+1'd post by Edward Morbius in EnergyAudi's Synfuels: Yes, there's potential, but a lot of hurdles yet I've been looking into similar prospects for synfuels for about a year. There's a lot of attraction to the idea, and after abandoning the possibility of biofuels which are the other viable path to creating liquid hydrocarbons from sunlight, CO₂, and minerals catalysts, this seems to be among the few viable options. First, a short look at oil and transportation, or why this is so crucial. Humans use a lot of oil. 33 billion barrels per year, or 5.3 km³. That is, a rectangular cube of oil 1 km square at the base and reaching 5.3 km (17,400 feet) into the sky. For the US, about 6.9 billion barrels per year. (Source: BP Statistical Review of World Energy, 2014: http://www.bp.com/en/global/corporate/about-bp/energy-economics/statistical-review-of-world-energy.html) There are a number of plant crops which can be tapped for biodiesel and/or ethanol: corn, canola, hemp, and sugarcane are the major alternatives. Plants manage to convert about 1-3% of sunlight to stored energy as cellose, carbohydrate, and lipids. Yields are on the order of 30 - 300 gallons per acre per year (280 - 2,800 liters/hectare), modulo energy density (alcohol has only 85% the energy of diesel, bio or otherwise, per unit volume). Ballparking with 100 gallons/acre is reasonably generous. The US Dept. of Ag tells us there's a bit over 400 million acres under cultivation in the U.S. If all of that produced biofuel at 100 gal/acre, the US would produce just under 1 billion barrels of fuel, less than 14% of present consumption. It would also not be producing any food. Relying only on biomass for fuels would almost certainly require vastly less fuel use per person, vastly fewer people, or both. Cost estimates for biodiesel run about $1,000/barrel, or $24/gallon. Another approach to the biofuel/biomass question looks at total global plant growth, which the eggheads call "primary production", and how much of that is already claimed, directly or indirectly, by humans. I ran across a great 2003 paper by Jeffrey S. Dukes a couple of years back which lays this out, in the context of present petroleum consumption. It's quite sobering (and was in fact what convinced me the biofuels option wasn't viable, at least not at present populations and levels of consumption): "Burning Buried Sunshine: Human Consumption Of Ancient Solar Energy" http://globalecology.stanford.edu/DGE/Dukes/Dukes_ClimChange1.pdf This conundrum has been termed "the photosynthetic ceiling" (by Jared Diamond) or HANNP, human appropriation of net primary production. The other side of this is what we use petroleum for. In a word, transport. Oil is hugely useful stuff: not very poisonous (you can splash a bit on your skin, inhale fumes in modest quantities), not terribly flammable, especially heavier oil weights, has among the very highest energy densities of all chemical fuels by either weight or volume, is liquid at a wide range of commonly encountered temperatures, from well below freezing to hot-desert conditions, isn't corrosive to most metals or other materials (though some plastics weaken in contact). Energy densities are about 100x that of most battery technologies by weight. It beats solids (hard to pump and transport), gases (hard to store, leak and explode easily). Burns with virtually no ash. Can be utilized in powerplants from a fraction of a cc to 25,500 liters (the RTA96-C ship's engine, with 109,000 horsepower output). In a word: handy. Globally, 64% of petroleum is used in transport, with other (11%, mostly residential/commercial heating), industrial (8.25%), and non-energy (15%, chemicals) use making up the remainder. U.S. use patterns are similar to overall global trends: ⚫ Light vehicles (passenger): 59% ⚫ Medium/heavy trucks (cargo): 22% ⚫ Air travel (cargo & passenger): 8% ⚫ Water/shipping: 4% ⚫ Rail: 2% ⚫ Bus: 1% ⚫ Other: 4% Globally air uses slightly less and shipping slightly more of the total, at about 6% each. Another interesting statistic: 30% of all ocean shipping is of oil itself. Much of that use is difficult to substitute for. Several modes would be largely impossible or hugely transformed without oil, most especially passenger and cargo air, and long-distance shipping. It's possible that we could return to sail or hybrid ships, but sizes would fall, schedule variability would increase, and shipping times would rise. Solid fuels, particularly pelletized biomass (wood and other chips) are possible, and appropriately produced function in many ways like liquids. Managing coal bunkers on ships -- physically transporting coal from one bin to another for both trim and fuel management, was among shipboard tasks. Air travel without cheap and abundant hydrocarbons would be exceptionally limited. About 40-50% of the take-off weight of a commercial jet is fuel. While the per-passenger mile efficiencies are high, the distances are vast. For a cross-country plane trip in the U.S., the typical passenger burns