A spatially-resolved inventory analysis of the water consumed by the coal-to-gas transition of Pennsylvania

Life cycle assessments (LCA) typically exclude spatial information in estimating the water consumptionassociated with a product, resulting in calls to improve regional detail to better reflect spatial variation. Inresponse to these calls, we have compiled a spatially-resolved inventory of changes in water con-sumption associated with the coal-to-gas transition in Pennsylvania at the resolution of watersheds from2009 to 2012. Results indicate that the total water consumption of the fuel extraction and power sectorsin Pennsylvania increased by 7.6 million m3(2 Bgal) over four years. At the state and watershed scales,we compare total water consumption for the coal-to-gas transition to a case where only the waterconsumed across two life cycle stages of electricity generation is considered e fuel extraction and use atthe power plant to generate electricity. The results for the latter indicate that water consumptiondecreased by over 15.1 million m3(4 Bgal). For both cases, watershed-level results showed water con-sumption generally increased in watersheds with growing shale gas activity or new natural gas capacity,while it decreased with diminishing coal-fired generation. Watershed-scale water consumption from2009 to 2012 may be reversed from a net increase to decrease (and vice versa) when the total waterconsumption is compared to the water consumed specifically for life cycle stages of electricity genera-tion, reinforcing the importance of further developing spatially-resolved inventories for LCA. Focusing onthe water consumption associated with only electricity generation and its fuel use does not capture thefull effects of fuel extracted for use in other sectors. We suggest that spatially-explicit inventories thatinclude multiple life cycle stages should be a critical component in the development of more compre-hensive, spatial LCA methodology. Spatial differentiation in inventories is necessary to adequatelycharacterize watershed-level impacts that can be normalized over a functional unit. The approach can beused as a complementary assessmen t to LCA that can inform policy-makers and investors about whereenergy developments may pose additional risks to water supply and availability