is needed (Steynberg and Nel 2004, van Dyk et al. 2006, Jiang et al. 2007a). Thus, the ratio between energy output and source-to-fuel energy input (energy payback) or source-to-fuel energy efficiency for synfuels is lower than for current conventional oil (see also Table 1).

Table 1. Ranking the energy payback for crude oils (sources: Steynberg and Nel 2004, van Dyk et al. 2006, Jiang et al. 2007a, Task Force on Strategic Unconventional Fuels 2007).

	Energy payback of crude oils
	←Relatively high		Relatively low→
Type of crude oil	Current conventional	Oil from bitumen	Coal liquids and shale oil

A corollary of the relatively low life cycle energy efficiency is that, more than in the case of conventional fossil fuels, major suppliers of energy should be in place to get to a stage comparable to crude oil. In practice, this can be a limitation to the exploitation of unconventional resources (Cohen 2007). Another corollary of relatively high upstream energy inputs is that the life cycle CO₂ emission of synthetic crude oil will (ceteris paribus) be larger than for liquid fuels based on current conventional mineral oil (Cohen 2007, Patel 2007, Jaramillo and Griffin 2008, Parks 2008, Tenenbaum 2009). The life cycle CO₂ emission of synfuels is also enhanced because the H/C ratio in the natural resources used is smaller than in crude oil. It is ∼1.8 in crude oil, ∼1.5–1.7 in raw shale oil and bitumen and ∼0.8 in bituminous coal (David 1981, Dung 1995). Thus, for instance life cycle emissions of oil products from oil shale are expected to have a life cycle CO₂ emission which is 50%–500% larger than in the case of conventional mineral oil (Sundquist and Miller 1980). The life cycle emission of CO₂ from coal-based diesel is about twice the well-to-wheel emission from conventional diesel (Vallentin 2008b). Also, the water input, problematic solid wastes and the emissions of non-CO₂ pollutants tend to be much increased in the production of synfuels, when compared with conventional fossil fuels (Metz 1974, Walker 1974, Purde and Rahu 1979, Mossop 1980, Kuljukka et al. 1996, Kok 2002, Hirsch et al. 2005, Puura and Puura 2007, Brecha 2008, Tenenbaum 2009).

Proposals have been made to reduce upstream CO₂ emissions by combining the production of synfuels with the sequestration of CO₂ in aquifers or abandoned gas and oil fields (Steynberg and Nel 2004, Kintisch 2008). Such sequestration however also leads to a further lowering of the life cycle energy efficiency (Rostrup-Nielsen 2004) and may lead to an increase in emissions of toxics and substances that may contribute to acidification, eutrophication, and photochemical smog (Koornneef et al. 2008).

3 Methane from methane hydrates

High pressure and low temperatures are conducive to the formation of methane hydrates (also called ‘methane clathrates’ or ‘methane ice’). These are icy solids with hydrogen bonded water molecules sequestering methane molecules (Buffett 2000). Thus, methane hydrates occur in the soils of high latitude continental regions and in sea sediments, the latter containing the largest stock (Rogner 1997, Buffett 2000, Fyke and Weaver 2006). The origins of the methane hydrates in seafloor sediments are under debate. Biogenic sources are likely and probably have at least a large share in methane hydrate formation (Fyke and Weaver 2006, Wagner et al. 2007, Colwell et al. 2008), but it may be that abiotic (thermogenic) processes have a substantial share in their occurrence too (Sleep et al. 2004). There is a substantial biogenic addition to the stock of methane hydrates (Colwell et al. 2008) and there are also substantial losses (Buffett 2000). It is not clear how additions and losses are balanced. Methane hydrates may become less stable when local temperature increases significantly and this fuels concerns about the stability of methane hydrates in view of expected climate change (Fyke and Weaver 2006).

Estimates of the amount of methane in hydrates are in the order of 10¹⁹ g C (Buffett 2000). Recoverable reserves are much lower. A main reason for this is that most hydrates occur in poorly consolidated formations of mud, silt and sand. In this case, permeability near boreholes is likely to be a major problem, allowing only for low methane flow rates, precluding commercial exploitation (Fischer 2008). Deposits of methane hydrates in sandstone however offer some hope of commercial production (Fischer 2008).