nearly 400 lb. of fuel, about 2.7 times their weight. For biofuel at $24/gallon, that's $1,470 in fuel costs alone. Overall passenger transport could probably be reduced, but this would require tremendous changes in land use. Co-locating housing, work, shopping, education, and recreation to not require individual personal motorized transport would be a huge efficiency gain. Even non-renewable energy regimes such as a fully-nuclear economy would fail of itself to provide liquid fuels. In fact the concept of electrically-powered fuel synthesis was first proposed with nuclear power as the energy source. No liquid fuels, far, far less (or more expensive) transport. As Richard Heinberg puts it, transport is the essence of commerce. So that's where we're coming from. What of synfuels? There are a number of attractive elements: ⚫ Effectively we're using hydrogen and carbon as chemical storage batteries, same as we've been doing, but taking care of the charging ourselves. ⚫ The substrates are both abundant and reasonably readily available. Hydrogen can be sourced from water, carbon from various sources, with proposals including the atmosphere, seawater, biomass or waste-stream feedstocks, and minerals such as limestone. ⚫ Drawing from the biosphere (air or oceans) means that the net biosphere CO₂ impacts are neutral, though the balance between, say, air and seawater might shift somewhat. ⚫ While the energy costs are high (more below), the resulting fuel is highly attractive: energy dense, stable, can be stored for years (or hundreds of millions of years) with minimal loss, and utilized in highly variable dispatchable energy systems for transport, heat, and generation. ⚫ The results are direct analogs of existing liquid fuels. Medium-chain liquid hydrocarbons resembling petrol or deisel fuel, with gas or heavier chain liquids also possible. That last point bears emphasis: These aren't "alternative fuels" in the sense that they're novel. They're the same stuff we're using now. They can be burned in the same engines or furnaces, stored in the same tanks, shipped in the same pipelines, dispensed from the same stations, and mixed with conventional fuels in all of the above during any possible transition period. Instead what's changed is the origin of the energy embodied in fuel. Where for many alternatives the transition "from here to there" is a huge challenge, this is an alternative with no switching barrier in this regard. That's attractive. On processing: ⚫ Biomass and wastestream sources are intrinsically limited though they might be of some use. As noted, the photosynthetic ceiling is a thing. Total potential is likely low-single-digit percentages of present fuel consumption. ⚫ In terms of energy requirements, the US Navy has estimated that a 240 MW marine nuclear reactor (of which it has several and some familiarity) could produce roughly 100,000 gallons of aviation-grade liquid hydrocarbon fuels (effectively: kerosene), per day. ⚫ Plant size: The Navy's model assumes a plant about the size of an existing aircraft carrier hull (of which it also has several). ⚫ Civilian equivalent: This is roughly the fuel consumption of a city of 100,000-200,000 people, depending on profligacy. ⚫ Costs are within reason: Navy's estimate is $3-$6/gallon. I've penciled this out for solar PV provisioning and come up with a $12/gallon value. Higher than today's fuel prices, yes, but solar prices should be stable (or falling), and with abundant substrates, the long-term costs should be stable. This would be an adjustment, but one to a stable and predictable cost basis. ⚫ Net efficiency: This is hard to establish, but hydrogen electrolysis is about 60% efficient (that is, you lose about 40% of input energy), ⚫ At national scale, substituting for present liquid fuels consumption would require scaling this out about 8,400 times, to 74,000 billion liters/day of seawater, 2 TW of electrical power, and a total plant volume of 210 million m³. Think of a structure 10m tall and 4.5 km on a side. Or multiple plants totalling this size. Power requirement is the biggie: this would require at typical solar production rates and efficiencies (1 kW/m₂, 17% efficiency, 20% capacity factor, 55% spacing factor, 90% inverter efficiency, 90% transmission efficiency), 132,000 km² of solar collection, a region a tad over 360 km on a side (or 51,000 mi² and 225 mi). That's on top of solar power requirements for existing electricity generation. Total costs are on the order of $14-25 trillion. Total present US GDP is about $14 trillion, so if the infrastructure is good for 40 years, it's about 2.5% of annual GDP per year. That may be optimistic: solar PV panel life is about 20 years.... This is discussed at more length here: https://www.reddit.com/r/dredmorbius/comments/22k71x/us_navy_electricitytofuel_synthesis_papers_and/ History and Viability The question that nags at me heavily is "well, if this is so all-fired awesome, why aren't we doing it yet?" It took me a while to become aware that this was