A recent analysis concluding that natural gas from shale poses more climate change problems than combustion of coal rests heavily on a problematic time frame. In the end, the widely reported article may do more to muddle understanding of natural gas vs. coal than to clarify it.

A recent article in the journal Climatic Change by Robert Howarth and his colleagues argues that natural gas from shale formations is worse for the climate than coal. This controversial finding has been widely reported in the media, but has also been subject to criticism from a wide variety of sources.

Critics say that the Howarth et al analysis used an unrealistically short time frame for estimating the warming effect of methane emissions, and they say Howarth did not take into account the relative efficiency of converting coal and natural gas into useful energy. When these factors are taken into account, electricity generation from natural gas emits one-third to one-half less carbon than coal, a range comparable to prior estimates.


Recent discoveries of massive natural gas reservoirs in shale formations in the northeastern United States have dramatically increased estimates of U.S. natural gas reserves. Because natural gas emits less carbon per unit of energy than any other fossil fuel, many environmentalists have focused on the potential for these natural gas reserves to replace more carbon-intensive fuels like coal for baseload electric generation. However, natural gas does have one potential problem: fugitive methane emissions.

Natural gas is comprised almost entirely of methane, a greenhouse gas much more potent than carbon dioxide in its ability to warm the planet. When natural gas is burned to generate energy or heat, however, virtually all the methane is converted into carbon dioxide. Concerns here involve the natural gas that escapes unburnt during production, transportation, and storage, known as fugitive methane emissions. The global warming potential (GWP) for fugitive methane emissions depends on the time frame considered, as methane in the atmosphere decomposes into carbon dioxide over time. According to the Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment Report, methane has a GWP potential 72 times higher than carbon dioxide over a 20-year period and 25 times higher over a 100-year period (over a 500-year period, methane’s global warming potential is only 7.6 times higher than CO2).

The Howarth article uses a slightly higher global warming potential for methane (105 over 20 years, and 33 over 100 years) than found in the IPCC report, based on more recent work that suggests that the indirect warming impact of methane through chemical interactions in the atmosphere is higher than previously estimated. Some critics point out that relying on that new work may be problematic as it hasn’t been subject to the same level of critical review as the numbers in the latest IPCC report. Nonetheless, the choice of higher global warming potential numbers does not significantly change the Howarth results.

Calculating the amount of fugitive methane released from natural gas production, storage, and transportation is fraught with uncertainty. The U.S. Environmental Protection Agency recently raised its estimate of fugitive emissions from 1 percent to 2.5 percent of total natural gas production. The Howarth report uses a wide-ranging estimate of fugitive emissions from conventional natural gas production — from a low of 1.7 percent to a high of 6 percent. Some say that high estimate is unrealistic, especially as the controversial conclusion of the paper — that natural gas from shale could be worse than coal — mostly relies on using the high estimate. Howarth and his colleagues estimate that natural gas from shale introduces more fugitive emissions comprising an additional 1.9 percent of total natural gas production. These increased emissions occur during the “flow-back” phase of well creation, when hydraulic fracturing fluids flow back to the surface immediately after fracturing and before well completion. These numbers are still highly speculative and based on only a few empirical studies, and the Howarth paper does not address the extent to which flow-back emissions could be captured or flared to reduce the releases of methane.

The major criticism of the Howarth paper involves the choice to focus on the 20-year horizon for calculating methane’s global warming potential. The question of what time frame to focus on has no clear answer; it depends in large part on what policy decisions one is trying to analyze. Most researchers tend to use a 100-year time horizon because societal effects of climate change are projected to be much more significant near the end of the century than over the next 50 years. Considering the implications of using shale gas for electricity generation, choosing coal over natural gas due to short-term 20-year GWP would result in considerably more warming over the 100-year period.

View larger image
Reproduction Figure 1a in the Howarth et al paper based on data presented in the paper and supplementary materials. Note that non-methane indirect emissions are excluded, but are mostly negligible.

Howarth and his colleagues do present the 100-year GWP calculations along with the 20-year calculations. Using the 100-year GWP, however, shows that natural gas is as carbon-intensive as coal only under the assumption of high fugitive emissions (8 percent).

View larger image
Reproduction Figure 1b in the Howarth et al paper based on data presented in the paper and supplementary materials.

Even this comparison is problematic, however. If used to generate electricity, natural gas is quite a bit more efficient than coal. New combined cycle natural gas plants, the type most likely to replace baseload coal generation, tend to convert the chemical energy in natural gas to electric energy at an efficiency of around 60 percent. Current coal plants in the U.S. average around 33 percent efficiency, and even new integrated gasification combined cycle coal plants can reach only 40 to 45 percent efficiency. Considering the generation efficiency differences, and using a more policy-relevant 100-year time horizon, natural gas is considerably lower-carbon than coal even under the most pessimistic assumptions for fugitive methane release.

View larger image
Figure created using data from the Howarth et al paper and supplementary materials. Generation efficiency for new combined cycle gas plants assumed to be 60%. New coal plant maximum efficiency assumed to be 45%. Current average U.S. coal generation efficiency is 33%. Average transmission losses for generation are taken as 6.5%.

Based on EPA’s estimate of  2.5 percent fugitive emissions for natural gas as a reasonable mean, and adding in Howarth’s 1.9 percent additional fugitive emissions from shale gas, a new high-efficiency natural gas plant emits 57 percent less carbon for energy generation than a conventional coal plant, and 41 percent less carbon than a new high-efficiency coal plant, over a 100-year GWP horizon. Even under Howarth’s high-end projection of 8 percent fugitive emissions from shale gas, and comparing natural gas generation to a new high efficiency coal plant, natural gas generation still emits 25 percent less carbon.

Howarth and his co-authors did briefly address the generation efficiency question, stating, “If fuels are used to generate electricity, natural gas gains some advantage over coal because of greater efficiencies of generation … However, this does not greatly affect our overall conclusion: the GHG footprint of shale gas approaches or exceeds coal even when used to generate electricity.” But that broad statement appears somewhat misleading, as it holds only if using a 20-year GWP horizon and using the high-end 8 percent estimates of fugitive emissions.

Howarth and his colleagues also note in their paper that about three-fourths of natural gas is not used for generation, but rather is used for space heating, water heating, fertilizer production, and other purposes. However, nearly 100 percent of coal is used for electricity generation, so a direct comparison between it and natural gas for electricity generation seems most appropriate. Even if coal-based electricity were used to produce the same heating services that natural gas generally provides, the average thermal efficiency of natural gas (75 percent for older furnaces, 90 percent or more for new ones) dwarfs that of coal (33 percent generation efficiency, 6 percent transmission losses, and nearly 100 percent conversion of electric energy into heat).

Ultimately, the paper published by Howarth and his colleagues makes a useful contribution in raising the issue of additional fugitive emissions from shale gas. However, by highlighting the controversial statement that shale gas is worse than coal, and basing the statement on somewhat dubious assumptions, the paper probably contributes more to confusing the issue than to helping to clarify it.

Zeke Hausfather

Zeke is an energy systems analyst and environmental economist with a strong interest in conservation and efficiency. He was previously the chief scientist at C3, an energy management and efficiency company,...