Methane hydrates increasingly attract attention as future energy resource (Bohannon 2008). It has been suggested to extract methane from methane hydrates by depressurization, thermal decomposition (including microwave irradiation) or inhibitor injection. In the latter case, substances are used which shift the hydrate equilibrium curve to higher pressures and lower temperatures, allowing for the release of methane at ambient temperatures and pressures. There has been limited exploitation of onshore methane hydrates in Russia (Messoyakha field) using inhibitor (methanol) injection (Englezos and Lee 2005).

A number of projects that seek to further unlock methane hydrates resources using the previously mentioned approaches are underway (Collett 2001, Servio et al. 2003, Taylor and Kwan 2004, Englezos and Lee 2005, Nazridoust and Ahmadi 2007, Rath and Marder 2007, Liang et al. 2008). However, whether methane hydrates will indeed be a major source of future fuels is still highly uncertain. As yet no feasible commercial exploitation strategies have been demonstrated and in case of thermal decomposition it is likely that energy input will be high, possibly exceeding energy in the methane output (Rogner 1997, Servio et al. 2003). The same probably holds for the use of microwave irradiation (Liang et al. 2008) when the fuel necessary for the generation of input electricity is taken into account. It is likely that energy inputs necessary to extract a specified amount of methane will be higher than in the case of current conventional natural gas. Moreover, exploitation of methane hydrates may have large environmental effects. Especially major inadvertent methane emissions linked to gas blowouts, gas leakage outside conductor casing and casing collapse are a matter of concern (Kvenvolden 1999; Chatti et al. 2005). All in all, it would seem likely that the life cycle environmental burden of methane from methane hydrates may be larger than in the case of conventional natural gas.

4 Should we aim at unconventional fossil fuels and methane hydrates?

Using conventional fossil fuels and methane hydrates has a major impact on fuel reserves, as indicated in Table 2. This table gives order of magnitude estimates for a variety of fuel reserves in terms of years of current input of primary energy into the world economy.

Table 2. Order of magnitude estimates of presumable fuel reserves in terms of years of current primary energy input into the world economy (∼500 EJ) (sources: Nederlof 1988, Smil 1994, Chatti et al. 2005, David 2005, NEA/IAEA 2006, BP 2008, Fischer 2008).

Type of fuel	Order of magnitude in terms of current input of primary energy into the world economy (∼500 EJ)
Conventional mineral oil	10^1
Conventional natural gas	10^1
Fuel from bitumen	10^1
Shale oil and gas	∼10^1
Coal	∼10^2
Methane hydrates	<10^3
Uranium from ore in light water reactors	10^1
Uranium and thorium from ore also in breeder reactors	10^2

Current geopolitical tensions about fossil fuels mainly originate from the concentration of much of the conventional oil and gas reserves in a limited number of countries. Using coal, shale oil, bitumen, and methane hydrates as a basis for liquid and gaseous fuel production will change this situation (Fyke and Weaver 2006, Patel 2007, Brecha 2008). Substantial amounts of methane hydrates may be found near the shores of countries that border deep seas and much of the methane hydrates is at the bottom of the oceans (Fyke and Weaver 2006), a global commons. Coal reserves are much more widespread than conventional oil and gas reserves (BP 2008). Relatively large reserves of oil shales are in a number of countries that have developed into major importers of conventional oil and natural gas, such as China and the United States (Gwyn 2001, Jiang et al. 2007a,b). So, should we take the road of unconventional fossil fuels and methane hydrates? Geopolitically it would seem a good choice. However geopolitics is not the only matter to be considered.

Firstly there is the matter of production cost. For instance, the production cost of synthetic crude oils is expected to be substantially higher than the current marginal cost of conventional mineral oil (Horn 2004, Persson et al. 2007). Similarly, the production cost for unconventional methane supply is expected to be on average substantially higher than for conventional natural gas (Nederlof 1988). The relatively large life cycle emission of greenhouse gases also gives much food for thought, as present CO₂ levels already commit to a major increase in temperature (Hansen et al. 2008). Indeed, climate policy in Germany has emerged as a major barrier to the implementation of coal-to-liquids technology (Vallentin 2008a). Increased costs and the increased life cycle environmental burden of unconventional fuels and methane hydrates are good reasons to think very hard about the choices about energy supply to be made.

