Greenhouse gas emissions from sanitation and wastewater management systems: a review

While sanitation is a fundamental basis for public health, there is lower awareness of the associated contributions to environmental pollution (Gwenzi et al. 2023). Sanitation systems, including all their components in the sanitation service chain from containment to conveyance, treatment, and disposal or reuse, are responsible for an often-overlooked share of greenhouse gas (GHG) emissions, especially nitrous oxide (N₂O) and methane (CH₄), whose global warming potential (GWP) over a 100-year period is 273 and 28, respectively (IPCC 2023). These emissions include direct emissions from decomposition of organic matter found in excreta, as well as indirect emissions due to, for example, energy and transport used for operating treatment facilities. Wastewater and sludge management are estimated to be responsible for 257 million tonnes of carbon dioxide equivalents (CO₂-eq) while non-sewered sanitation is responsible for 267 million tonnes CO₂-eq annually (Lutkin et al. 2022). This altogether accounts for approximately 1.3% of global GHG emissions (Ritchie et al. 2020), roughly similar to the global aviation sector. With regards to methane, non-sewered sanitation is estimated to be contributing up to 4.7% of global anthropogenic methane emissions (Cheng et al. 2022). At a city level, recent modelling estimates of direct and indirect emissions from both sewer-based and non-sewered sanitation technologies in Kampala, Uganda, indicated that sanitation could be responsible for more than 50% of the city's emissions (Johnson et al. 2022).

While emissions from sanitation systems have been small relative to those from other sectors, recent trends point to increases in relative contributions globally. In the United States, a growing proportion of CH₄ emissions arise from wastewater treatment (10% in 1990 to 14% in 2019) (Moore et al. 2023). In regions such as South Asia and sub-Saharan Africa, which still have a huge sanitation gap and which are also experiencing rapid population growth and urbanization, GHG emissions from sanitation systems are projected to increase, mainly due to a reliance on pit latrines (Reid et al. 2014). Globally, about 3.3 billion people are dependent on non-sewered sanitation technologies today and this number will possibly reach 5 billion by 2030 (Strande et al. 2014; Cheng et al. 2022). This will involve the construction of many new sanitation facilities, and thus may result in an increase in GHG emissions (Orner & Mihelcic 2018; IPCC 2022). It is therefore important to consider options for sanitation technologies with lower emissions, while also ensuring that users themselves, particularly in low-income countries, should not have to bear the burden of the potential extra costs of lower-emissions sanitation systems entirely alone.

One key challenge to improving identification of sanitation options with lower emissions is that GHG emissions accounting across the sanitation service chain remains limited due to barriers rooted in the inherent complexity around designing, implementing, and operating sanitation systems (see e.g. Spuhler et al. 2020). Relatively little research has been focused on trying to systematize the relevant information concerning GHG emissions throughout the entire sanitation chain, with most work typically addressing isolated components, especially in the treatment part of the chain, without integrated systems approaches as indicated by Shaw et al. (2021). Therefore, it is critical to understand the main factors influencing emissions and the current evidence gaps in the sanitation and wastewater sector across the entire sanitation chain.

In this paper, we aim to contribute to better consideration of climate change in planning sanitation services by reviewing and synthesizing information on GHG emissions from the sanitation and wastewater management chain. We categorize results according to the type of sanitation system, either sewer-based with wastewater treatment or non-sewered systems, describe the key emissions sources at each stage of the sanitation service chain, and discuss three key aspects related to GHG emissions from the sanitation chain: systems configuration; systems operation; and energy consumption.

Estimating GHG emissions from sanitation technologies can be highly uncertain (Doorn et al. 2006). The IPCC (Intergovernmental Panel on Climate Change) approach for estimation of GHG emissions is based on the application of emissions factors, i.e. a maximum emission capacity for each treatment technology. Such parameters are always established based on measurements of some kind, although those can be challenging and difficult to validate (Hobson 2002; Johnson et al. 2022). The IPCC differentiates three possible tiers of methods (Eggleston et al. 2006a), which vary according to the availability and quality of data, progressively. The first tier involves using default values in countries with limited data, the second tier involves using default values but with the incorporation of country-specific emission factors and activity data, and the third tier involves using local measurements and specific methods developed at the country level in those contexts with advanced methods and sufficient data.

Despite rising concerns regarding GHG emissions and the sustainability of sanitation systems, estimates of GHG emissions are still mostly focused on large centralized treatment facilities (Orner & Mihelcic 2018; Johnson et al. 2022). There has been relatively less interest in the GHG production dynamics of non-sewered sanitation technologies, and therefore a limited understanding of emissions from these technologies persists. The lack of focus on comprehending and quantifying GHG emissions from non-sewered sanitation systems pervades both low- and middle-income countries, where these systems are prevalent, and high-income countries. In Switzerland and Sweden, GHG emissions from non-sewered sanitation systems are not included in the national GHG inventories due to the assumption that low average temperatures imply negligible production and release of gases like CH₄ (SEPA 2022; Ulrich & Etter 2023). However, rising temperatures due to global warming challenge this assumption, and a recent report found that CH₄ emissions from on-site sanitation systems in Switzerland could be contributing as much as 15% of the country's wastewater-related CH₄ emissions (Ulrich & Etter 2023). Even while quantification of GHG emissions from centralized systems is a common practice, emissions sources and mechanisms connected to biological processes are not yet fully understood, especially given the complexity of such processes and the inadequacy of applicable data (Mannina et al. 2016). Achieving sustainability in the sector will require deepening our knowledge on emissions from the perspective of the whole sanitation chain, as well as from different types of sanitation systems.

A challenge to the dissemination of adequate information is the complexity of the sanitation chain, which by combining a growing number of possible technologies and system configurations, can reach thousands of potential combinations (Spuhler et al. 2020). However, this complexity also means there are a wide range of possibilities for mitigation of GHG emissions and hence opportunities to contribute to more sustainable sanitation systems. It is important to note that sanitation systems consist of more than just the technological choices along the chain, and include institutional and governance arrangements for operating and managing the systems (Ddiba et al. 2020). However, an analysis of governance arrangements and their role in mitigating GHG emissions is beyond the scope of this paper.

Given the complexity and multiplicity of system configurations within the sanitation service chain, we used a narrative literature review approach (Benoot et al. 2016) but integrated some insights from systematic review methodology in our search process (see e.g. Haddaway et al. 2015). To identify literature sources for this review, we searched in Web of Science, Scopus, and Google Scholar for eligible studies based on titles, abstracts, and keywords, enabling us to cover a wide range of sources for scientific and grey literature (see Gusenbauer & Haddaway 2020). The search strings comprised a combination of different words and synonyms to retrieve relevant results, including four central terms as the main components: sanitation, wastewater, GHGs, and emissions. For the first two databases, the resulting articles were exported to reference management software. In Google Scholar, only literature from the first ten pages of results was considered. The retrieved search results were analysed to remove duplicates and non-relevant articles. The sampling process was iterative, with ongoing identification, assessment, and synthesis of relevant information.

