Methane leakage in North China contributing to global warming

Natural gas is a reasonably clean-burning fossil fuel that emits less pollution into the atmosphere than coal and is widely used around the world. Recent studies have demonstrated that natural gas leaks from production, supply chain, and end-use facilities are a significant source of atmospheric methane (CH4) and that bottom-up inventories often underestimate the leaking budget. With a somewhat shorter lifetime than carbon dioxide (CO2), CH4 is the second most important greenhouse gas (GHG) contributing to global warming, making the reduction of CH4 emissions a good target for adopting swift and achievable mitigation solutions under the Paris Agreement.

Natural gas has become the fastest-growing fossil energy source in China during the previous decade, thanks to a government drive to minimize air pollution and CO2 emissions. Natural gas usage has skyrocketed from 108.5 billion standard cubic meters (bcm) (4% of primary energy consumption) in 2010 to a new high of 280 bcm (7.6% of primary energy consumption) in 2018. Furthermore, according to China's energy strategy, the share of primary energy from gas will continue to rise and is expected to reach 15% by 2030, while coal and oil consumption would fall. From 2010 to 2018, the length of gas supply pipes in China's urban areas than tripled, from 298 to 842 thousand kilometers.

However, leakage of CH4 from those pipelines has not been actively recorded, and there is limited publicly available data in China on upstream emissions and local distribution of natural gas emissions.

In a study published in the journal 'Scientific Reports', nine years have been used (2010-2018) of CH4 observations by the Greenhouse gases Observing SATellite "IBUKI" (GOSAT) and surface station data from the World Data Centre for Greenhouse Gases (WDCGG) to estimate CH4 emissions in different regions of China. GOSAT observes the column-averaged dry-air mole fractions of CH4 in the atmosphere, and the surface stations monitor CH4 concentrations near surface. The observation data were used for simulations by the high-resolution inverse model NTFVAR (NIES-TM-FLEXPART-variational) to infer the surface flux of CH4 emissions. Inverse modelling optimizes prior flux estimates, which are constrained so that an acceptable agreement between the simulated and observed atmospheric concentrations is achieved.

Figure 1 shows the model-estimated CH4 fluxes in four regions of China. The four regions, North China (NE), South China (SE), North-west China (NW), and the Qinghai-Tibetan Plateau (TP), vary with respect climate, geographical features, types of agriculture, major economic activities, and CH4 emission sources. The model-estimated average CH4 emissions from the four subregions over the period 2010-2018 are 30.0+-1.0 (average +- standard deviation) Tg CH4 yr-1 from the SE region, 23.3+-2.7 Tg CH4 yr-1 from the NE region, 2.9+-0.2 Tg CH4 yr-1 from the NW region, and 1.7+-0.1 Tg CH4 yr-1 from the TP region. The trends in CH4 emissions have varied in the different regions of China over the last nine years, with significant increase trends detected in the NE region and the whole China.

We focused our analysis on the NE region where natural gas production and consumption have increased dramatically and are likely one of the main contributors to the increase estimated in regional total CH4 emissions. The CH4 emissions from natural gas, including leakage from fuel extraction, processing, transport, and the end-use stage, were estimated using an approach that combined data for the province-level emissions inventory and published inverse model studies. The model-estimated total CH4 emissions and the estimated natural gas emissions both increased significantly during 2010-2018 (Figure 2). The total amount of natural gas emissions due to leakages constitutes a significant waste of energy and value. For example, in 2018, natural gas consumption in the NE region was 101.5 bcm and the estimated total natural gas emissions were 3.2%-5.3% of regional consumption.

Figure 3 shows the changes in estimated CH4 emissions from natural gas and the model-estimated total CH4 emissions for 2010-2018 compared to previous years in the NE region. The year-over-year change in the model-estimated total CH4 emission closely follows the changes in CH4 emissions from natural gas. In January 2016, record cold wave hit the region causing a sudden increase in natural gas use, and natural gas suppliers recorded an increase in natural gas loss (i.e., the difference between the amount of gas purchased and the amount of gas sold). Simultaneously, the atmospheric observations also captured the emission changes, as reflected in our inverse estimates (Figure 3). The analysis shows a strong correlation between trends in natural gas use and the increase in the atmospheric CH4 concentration over the NE region, which is indicative the ability of GOSAT to monitor variations in regional anthropogenic sources.

