Date of Award

Spring 4-2016

Embargo Period


Degree Type


Degree Name

Doctor of Philosophy (PhD)


Engineering and Public Policy


W. Michael Griffin

Second Advisor

H. Scott Matthews

Third Advisor

Constantine Samaras


Natural gas production in the United States has increased significantly over the past decade. This is largely due to advancements in hydraulic fracturing, horizontal drilling, and seismic monitoring capabilities that have enabled extraction of natural gas from shale resources to be economically viable. While natural gas is an important global energy resource and may result in fewer emissions than coal for electricity generation, it is important to recognize that extraction of natural gas has the potential to cause local and regional environmental damages. Successfully managing these risks is critical in order to ensure natural gas consumption has a net positive environmental footprint. The second and third chapters of this dissertation use quantitative modeling to assess how policies can address and mitigate potential environmental impacts in a cost-effective manner. Specifically, this work focuses on minimizing incremental fragmentation in critical core forest ecosystems resulting from natural gas infrastructure and on managing wastewater byproducts from natural gas extraction. The second chapter finds that in the case study of a core forest region in Bradford County, Pennsylvania, the number of core patches of forest, an indicator of fragmentation, could as much as double throughout the life time of the Marcellus Shale play (from 80 to 160 core patches) without any regulatory intervention. However, through unitization and collaborative planning, and by designating that gathering pipelines must follow the route of pre-existing roads in forests whenever possible, natural gas infrastructure can be developed in a manner that would both prevent incremental fragmentation from occurring and reduce pipeline construction costs for operators as a result of reduced infrastructure redundancies. The third chapter finds that approximately 1.3 million gallons of wastewater, called produced water, are generated by each well. Across Pennsylvania, 67% of the time Class II disposal is the least cost option, 25% of the time CWT is the least cost option, and 8% of the time on-site treatment is the least cost option. The corresponding average costs are $5.80/bbl ($0.015/Mcf), $7.80/bbl ($0.020/Mcf), and $8.40/bbl ($0.021/Mcf), respectively. In addition to cost, however, there are several technical, ecological, regulatory, and logistical issues that also affect the relative feasibility of these three produced water management strategies. If regulators could capture producers willingness to pay to dispose of water rather than treat the water, that money could be invested in treating other water quality issues in Pennsylvania such as coal mine drainage, which can be treated for $0.064/bbl on average, or agricultural runoff, which could be prevented at an average cost of $0.08/bbl. The last two chapters in this dissertation explore how quantitative modeling can inform policy making on a national and global scale. Chapter 4 does this by characterizing the life cycle greenhouse gas impact of United States natural gas exports. This study finds that mean landed (pre-combustion) life cycle GHGs for exported U.S. LNG after regasification at the importing country were found to be 37 g CO2-equiv/MJ with a range of 27 to 50. The net global impact of these emissions depends on the global warming potential time scale, methane leakage rate, end use, and the fuel it displaces. On a 100 year time scale, life cycle emissions from exported LNG were found on average to be 655 g CO2-equiv/kWh for electricity generation, a 45% reduction over life cycle emissions from coal consumption. However, because of the spatial shift in emissions generation, although there is a global GHG benefit to US natural gas exports, the United States should consider the implications of this given that emissions calculations are based on CO2-eq emitted within a country’s borders rather than based on the net global impact of those emissions. The fifth chapter continues to explore international trade policy by focusing on the global crude trade as a case study. This chapter considers how shifting trade patterns can influence global costs and greenhouse gas emissions using a linear optimization model. The baseline 2014 crude trade system had a global cost of $3T and resulted in 16.5 Gt of CO2. Minimizing by cost would save $6T and increase emissions by 4 Gt CO2, while minimizing by emissions would increase cost by $0.5T and decrease emissions by 5.4 Gt. This chapter then explores the interaction between climate policies including carbon accounting methods, a designated global carbon cap, and unilateral country specific emissions allocations. There is a 40% higher allowable consumption under a strict global carbon cap without country-specific emissions allocations (1100 Mmt) than with country-specific emissions caps (770 Mmt). These results illustrate cooperative international climate policy could be more cost-effective in mitigating carbon emissions than countries acting individually.