The narrative review approach drew on a combination of three purposeful sampling techniques as proposed by Suri (2011): maximum variation sampling, criterion sampling and theoretical sampling. Maximum variation sampling sought to understand the phenomenon, in this case emissions from sanitation services. This step aimed to delineate the key dimensions of the problem and its variations across different contexts. Papers were retrieved considering their broad relationship with the topic, such as describing sanitation-related emissions and without considering further details related to specific study design or approach. A first screening of the results showed that most articles were centred around centralized sanitation systems, especially in the treatment phase. Therefore, focused searches were conducted to retrieve articles on other parts of the sanitation chain, such as conveyance and on-site treatment technologies.

The criterion sampling step involved retrieving articles with information-rich cases, providing further detail into emissions along the sanitation chain. The criteria used at this stage were that each paper should: (i) describe part of the sanitation chain and/or technology, (ii) describe GHG emissions from the sanitation technologies, and (iii) discuss the factors influencing emissions. In the last stage, theoretical sampling was used, and the collected evidence was placed along the sanitation chain. This allowed the identification of knowledge gaps that then led to new focused searches given that there were gaps related to non-sewered sanitation systems and treatment technologies. This step was implemented to ensure a diverse range of studies for mapping of emissions throughout the sanitation chain in line with the study objectives. Additional references were also identified through snowballing and citation tracing (Dixon-Woods et al. 2005). At full text review, evidence was coded and extracted into spreadsheets, including information about type of sanitation system, system configuration, energy consumption, emissions' sources, wastewater characteristics, and system operation.

This section explores how various functional groups within centralized sewer-based sanitation systems have been addressed in the context of GHG emissions accounting.

In the context of sewer-based sanitation systems, the user interface functional group of technologies, such as cistern flush toilets, has been largely overlooked in the assessment of GHG emissions. This oversight may stem from the prevailing belief that these interfaces merely act as points of entry to the broader sanitation network, with the assumption being that waste does not remain within this segment long enough to produce notable emissions. Consequently, any emissions originating from these user interfaces are typically considered to be encompassed within the overall emissions attributed to the sewerage system, specifically under the conveyance category. However, it is essential to recognize the indirect emissions associated with flush toilets, which arise from the processes involved in water extraction, treatment, and distribution for flushing purposes. This aspect, underscored by several studies (see e.g. Reffold et al. 2008; Hackett & Gray 2009; Shimizu et al. 2012) necessitates a more comprehensive evaluation of emissions in this area.

Within the conveyance functional group, sewer networks may be of the simplified type, solids-free type or gravity-driven sewers (Tilley et al. 2014). One of the significant sources of CH₄ emissions in centralized sanitation systems is sewer lines. Improper construction and insufficient maintenance of sewer lines leads to stagnation of wastewater in a closed environment and results in anaerobic conditions which can produce significant amounts of CH₄ (Chaosakul et al. 2014). Foley et al. (2009) estimated an additional 6%–9% increase in the annual GHG inventory for a WWTP due to CH₄ emissions from the sewer network, while Liu et al. (2015) estimate that CH₄ from sewers can contribute up to 18% of the carbon footprint of wastewater management systems.

A review by Liu et al. (2015) indicates that CH₄ in sewers is mainly produced by methanogens from acetate and hydrogen under anaerobic conditions in sewer biofilms and sediments, both in rising and gravity sewers, with gas release occurring in ventilated spaces under turbulence. However, there is substantial spatial and temporal variation in the production of CH₄ within sewers and while there have been some indications of CH₄ sinks in sewers, their potential is not yet well understood or quantified, limiting our understanding of the full extent of CH₄ emissions from sewers. Factors that influence the generation of CH₄ in sewers include the hydraulic retention time, temperature, COD loading and the pipe area-to-volume ratio. Given that several field studies have acknowledged this source of CH₄ emissions (Noyola et al. 2018), it cannot be disregarded in GHG emissions accounting. The 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories (hereinafter, the 2019 refinement to the IPCC guidelines) acknowledges this issue but only includes emission factors for stagnant sewers which are assumed to have anaerobic conditions, while sewers in which wastewater is freely flowing are assumed to have negligible CH₄ emissions.

Some studies have also pointed to the fact that underground gravity sewers are a likely source of N₂O emissions due to nitrification and denitrification processes in biofilms (see e.g. Short et al. 2014; Eijo-Río et al. 2015). Particularly, sewers with bigger head space, i.e. lower wastewater levels, tend to generate higher emissions of N₂O. However, more research is still needed to characterize the spatial and temporal variabilities of N₂O production and release, particularly at hotspots like pumping stations, gas relief valves and other turbulence zones (Short et al. 2014).

Most of the research on fugitive emissions from sanitation systems has been focused on technologies that fall within the centralized treatment functional group (Johnson et al. 2022). Wastewater treatment units are the biggest source of N₂O emissions in centralized systems, particularly in the biological treatment and nutrient removal processes (Lutkin et al. 2022), with even low quantities of N₂O possibly accounting for up to 90% of the total GWP from these systems in some instances (Préndez & Lara-González 2008). This implies that even low emissions of N₂O cannot be disregarded. In that sense, it is necessary to understand the factors behind N₂O production and release to the atmosphere. In studying N₂O production in a laboratory-scale activated sludge process, Schneider et al. (2012) observed how the availability of nitrite and organic carbon had a direct role in denitrification and hence the formation of N₂O. Their analysis indicated that if an overload of nitrate and organic carbon sources can be avoided, substrates for denitrification would become limited and consequently lead to decreased production of N₂O. Brotto et al. (2015) reached similar conclusions regarding nitrite accumulation. Similarly, Adouani et al. (2015) investigated N₂O emissions in a batch reactor and observed that temperature is critical for nitrogen abatement, with low temperatures inducing more emissions of N₂O. They, however, point out the necessity of fully understanding the mechanisms behind such phenomena.

Additionally, per capita protein and water consumption also influence N₂O formation, given that together they affect the properties and volume of wastewater being treated (Ramírez-Melgarejo et al. 2019). The 2019 refinement to the IPCC guidelines includes two supporting equations to determine the total nitrogen in domestic wastewater, which show that N₂O emissions are functions of protein consumption per capita, the fraction of nitrogen in protein, the additional nitrogen from household products, the portion of nitrogen in non-consumed protein that is disposed in sewers via food waste, the portion of nitrogen in industrial and commercial co-discharged protein, the fraction of total nitrogen removed during wastewater treatment and the degree of utilization of the treatment system (Bartram et al. 2019, pp. 6.40, 6.42).

Additionally, it is stressed how the 2019 refinement to the IPCC guidelines properly considers emissions from discharges only when the environment is sufficiently clean and well-oxygenated. In the case of eutrophic or stagnant conditions however, emissions would be significantly higher than those estimated. Conversely, not only biological processes will influence emissions, but so also will operational conditions. Rodriguez-Caballero et al. (2014) analysed data from aerated and non-aerated zones in WWTP with activated sludge and concluded that process perturbations such as highly irregular aeration flow and nitrification instability can affect N₂O emission patterns, which provides more evidence to the hypothesis of N₂O formation being strongly related to process design and operation as stated by the IPCC (Bartram et al. 2019).

