What We Know and Don’t Know on Methane Leakage from the Natural Gas Fuel Chain

Author: Joel Swisher

We know that substituting natural gas for coal in electricity generation can reduce CO2 emissions. Overall national CO2 emissions in the first half of 2012 were about 6% lower than 2011 emissions, and about 15% lower than in 2007.

How did these dramatic reductions occur even before pending Environmental Protection Agency (EPA) air pollution rules take effect despite the complete absence of a national greenhouse gas (GHG) policy? The steep increase in the use of natural gas in place of coal for power generation is the single key contributor to these reductions. In April 2012, for the first time ever, electric generation from natural gas was comparable to generation from coal, each about one-third of the US total.

The substitution of some coal plants to natural gas has lowered U.S. carbon emissions by 6% since 2011, and 15% since 2007.

Switching from coal to natural gas clearly reduces CO2 and other local emissions. Nevertheless, substantial uncertainty remains over the baseline rates of methane leakage from natural gas production and delivery. Leakage is a concern because methane is a more powerful GHG than CO2 in terms of its radiative forcing (heat trapping effect).

We can be confident that fuel chain GHG emissions – from fuel production, delivery, and use – associated with gas-fired power generation are lower than emissions from coal-fired generation. To make this comparison based on the existing literature, I compared eleven studies of natural gas fuel chain emissions. I checked each study’s inputs and harmonized their logic so that it became clear which inputs and assumptions drive the variations in results.

Ten of the eleven studies found that methane leaks and other GHG emissions from the natural gas fuel chain that are not directly resulting from the combustion process add 19-36% to the total GHG emissions. Upstream and non-CO2 GHG emissions also add about 5% to coal fuel chain emissions. At these emission rates, power generation from combined-cycle gas turbines (CCGT) is about half as GHG-intensive as coal-fired generation (see chart).

The outlier among the ten studies is from Howarth, et al. at Cornell, which has been widely cited and criticized in the literature. The emissions estimated in the Cornell study are much higher than the others because the authors assume stronger radiative forcing by methane, compared to CO2, as well as high methane leakage rates from unconventional gas production (use of fracturing to extract gas from less porous rock).

Fuel Chain GHG Emissions

Fuel-Chain GHG Emissions of Gas-Fired CCGT and Coal-Fired Power Plants

The variation in assumptions and results among these studies is an indication of the uncertainty over the methane leakage rates and GHG emissions from the natural gas fuel chain. Existing data compiled by the EPA and Department of Energy are also incomplete and uncertain. For example, in its 2011 national GHG inventory, the EPA increased its estimate of methane emissions from natural gas systems by more than 100% from its 2010 inventory. The EPA noted great uncertainty and applied a modified methodology that resulted in higher estimates of emissions from upstream production.

For conventional production, the major sources of uncertainty in leakage are emissions from cleanups and hydrocarbon liquids unloading. For unconventional sources of gas, the emissions from flowback during well completions and workovers are most uncertain. Additional uncertainty stems from variation in the use and effectiveness of leakage reduction methods.  These include artificial lift (capturing gas while removing accumulated liquids from the wellbore) and “green completions,” which employ temporary processing equipment to capture gas from flowback in the early stages of production.

Gas industry sources have the most detailed performance data, but it is unclear if such data are representative of industry-wide operations. Producers that actively monitor and report leakage rates and reduction actions are likely to be better performers than their peers.  Given their vested interest in reporting low emissions, scientific teams within the industry need to join independent research efforts (from academia, government, and non-profits) to ensure the quantity, accuracy, and representativeness of emissions data from the natural gas fuel chain.

Such joint research is needed to monitor and document adoption of emission reduction practices. Additional collaborative work is also needed to quantify baseline emission rates and reductions in the downstream segments of the natural gas vehicle (NGV) fuel chain; specifically, refueling and storage infrastructure.

A consortium of academic and non-profit scientists led by the Environmental Defense Foundation (EDF) has launched a promising in­dustry collaboration to begin closing the data gaps on natural gas fuel chain emissions. The EDF consortium is working with nine natural gas producers to collect new field data on methane and other emissions from gas production. They are collecting data from gathering and processing facilities as well as transmission and storage. They are also working with local distribution utili­ties to collect downstream leakage data and with gas distributors and fleet man­agers to collect leakage data on stor­age and refueling infrastructure for NGVs.

This work is underway and expected to produce results in 2013. By that time, the first two years of GHG data from oil and gas production should be available from the EPA annual reporting program. The resulting data from all of these activities will be more credible due to the role of EDF and academic experts as independent contributors and reviewers.

The EDF research and emissions studies I reviewed demonstrate the uncertainty that remains regarding methane leakage and the environmental impact of natural gas production and delivery. Given this uncertainty, the gas industry, with the help of independent collaborators and reviewers, should feel the onus to document emission rates and steadily reduce emissions while also reducing the uncer­tainty about leakage and byproducts.


Dr. Joel Swisher is a Consulting Associate Professor of Civil and Environmental Engineering at Stanford University, and an independent consultant in clean energy technology and business strategy. He was also formerly Managing Director at Rocky Mountain Institute.