The findings of our study highlight that the increase in natural gas use threatens China's carbon reduction efforts. The increase in CH4 leaks from natural gas production and the supply chain will adversely affect the interests of diverse stakeholders, despite the introduction of carbon reduction measures. Given that the large natural gas distribution pipelines span more than 900 thousand kilometers in China, natural gas leaks constitute a significant waste of energy and value. The year-over-year changes in regional emissions and trends were detected by satellite and surface observations in this study. In the future, additional observations using high-resolution satellites will help to more accurately quantify emissions and provide scientific directions for emission reduction measures. There is also a need to further detect and locate such leaks using advanced mobile platforms in order to effectively mitigate CH4 emissions in China and bring about economic, environmental, and health benefits.

The Greenhouse Gases Observing Satellite "IBUKI" (GOSAT) is the world's first spacecraft to monitor the concentrations of the two major GHGs CO2 and CH4 from space. NIES has promoted the GOSAT series projects for GHG observation from space, together with the Ministry of the Environment, Japan (MOE) and the Japan Aerospace Exploration Agency (JAXA). GOSAT (IBUKI) is the first satellite in the series and has been observing column-averaged concentrations of CO2 and CH4 for more than 13 years since its launch in 2009. The second satellite, GOSAT-2 (IBUKI-2) was launched in 2018 and started observing carbon monoxide in addition to CO2 and CH4. Furthermore, the third satellite, Global Observing SATellite for Greenhouse gases and Water cycle (GOSAT-GW) is under development and due for launch in Japanese fiscal year 2023.

Methane is the second most important well-mixed GHG contributing to human-induced climate change after CO2. The lifetime of CH4 in the atmosphere refers to the time that CH4 stays in the air after being emitted from a variety of sources. CH4 is removed from the atmosphere mostly by chemical reactions. The atmospheric lifetime of CH4 is 10 +- 2 years, which is relatively shorter than that of CO2 (approximately 5 to 200 years) (IPCC, 2013).

Methane is emitted from a variety of anthropogenic and natural sources. Approximately 60% of all CH4 emissions come from anthropogenic sources, such as agricultural activities, waste treatment, oil and natural gas systems, coal mining, stationary and mobile combustion, and certain industrial processes. Natural emissions include wetlands, freshwater bodies such as lakes and rivers, and geological sources such as terrestrial and marine seeps and volcanoes. Other smaller sources include ruminant wild animals, termites, hydrates and permafrost.

Methane can leak into the atmosphere from upstream/downstream natural gas operations (i.e., extraction and gathering, processing, transmission and storage, and distribution) and end-use combustion. Atmospheric measurement studies have shown that a large amount of CH4 emissions from oil and gas production are unaccounted for in bottom-up inventories. Using high-resolution satellite observations, Zhang et al. (2020) estimated a leakage equivalent to 3.7% (~60% higher than the national average leakage rate) of all the gas extracted from the largest oil-producing basin in the United States. Chan et al. (2020) reported eight-year estimates of CH4 emissions from oil and gas operations in western Canada and found that they were nearly twice that from inventories. Weller et al. (2020) used an advanced mobile leak detection (AMLD) platform combined with GIS information of utility pipelines to estimate CH4 leakage from pipelines of local distribution systems in the United States. They found that the leakage from those pipelines was approximately five times greater than that reported in inventories compiled based on self-reported utility leakage data.

Inverse modeling is an important and essential method for estimating GHGs emissions. The model uses atmospheric observation data as a controller in atmospheric models to optimize bottom-up emission inventories (prior fluxes).

The NIES-TM-FLEXPART-variational (NTFVAR) global inverse model was developed by Dr.Shamil Maksyutov's group at NIES. NTFVAR is combined with a joint Eulerian three-dimensional transport model, the National Institute for Environmental Studies Transport Model (NIES-TM) v08.1i, and a Lagrangian model, the FLEXPART model v.8.0. The transport model is driven by JRA-55 meteorological data from JMA. The prior fluxes include gridded anthropogenic emissions from the EDGAR database, such as energy, agriculture, waste and other sectors; wetland emissions estimated by the Wetland emission by the VISIT model; biomass burning emissions estimated by GFAS; and climatological emissions from oceanic, geological, and termite sources. The inverse modeling problem is formulated and solved to find the optimal value of corrections to prior fluxes minimizing mismatches between observations and modelled concentrations. Variational optimization is applied to obtain flux corrections to vary prior uncertainty fields at a resolution of 0.1° x 0.1° with bi-weekly time steps. A variational inversion scheme is combined with the high-resolution variant of the transport model and its adjoint described by Maksyutov et al. (2021).