The potential for CH₄ production from WWTPs is highly dependent on the chosen treatment technology, and its release to the atmosphere is related to the place of its occurrence in the treatment process and where it is stripped out to the surface (Préndez & Lara-González 2008). Rodriguez-Caballero et al. (2014) observed the release of dissolved CH₄ during aeration of inflow and reject wastewater entering the bioreactor. Préndez & Lara-González (2008) estimated that sludge digestion could account for up to 98% of the emissions in WWTP. Similarly, direct measurements of CH₄ emissions from anaerobic sludge digestion by Daelman et al. (2012) indicated that the emissions could correspond to three-quarters of the total CH₄ produced by WWTP. In addition, Daelman and colleagues bring attention to how digested sludge still has a high potential for producing residual CH₄ during storage, a factor that has been broadly overlooked.

Parravicini et al. (2016) and Daelman et al. (2012) observed how certain conditions can make anaerobic digestors and anaerobic sludge storage tanks significant sources of CH₄ and why controlling such emissions' sources is paramount. Measurement of GHG emissions at a Swedish wastewater treatment plant serving 145,000 people indicated large quantities of CH₄ emissions from the storage piles of digested sludge (81,500 ± 3,800 kg CH₄/year) and that it was the largest source of CH₄ emissions at the entire plant (Gålfalk et al. 2022). Rodriguez-Caballero et al. (2014) indicated the possibility of converting CH₄ before aeration, while Daelman et al. (2012) observed the potential to aerobically oxidize CH₄ in the activated sludge tanks, both aiming at decreasing emissions. Therefore, it becomes clear how CH₄ emissions are dependent both on operational conditions and the system's configuration. As an example, when comparing anaerobic and aerobic treatment options for sewage sludge, Parravicini et al. (2016) estimated that the carbon footprint of an activated sludge WWTP using anaerobic digestion for sludge treatment is close to those using simultaneous aerobic stabilization of sewage sludge (SAS), resulting in a similar total nitrogen removal while CH₄ emissions from sludge stabilization are avoided.

Furthermore, aerators and the circulating flow of dissolved gases in wastewater treatment processes have a great impact on GHG emissions. The maximum releases of CO₂, N₂O and CH₄ happen in aerobic areas independent of the treatment technologies used (Yan et al. 2014; Liu et al. 2015). This indicates that dissolved gases in the wastewater together with turbulent processes play an important role in stripping out GHG to the surface. Parravicini et al. (2016) noticed that activated sludge tanks in conventional treatment systems dominate the carbon footprint of WWTPs with moderate nitrogen removal, mainly due to direct N₂O emissions.

Turning attention to the end-use and/or disposal functional group, sewer-based wastewater management systems yield two main products: effluent and sludge. Emissions that occur during effluent discharge into receiving water bodies are important to consider for accurate modelling of the complex processes governing GHG release into the atmosphere (Koutsou et al. 2018). The quality of effluent, local environmental conditions like temperature variations and ambient pressure, and various other factors intricately influence these emissions, contributing to uncertainties in their quantification. For example, the discharge of wastewater increases nutrient concentrations in recipient waters, leading to eutrophication which enhances the production and release of N₂O. The quantities of these emissions fluctuate significantly between summer and winter seasons (Masuda et al. 2018). The IPCC (Bartram et al. 2019; Hergoualc'h et al. 2019) provides default emission factors for N₂O and CH₄ emissions, as well as methane correction factors (MCF), differentiated by effluent discharge pathways such as discharge to aquatic environments and discharge to soil. The concentration of organic matter and total nitrogen in wastewater effluent significantly influences CH₄ and N₂O emissions, respectively. Notably, nitrogen, mainly concentrated in the effluent after treatment, contributes to N₂O emissions when applied to land through irrigation or groundwater recharge, where denitrification of nitrate can release N₂O (Rivett et al. 2008).

Emissions from the discharge of untreated wastewater are also significant sources of CH₄ with some studies estimating that they could be as high as three times the emissions from WWTPs (Giné-Garriga et al. 2022). A recent study based on remote-sensing data indicated that emissions from untreated wastewater in urban areas could range from 3 to 7 kg CH₄/m³ of untreated wastewater (de Foy et al. 2023). However, there are some uncertainties around these estimates, the processes for CH₄ formation in post-discharge wastewater and about how increasing the proportion of wastewater that is treated could influence the overall emissions from the sanitation sector. Miller-Robbie et al. (2017) estimated that there would be 32% less emissions due to the implementation of centralized systems and reduction of untreated wastewater in recipient streams in growing cities. In contrast, Singh et al. (2017) estimated that the construction of WWTPs for the wastewater that is currently not treated in India could increase GHG emissions by up to 269%, depending on which treatment technologies are deployed.

With regards to sludge, studies indicate that CH₄ emissions from agricultural land after applying excreta-derived fertilizers e.g. sewage sludge pellets, composted sewage sludge and digested sludge, are negligible (Ball et al. 2004; Jones et al. 2006). Most of the GHG emissions when these kinds of excreta-derived fertilizers are applied to land are a result of the release of N₂O, especially in warmer conditions and soils with poor drainage (Brown et al. 2010). Excreta-derived fertilizers, when applied to land, also contribute to the release of carbon dioxide (CO₂), but this is considered to be of biogenic origin and hence does not contribute to global warming. However, energy-related emissions can be generated due to the use of vehicles and other machinery when applying the fertilizers.

Field experiments in Canada indicated that the average N₂O emissions over a two-year period were much higher for soils where digested sludge (biosolids) was applied than for soils where composted sludge or alkaline-stabilized sludge was applied. In one year, for example, 6.3% of the total N applied in digested sludge was emitted as N₂O, as opposed to 0.24% for alkaline-stabilized sludge and 0.17% for composted sludge (Obi-Njoku et al. 2022). For comparison, the N₂O emissions factor for all kinds of sludge, wastewater and animal manure in the IPCC guidelines (Hergoualc'h et al. 2019) is 1%, while a review by Charles et al. (2017) recommends an emissions factor of 1.21% ± 0.14% for the same three amendments. However, these N₂O emissions are highly influenced by other factors like the C:N ratio, moisture content, soil texture and climatic factors like precipitation (Charles et al. 2017).

In some cases, surface disposal or storage is used as an option for sludge management but this is associated with significant CH₄ emissions, mainly due to anaerobic conditions created in the pile of material when it is disposed of (Roy et al. 2011; Majumder et al. 2014; Bora et al. 2020). Significant CH₄ emissions are generated when sewage sludge is disposed of at landfills, but also when it is placed in temporary storage, as found in a recent study in Sweden where sludge storage at a WWTP was found to be the single biggest source of CH₄ emissions (Gålfalk et al. 2022). In another study, measurements at landfill sites in Sweden where sludge was disposed of indicated that there were significantly higher N₂O emissions compared with those where sludge was not disposed of (Börjesson & Svensson 1997). This also suggests that landfilling of excreta-derived waste streams can also be a significant source of N₂O emissions, and not only CH₄.

