© The Author(s), published by EDP Sciences, 2025

1 Introduction

Many human activities (agriculture, burning fossil fuels, cutting down forests, raising livestock, etc.) add huge quantities of greenhouse gases to those naturally present in the atmosphere, increasing the greenhouse effect and global warming [1]. Carbon dioxide (CO₂) is the most emitted anthropogenic greenhouse gas: it is responsible for more than half of global warming. Methane (CH₄) comes next with a contribution of 20–30% [2, 3]. Two properties characterize the impact of greenhouse gases on climate: the time they remain in the atmosphere and their ability to absorb energy. Methane has a much shorter lifespan in the atmosphere than CO₂ (around 12 years versus centuries for CO₂ [4]. However, it is more potent as it absorbs much more energy while in the atmosphere: its global warming potential (GWP) is about 30 times that of CO₂ over 100 years, or 80 times over 20 years [2]. For both reasons, removing methane from the atmosphere should limit warming more rapidly than removing CO₂ [5, 6]. According to the 27th Conference of the Parties (COP27), “controlling methane is the cheapest and fastest way to slow global warming in the years ahead” [7]. Mitigating anthropogenic methane emissions is therefore, a key factor in achieving the climate targets set out in the Paris Agreement, i.e., a global average surface temperature in 2100 no more than 1.5 °C above pre-industrial levels. This lever could contribute to a temperature reduction of around 0.1 °C by 2050 [8, 9]. In addition, this would have a positive impact on human health as methane is a precursor of tropospheric ozone (O₃), and thus plays a part in air pollution worldwide [10].

Although estimates of methane emissions are highly uncertain, the annual volume of methane rejected in the atmosphere worldwide is considered to be around 570 million tons (Mt) in 2020, of which 40% is of natural origin and 60% of anthropogenic origin [11]. The primary sources of anthropogenic methane emissions in 2022 are agriculture (140 Mt), the energy sector (128 Mt) with around 40 Mt for coal, the same for oil and gas, and waste (70 Mt) [3, 4]. Clearly, efforts need to be made in all sectors to reduce methane emissions to help meet the objectives of the Paris Agreement. However, it may prove simpler and quicker to act on the energy lever. According to the IEA [3], global methane emissions from the oil and gas sector could be reduced by around 75% with existing low-cost technologies (leak detection, repair programs, upgrading of leaking equipment), by investing around $US 100 billion between now and 2030. The coal sector is often disregarded, as it is generally more difficult to achieve significant reductions in methane emissions than in the oil and gas sector, and because the most effective strategy for reducing emissions from coal mines is basically to reduce coal consumption.

States have become aware of the stakes involved in lowering methane emissions. Under the Global Methane Pledge launched at the 26th Conference of the Parties (COP26) in 2021, the European Commission, together with 159 countries, committed to reducing global anthropogenic methane emissions by 30% in 2030 compared to 2020 levels [12]. More recently, in April 2024, the European Parliament passed a law to reduce methane emissions from the energy sector within Europe as well as across global supply chains, with the intention of influencing practices beyond Europe’borders. While the impact of this legislation remains to be seen, it targets direct methane emissions from the oil, fossil gas, coal, and biomethane sectors, once injected into the gas network.

Gas transportation is part of the natural gas supply chain [13]. It can occur via pipelines or, when gas is liquefied, by ships. The case of liquefied natural gas (LNG) falls outside the scope of this study. Pipeline networks, on the other hand, perform both transmission and distribution functions. They can be a source of methane leaks in mainly two respects: by fugitive emissions and by venting. Fugitive methane emissions are caused by unintentional releases resulting from leaking components, such as a faulty seals or leaky valves. Vented methane emissions result from intentional releases, often for safety reasons, due to the design of the facility or equipment (e.g., pneumatic controllers) or operational requirements (e.g., venting a pipeline for inspection and maintenance).

