Life-cycle analysis of environmental loads from household septic systems in Japan focusing on effluent water discharge

Life-cycle assessment (LCA) is extremely useful in the evaluation of modern water treatment-related infrastructure. Since the 1990s it has been used in several studies to evaluate the environmental impacts of urban wastewater treatment systems, particularly sewage treatment systems (Emmerson et al. 1995). These studies had different aims, scopes, and assumptions, and so the methods employed were also different. Wastewater treatment plants (WWTPs), for example, generally involve several processes, including infrastructure construction, wastewater treatment, sludge treatment, and sludge disposal. Most previous research on WWTPs evaluated only some of these processes. After reviewing 45 papers, Corominas et al. (2013) identified 22 studies in which the construction process was evaluated and 10 in which the sludge disposal process was excluded. Morera et al. (2017) developed a detailed inventory analysis of WWTP construction and investigated how the construction process contributes to the total environmental load in various impact categories. Comparison of the results obtained with those of seven previous studies showed that, according to most of these studies, the contribution of the construction process to climate change and freshwater eutrophication was 5–30 and <1–6%, respectively. Garfi et al. (2017) targeted relatively small-scale domestic wastewater treatment systems that served 1,500 person equivalents (PE), and assessed the environmental impact of three such systems: the conventional WWTPs (activated sludge systems), hybrid constructed wetlands, and high-rate algal pond systems. Of the four processes related to a conventional WWTP (i.e., construction, operation, sludge treatment and disposal, and emissions to air), the environmental impact of the operational process was most significant, being 85–97% of the total impact among all seven impact categories considered. These studies suggest that operation-related stages, including wastewater treatment, sludge treatment, and sludge disposal, are the main contributors to the environmental impact of conventional WWTPs (>90%). In addition to these studies, an increasing number of LCA studies are being conducted on large-scale WWTP systems and related processes (Hong et al. 2009; Lam et al. 2020; Zawartka et al. 2020; Liu et al. 2021).

Although LCA studies are not limited to wastewater treatment, they usually involve life-cycle impact assessment (LCIA) to compare the environmental loads of different impact categories. Recently developed LCIA methodologies, for example IMPACT World+ (Bulle et al. 2019), ReCiPe (Huijbregts et al. 2017), and LIME 2 (Itsubo et al. 2012), are damage-oriented approaches, wherein the damage at an endpoint (the areas of protection) is calculated. Multiple LCIA methodologies, including predecessors of these methodologies, have been applied to LCA case studies of wastewater treatment (Corominas et al. 2013; Sabeen et al. 2018; Gallego-Schmid & Tarpani 2019).

As expected, some existing studies in which the impact assessment step was taken into consideration illustrate that effluent quality is a key factor in the LCA of the environmental impact of wastewater treatment systems. Larsen et al. (2010) compared the environmental impacts of several wastewater treatment options, including conventional WWTPs, for multiple impact categories using the EDIP97 method (Wenzel et al. 1997). The wastewater treatment and sludge digestion process of the conventional WWTP was evaluated, and the decrease in the environmental impact due to nutrient removal was found to be approximately 18-fold that of similar processes without nutrient removal. Limphitakphong et al. (2016) evaluated the life-cycle environmental impact of WWTPs in Thailand in three impact categories: climate change, eutrophication, and acidification. The environmental impact that was normalized using LIME (Itsubo et al. 2004) showed that eutrophication accounted for over 90% of the total environmental impact. Therefore, it is important to assess effluent water quality thoroughly when comparing treatment options for wastewater with different effluent qualities.

Wastewater treatment in Japan includes sewerage systems in highly populated areas. However, in rural areas, septic systems (known as Johkasou in Japan) including Tandoku-shori septic systems and night-soil storage systems are used. Tandoku-shori septic systems and night-soil storage systems can only treat or store human waste from toilets; hence, they cannot treat gray water from sources such as kitchens, laundering, or baths. The use of night-soil storage systems is very limited and, in 2001, the installation of new Tandoku-shori septic systems was forbidden to reduce the environmental impact of gray water on the water environment. Therefore, at present, the septic systems (Johkasou) which can treat all domestic wastewater are the main individual, decentralized, and on-site wastewater treatment system employed in rural areas. Ebie et al. (2014) conducted a greenhouse gas (GHG) emissions survey on septic systems and clarified their emissions. Fujimura et al. (2017) studied septic systems and reported their biochemical oxygen demand (BOD) and nitrogen treatment performance, and the relationship between the recycle ratio and nitrogen treatment of the system. The technical aspects of the septic systems have been widely examined; however, there are few studies on the use of LCA to investigate their environmental impact. Thus far, LCA studies on large-scale wastewater treatment systems in Japan have focused primarily on assessing the total cost or GHG emissions of actual large-scale sewage treatment plants.