a flag, as there's 1) Very little discussion of the process in most alternative / renewable energy literature. I've found a very few brief mentions: a paragraph in the thousand-plus pages of the IPCC's renewable energy report, and a handful of mentions elsewhere, and 2) Many recent storie and even technical reports and papers fail to document the full history. It's as if a report on evolution failed to credit Darwin. http://srren.ipcc-wg3.de/report It doesn't help that most recent discussion, even papers with references, don't mention the foundational work on the method (the USNRL's papers are particularly at fault for this). It would be like a molecular biologist failing to credit Darwin for evolution. I've compiled a set of studies from 1964 to present here from BNL, M.I.T., and US Navy Research Lab. Much of this was performed by Meyer Steinberg (BNL) and Michael J. Driscoll (M.I.T.). The U.S. Navy, which is hugely fuel dependent for both its ships and aircraft, has been particularly interested. In short: serious research from non-crackpot institutions. https://www.reddit.com/r/dredmorbius/comments/28nqoz/electrical_fuel_synthesis_from_seawater_older/ The variant I'm most interested in is the angle being pursued recently by the US NRL which sources its carbon from seawater. Short version: dissolved CO₂ gas is a small fraction (~2-3%) of the total available carbon, most is in the form of dissolved carbonate and bicarbonate (97-98%, and can be released through both reduced pressure and acidification of the water. Energy costs for this are far lower than CO₂ extraction from the atmosphere. A number of technologies are being pursued to this end. Otherwise, there are two well-established procedures, both of which have operated at industrial scale for decades: hydrogen electrolysis, dating to the 19th century, and Fischer-Tropsch fuel synthesis, dating to the 1920s. The water-handling requirements are pretty immense. I'm not a hydraulic flow engineer, but I suspect this is a considerable challenge. The environmental impacts are also nontrivial: thousands of billons of liters of seawater will contain a tremendous amount of sealife, from plankton to apex predators, and even the most gentle filtration system will tend to be fouled and/or kill many of these. A key enabling technology may well be devising some way to effectively filter water from a very large area at low incremental but high overall flow rates. Seawater desalination faces a similar challenge though at a much smaller scale. For electrical energy storage -- say, as fuel for standby generation using surplus power, this is a fairly rough sell. You'd be better of with a number of other options largely due to the efficiency losses in conversion. Net generating return on storage from initial electrical generation is likely on the order of 17-22%,though the net storage capacity and dispatchability of storage are large. Considered in terms of capture of incident solar energy it's far less still. Assuming solar inputs, we're starting with 1 kW of solar energy per square meter at Earth's surface, nominally. Plants as noted above convert this at about 1-3% efficiency into cellulose and lipids, requiring water, soil, fertilizers, and pest control. While the Green Revolution has resulted in much larger crop yields -- on the order of 3-5x over existing, it's done so by greatly increasing the inputs, including energy inputs, of agriculture. In much of Europe food on the table provides 1 calorie for every 5 calories of input and processing energy. In the U.S. that value is 1 for every 10 calories. So there would seem to be a considerable room for improvement. But the problem turns out to be hard. Solar PV has a maximum theoretical efficiency of about 86%, though that relies on an cell of infinite layers and concentrated sunlight. Single-layer cells have a theoretical maximum efficiency of 33.7%. In practice, achieved efficiencies are closer to 17%. While this can be boosted, this is about 50% of the single-layer max, and greater effieciencies tend to increase costs significantly. In solar, costs rather than efficiency are a key concern. There are two battles underway for solar PV, one for higher efficiency, one for lower costs. We're seeing progress on both fronts, though most sources put the emphasis on cost. My general sense is that once you've hit a minimum efficiency threshold, they're likely right, though I wouldn't mind seeing that baseline rise a bit. Another issue is that as the cost of cells and panels themselves falls, other costs dominate, including mounting systems (platforms, foundations, hardware), electrical conversion equipment, and most especially labour. Another factor is that panels have, as noted above, a practical life of about 20 years. They degrade in a number of ways in which the power output falls to about 80% of nominal. Extending panel lifetime seems to me another possible area for research. http://en.wikipedia.org/wiki/Solar_cell#Efficiency So we start with 17% conversion