5 A hydrogen economy

Since 2000 there has been considerable furore about another fuel: hydrogen (H₂). Much has been made of the benefits of hydrogen, with some expectations achieving utopian dimensions (e.g. Rifkin 2003). Hydrogen is not in the category of primary energy or fuels, like coal or oil shale are. Hydrogen has to be produced while consuming primary energy and is thus comparable to secondary energy carriers such as electricity: it is a secondary fuel. As in the case of electricity, primary energy is needed for its generation and the ‘energy content’ of the hydrogen produced is substantially lower than the energy input in its production.

H₂ can be applied in fuel cells (that generate electricity) or in engines. In mobile applications, the former appears to have a better life cycle energy efficiency than the latter (Reijnders and Huijbregts 2009). Switching to hydrogen on a major scale will require a major overhaul of the energy infrastructure, including the supply system, and a major effort to limit fire and explosion risks (Agnolucci 2007, Markert et al. 2007, Melaina 2007, Ng and Lee 2008). Also major adaptations of means of transport and other fuel consumers will be necessary (e.g. Hoyer 2008).

Currently H₂ is mostly generated from fossil fuels though there is also some production based on hydro powered electrolysis (Ferreira-Aparicio et al. 2005, Heiman and Solomon 2007). Steam reforming of methane (natural gas), producing H₂ and CO₂ is most commonly used (Koroneos et al. 2004, Armor 2005). H₂ is separated from the H₂/CO₂ mixture by pressure swing absorption, membrane separation, or cryogenic separation (Ferreira-Aparicio et al. 2005, Barelli et al. 2008). The energy efficiency of the methane to hydrogen conversion by steam reforming is about 64% (Bargigli et al. 2004). Dufour et al. (2009) have suggested that, from a life cycle perspective, autocatalytic methane decomposition may have a better energy efficiency and a lower environmental burden than steam reforming. When H₂ is produced in this way, the upstream energy efficiency would still be poorer than in case of its parent compound; natural gas.

The ‘well-to-tank’ energy efficiency of hydrogen produced from methane by steam reforming is lower than in the case of current conventional natural gas or gasoline (Colella et al. 2005). And storage of H₂ in tanks requires, when compared to conventional fuels, a relatively high input of energy (Neelis et al. 2004). Also the well-to-tank water intensity of H₂ produced by steam reforming of methane may well be higher than in the case of conventional fuels (Webber 2007). On the other hand, the use of H₂ in fuel cells to generate electricity or (electric) traction is relatively energy efficient if compared with the use of conventional fossil fuels (Blok 2005, Osman and Ries 2007). The net effect thereof is that, with present technology, the well-to-wheel global warming potential of hydrogen produced by steam reforming of methane is roughly similar to diesel fuel, though such hydrogen does worse concerning emissions contributing to acidification and photochemical ozone (Ally and Pryor 2007). The production of hydrogen by the use of hydropower leads to much lower life cycle emissions than in the case of steam reformed methane (Koroneos et al. 2004).

H₂ can also be produced in other ways, for instance from biomass (Iwasaki 2003). In case of H₂ production by steam reforming of biomass about 2.4 J is needed to generate 1 J of H₂ (Koroneos et al. 2008). When cultivated biomass is used for this purpose life cycle emissions are unlikely to be better than in case of steam reformed methane (Huijbregts and Reijnders 2009). H₂ production by algae is with present technology unlikely to have a positive energy balance (Burgess and Velasco 2007).

Furthermore, it is possible to produce H₂ from water by wind electricity or solar power, (Armor 2005). H₂ produced by solar thermal energy or wind energy has lower life cycle emissions than steam reformed methane (Koroneos et al. 2004). When H₂ is produced from water by electricity generated by state of the art photovoltaic modules life cycle emissions will probably be lower than in the case of production based on conventional fossil fuels (Jungbluth et al. 2005, Fthenakis et al. 2008, Mohr et al. 2009). So, the attractiveness of hydrogen will be to a large extent determined by the primary energy and conversion technology used for its production. These in are the main determinants for natural resources available for H₂ generation (see Table 2) and for the life cycle impacts of hydrogen (Pehnt 2003, Dufour et al. 2009). Thus, for instance, H₂ made from unconventional fossil fuels and methane hydrates will (ceteris paribus) give rise to a larger life cycle environmental burden than H₂ made from conventional mineral oil or natural gas (Pehnt 2003).