With respect to two impact categories, the integrated environmental loads of WWTPs were evaluated and compared using the LCIA method based on endpoint modeling ver. 2 (LIME2; Itsubo et al. 2012) and the ecotoxicity estimation model (Mishima et al. 2016). This approach was very useful in the evaluation of the integrated environmental load on climate change, eutrophication, and biodiversity. Wastewater treatment facilities consume large amounts of energy to improve pollution and discharge treated water. Because some pollutants remain in the treated water, it is necessary to evaluate the energy consumption and environmental load of the treated water comprehensively. Therefore, LCA can be used to evaluate the integrated environmental load, including effluent discharge to water environments, from septic systems to quantify the environmental performances of these systems. Today, several types of septic system such as the standard-structure type, BOD-removal type, nitrogen removal type, phosphorus removal type, and very compact type are operated to treat all domestic wastewater from individual houses.

When considering the aim of reducing environmental impact, it is desirable to select the type that improves water quality in areas where local water pollution is likely to be a problem. One that ensures a certain level of water quality and has less global environmental impact should be selected in areas where there is less demand for water-quality improvement. To promote this policy, economic measures such as a change in the subsidies for septic systems by local governments are possible. However, currently there is little quantitative information on the environmental impact that can be used as a basis for such measures. The objective of this study was to determine the integrated environmental performance from several types of septic systems, while quantitatively considering the specific features of each system using LCA.

The quality of wastewater from operational septic systems in Japan was investigated and LCA was performed on the water-quality data. For both procedures, septic systems were classified under five categories (Table 1). Septic systems consist of the sedimentation tank, anaerobic tank, aerobic tank, clarification tank, and disinfection tank. These systems can be classified into two types (K and C) based on their structure. The K-type, which is known as the standard-structure type, comprises systems designed and manufactured in compliance with the Structural Standards for Johkasou as prescribed by the Japanese government. C-type (C1–C4) septic systems, which are known as certified structure types, are freely designed and manufactured by septic system manufacturers and certified with the authorized effluent water quality based on the treatment performance test. These classifications and types are well known in the environmental administration and small-scale wastewater treatment industry in Japan. In general, the K-type septic systems are larger than the C-type systems. Most septic systems installed in the last decade have been the C-type. These are further divided into four subtypes (C1–C4) based on the type, authorized effluent water quality, and other functions (Table 1). The average volume of typical aeration tanks and average total volume of typical tanks, mainly based on data from the Johkasou System Association (2015), are also shown in Table 1. The quality of the effluent water from C1-type septic systems conforms to the BOD standards. C2 complies with a stricter standard than C1; in addition to the BOD standards, it complies with the standards for the removal of total nitrogen (T-N). C3 includes a phosphorus removal function with the electrocoagulation method using iron electrodes (Mishima et al. 2017, 2018) and complies with the strictest effluent standards, especially with respect to phosphorus concentration. Although the application of aluminum-containing pellets for phosphorus removal in septic systems has been explored (Fujimura et al. 2019), electrocoagulation using iron electrodes is the commonly adopted commercial technology. Therefore, among the C-type septic systems, the C2- and C3-types are larger and have long retention times for wastewater treatment. C4 is a compact septic system, and has a smaller tank than the conventional C-type septic systems. The average volume of typical aeration tanks, the average total volume of typical tanks and the average total retention time for five PE systems are also shown in Table 1.

The quality of the effluent treated using five categories of septic systems (Table 1) was investigated. The septic systems for five, seven, or 10 PE are planned and installed depending on the size of the house. Five and seven PE septic systems are commonly used; however, the actual size of households that use the septic systems varies. The actual/planned population ratio was estimated from the ratio of the actual number of occupants to the planned number of occupants to evaluate the influent load into each septic system (Ebie et al. 2014). The average of the actual/planned population ratio for approximately 12,000 existing septic systems in Saitama Prefecture was 0.56 ± 0.27. Therefore, to obtain effluent samples from septic systems with a standard load of wastewater, five PE systems that were actually used by two to four people and seven PE systems that were actually used by three to five people were selected. The sampling of the effluent from the selected septic systems was accompanied by annual legal inspection of the septic systems by a designated inspection organization in Saitama Prefecture, located around the center of Japan. The data from noncompliant septic systems were excluded from this investigation. Since these are individual septic systems, strictly speaking, the influent water quality and loadings are different. Therefore, we designed the study to investigate a large number of septic systems and to obtain an average value of effluent water quality. Because the number of C3 septic systems is limited, two to five PE C3 septic systems were selected and investigated once a month for approximately one year. The water quality of the sampled effluent was measured according to JIS K 0102 which is a Japanese industrial standard and used for water quality measurement. Chemical oxygen demand (COD) was not measured in the C3 survey. Therefore, COD was estimated by performing multiple regression analysis based on the data of BOD, NO₂-N and T-P obtained from other water-quality surveys (adjusted R = 0.854, p < 0.001). It had been verified that this COD estimation had no impact on the subsequent analysis of environmental impact assessment.