Sludge from WWTPs can also be used to generate energy through biogas production and via incineration. Incinerating sewage sludge contributes to significant emissions of N₂O, as well as CH₄ in some instances (Chen & Kuo 2016; Piippo et al. 2018). Biogas production through anaerobic digestion and the use of the biogas can also contribute to GHG emissions through, for example, CH₄ leakages and incomplete combustion (Brown et al. 2010). Further details about anaerobic digestion are described in Section 3.1.4.

Falk et al. (2013) evaluated five different levels of wastewater treatment according to treatment objectives with technologies ranging from conventional activated sludge to more advanced tertiary biological and nutrient removal processes, and concluded that GHG emissions linearly increase together with the number of additional facilities and chemical demand. Besides that, if nitrogen/phosphorus removal processes are included, then emissions increase exponentially. Similarly, Mamais et al. (2015) concluded that while indirect emissions increase together with the treatment capacity of WWTPs, direct emissions decrease. However, this does not mean that fugitive emissions are not significant in large-scale treatment plants. Emissions related to energy and chemical use, for example, can increase at a faster rate depending on the system's design, such as features related to removal of pollutants, and the volume of influent wastewater (Gémar et al. 2018).

For treatment processes that include constructed wetlands, Laitinen et al. (2017) highlighted the complexity of the processes governing the exchange of gases with the atmosphere, such as the chosen features of the wetland, the vegetation implemented, seasonal variations, and regional characteristics. The natural cycles of a wetland, for example, nutrient recycling processes, act as a biomass sink (Machado et al. 2007), possibly avoiding GHG emissions by up to 20% (Laitinen et al. 2017). Indeed, comparing activated sludge processes, sequencing batch reactors, up-flow anaerobic sludge blanket reactors, and constructed wetlands in an Indian context, Kalbar et al. (2013) observed that constructed wetlands have the lowest overall environmental footprint and even have the potential to mitigate GHG emissions naturally by sequestering atmospheric carbon. Regarding energy consumption, chemical use, and investment costs, constructed wetlands are two-fold better and can be a viable alternative to conventional WWTP (Laitinen et al. 2017). Similarly, Singh et al. (2017) indicated that oxidation ponds have lower GHG emissions than many other WWTPs since they require almost no energy input and have low operation and maintenance requirements. A conventional WWTP has an environmental impact between two-and five-fold higher than a nature-based solution (Garfí et al. 2017).

Life-cycle assessments (LCAs) frequently quantify indirect emissions, encompassing sources within the construction and operation phases of wastewater treatment systems. It has been indicated that a significant quantity of GHG emissions result from energy consumption in the operational phase of centralized sewer-based sanitation systems. For example, Singh & Kansal (2018) analysed 35 WWTPs in Delhi and observed that operating the infrastructure accounted for 70% of the total energy demand for wastewater treatment.

Efficiency in the sanitation system also plays an important role in determining total GHG emissions. From the analysis of 1,079 WWTPs in China, Zeng et al. (2017) concluded that, depending on specific configurations and operational conditions, emissions could be mitigated by up to 32% if all plants were fully efficient. Large-scale plants often outperform smaller ones in efficiency, particularly those employing bioreactors or anaerobic–anoxic processes, though tertiary treatment steps may lower efficiency. Similarly, Molinos-Senante et al. (2014) analysed 60 WWTPs in Spain and found that maintaining WWTPs in Spain at full capacity and optimal conditions could directly or indirectly mitigate GHG emissions, potentially reducing them by 30%. In contrast, if a WWTP does not work properly, it can result in a higher carbon footprint per unit volume of wastewater treated (Singh & Kansal 2018).

In the 2019 refinement to the IPCC guidelines, centralized aerobic WWTPs are recognized as a possible source of CH₄ given the formation of anaerobic pockets in poorly designed or managed facilities (Bartram et al. 2019). However, Noyola et al. (2018) assert that even well-managed treatment facilities are actual, not just potential, sources of CH₄ emissions. Noyola et al. (2018) do not support the assumption of CH₄-neutral plants, and instead propose a standard MCF of 0.06, focusing on intertropical countries. Studying two facilities in Mexico, they also discuss how sewers in such climate zones possibly produce higher concentrations of CH₄. The 2019 refinement to the IPCC guidelines incorporated new MCF to account for such emissions (Bartram et al. 2019). Although climatic factors significantly affect GHG emissions in sanitation systems, studies seldom include them as contributing factors in total emissions analyses, particularly in low- and middle-income countries. This has been pointed out by several authors such as Chaosakul et al. (2014) with regards to low- and middle-income countries within the tropics.

Noyola et al. (2018) further observed large amounts of CH₄ emissions from two municipal Mexican WWTPs, indicating their poor operational conditions. However, by assuming full WWTP efficiency, national inventories in Mexico likely underestimate CH₄ emissions, a situation that may be common worldwide. A similar picture can be observed for N₂O accounting. Brotto et al. (2015) reported higher emission factors for full-scale activated sludge plants in Brazil than recent values observed in international inventories. The deviation is linked to variations in temperature. Furthermore, malfunctioning plants were linked to higher N₂O emissions, underscoring the need for improved efficiency as a means to control GHG emissions in similar contexts.

However, some studies have indicated that GHG emissions from the operation of wastewater treatment facilities (i.e. electricity and fuels to run the operations and equipment) could be much lower than the direct emissions of CH₄ and N₂O from the treatment processes in some cases. In Kampala for example, the direct emissions from wastewater treatment were estimated at 29,629 tonnes CO₂-eq annually, while operational emissions were only 2,950 tonnes CO₂-eq annually (Johnson et al. 2022). The ratio of operational to direct fugitive GHG emissions from wastewater treatment hinges on the treatment technologies' energy intensity, the required chemical inputs, and the carbon footprint of the local or national energy mix.

Although CH₄ emissions contribute significantly to higher GWP in centralized WWTPs, biogas production and use at WWTPs often meet up to 40%–60% of energy requirements (Daelman et al. 2013; Singh & Kansal 2018). Biogas production and use has been found to possibly reduce GHG emissions by around 5% to 12% according to Singh & Kansal (2018) and Lahmouri et al. (2019), respectively. Daelman et al. (2012) also estimated that additional biogas valorization from residual CH₄ in storage tanks could further decrease CH₄ emissions from a WWTP by around 22% to 48%.