This paper provides estimates of the losses along natural gas transmission and distribution networks. The first two sections are dedicated to the quantification of “Unaccounted-for gas” (UFG) all over the world: the first one describes the methodology followed, and the second one focuses on the results. UFG usually refers to the difference between the measured quantity of natural gas entering the gas network and the measured quantity of natural gas flowing out of the same network [14]. The reasons for gas loss and unaccountability are manifold: leaks (i.e., fugitive emissions, incomplete flaring, and venting), theft, inaccurate metering, and gas used to operate network equipment that is known as “own use gas” (it is not accounted for). Then, the third section aims to deduce the methane released by each country into the atmosphere along gas networks. The quantification of UFG enables methane emissions to be estimated. Since natural gas is a mixture of different hydrocarbons, of which methane is the main constituent (usually 87–97%), a non-zero UFG is an indicator of potentially high methane emissions.

2 Methodology to quantify unaccounted-for gas

Various techniques, usually defined as “top-down” and “bottom-up”, can be used to estimate methane emissions. “Top-down” techniques use remote measurement devices to measure atmospheric methane concentration. For instance, methane concentration can be measured locally in ambient air with gas analyzers [e.g., 15–17] or regionally in the atmosphere using satellite, drone, and other imaged-based techniques [18, 19]. Then, the measured concentrations are inverted to evaluate annual methane emissions. “Bottom-up” techniques approach the question the other way around with measures directly performed at the source, thus providing more accurate results. The approach developed below for estimating methane emissions from natural gas networks is of the “bottom-up” type.

Immediately after production at the field, the gas produced is purified and prepared for pumping. Pumped by compressor stations to high pressures, it enters the transmission network designed to transport substantial volumes of gas across extensive distances. Gas flows into the high-pressure pipeline network via a number of entry sites, delivered either via pipelines from offshore or onshore gas fields or via international transit pipelines and, in some cases, LNG import facilities. The LNG chain is not considered in its entirety; only the gas that entered the transmission and distribution networks is taken into account. At all entry points, the gas must be measured with great accuracy for fiscal reasons. At some point, the gas leaves the gas transmission networks and enters the distribution networks. The interface between both systems requires accurate gas metering. The distribution networks, which are adjusted to gradually reduce pressure, then transport gas from transmission networks to end consumers. Again, gas metering is essential for billing purposes. Many measurements are carried out all along the natural gas networks to monitor their operation. Most of them are disclosed in public reports or on the web by various organizations (transmission system operators, distribution system operators, companies, national and international agencies).

The imbalance between gas entering the network and gas exiting corresponds to the unaccounted-for gas (UFG). Since there is no universal definition of UFG in the energy industry, we have adopted the definition provided by [20], which aligns with the commonly understood meaning:$$\mathrm{UFG }= \mathrm{Leakage} + \mathrm{Own} \mathrm{use} \mathrm{gas} + \mathrm{Theft}$$

“Leakage” refers to gas losses due to daily operations, venting, accidentally damaged pipes, or metering errors. Leaks can occur on various devices, such as compressors. “Own use gas” (OUG) is the gas used by the network utility to deliver the gas it transports to the customers. It is then used to support network operations, in the form of preheating and fuel for compressor stations. Indeed, gas must be heated prior to pressure reduction through the preheating process to mitigate the negative effect of a temperature drop on equipment during pressure reduction. Besides, “Theft” represents the gas withdrawn from the network by unauthorized third parties.

Leaks from high-pressure gas transmission pipelines are generally very rare and very low, as the high-pressure gas transmission network consists of welded steel high-pressure pipelines, which do not normally leak. Under normal circumstances, the percentages of UFG in gas transport systems are therefore quite low, typically less than 0.5% of total flows [21]. In distribution networks, on the other hand, they can be much higher, usually in the order of 1–5%.