efficiency. Next, spacing. If you shade a portion of a solar panel circuit, you lose power over the entire circuit. This means that you don't want your panels to be in the shade, including being shaded by other panels. This results in a maximum spacing density which varies by latitude and panel incident angle. For most of the United States, this is about a 55% maximum density factor. http://www.solarabcs.org/about/publications/reports/aoi/pdfs/ABCS-29_AOI-SR-3-4.pdf Capacity factor describes the fraction of rated "nameplate" capacity of a generating system that's achieved over time. For solar power, this is based on incident sunlight, hours of sunlight per day, and occlusion due to clouds or other weather. On average it was 27.8% for 2014 (that's higher than the 20% I've been using). http://www.eia.gov/todayinenergy/detail.cfm?id=14611 http://www.eia.gov/electricity/monthly/epm_table_grapher.cfm?t=epmt_6_07_b http://www.eia.gov/electricity/monthly/epm_table_grapher.cfm?t=epmt_6_07_a Inverters are electrical equipment which converts the direct current (DC) flow PVs produce to the alternating current (AC) the power grid requires. We could skip this with dedicated solar plant, but that's somewhat unlikely. Inverters are about 90% efficient. There's a further 10% loss in long-distance transmission, so another 90% factor. Put that together, and we find we've cut into our available energy considerably: 1 kW/m^2 * 0.17 * 0.55 * 0.278 * 0.9 * 0.9 = 21 watts/m^2 Of this, we can create fuel equivalent to about half of this, so 10 watts/m^2 If we were to use this for generation, the effective efficiency is reduced by Carnot's law, which describes heat engines. Typical thermal efficiencies are about 35-45%. We'll be generous and use the high end. So on a rainy day, we can generate 4.725 watts of energy per meter of source feed on a 24/7 basis. And it turns out we're not doing a whole lot better than plants. Even taking the capacity factor out of the equation, that's 17 watts delivered electricity on storage per 1000 watts incident sunlight, or about 1.7% net efficiency. The advantage is that solar panels don't generally need water, weeding, fertilizer, and pest control. They can be placed on marginal land, or buildings, or over car parks, and the like. But in terms of beating mother nature at her own game, it's a tough call. By contrast, plants provide a lot of what I call "plant services": they build themselves, provide structural integrity and foundations, have rudimentary protections against pests and disease, and provide for transport of water, air, and structural minerals. The infrastructure requirements are modest: tilled, or frequently untilled soil, and irrigation water. Replacing that with exceptionally complex industrial infrastructure on massive scale is a challenge. EROI, prices, and markets Another way to look at this is in terms of EROEI. This section starts wandering into some adjoining conceptiual neighborhoods, so a warning and apology in advance. I think it holds together moderately well.... Conventional fossil fuels deliver a huge amount of energy for not too much work. Or at least they used to. EROEI stands for energy returned on energy invested, and was coined by ecologist Charles A.S. Hall. His first application was for animals in nature, but he realized it could also be applied to human energy systems. Values for a number of sources have been computed. For "easy" coal it was as high as 200:1, for easy oil, 100:1. More recently both are closer to 40:1 as a global average, with recent oil and gas ranging from 20:1 to 10:1. Again, the first number tells you how much you receive, the second what it costs you. https://upload.wikimedia.org/wikipedia/commons/f/fe/EROI_-_Ratio_of_Energy_Returned_on_Energy_Invested_-_USA.svg https://en.wikipedia.org/wiki/Energy_returned_on_energy_invested http://www.kcet.org/news/redefine/rewire/explainers/explainer-energy-return-on-energy-invested-eroei.html Seawater based Fischer-Tropsch fuel synthesis has an EROEI of 1:2. You get one unit of energy for every two units invested. That's a negative. What you gain is that you're trading energy now for energy later. That's the case with all storage. The effect is that liquid hydrocarbons' cost in energy terms has just increased 80x over present sources. The net cost in financial terms isn't as clear to me, and I've cited the US Navy Research Lab's values above ($3-$6/gallon) which look fairly reasonable. I've got concerns over the assumptions baked into that and my own higher estimate of $12/gallon as both assume a world in which existing fossil-fuel subsidized costs are in effect. Which gets into another area I've been thinking a lot about but haven't yet reached conclusions. It's based also in part on some of Hall's work, this on what's called the Cobb-Douglas production function, and also on the question: what's the correct price for fossil fuels? Briefly, the Cobb-Douglas function explains economic productivity as the result of capital, labour, factor productivity (better seen as technological progress), and factors for labour and capital elasticity (sensitivity to price changes). Hall has shown that virtually all "technological improvement" can actually be explained by additional energy supplied, and that as more energy is supplied, a lower labour input (and higher measured economic productivity of labour) is observed. But the actual "efficiency" gains are largely (though not entirely) due to added energy. And we're under-pricing those energy inputs. I got into this a few months back with David Friedman (son of Milton), on a post of John Baez's. My contention was this: "There's another and more insidious manner in which fossil fuel "prices" don't reflect full value or costs. "By way of an analogy: you're born into a wealthy family who's accumulated financial wealth over decades. It's stored in various bank accounts. You and your many siblings don't actually work, but simple draw off the funds. Your "cost" of those funds is determined to be how much it costs to go to the bank to make a withdrawal. There's no provisioning for replenishing the accounts. "That's how accounting for the cost of fossil fuels works." (I should have said "pricing", more strictly, but close enough...) Freidman responded with an old economic principle, Hotelling's Law, which provides a theoretical basis for accounting for extractive resource costs, and pointed to a paper of his own as a reference. The main problem is that pricing data for fossil fuels -- and we've got an excellent price history for oil dating to 1860 -- don't reflect this at all. Included in BP's annual review cited above. And in fact if you turn to Friedman's own paper which he presented in support of his argument in this discussion, you'll find: Changes in its price over time will be almost entirely determined by changes in production cost. The good is, strictly speaking, depletable, but that fact has no significant effect on its price. The pattern of oil prices over the past ninety years or so suggests that that may well be how the market views petroleum. http://www.daviddfriedman.com/Academic/Price_Theory/PThy_Chapter_12/PThy_Chapter_12.html Hotelling's Law is also soundly dismissed by others, including the generally cornucopian M.A. Adelman. One possibility is that we've been hugely underpricing fossil energy, and that the proper basis should have been to 1) price it at replacement, not access, cost, and 2) allocate it entirely to creating a sustainable or renewable energy infrastructure. There are shades of this concept going back through over 200 years of economic history, including in William Stanley Jevons The Coal Question (1865), earlier works dating to 1789, quite explicitly in US Navy Admiral Hyman G. Rickover's 1957 speech on the future of energy (Rickover is also known as "Father of the Nuclear Navy). It's the idea that underpinned the first suggestions of a nuclear-sourced synthetic fuels program by M. King Hubbert in 1963 (Steinberg's research was published the following year). I've got to admit I'm partial to the idea that oil, coal, and gas should be priced far higher than they are, though I'm not yet convinced. But if it were, it would mean that we should have been pricing energy at far higher rates. Electricity at the net of hydro, wind, solar, and geothermal sourcing. Fuels at the costs associated with biomass or synthesis. And those costs are baked into the prices of everything in the economy as their basis. Increasing transport, heat, and industrial energy costs 80-160-fold would have a massive impact on what's considered to be economically viable. It's also worth noting that throughout the history of petroleum, the market for it has virtually never been free-floating, but that price and or production quantities have been effectively controlled by some entity. There was Standard Oil, the As-Is agreement, the State Extractors (national oil companies), OPEC, the US-Saudi price-setting agreements (used to great effect against the Soviet Union in the 1980s), and commodities traders. Or consider this passage from a den of pinko academics who've blown the cover on the whole party: From 1948 to 1972, the price of oil produced [that is: extracted] in the U.S. was influenced by the production quotas set by the Texas Railroad Commission[1] (RRO). Each month, the RRO (and other state regulatory agencies like it) made forecasts of petroleum demand for the upcoming month and set production quotas to meet the forecasted demand. Source? Keith Sill. He's ... oh. Chief economist for the Federal Reserve Bank of Philadelphia. Page 1, footnote 1 of "Macroeconomics of Oil Shocks" http://www.phil.frb.org/research-and-data/publications/business-review/2007/q1/br_q1-2007-3_oil-shocks.pdf Daniel Yergin's comprehensive history of petroleum, The Prize, describes various monopolies and cartels generally throughout the work, and the RRO in detail in chapter 13. Very strongly recommended. Imagine the governors of Texas, Oklahoma, and, say, North Dakota calling out the national guard, seizing control of production wellheads, establishing production quotas, and working with the U.S. Department of Interior to enshrine these for 40 years. Because that happened in 1930-31. There's some discussion of this at Wikipedia: https://en.wikipedia.org/wiki/Railroad_Commission_of_Texas This also raises a few further questions, key of which are: 1. What's the true meaning of "price", and how should it be established, especially in the case of forseably finite extractives? 