However, leakages of CH₄ from anaerobic digestion and biogas valorization processes can significantly diminish the GHG reduction potential of biogas-based energy recovery (Daelman et al. 2013). For example, Koutsou et al. (2018) analysed data from 128 WWTPs in Greece and concluded that biogas use was the biggest contributor to CH₄ emissions, after sludge disposal to landfills. A study using ground-based remote-sensing methods identified previously unknown leakages of large quantities of CH₄ from biogas production facilities at a wastewater treatment plant in Sweden (Gålfalk et al. 2022). According to Noyola et al. (2018), around 8% of the captured CH₄ from anaerobic sludge digesters at WWTPs can end up in the atmosphere due to leakages. Therefore, it can be observed that while biogas production at a WWTP can reduce the need for external energy sources significantly, fulfilling 40%–60% of the internal demand (Daelman et al. 2013; Singh & Kansal 2018), in some circumstances it offers no benefits regarding GHG mitigation due to methane leakages (Daelman et al. 2012).

Non-sewered sanitation systems include on-site sanitation technologies like pit latrines, septic tanks, CBS, composting toilets etc, as well as the infrastructure for managing and treating the waste streams emanating from these technologies. They are mostly used in low- and middle-income countries, but significant portions of the population in high-income countries also use non-sewered sanitation systems including 12% to 15% of the population in some Nordic countries (Laukka et al. 2022) and a fifth of the population in the USA and the European Economic Area in general (Somlai et al. 2019). In some areas of the world where open defecation is still prevalent, non-sewered sanitation systems are seen as an ideal intervention. While open defecation has minimal contribution to GHG emissions compared with improved sanitation solutions, its elimination is a necessity for human dignity, public health and well-being (Shaw et al. 2021).

Field-based measurements of GHG emissions from non-sewered sanitation remain relatively sparse (Ryals et al. 2019; Poudel et al. 2023). This is despite the fact that some studies have indicated that non-sewered sanitation systems might have relatively higher contributions to GHG emissions, for example in the USA where septic tanks are estimated to contribute to about 65% of GHG emissions from sanitation systems even though they account for only a quarter of the wastewater treatment capacity in the country (US EPA 2012).

Just like the case of sewer-based sanitation systems, the user interface functional group has not received much attention in the literature on GHG emissions accounting within non-sewered sanitation technologies, except for emissions related to the use of water for toilet flushing where applicable (as in Section 3.1.1.1).

Most research on GHG emissions from non-sewered sanitation technologies focuses on the containment and storage functional group, i.e. pit latrines, septic tanks, and CBS, as described in Sections 3.2.2–3.2.4. There are other technologies within this functional group such as composting toilets, dehydration vaults and biogas toilets (see Tilley et al. 2014, pp. 56–81), but there seem to be no studies quantifying their GHG emissions. This is also reflected in the IPCC guidelines, which only include emission factors for septic tanks and four pit-latrine types (Doorn et al. 2006).

With regards to the conveyance functional group, no literature was found with data on fugitive emissions from the transportation of faecal sludge or other excreta-derived waste streams within jerrycans or containers, as well as manual or motorized emptying and transport options, although the motorized options come with indirect transport-related emissions. Recent studies based on modelling have considered these emissions to be negligible due to the relatively short time that faecal sludge spends during transportation processes in comparison with the containment and treatment processes (Johnson et al. 2022). For transfer stations, the emissions could likely be similar to those from containment or storage depending on the presence of anaerobic conditions and the residence time of the excreta under those conditions.

With regards to the treatment functional group, there are some overlaps between non-sewered sanitation systems and sewer-based sanitation systems as far as emissions sources are concerned. In many instances where faecal sludge in pit latrines and septic tanks is not emptied, then treatment occurs in the containment and storage phase in which case the emissions are as described in Sections 3.2.2–3.2.4. In some instances however, faecal sludge is emptied and transported to decentralized or centralized facilities for treatment as shown in Figure 1 – System C (see also Strande et al. 2014). Depending on the technology configuration and environmental conditions, the GHG emissions from these facilities can be governed by similar mechanisms to those in treatment technologies within sewer-based sanitation systems, with anaerobic conditions facilitating CH₄ emissions while biological treatment and nutrient removal processes facilitate N₂O emissions (see Section 3.1.1.3).

Non-sewered sanitation systems produce liquid products, including effluent, urine, and digestate, as well as semi-solid products like compost, sludge and pit humus. The mechanisms influencing GHG emissions from effluent and digestate disposal in non-sewered sanitation systems are similar to those in centralized sewer-based systems (see Section 3.1.1.4). This also includes instances where sludge from pit latrines or septic tanks is released untreated into urban drains or other surface waters. In the USA, studies suggest that leach fields and drainage fields for septic tank effluent are significant sources of N₂O emissions (Leverenz et al. 2010; Truhlar et al. 2016), due to the nitrification and denitrification processes that occur in the soil where the effluent is dispersed. In contrast, non-sewered sanitation technologies that dispose of effluent in soak pits mainly emit CO₂ and CH₄, as observed in Irish field investigations (Somlai-Haase et al. 2017; Somlai et al. 2019), likely due to the presence of anaerobic conditions in the soak pit. The CO₂ is deemed to be biogenic.

With regards to urine, negligible amounts of CH₄ are released when urine is applied to soil (Tidåker et al. 2007). There are relatively few experimental studies about N₂O emissions from fields where urine or urine-derived fertilizers have been applied (Martin et al. 2020). An incubation study in Denmark indicated that N₂O emissions were lower for urine-applied soil than from that where mineral fertilizers had been applied (Gómez-Muñoz et al. 2017), while a pot experiment in Germany indicated higher N₂O emission rates for urine than for mineral fertilizers (Simons 2008). LCA studies involving the application of stored urine and urine-derived fertilizers on agricultural land indicate that the majority of the GWP impact arises from N₂O emissions after spreading on land (Spångberg et al. 2014; Medeiros et al. 2020; Martin et al. 2023). These N₂O emissions are partly resulting from a portion of the total nitrogen added to soil that is later emitted into air, and partly from ammonia-nitrogen that is emitted into air through volatilization (Eggleston et al. 2006b). Mineral fertilizers, when applied to agricultural land, also release large amounts of N₂O due to the action of soil micro-organisms breaking down excess nitrogen that is not absorbed by plants. Unlike urine-derived or organic fertilizers however, mineral fertilizers also have significant emissions released from their extraction, production and distribution chains (Walling & Vaneeckhaute 2020).

Solid or semi-solid products like compost, dehydrated faeces, sludge, and insect frass, applied to arable land, have GHG emissions mechanisms similar to sewage sludge from centralized sewer systems (see Section 3.1.1.4). Incubation experiments with black soldier fly frass as an organic amendment to arable soil indicate significant N₂O and CH₄ emissions, especially immediately after application of the amendment (Rummel et al. 2021). However, more extensive field-based studies are yet to be done to determine the extent of emissions from BSF frass resulting from larvae fed on human excreta-derived feedstock.