The evaluation of UFG volumes calls for data on UFG percentages, gas demand, and pipeline network sizes. Gas demand data were extracted from the 2021 Cedigaz database. Cedigaz is an association founded in 1961 that aims to collect data on gas worldwide. Year after year, it has collected a large amount of data with the same rigor, creating a database covering more than 120 countries with gas reserves, gross production, gas flared, reinjected, or otherwise lost, marketed production, imports, exports, and consumption. The pipeline lengths were sourced from transmission system operators (TSO) (e.g., GRTGaz in France¹), but also from the Central Intelligence Agency world factbook.² The UFG percentages used at the national level, defined as the ratios of UFG volumes to national gas demands, were those provided by national regulators (e.g., Gas Industry Co. for New Zealand³), governments, or TSOs, where available and reliable. Otherwise, there were two possible cases. First, UFG percentages were calculated as a weighted average of the UFG percentages reported by individual TSOs and distribution system operators (DSOs) (e.g., Gas Networks Ireland⁴). Second, in the absence of any data, they were estimated on the basis of the age and physical characteristics of the pipeline network and the number of customers. The Cedigaz database lists 108 countries with gas demand above 0, ranging from 0.05 billion m³ to 868 billion m³ (US). This list has been reduced to 82 countries based on the following criteria: annual gas demand greater than 1 billion m³ existing gas transmission/distribution networks data of sufficient quality.

3 Unaccounted-for gas percentages and volumes

The UFG percentages by countries derived from the methodology described above are displayed in Figure 1. They range from 0.01% to 15%, with an average value of 1.7%. A low percentage of UFG is a sign of a recent or well-maintained gas network or more recent pipeline materials.

Figure 1 UFG percentages by country.

The countries with the highest UFG percentages are Myanmar (15%), Syria (13.8%), Pakistan (11.7%), South Africa (8%) and Australia (4%). It is worth mentioning that no data were reported for China, as it was not possible to collect sufficiently reliable values for UFG percentages. At the regional level (Table 1), the largest UFG percentages are in Asia-Oceania (2.04%) and the Commonwealth of Independent States (CIS) (1.95%), followed by North America (1.88%), the Middle East (1.64%), Africa (1.35%), Central and South America (1.32%), and Europe (0.52%). The UFG percentages for CIS (2.0% for Russia) and North America (2.1% for the US) are above the world mean (1.67%). They are also the two leading producers of natural gas with the oldest and largest natural gas networks. In CIS countries, outdated leak detection technologies and inadequate measurement practices create significant uncertainties regarding the exact volume of natural gas leakage, which could reach up to 5% in some areas.

However, UFG percentages alone are not enough. They must be set against UFG volumes, which depend on the size and maturity of the natural gas market. For example, Myanmar’s UFG percentage and volume are 15% and 0.615 billion m³, respectively, compared with 2.10% and 17.997 billion m³ for the US, and 0.75% and 0.325 billion m³ for France. The two countries with the largest UFG volumes are the US and Russia (Fig. 2). Methane leaks in the US are primarily due to the aging infrastructure of the gas network. Many pipelines, some dating back to the early 20th century, are susceptible to corrosion and cracking. Additionally, the US has one of the largest gas networks in the world, significantly increasing the potential sources of leaks. Regulatory frameworks also vary widely between states, with some enforcing stricter standards than others. Finally, gas transportation is handled by a large number of small companies, whose practices are not always of the highest standard. Similar to the United States, Russia’s gas infrastructure is vast but aging. A large part of the pipeline network was built during the Soviet era, and many of these pipelines are now outdated, prone to leaks, and in need of modernization. The long distances over which gas is transported, particularly from remote production areas to major consumption centers, further amplify the risk of UFG. Last, Russian regulations regarding methane emissions are relatively lax, and leaks in natural gas networks are not always treated with the attention they deserve.

Figure 2 Countries in the top 10 UFG volumes.