6 Biofuels

Biomass can be used as fuel or serve as feedstock for biofuels. Biofuels currently have a somewhat over 10% share in the energy input of mankind in the economy (Lysen and van Egmond 2008). Most of that input is firewood (also called ‘traditional biofuel’) for household applications. 1–2% of the energy input of mankind into the world economy is called modern biomass or modern biofuels (Lysen and van Egmond 2008). Modern biofuels include solid ones (for use in e.g. power stations and industry), liquid ones (for use e.g. in transport) and gaseous conversion products derived from biomass (Lysen and van Egmond 2008). It has been argued that modern biofuels may be fit for a large expansion in the 21st century, with potential worldwide supply exceeding expected worldwide demand for fuels (Hoogwijk et al. 2003; Moreira 2006, De Vries et al. 2007).

A number of countries have developed programs for the expansion of modern biofuel supply, especially for transport. They have done so for a variety of reasons, including agricultural policy, improving energy security and improvement of environmental performance especially regarding air quality and greenhouse gas emissions (Di Luca and Nilsson 2007; Tyner 2007). This development has contributed to increasing food prices, as biofuels compete with food and feed for good quality arable soils (Johansson and Azar 2007, Naylor et al. 2007) and agricultural inputs such as phosphate ore (Cordell et al. 2009). Under market conditions, further expansion of biofuel production is expected to link prices of crops which serve as biofuel feedstocks to fossil fuel prices, when corrected for the energy content of biofuels (Johansson and Azar 2007, Naylor et al. 2007).

Because biofuels are discussed here as displacement for conventional fossil fuels, energetically speaking cumulative fossil fuel inputs in biofuel production should be substantially lower than biofuel output. There are cases that cumulative fossil fuel inputs at least equal the energy content of biofuels. Such cases include: current algal oil production in bioreactors (Wijffels 2008) and presumably also in open ponds (Reijnders 2008a), CH₄ generation from swine and cattle manure in the Netherlands (Zwart et al. 2006), and the incineration of wastewater treatment sludges for the production of heat or electricity (Wang et al. 2008).

However, there are also cases where energetic output much exceeds fossil fuel input. Examples are palm oil and ethanol from sugarcane (Reijnders 2008a).

A major characteristic of biofuels is the limited solar energy to biomass conversion efficiency. The best performing commercial terrestrial plant (sugarcane) has in practice an efficiency of ∼0.9% in converting solar energy into biomass. For modern liquid transport biofuels the conversion efficiency, after correction for fossil fuel inputs, tends to be below 0.2% (Reijnders and Huijbregts 2009). A large part of the biomass feedstock is consumed or turned into waste in generating a useful transport fuel, giving rise to a low life cycle energy efficiency for transport biofuels, when compared with conventional fossil transport fuels (Reijnders and Huijbregts 2009). However, there are also biofuels which have life cycle energy efficiencies that are similar to fossil fuels, such as wood pellets for use in power stations (Reijnders and Huijbregts 2003).