Pit latrines are used by approximately 1.6 billion people globally (Cheng et al. 2022). Lutkin et al. (2022) estimated that pit latrines globally contribute about 42 million tonnes of CO₂-eq/year while Reid et al. (2014) estimated that they are responsible for about 3.8 million tonnes of CH₄ emissions annually. Pit latrines can also be significant sources of N₂O emissions, depending on the presence of nitrate to facilitate denitrification (Doorn & Liles 1999; Poudel et al. 2023). Pit latrine design and construction typically promote anaerobic degradation of excreta, facilitating GHG production in the pit. Analysing samples from 45 different pits in Tanzania, van Eekert et al. (2019) identified anaerobic digestion as the main pathway for organic degradation in pit latrines, with biogas production being observed in 73% of the pits within 40 days, although Bourgault et al. (2019) contend that there is a significant extent of aerobic decomposition. Van Eekert et al. (2019) also identified higher moisture content as a crucial factor for increasing anaerobic digestion and hence CH₄ production. This indicates the influence of the characteristics of the faecal sludge in pit latrines, which is determined by factors like whether greywater is also disposed of in the pit. Moreover, it aligns with modelling results from Trimmer et al. (2020) which suggest that fugitive emissions from pit latrines with 50% anaerobic conditions can contribute to between 6% and 22% of the total per capita emissions in Kampala. If the latrines are further submerged due to water table rise, direct emissions may increase by 58% (Trimmer et al. 2020).

Besides aeration status, temperature variations and the number of users are important factors in estimating pit latrine GHG emissions. Trimmer et al. (2020) conducted a detailed analysis of how single-household facilities can reduce CH₄ emissions through improved management practices. They also estimated that the widespread construction of pit latrines for households without access to improved sanitation worldwide could lead to an increase in emissions of up to 34%. The trade-offs between improving access to basic sanitation and reducing GHG emissions raise concerns regarding the use of pit latrines to close the sanitation gap instead of employing other technologies. This is because a drastic increase in the number and use of pit latrines would more than double the GHG emissions contributions from this source, currently estimated around 0.3% of the global emission according to the experimental measurements of van Eekert et al. (2019). Based on these considerations, a crucial challenge in climate mitigation within the sanitation chain is addressing the significant source of CH₄ emissions represented by pit latrines. However, there is potential for reducing GHG emissions from pit latrines by promptly and routinely emptying the pits (Johnson et al. 2022), and directing the collected sludge to anaerobic treatment facilities (Trimmer et al. 2020).

About 1.7 billion people use septic tanks globally (Cheng et al. 2022). This prevalence is primarily in low- and middle-income countries, though significant usage is also observed in high-income countries, including about 20% of the USA population (Diaz-Valbuena et al. 2011). Estimates by Lutkin et al. (2022) suggest that septic tanks globally contribute about 210 million tonnes of CO₂-eq annually.

GHG emission rates from septic tanks are influenced by their design and the components to which they are connected. Diaz-Valbuena et al. (2011) measured GHG emissions from different compartments of septic tanks in the USA and found that they varied significantly, suggesting that the number of compartments affects the overall emissions. Moreover, the discharge of septic tank effluent to leach fields, soak pits, or sewers affects emissions, as varying biogeochemical processes occur in the soil or water. Leverenz et al. (2010) and Truhlar et al. (2016) showed that nitrification and denitrification in the soil after effluent dispersal contribute to N₂O and CO₂ emissions. Huynh et al. (2021) studied septic tanks in Hanoi, Vietnam, that discharged their effluent to the sewerage network, which then released most of it to water bodies. While direct discharge emissions were not measured, other studies have documented high N₂O emissions from wastewater effluent entering surface water bodies (Kampschreur et al. 2009; Wang et al. 2022).

With regards to emissions sources from septic tanks, Leverenz et al. (2010) conducted experiments on eight septic tank systems in the USA and indicated that CH₄ emissions come mainly from the septic tank while CO₂ and N₂O emissions originate mainly from the soil dispersal system due to nitrification and denitrification of the effluent as it disperses. This is in agreement with Truhlar et al. (2016), whose experiments on septic tanks in New York indicated that N₂O emissions mainly originated from the leach field and the roof vent, as well as with Huynh et al. (2021), whose measurements on septic tanks in Vietnam indicated that N₂O emissions from the first compartment of the septic tank were negligible compared with CH₄ and CO₂ emissions. In septic tanks, CH₄ emissions arise from anaerobic reactions in the sludge layer, whereas CO₂ emissions stem from anaerobic, facultative, and aerobic reactions within the tank. Higher temperatures and influent loading rates tended to result in higher CH₄ emissions from septic tanks. However, no correlation was found between the thickness of scum layers in septic tanks and the rates of CH₄ emissions according to Diaz-Valbuena et al. (2011). Leverenz et al. (2010) also indicated that septic tank installations linked to soft water supply tend to result in higher CO₂ fluxes. In contrast to the IPCC, which indicates an MCF of 0.5, Diaz-Valbuena et al. (2011) suggest an MCF of 0.22 based on results from their field measurements.

Characterization of waste is another important factor that influences GHG emissions from septic tanks. If influent wastewater to septic tanks contains sulphate compounds, then hydrogen sulphide can also be released in septic tanks, according to Leverenz et al. (2010). Diaz-Valbuena et al. (2011) suggest that other factors such as COD loading and their variability may have more influence over CH₄ emission rates in septic tanks compared with the temperature of the influent wastewater. Septic tanks in some areas in low- and middle-income countries receive only black water since greywater is commonly discharged into open drains. This implies that the influent wastewater has lower organic loading rates, compared with septic tanks in high-income countries which receive black water, greywater and even food waste from macerators in sinks. Considering that anaerobic processes primarily decompose organic pollutants in septic tanks, the variations in influent wastewater composition may account for the observed differences in GHG emissions across studies in the USA (Leverenz et al. 2010; Diaz-Valbuena et al. 2011; Truhlar et al. 2016) and Ireland (Somlai et al. 2019) in comparison with Vietnam (Huynh et al. 2021).

How septic tanks are operated throughout the seasons also plays a significant role in determining GHG emissions, given that microbial activity in a septic tank varies over the seasons, especially in places with significant seasonal temperature variations. During the spring and early summer, increasing temperatures result in an increase in microbial activity in the sludge layer of septic tanks, which leads to more gas production, according to Leverenz et al. (2010). The solubility of the gases also decreases and hence results in higher gas release. Furthermore, septic tanks that have been de-sludged tend to have an initiation period before production of CH₄ and CO₂ fully resumes, according to Diaz-Valbuena et al. (2011). Huynh et al. (2021) and Moonkawin et al. (2023) found that septic tanks with longer de-sludging intervals had higher CH₄ emission rates, which is consistent with the findings of Johnson et al. (2022) in Kampala. They also expected that the tropical temperatures in Vietnam would increase the emissions from septic tanks compared with colder climates in high-income countries of the Global North. However, they did not observe a significant influence of influent temperature on emission rates, which is similar to the results of Leverenz et al. (2010) in the USA. They suggested that this could be due to the low variability of influent temperature in their study. Furthermore, they reported that the septic tanks had low treatment efficiency and produced effluent with high concentrations of organic pollutants, which could lead to more GHG emissions in the post-treatment components of the septic tank system. This demonstrates the importance of effective treatment and the influence that it can have on GHG emissions from septic tanks. Addressing the often-ineffective performance of on-site treatment technologies in many low- and middle-income countries, as discussed by Strande et al. (2014), is therefore essential for climate mitigation efforts.