4 Resulting methane emissions

Non-zero volumes of UFG mean that some gas is lost along the transmission and distribution networks and that these losses result, at least in part, in methane emissions into the atmosphere. However, it is quite difficult to estimate the amount of methane emitted. The contribution of leaks is clearly direct. As for the stolen volumes, some of them also lead to methane emissions, as non-standard connections to the system are highly susceptible to leakage. Most of the time, network operators do not know with any real certainty the actual proportions of own-use gas (OUG) associated with leaks and theft, so these figures can only be estimated. In some cases, these estimates are to be linked to concerns of the gas network operator. For example, estimates of high levels of leakage could be used to justify demands for further investment in infrastructure, while estimates of high levels of theft could be used to conceal poor operation and maintenance practices. Consequently, since the relative proportions of leakage and theft in the UFG volumes are unknown, the resulting methane emissions are assumed to correspond to 50% of the UFG. This is obviously a crude approximation, but it reflects the poor quality of the data displayed on the UFG. We also assume that natural gas contains 95% methane.

Given these assumptions and the data available, we estimate that the total amount of methane emitted along natural gas transmission and distribution networks in the world was around 18 Mt in 2021 (Table 1). Comparing our results with other studies is challenging, as the segments analyzed are never the same. Of the 128 Mt of methane emissions reported by the IEA for the energy sector, 7 Mt are attributed to gas networks and LNG-related infrastructures [3], meaning that emissions due from gas networks alone are lower. This suggests that our estimate of 18 Mt is significantly higher than the IEA’s inferred figure for the same scope. Another source mentions methane emissions between 14.5 and 48.2 Mt for the major natural gas supply chains [22], which include gas production, transport by pipeline or as LNG, and consumption. Once again, the segment considered is broader and the range provided is quite large. As stated above, our estimate also carries uncertainties due to errors in the collected data and the assumption made regarding the proportion of UFG ultimately released into the atmosphere. We assumed this proportion to be 50%. However, in countries where OUG is well quantified, and theft is minimal, this proportion is likely closer to 100%. In contrast, in other regions, it can be lower. Anyway, the strength of the methodology developed in this paper lies in its ability to be repeated over time, making it possible to monitor the evolution of the 18 Mt of methane released. Focusing on mitigating these emissions would make a significant contribution to the fight against global warming. Adopting the best available methane abatement technologies can actually lead to an 80% reduction in methane leakage along the natural gas value chain, capping the additional environmental burden to 8% of direct CO₂ emissions (compared with 35% today) [23].

As shown in Figure 3, the greater the demand for gas, the greater the methane emissions. However, the data are widely dispersed around the regression line. This is due to UFG percentage effects: countries with a high UFG percentage are above the line, the others below. The results by countries are mapped in Figure 4. The 10 countries with the highest methane emissions from their gas transmission and distribution networks are: the US (5.62 Mt), Russia (3.06 Mt), Pakistan (1.76 Mt), Iran (1.45 Mt), Australia (0.65 Mt), Saudi Arabia (0.59 Mt), Argentina (0.49 Mt), Turkey (0.45 Mt), Canada (0.43 Mt), and Qatar (0.35 Mt). These top 10 countries include major gas producers, some with extensive pipeline networks, as well as consumers and transit countries. The largest producing countries (US, Russia, Canada, Iran) are the most mature, as gas networks were first built in the vicinity of major gas fields. They, therefore, have the oldest infrastructure networks. The main producing countries are among the top 10 emitters, with the exception of China and Norway, but these two countries have not been analyzed here. China, the world’s leading methane emitter due in part to coal mining, also records a significant amount of fugitive and vented methane emissions from downstream gas (0.85 Mt from pipelines and LNG facilities, according to the IEA’s Methane Tracker database 2023 [3]). Norway is a special case, as its production fields are export-oriented, while the volume of natural gas transported for domestic use is marginal. On the other hand, more recent markets and “modern” importing economies also have very dense but better-maintained infrastructure networks and, therefore, lower emissions. This is generally the case in Europe with Italy (0.16 Mt), France (0.10 Mt), United Kingdom (0.08 Mt) or Germany (0.02 Mt) and in the advanced Asian countries with South Korea (0.03 Mt), Japan (0.01 Mt) or Taiwan (0.0008 Mt). For the former, the results can be attributed to good practices. For the latter, they mainly import LNG and have smaller networks. This choice can also be explained, particularly in the case of Japan, by the high seismic risk.