2. How do we interpret "the market" and its price discovery process in light of this? The conventional wisdom is that "the market is right", and that market-set prices are inherently correct and efficient. That flies in the face of numerous and widely-known market failures in which full costs (or occasionally benefits) aren't fully priced in, and allocation of goods and services isn't efficient. My view increasingly is that the market is powerful, in that the price-discovery process it entails is exceptionally difficult to fight. But that doesn't make it correct. I've also been intrigued by how freqently Gresham's Law turns up in different corners of economics, and am suspecting that more generally, bad price-discovery mechanisms drive out good. Especially in the classic cases of market failure: time-inconsistent preferences, information asymmetries, non-competitive markets, principal–agent problems, externalities, or public goods. https://en.wikipedia.org/wiki/Market_failure 3. What's the relationship between price, cost, and value, particularly as considered with analogs of metabolism (within organisms) or total energy throughput (within populations), and from there to money. The implication is that the very concept of price and cost basis has been tremendously distorted by use of fossil fuels, which is to say, over much of the past 200 years. And, yes, this is ran a bit wide of the whole concept of fuel synthesis, but writing this is helping me formulate some thoughts I've been having on these topics, and there is some relation. Viability So, there remains the question of why there hasn't been more progress or publicity over fuel synthesis. The price/market digression above suggests that a key problem is that the process actually does represent the true real price of fossil fuels, but that exceptionally pervasive (but perverse) price-setting mechanisms. Despite being technically viable and cost-justified on a rational basis, it simply cannot compete against (flawed) market logic. Note that the Navy's interest is an interesting one. Military supplies have a cost basis not based on the market price at source but with the effective delivered cost of materiel. The cost of a gallon of gasoline delivered to Afghanistan is $400 after provisioning and convoying it are accounted for. The Navy's need to supply carrier task forces via oilers across oceans similarly mean that it's not dealing just with a market price of oil but one substantially greater, and the option of provisioning on-site is highly attractive. In a similar example, a proposal to update the engine of the B-52 bomber fleet was revisited when analysts realized that the correct cost basis wasn't of aviation fuel on the ground, but the realized cost when delivered by tanker aircraft for in-air refueling. http://thehill.com/homenews/administration/63407-400gallon-gas-another-cost-of-war-in-afghanistan- It could simply be that technical complexities are too great. Fischer-Tropsch itself has been used, but only in Germany (under wartime pressures) and in South Africa. Exploration of the concept in the U.S., during the 1920s, 1930s, and again in the 1950s. None of the efforts were successful, and the 1950s experiment left a record of significant problems. http://web.archive.org/web/20120111183405/http://fossil.energy.gov/aboutus/history/syntheticfuels_history.html https://www.reddit.com/r/dredmorbius/comments/29ihl7/us_doe_synthetic_fuel_history_coaltoliquids_and/ Electrolysis seems more viable, it's actually used extensively by the Navy today to provide oxygen aboard nuclear submarines, using the vessel's reactor for electricity generation (I'm not sure what's done with the produced hydrogen). CO₂ separation remains a process under development, and might also present challenges at industrial scale. The thought of processing billions of liters of only marginally filtered seawater through sensitive chemical processes strikes me as prone to multiple issues (fouling, contamination, corrosion, mineral deposits, acquatic growth, etc.). Still, in a world where other energy alternatives, particularly those aimed at delivering energy-dense, convenient, and viable liquid fuels fail to land anywhere near "not obviously batshit crazy", this does strike me as an option that falls within the realm of plausibility. In a world which offers little by way of real promise, this is the most interesting possibility I've run across in decades. http://www.sciencealert.com/audi-have-successfully-made-diesel-fuel-from-air-and-water