Harmonization of initial estimates of shale gas life cycle greenhouse gas emissions for electric power generation

Garvin A. Heath; Patrick O’Donoughue; Douglas J. Arent; Morgan Bazilian

doi:10.1073/pnas.1309334111

New Research In

Edited* by Hans Joachim Schellnhuber, Potsdam Institute for Climate Impact Research (PIK), Potsdam, Germany, and approved April 21, 2014 (received for review May 17, 2013)

Significance

Previously published life cycle assessments (LCAs) of greenhouse gas emissions from the production and use of shale gas have come to widely varying conclusions about both the magnitude of emissions and its comparison with conventionally produced natural gas and coal for electricity generation. We harmonize estimates from this literature to establish more consistently derived and robust summary of the current state of knowledge. Whereas median estimates for both gas types appear less than half that of coal, alternative assumptions may lead to emissions approaching best-performing coal units, with implications for climate change mitigation strategies.

Abstract

Recent technological advances in the recovery of unconventional natural gas, particularly shale gas, have served to dramatically increase domestic production and reserve estimates for the United States and internationally. This trend has led to lowered prices and increased scrutiny on production practices. Questions have been raised as to how greenhouse gas (GHG) emissions from the life cycle of shale gas production and use compares with that of conventionally produced natural gas or other fuel sources such as coal. Recent literature has come to different conclusions on this point, largely due to differing assumptions, comparison baselines, and system boundaries. Through a meta-analytical procedure we call harmonization, we develop robust, analytically consistent, and updated comparisons of estimates of life cycle GHG emissions for electricity produced from shale gas, conventionally produced natural gas, and coal. On a per-unit electrical output basis, harmonization reveals that median estimates of GHG emissions from shale gas-generated electricity are similar to those for conventional natural gas, with both approximately half that of the central tendency of coal. Sensitivity analysis on the harmonized estimates indicates that assumptions regarding liquids unloading and estimated ultimate recovery (EUR) of wells have the greatest influence on life cycle GHG emissions, whereby shale gas life cycle GHG emissions could approach the range of best-performing coal-fired generation under certain scenarios. Despite clarification of published estimates through harmonization, these initial assessments should be confirmed through methane emissions measurements at components and in the atmosphere and through better characterization of EUR and practices.

Footnotes

↵¹To whom correspondence should be addressed. Email: garvin.heath{at}nrel.gov.

Author contributions: G.A.H. designed research; G.A.H. and P.O. performed research; G.A.H. and P.O. analyzed data; and G.A.H., D.J.A., and M.B. wrote the paper.
The authors declare no conflict of interest.
ⁱWeber and Clavin (25) is a Monte Carlo-based synthesis of results from six unconventional gas LCAs and thus included for perspective but not harmonization.
ⁱⁱFor rhetorical simplicity, these eight LCAs will sometimes be referred to by the first author’s last name.
ⁱⁱⁱRestimulation is a synonymous term to recompletion; workover can refer to well maintenance without hydraulic fracturing (29, 30), or mean recompletion (31).
^ivHeath et al. (15) didn’t create their own comparison case, but rather compared Barnett shale results to harmonization of 42 references (26) that collectively focus on a diverse set of conventional or domestic gas types. Laurenzi and Jersey (16) compared their results to coal only; we use the results of O’Donoughue et al. (26) for comparison with Laurenzi and Jersey where necessary.
^vNote that each reference could report more than one estimate.
^viThe findings reported here are robust to use of the probability distribution of EUR for wells in all shale gas plays in the United States. Both distributions are described in SI Text and Table S4.
^viiThere is also the possibility of differences in definition of methane leakage, where some might only include methane contained in natural gas that is unintentionally released to the atmosphere (often referred to as fugitive emissions), others might additionally include methane in natural gas intentionally leaked to the atmosphere (often referred to as vented emissions), and still others might additionally include methane not just contained in leakage of natural gas but also emissions of methane from combustion or even from tanks storing coproducts (condensate or oil). For instance, Heath et al. (15) included fugitive, vented, and combustion-emitted methane in their leakage estimate, whereas Skone et al. (11) and Burnham et al. (13) only included fugitive and vented methane emissions. Harmonization to a common definition of methane leakage was beyond the scope of this study because the data were not reported at this resolution.
^viiiA leakage rate reported in these units enables rapid estimation of methane emissions based on a known amount of produced natural gas.
^ixFor gas produced from oil wells, only GHG emissions starting with gas processing are assigned to the natural gas industry; the EPA assigns oil production GHG emissions, including those related to associated gas, to the oil industry (48). This approach is consistent with what is known as the product-purpose coproduct allocation philosophy (47).
^xSee www.nrel.gov/harmonization for a complete list.
^xiThe unconventional gas well EUR used here is the median for active shale plays in the United States (2.2 bcf) as determined through our analysis of EIA data (41), and well lifetime is assumed to be 30 y. Conventional gas well EUR is assumed to be 1 bcf (a central estimate from LCAs considered in this study), and well lifetime is 30 y.
↵*This Direct Submission article had a prearranged editor.
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1309334111/-/DCSupplemental.