Figure 3 Volume of methane emitted against gas demand. Bubble’s sizes and colors are determined by UFG percentages.

Figure 4 Estimated methane emissions due to UFGs by country in 2021.

In the near future, methane emissions from the two largest emitters mentioned above could take a turn for the worse. In the US, recently re-elected D. Trump strongly supports the expansion of fossil fuel production. Increased gas production or the rollback of regulations introduced by previous administrations, such as the Methane Emissions Reduction Program under the Inflation Reduction Act (IRA), could increase the risk of methane leaks. In addition, Western sanctions on Russia following its invasion of Ukraine have reshaped Europe’s gas supply flows. Russian gas imports by pipeline have been largely replaced by deliveries of US LNG. Yet LNG is associated with significant methane emissions. Indeed, transporting natural gas in LNG form tends to generate more methane emissions than pipeline transport for the same volume of gas transported due to the additional processes involved: liquefaction, boil-off gas, and regasification. On the other hand, focusing on Russia, the invasion of Ukraine may also lead to an increase in methane emissions. Before the war, Europe was Russia’s primary market for gas exports and put growing pressure on Moscow to reduce its methane footprint. In response, Russia began to take the issue more seriously, especially after the European Commission launched its methane strategy in 2020 and discussed a carbon border adjustment mechanism (CBAM). At a climate summit convened by J. Biden in April 2021, Vladimir Putin even acknowledged methane’s impact on global warming and called for international cooperation to curb emissions. However, since the invasion, methane reduction has taken a backseat. Russia shifts toward new markets in the East, thus bypassing European pressures. In addition, the country’s wartime economic restructuring is likely to severely limit its ability to monitor and mitigate methane emissions.

5 Conclusion

With growing environmental and financial concerns about potential levels of methane emissions, the issue of methane emissions in general and UFG in particular is increasingly on the agenda. This is a key issue for the production, transport, and distribution of natural gas. It also concerns the future production of blue hydrogen. This hydrogen is formed from the steam reforming of methane. For it to be truly low-carbon, it needs to be combined with carbon capture and storage technologies downstream, while any methane leaks upstream need to be as low as possible. Reducing UFG would not only reduce financial losses but also, above all, help contain global warming. The persistence of methane in the atmosphere is much lower than that of CO₂: a decrease in the quantities of methane emitted would have a rapid beneficial effect.

Using data published by TSOs, DSOs, and government institutions on gas networks as well as the Cedigaz database, we first estimated the UFG and then calculated the associated methane emissions. The average UFG for the world is 1.7%, but with significant disparities between regions, ranging from 0.01% to 15%. The resulting methane emissions are provided by country and correspond to 18 Mt worldwide. The results point to two types of countries. On the one hand, the main producing countries are characterized by high UFG volumes and high methane emissions. The reason for this is the aging of their gas networks and materials. By comparison, networks in more recent markets or advanced countries have fewer leaks, leading to more moderate methane emissions. The state of the environmental regulation and its implications for players are also essential.

This suggests that it is possible to cut methane emissions. The complete elimination of gas leaks is impossible, but the implementation of robust maintenance and inspection strategies for pipelines and equipment can help reduce them. An additional advantage of the counting methodology developed in this paper is that it can be repeated every year, enabling future emission reductions to be monitored.

Finally, regulations encouraging the disclosure of shrinkage and Own Use Gas data will be a step in the right direction. These practices will make operators publicly accountable. Greater transparency of UFG data on the part of the stakeholders concerned is necessary to better identify actions to be taken to curb methane emissions. In addition, it will prevent operators from potentially passing on the cost of UFG to the end consumer.

References

All Tables

All Figures

	Figure 1 UFG percentages by country.
In the text

	Figure 2 Countries in the top 10 UFG volumes.
In the text

	Figure 3 Volume of methane emitted against gas demand. Bubble’s sizes and colors are determined by UFG percentages.
In the text

	Figure 4 Estimated methane emissions due to UFGs by country in 2021.
In the text