CBS systems are made of sealed, waterless, and portable containers that capture human excreta. Once filled, they are replaced with clean containers, and the filled ones are taken for treatment, such as thermophilic composting, and resource recovery (Russel et al. 2019). CBS is considered as an integrated solution for meeting the sanitation gap in some contexts of low- and middle-income countries. In CBS, human excreta stays on-site for a limited period and in a confined impermeable container. Therefore, fugitive emissions are reduced, nutrient leaching in soil is avoided, and the potential for nutrient recovery is increased (Trimmer et al. 2020). In contrast, using a large container for a longer period could favour anaerobic conditions and consequently increase GHG emissions (Johnson et al. 2022).

Moreover, off-site composting of excreta also emits GHGs where the intensity depends on operational decisions, such as turning of compost piles and the permeability of the pile lining-material. Research by Ryals et al. (2019) in Haiti indicated a three-fold reduction of N₂O and a four-fold reduction of CH₄ emissions when using permeable soil lining instead of cement lining for compost piles of excreta from CBS. Conversely, pile-turning doubled NO₂ and CO₂ emissions while nearly eliminating CH₄ production (Ryals et al. 2019). McNicol et al. (2020) observed that CBS along with composting of the excreta presents lower CH₄ and higher NO₂ emissions overall than systems such as pit latrines, and considerably smaller GWP than other sanitation solutions typically used in the Global South.

Considering the full cycle of sanitation services, CBS along with off-site composting have the potential to mitigate up to 92% of the sanitation-related emissions for the global population living in slums, mainly by reducing up to 99% of CH₄ emissions compared with the emission levels from typical pit latrines and similar sanitation technologies used by this demographic (McNicol et al. 2020). Consequently, implementing CBS alone for 1 billion slum residents could mitigate between 13% and 44% of the CH₄ emissions from the global sanitation sector. However, uncertainties remain due to the paucity of emissions data from composting excreta in diverse geographical contexts, despite composting's reliance on aerobic processes that primarily produce climate-neutral CO₂ emissions (McNicol et al. 2020). Also, few studies have quantified the flows and transformation of organic compounds within non-sewered technologies (Lourenço & Nunes 2020). This raises questions about the actual fraction of possible non-biogenic carbon in these non-sewered sanitation systems, and how this could be accounted for in CBS more specifically.

Furthermore, not all CBS toilets currently include urine collection due to higher costs of conveyance and treatment (Russel et al. 2019). This complexity affects how emissions from urine management, disposal, and waste transportation might negate anticipated mitigation benefits (Reid 2020). Transportation and conveyance are usually a significant source of indirect emissions from non-sewered sanitation in some contexts (Lehtoranta et al. 2013), even though they may be relatively low in comparison with fugitive emissions from faecal sludge management systems (Johnson et al. 2022). Anastasopoulou et al. (2018) compared sanitation systems in South Africa and concluded that for both urine-diverting dry toilets (UDDTs) and pour-flush toilets, transportation was among the biggest contributors to the overall climate impact of these technologies.

To mitigate the climate impacts of sanitation systems, non-sewered sanitation systems that are dry and controlled offer a promising alternative to conventional centralized systems that consume more energy and resources (Lourenço & Nunes 2020). Several studies have compared different scenarios of sanitation systems and their contributions to GHG emissions and resource recovery. For instance, Prado et al. (2020) found that centralized systems had higher overall climate impacts due to their higher energy consumption. Zhou et al. (2010), compared scenarios from traditional centralized treatment to decentralized ecological sanitation (EcoSan) systems, noting that EcoSan systems emit fewer GHGs due to lower gasoline and electricity consumption. Xue et al. (2016) demonstrated that dry composting toilets when used along with septic tanks for greywater treatment were less energy-intensive than centralized sewer-based systems.

The context and configuration of the system also influences the energy consumption and GHG emissions. Remy & Jekel (2008), noted that, in small urban settlements in Germany, source-separating systems with composting might require similar or greater energy due to the construction of parallel pipe networks and associated greywater treatment, compared with conventional combined-flow sewer networks with centralized treatment. However, they could also reduce GHG emissions by 15% to 30% compared with activated sludge treatment. Fan et al. (2017) observed that waterless urinals and vacuum source-separation toilets had 73% lower climate impact than conventional systems in China, mainly due to their capacity for resource recovery. Additionally, by analysing four different scenarios of sanitation systems in Tanzania, Krause & Rotter (2017) concluded that by implementing UDDTs, emissions could be significantly reduced in comparison with those from pit latrines and septic tanks, while reaching higher levels of carbon, nitrogen, and phosphorus recovery. UDDT systems had up to 55% lower GHG emissions than pit latrines and even lower than septic tanks. Pit latrines had higher carbon recovery potential than septic tanks, but this was not realized as sludge was often left in the ground. Therefore, dry controlled systems can present lower GHG production at different stages of the sanitation chain by being less resource-intensive or promoting higher resource recovery (Reid et al. 2014; Lourenço & Nunes 2020).

While most studies identified in this review have predominantly emphasized sewer-based sanitation and WWTPs, non-sewered sanitation systems exhibit a paucity of empirical investigations. While some studies have focused on pit latrines, septic tanks, and CBS, a broader understanding of other non-sewered technologies like composting toilets, biogas toilets, and fossa alternas, their GHG footprint, and emissions variability across geographical contexts is crucial. Knowledge gaps about these other types of non-sewered technologies impedes the effective translation of findings from experimental studies into comprehensive model-based studies, as regional variations in environmental factors such as climatic conditions and seasonal temperature variations, water-table depth and soil drainage properties are possibly overlooked. As a corollary, the emphasis on WWTPs within experimental studies engenders model-based studies that are skewed towards these systems, leaving a critical void in the representation of emissions from non-sewered sanitation across diverse locations. There are ongoing efforts through research projects like SCARE (University of Bristol 2023), but these are still insufficient to cover the wide variety of non-sewered sanitation technologies and systems spread across the globe. Given the extensive use of non-sewered sanitation globally, significant investment in research to address these gaps is imperative for a robust empirical foundation in emissions modelling. Innovative approaches to indirectly measure emissions, such as assessment of seasonal changes in groundwater to account for the changing inundation of pit latrines and associated CH₄ emissions, can also play a role in low-income regions where there are limited resources for monitoring, but still rely on empirical data for validation (Reddy et al. 2022).

