Evaluating the Economic, Environmental and Policy Impacts of Ethanol as a Transportation Fuel in Pennsylvania

Natural gas is a growing energy source in the US for various end-uses, and its potential future as a transportation fuel has been the focus of recent policy discussions. Nationally, ethanol is blended with gasoline up to 10% for conventional vehicles, and up to 85% (E85) for use in Flexible Fuel Vehicles (FFVs). Federal mandates require increasing ethanol use in the transportation sector. Meeting the mandates could mean increasing the blend in conventional gasoline, or increasing the use of E85 in FFVs. This dissertation explores the economic, environmental and policy effects from producing ethanol from natural gas, and generally expanding access to ethanol as a transportation fuel (feedstock agnostic). Three processes are considered for producing ethanol from natural gas: (1) autothermal reforming (ATR) with catalytic conversion, (2) TCX, a process that produces intermediate products of methanol and acetic acid, developed by Celanese Corp., and (3) a fermentation process developed by Coskata Inc. I first estimate the cost of producing ethanol from natural gas to power light-duty FFVs in Pennsylvania (PA). Relying on production cost estimates provided by developers and assuming recent natural gas and gasoline prices are good proxies for future prices, I conclude that the cost of producing ethanol with either the Coskata or ATR processes would more likely than not be cheaper than gasoline and corn-based ethanol. However, capital costs from these emerging processes and future natural gas and gasoline prices are highly uncertain. The NGLF ethanol must also have acceptable greenhouse gas (GHG) emissions, for which an estimate is not currently available in the literature. I find the average life cycle GHG emissions for a 100-yr global warming potential (GWP) are 137 g CO2-eqiuv/MJ (ATR Catalytic), 119 g CO2-eqiuv/MJ (Celanese TCX) and 156 g CO2-eqiuv/MJ (Coskata fermentation), given the uncertainty in some parameters the estimate could be slightly higher or lower. All processes have life cycle emissions well above gasoline, and the 20% reduction from gasoline required by the Renewable Fuel Standard (RFS2). Even in the unlikely scenario of zero emissions from the upstream processes, NGLF ethanol process and combustion emissions are still larger than gasoline, although with more overlap in the error bars. More detailed life cycle assessments with process modeling could refine the emissions estimates. Existing policies incentivize ethanol produced from renewable sources, but no current policy provisions specifically incentivize the use or production of ethanol produced from natural gas. I conclude the dissertation with estimates of additional refueling costs for an FFV driver and infrastructure costs for expanding E85 access in Pennsylvania. The state recently received government grants for biofuels infrastructure. I find that even with a subsidy to cover average infrastructure costs of $0.03 to $1.48 per gasoline gallon equivalent (gge) for the retailer, the consumer would still incur additional costs for refueling more often with E85. A refueling cost subsidy of $3.60/gge to cover the additional costs is also higher than historical ethanol subsidies. Additionally, a subsidy to encourage E85 use could reduce emissions at a cost equivalent to $1,320/metric ton CO2, which is approximately two orders of magnitude above the average social cost of carbon. Therefore, reducing emissions through more ethanol fuel use is not a cost-effective mitigation strategy.