Though energy in the overall solar irradiation of the Earth is about 10⁴ times the fossil fuel input into the human economy (Smil 1994; Lewis and Nocera 2006), the low solar energy conversion efficiency which is achieved by current biofuels means that very large areas are required to give modern biofuels a significant share in fuel supply, which currently totals about 500 EJ (Lysen and van Egmond 2008). A study of Gurgel et al. (2007) estimated that 2.5 Gha of land would be needed to supply 368 EJ of biofuel energy. The current worldwide area cropland is about 1.6 Gha and the area that may be fit for expansion of cropping has been estimated at between 0.4 and ∼1.2 Gha, with much of the latter already needed for increased food production in the period up to 2050 (Renewable fuels Agency 2008). Also, there is only limited scope for biofuel use to tackle climate change. Under market conditions, expansion of biofuels production tends to lead directly or indirectly to substantial clearing of natural ecosystems, which have a higher C content than crop ecosystems (Fargione et al. 2008, Searchinger et al. 2008). Also there tend to be substantial inputs of fossil fuels and emissions of the greenhouse gas N₂O associated with biofuel production, whereas many systems producing biofuel feedstocks are prone to loss of soil carbon (Reijnders and Huijbregts 2007, 2008). The combined effect thereof is that many current transport biofuels are not suitable for tackling climate change (Fargione et al. 2008, Gibbs et al. 2008, Searchinger et al. 2008, Reijnders and Huijbregts 2009). And only limited use of crop and processing residues for biofuel production is possible, as otherwise soil carbon stocks will deteriorate which in turn can have a negative impact on soil fertility and erosion (Reijnders 2008b). Regarding other aspects of the environmental burden (such as contribution to acidification, eutrophication excotoxicity and oxidizing smog) biofuels rather often do worse over their life cycle than fossil fuels (Kaltschmitt et al. 1997, Sheehan et al. 2003, Zah et al. 2007, Kim and Dale 2008).

It may be argued that to limit the upward effect of food prices and to lower the emission of greenhouse gases, feedstock production for biofuels should probably be restricted to abandoned lands that currently sequester little C. Sustainable production of biofuels from such land is unlikely to exceed 23–29 EJ (Reijnders and Huijbregts 2009). More in general, the sustainable supply of biofuel would seem limited, when also large numbers of people have to be fed (Renewable Fuels Agency 2008).

7 Why fuels at all?

It has emerged that the unconventional fossil fuels, methane hydrates, and biofuels discussed here have substantial weak points. This raises the question what the options are for a drastic reduction in fuel use per se. The share of fuels in catering for energy demand is not immutable. And technically it may be possible to in the future adequately supply energy by non-fuel energy sources such as solar- wind, hydro and geothermal energy, when there are large facilities for storage of heat and power, grids become ‘smarter’ and ‘super grids’ are installed which allow for energy efficient long distance transport of electricity (Hammons 2001, Anderson and Leach 2004, Marris 2008). Interestingly, to the extent that solar energy is directly converted into electricity or heat the conversion efficiency thereof is much higher than in case of biofuels. For instance, when corrected for fuel inputs, conversion of solar energy by state of the art Si photovoltaic modules for traction is about a factor of 50 times better than for the most efficient current biofuel: ethanol from sugarcane (see Table 3).

Table 3. Net energy yields in giga joules (GJ) per hectare for selected biofuels and photovoltaic modules.

Crop	Location	Product	Net energy yield in GJ ha^−1 year^−1	Source
Sugarcane	Brazil	Ethanol	160–175	Macedo et al. 2008
Starch crops	USA, Europe	Ethanol	35–50	Hill et al. 2006, Reijnders and Huijbregts 2007
Multicrystalline photovoltaic modules	Brazil	Electricity	76×10^2–87×10^2	Fthenakis et al. 2008

In view of the major overhaul necessary for a fuel-less supply of energy, the full implementation thereof is not a near term matter, but near term changes in that direction may be conducive to its long term implementation.

In considering fuel-less options, the impacts thereof should be considered as carefully as in case of fuels. Hydropower provides for a sobering lesson in this respect. It provides relatively low-cost electricity but has also led to major involuntary displacements of people and major environmental impacts (Gutman 1994, Sternberg 2008).

One of the obvious matters to consider in this context is cost. In the 19th and 20th century, there have been predictions that conventional fossil fuels would be out-competed within those centuries by solar energy conversion, for instance by J.A. Etzler (in his 1836 book: Paradise within Reach of All Men) and by H. Wiley (president of the American Chemical Society) in 1902 (Nader 2004). Such predictions have been invalidated by the apparently low costs of fossil fuels and the neglect of their external costs. However, the future may be different in this respect. Currently wind and solar power are on a downward cost curve and it is expected that they will break even with conventional fossil fuels during this century (Chakravorty et al. 1997, Martinot 2006, Braun 2008, Lior 2008). Solar and wind power have probably also substantially lower external costs than unconventional fossil fuels and methane from hydrates. And though a fuel-less supply requires a major and costly overhaul of the current energy infrastructure, it may well be that it will turn out to be less costly than reliance of the unconventional fuels considered here. To the extent that a fuel is more convenient in relying on non-fuel primary energy, e.g. because a fuel can be more cheaply stored than electricity, one may consider the use of hydrogen generated by e.g. solar power (Anderson and Leach 2004, Armor 2005).

