PEFTEC - Abstract

Abstract Information :

Atmospheric methane concentrations rose by about 3% during the decade from 2006 to 2016 following a previous decade of near-constant concentrations. Studies based on ethane measurements suggest that much of this global rise has likely been caused by increasing oil and gas methane emissions [1,2], but a recent analysis of 13CH4 data proposed a biogenic source [3]. This issue remains controversial, mainly as a result of differences in the physical assumptions made by the respective research groups.

This work achieves a substantial step towards resolving these disagreements. Using a synthesis analysis of both ethane data [1,2] and isotopic (13C and 1H) CH4 data [4], it is possible to more tightly determine the contributions of oil and natural gas sources to global methane emissions. As a result, it is shown that about half of the recent rise in atmospheric methane has been caused by the rapid growth of the shale gas industry, with fugitive emission ratios from natural gas systems supplied by shale or tight gas averaging 4-6% of production, compared to 2.5-3.2% from systems supplied by conventional natural gas.

The approach used here utilises the time variation of the methane/ethane ratio (MER), first to separate emissions from all oil and gas sources into fluxes arising from oil-associated gas (AG) and non-oil natural gas (NG). The NG contribution is then compared to NG production to estimate a fugitive emission ratio, FERNG. Remarkably, it is found that FERNG has slowly risen over the last 45 years, with a more rapid increase during the last decade, despite improved engineering methods such as reduced emissions completion and replacement of pressure-driven valves.

This result is modelled by further separating natural gas emissions into contributions from conventional and unconventional gas with different FER values. Since unconventional NG production has risen much more rapidly than conventional NG production, a higher FER value for unconventional NG recovers the observed increase in overall FER. This result is consistent with numerous reports from top-down (aircraft and satellite) measurement studies, for example, Refs. 5 and 6, and with more recent bottom-up studies [7], showing relatively higher emissions from unconventional than conventional sources. Finally, in the US, where in the last decade conventional NG production significantly fell while unconventional production strongly increased, we show that a substantial portion of the observed ˜30% increase in US methane emissions from 2002-2014 [8] is likely a result of unconventional gas production. Clearly, the present conclusion that natural gas systems are a major contributor to the recent increase in atmospheric methane has to be robustly evaluated. The paper will therefore conclude with a discussion of merits and weaknesses of model assumptions in this and other recent analyses. Key points to be made are:

(1) The analysis in Ref. [3] is impacted by likely incorrect assumptions that biomass burning emissions are constant, and that the stable carbon isotope ratio for oil and gas methane can be lumped with that of coalbed methane, with a value of δ13C ˜ -35°/00
(2) Historical variations in the OH sink [9] have little effect on the FER estimate of 4-6 % for unconventional natural gas, extracted in this work.