Moreover, the methods for empirically quantifying GHG emissions necessitate more nuanced scrutiny. A multitude of methodologies exist, ranging from static chambers for localized measurements (see e.g. Reddy et al. 2022) to remote-sensing techniques for capturing emissions over expansive regions (Bastviken et al. 2022). Recent studies by, for example, de Foy et al. (2023) and Gålfalk et al. (2022) have unveiled potential disparities in results stemming from different methods, underlining the imperative of methodological rigour. The interplay between empirical measurements and model-based inferences highlights the need for precision in emissions quantification and a critical examination of methodological variations. However, there is also a need for empirical methods that are suitable for low-income contexts, where conventional methods that rely on access to laboratory facilities may be limited (Poudel et al. 2023).

Our findings indicate that prior research often focuses on specific technologies, lacking a comprehensive view of the entire sanitation chain. Even ground-breaking studies such as the city-wide assessment of sanitation emissions in Kampala (Johnson et al. 2022) exclude crucial aspects like resource recovery or end-disposal. Spuhler et al. (2020) underscored that the environmental ramifications of sanitation systems are intrinsically linked to the configuration and synergistic amalgamation of diverse technologies. This implies that GHG emissions from sanitation systems are also influenced by the configuration of the technology combinations across the entire sanitation chain, and hence a whole-system perspective is necessary to achieve a good understanding of the carbon footprint thereof.

Exemplifying this perspective, comprehensive assessments of GHG emissions have been undertaken for solid waste management systems as part of LCA studies, covering everything from waste collection bins at household level through transportation and treatment up to end-use of resource recovery products or disposal of the waste residues (see e.g. Laurent et al. 2014a, 2014b). The integration of this approach within the sanitation discourse is not only feasible but also essential. A comprehensive understanding of emissions necessitates the monitoring of emissions all the way from user interface through treatment to end-use and disposal. Moreover, the diverse juxtaposition of non-sewered sanitation alongside sewer-based centralized systems in heterogeneous infrastructure configurations (Lawhon et al. 2018) calls for an integrative approach that considers the multiplicity of sanitation configurations. Furthermore, a comprehensive emissions assessment must encompass resource recovery, recognizing the varying mitigation potential associated with different recovery options.

Although sanitation contributes only 1.3% of global emissions (Ritchie et al. 2020), recent research unearths a contextual variance. For instance, findings from Kampala indicate that the sanitation sector's emissions could possibly account for about 50% of the city's total emissions (Johnson et al. 2022). This creates an imperative to explore avenues for reducing the carbon footprint of sanitation systems and hence contribute to climate mitigation efforts. Our findings in this review point to three avenues to explore for the development of mitigation strategies in the sanitation sector, as described in the following section.

The strategic role of technology choices:Zeng et al. (2017) underscores the efficiency of bioreactors and anaerobic–anoxic processes in WWTPs, indicating that targeted technology choices can yield substantial emissions reductions. The implications span across various sanitation and wastewater management processes, from CH₄ emissions in technologies that foster anaerobic conditions to N₂O emissions in aeration-based processes, and constructed wetlands which may act as carbon sinks despite their large land-footprints. With regards to non-sewered sanitation, studies have highlighted how efforts to increase access to sanitation through massive construction of pit latrines could impede climate mitigation efforts (Reid et al. 2014; Shaw et al. 2021; Cheng et al. 2022), hence pointing to the need to consider alternative lower emission technology options. At the end of the sanitation chain, the various options for resource recovery have different emissions profiles, thus advocating the need for careful technology selection.

Operational conditions and continuous monitoring: As highlighted in our findings, studies by, for example, Molinos-Senante et al. (2014) and Noyola et al. (2018) indicate a correlation between efficient WWTP operations and reduced GHG emissions, while the influence of operational conditions on sewer systems' emissions underscores the significance of this aspect. In the realm of non-sewered sanitation, operational optimization is crucial under varying environmental conditions such as temperature and water-table depth. This approach is particularly emphasized in maintaining aerobic conditions within systems, for instance, through frequent emptying so as to limit CH₄ emissions. This is particularly important in contexts where climate change is worsening flooding or water-logging from sea-level rise. The dynamic nature of sanitation systems also underscores the indispensability of continuous real-time monitoring to identify emergent emissions hotspots, hence the need for comprehensive data collection.

Resource recovery: Recovering and reusing resources like water, energy and nutrients which are embedded in excreta-derived waste streams in sanitation systems mainly contributes to mitigation through the avoidance of emissions from products that are replaced by resource recovery products. The mitigation potential of anaerobic digestion and subsequent use of biogas for energy has been extensively covered in the literature, as well as that of using various excreta-derived fertilizers and soil amendments as an alternative to artificial fertilizers (see e.g. Otoo & Drechsel 2018). Other options include water reuse and fly larvae composting for animal feed production, all of which contribute to climate mitigation by reducing the need for primary production (Ddiba et al. 2022). These strategies not only circumvent the emissions associated with conventional products but also contribute to sustainable resource management.

It is important to highlight that we aimed to identify literature that is relevant to understand the emission sources across the sanitation chain, not to comprehensively cover all literature that exists on the topic. As a result, certain studies may not have been incorporated, which could impact some of the insights presented in this review. In addition, the accuracy and consistency of data across different studies also pose a limitation. Comparability can be challenging due to the diverse range of methods, data sources and techniques employed for GHG measurements and the variations in assumptions and reporting standards.

This paper reviews evidence on emission sources and hotspots across the sanitation chain for both non-sewered and sewer-based sanitation systems, covering all the stages of the sanitation service chain. We review existing literature on key GHGs like CH₄ and N₂O in both types of systems and present the current state of knowledge, providing a comprehensive overview for understanding the factors that influence the quantity of emissions including system design and configuration, system operation, characterization of waste streams, energy consumption and resource recovery.

Our findings underscore significant gaps in the literature and the imperative for further research on sanitation–climate change linkages. First, we found few empirical studies on GHG emissions from non-sewered sanitation. While there are some studies focused on pit latrines, septic tanks and CBS in a few locations, there is inadequate attention to other types of non-sewered sanitation technologies such as composting toilets, biogas toilets and fossa alternas and their variations across diverse geographical contexts. Secondly, we report that contextual factors such as climatic conditions, soil properties and the status of operation of sanitation infrastructure are sometimes not considered in GHG emissions accounting, despite their significant impact on emissions. Thirdly, few studies account for GHGs across the whole sanitation chain, a holistic approach long established in the solid waste management sector, through empirical and LCA-based research. Finally, we highlight variations in GHG measurement results from methods like flux chambers and remote-sensing, underscoring the need to explore and clarify methodological discrepancies. The above gaps point to areas where further research efforts could be directed.

This paper's comprehensive review can serve as a valuable reference in the design of climate mitigation strategies in the sanitation sector by practitioners and policy-makers. For instance, we provide insights to inform the design, implementation and operation of so-called low-emission or climate-smart sanitation infrastructure. Practitioners should also pursue real-time GHG emissions monitoring in the sanitation and wastewater systems they operate, to bolster evidence and hence advance synergies between sustainable development goal 6.2 (sanitation and hygiene) and 13 (climate action).

Funding for this research was provided by the Swedish International Development Cooperation Agency (Sida), through core support to the Stockholm Environment Institute.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.