8 Sustainability of options for energy supply

Table 4 gives order of magnitude estimates of the future availability of several types of primary energy during the next million years.

Table 4. Order of magnitude estimates of primary energy available during the next million years in years of current primary energy input into the world economy (∼500 EJ) (sources as in Table 2).

Type of primary energy	Availability in the next 10^6 years in years of current input of primary energy into the world economy (∼500 EJ)
Solar irradiation of earth	10^10
Wind	10^7
Fossil + nuclear fuels (using breeder technology)	10^3
Methane hydrates	<10^3

Table 4 shows that energy options widely differ as to future availability. The future availability of natural resources is often one of the elements included in the concept of sustainability. However, sustainability is also a term that currently has many meanings. Here, the term will be used in the original meaning in the modern environmental debate, linked to a steady state economy (Daly 1973; Hueting and Reijnders 1998). Thus sustainable use of fuels (or energy) may be defined as a type of use that can be continued indefinitely, without jeopardizing the supply of natural resources, which does not lead to the accumulation of pollutants which may negatively affect future generations and maintains available ecosystem services. Unconventional fossil fuels do not belong in this category of sustainable fuels, as their underlying natural resources have a finite character and their use leads to the accumulation of pollutants which may negatively affect future mankind. The finite character of unconventional fossil fuels clearly emerges when they are energetically compared with solar irradiation of the earth. The total presumable stocks of fossil fuels are, energetically speaking, smaller than 1 month of solar irradiation, and the latter will continue with at least the same intensity for ∼5 × 10⁹ years (Smil 1994). The stock of methane hydrates may be larger than the stock of fossil fuels (Chatti et al. 2005), but still the stock is finite and its large scale use will lead to the accumulation of carbonaceous gases such as CO₂ (a conversion product of methane) in the atmosphere which will be conducive to climate change affecting future generations. Full removal of emitted CO₂ from the atmosphere has been estimated to take 30,000–35,000 years (Archer 2005).

Biofuels may be used as long as photosynthesis remains operational, which is for a species such as Homo sapiens an indefinite time. However, there are constraints on its supply, following from the demand for food and the requirement to maintain ecosystem services and photosynthetic productivity. For terrestrial biofuels, the latter for instance requires maintaining the stocks of freshwater, nutrients, and soil organic carbon. This in turn strongly limits the amount of terrestrial biofuels which can be supplied in a sustainable way in the case that also large numbers of people have to be fed (Renewable Fuels Agency 2008).

There would seem to be more scope for a sustainable physical conversion into energy for economic use of solar irradiation and derivatives thereof such as wind energy. As pointed out in section 4, the physical conversion of solar irradiation into electricity is relatively efficient, if compared with photosynthesis. Also, such conversion can be partially integrated in the built environment. Both characteristics are beneficial for maintaining ecosystem services. The focus of sustainability within the framework of physical energy conversion technologies should be on use of materials which allows for indefinite supply thereof and the prevention of the accumulation of pollutants. Life cycle assessments of current conversion systems suggests that this may well be within the realm of what is possible, if such systems and the components thereof are designed for recyclability and if there is proper recycling (Ardente et al. 2005, Jungbluth et al. 2005, Pehnt 2006, Fthenakis et al. 2008, Mohr et al. 2009).

Conclusion

There are unconventional fuels that may be major replacements of conventional mineral oil and natural gas. Often they are characterized by a relatively poor life cycle energy efficiency if compared with current conventional mineral oil and gas. They are also often characterized by relatively large life cycle water inputs, emissions and wastes. The unconventional fuels shale oil, bituminous oil, liquids from coal and methane from methane hydrates have underlying natural resources which are practically finite. The sustainable supply of biofuels is limited when also large numbers of people should be fed. In view thereof there is a case to consider fuel-less ways to supply usable energy, such as the generation of solar and wind energy which interestingly show a decreasing cost curve.

Acknowledgement

The comments of two anonymous reviewers are gratefully acknowledged.