The economic sustainability of any biofuel is dependent on the existence of a mar­ket, which is largely determined by the cost of production relative to alternative approaches. Such calculations are fraught with uncertainties regarding future costs of scarce resources and without proper assignment of various “hidden” costs, or costs not explicitly associated with or assigned to a particular fuel. To be fully com­parable, such costs should include societal costs such as the military and political costs associated with the import of foreign oil and the opportunity costs associated with diverting agricultural resources to the production of fuel. Additionally, at this early stage of development the final configuration of a biofuels production facility based on a nonphotosynthetic autotrophic organism is unclear. Still, it is possible to consider at least the factors of production that will drive costs and to make at least estimates regarding limits to costs of production.

The sustainable synthesis of any liquid fuel requires inorganic carbon, hydrogen, and energy from a renewable source. In the approaches under development through the Electrofuels program a renewable source of energy is converted to either hydro­gen or electricity, which in turn serves as the energy input for carbon fixation and fuel synthesis. The energy density of liquid fuels is extraordinary-gasoline con­tains roughly 13 kWh/kg or 34 MJ/L-and the entire electricity generating capacity of the USA represents roughly half of US oil consumption, assuming a conversion efficiency of 65% (Supp Calc 7). Given these levels, dedicated on-site electricity generation will almost certainly be the most cost efficient design. A cost comparison of chemolithoautotrophic approaches to corn-based fuels thus turns primarily to a comparison of feedstock costs: electricity vs. corn (Fig. 3).

In recent years, a variety of factors, including a rising demand for carbon-neutral biofuels, has driven sugar costs significantly above historical levels. At the same time, advances in renewable energy production have driven down the cost of renew­able electricity, especially from wind, but also from solar and other resources. Electricity costs are very low for on-site generation (Fig. 4), especially for fully depreciated installations; arbitrage opportunities may further diminish effective electricity costs by balancing electricity sales vs. fuel production. West Texas, for example, possesses a significant wind resource that is largely under-utilized at night, when electricity demand is low. This lack of demand leads to significant wind cur­tailment—3.9 TWh in 2009 and 2.1 TWh in 2010 of essentially “free” electricity

[66] —that would have supported production of1.5MBOE and 0.8MBOE in 2009 and 2010 respectively. This combination of effects leads to periods in which elec­tricity is a significantly cheaper feedstock than sugar for the domestic production of biofuels, a trend that could amplify in the future (Figs. 3 and 4).

To project the cost of a mature Electrofuels technology, we consider an Electrofuels facility operating at the production level of a typical corn-to-ethanol facility, producing roughly 35 million GGE annually. Feedstock costs include the cost of electricity, CO2, and water, with oxygen production considered as a coprod­uct when using an electricity feedstock and assuming a 65% overall energy conver­sion efficiency (Supp Calc 6). Wind electricity costs in the US Midwest currently range from $30 to $70/MWh [66], although these costs will diminish in the future as capital is fully depreciated. Time-of-day pricing is also especially significant for wind resources: during the cheapest 6 h of the day prices seldom rise above $20/ MWh, and wind curtailment produces sustained periods of negatively priced elec­tricity [16]. The cost of the CO2 feedstock could approach zero where concentra­tion and/or purification from waste streams is not necessary. Operating costs include maintenance and taxes, labor and overhead, materials and waste. The cost of capital is for the nth plant cost and based on capital costs for a typical corn-to — ethanol plant, using standard interest rates, construction periods, and depreciation schedules. Additional assumptions are described as footnotes to the calculations (Supp Calc 8).

Not surprisingly, the price of fuel from an Electrofuels approach is extremely sensitive to the price of the energy feedstock, and both conversion efficiencies and the choice of carbon fixation platforms are crucial. Based on realistic assumptions and a cost of electricity of even $40/MWh, the projected cost of fuel is close to $3/ GGE (Table 3); at $20/MWh—the current cost of electricity in the cheapest 6 h of the day—costs drop to $2/GGE. Significant efficiency losses included our calcula­tions include 18% from projected voltage overpotential and 14% for electricity con­version to hydrocarbon fuels. It is important to note, however, that these estimates still largely lack experimental underpinning, and more accurate estimates await additional data.

Both electricity and hydrogen feedstocks are clearly compatible with modern industrial infrastructure and both can supply the energy needed for an Electrofuels process (Fig. 5a). While hydrogen can be produced via electrolysis or perhaps from catalytic water oxidation in the future, it might also be produced from natural gas or biomass reforming, offering gas-to-liquids or biomass-to-liquids (GTL/BTL) pos­sibilities (Fig. 5b). For example, steam reforming of industrial tail gases or low value hydrocarbons can produce hydrogen at less than $1.00/kg [32] with concomi­tant production of free, clean, concentrated inorganic carbon. These values would reduce feedstock costs to $1.34/GGE and overall costs to $2.25/GGE, even consid­ering a likely increase in capital costs (Supp Calc 9).

Cost components and base values are tabulated to determine the individual cost of specific compo­nents as well as the overall cost of fuel production through electrofuels. With each cost item, a sensitivity analysis is provided in the Tornado chart on the right to illustrate how the variation in a single parameter influences the overall cost. The top of the table/chart itemizes standard engineer­ing parameters, whereas the bottom of the table/chart itemizes biological constraints, the latter which each of the Electrofuels projects address

Electrofuels

Fig.5 (a) Paths to Electrofuels using either direct current or dihydrogen as the energy source. (b) Proposed Electrofuels pathway as a gas-to-liquids technology where methane first passes through a steam methane reformer (SMR reactor) and then through a water-gas shift reaction, before the effluent is cooled and fed to an Electrofuels reactor