For the operation stage, the average individual water usage in Japan was assumed to be 250, 50, and 200 L/p/d of all wastewater, black water, and gray water, respectively (Fujimura 2006). Additionally, the household size was assumed to be 3.03 people per household in 2015 (The Statistics Bureau of Japan 2017). This statistical value was obtained from a survey of the average household size of households with two or more people. The quality of the gray water influent was assumed to be: BOD, COD, T-N (100% as NH₄–N), and T-P of 145, 65, 10, and 1.5 mg/L, respectively (Fujimura 2006). The average amount of sludge generated was considered to be 1.2 L/p/d (Ministry of Health and Welfare 1988). The functional unit was considered to be 1-year use of a five PE septic system by one household, and the lifespan of a septic system was assumed to be 40 years. Throughout the life cycle of the septic system, 10 impact categories were considered: climate change, eutrophication, biological toxicity, toxic chemicals, urban area air pollution, acidification, ozone layer destruction, photochemical oxidant, resources depletion, and waste.

The technical parameters of the septic systems (i.e., rated electricity consumption and system components) were obtained from reports (Johkasou System Association 2015) and interviews with manufacturers. Information on 38 models, i.e., three, 18, 10, four, and three models from K, C1, C2, C3, and C4 categories, respectively, was collected. In addition to the principal model input data (Table 2), process data with parameters of the five categories with respect to manufacturing, transportation, and installation were mainly obtained from the Johkasou System Association (2015). These data were used to make estimates for systems with unreported process data. The disposed sludge is usually treated in excreta disposal treatment plants. Inventory data on the sludge disposal process were based on the existing reports (Matsui et al. 2002; Morikawa 2010; Ministry of Environment 2016; National Institute for Environmental Studies 2021). Table 3 presents detailed data of the sludge treatment process. Unit GHG emissions data from the septic systems were obtained from Yamazaki et al. (2014), which reported by septic system type: 2,477 (for K), 1,984 (for C1 and C4), 1,044 (for C2 and C3) g-CH₄/person/year and 71.7 (for K), 54.5 (for C1 and C4), 123.2 (for C2 and C3) g-N₂O/person/year in each septic system, respectively. GHG emission factors from discharged water and disposed sludge were referenced from National Institute for Environmental Studies (2021). The emission factors of natural decomposition of domestic wastewater were 0.06 kg-CH₄/kg-BOD and 0.0079 kg-N₂O/kg-N. The emission factors of septic system sludge treatment in a human waste treatment plant were 0.005 kg-CH₄/m³ and 0.0029 kg-N₂O-N/kg-N. The Inventory Database for Environmental Analysis (IDEA) ver. 2 was used as a background data source.

Finally, in accordance with the Japanese impact assessment model LIME2 (Itsubo et al. 2012) and the ecotoxicity estimation model (Mishima et al. 2016), each environmental load was converted into a monetary value (JPY) by multiplying it with the conversion factor. Although the LIME2 model considers 15 impact categories and many pollutant characterization factors reflecting Japanese conditions, it excludes some pollutants associated with the discharge of treated water. The ecotoxicity estimation model was used to evaluate the ecotoxicity effect of NH₄–N emissions into environmental water, a factor that can significantly affect aquatic life, including fish and crustaceans. This model was designed to be consistent with the LIME2 framework; therefore, both models were used to estimate the environmental impact in monetary terms. In addition, the ozone depletion potential of N₂O (Lane & Lant 2012) was considered.

The T-N and T-P levels after treatment were found to be lower than their standard values in classifications targeting their removal for C1–C4. Based on these observations, the studied septic systems mostly operated properly, and a representative water quality for each type of system was obtained from this survey. The squares in Figure 2 indicate the average values that were used for the calculation of the environmental impact in effluent discharge.

Direct emissions were the CH₄ and N₂O emitted directly from septic systems. CH₄ emissions ranged from 3.16 (C2 and C3) to 7.51 kg-CH₄/year/house, which accounted for 41–74% of the total GWP of this process. Direct GHG emissions were lower in the systems with higher-quality effluent (C2 and C3) than in the others (K, C1, and C4). Indirect emission refers to CH₄ and N₂O emitted from the natural decomposition of effluent in a water environment. This emission depends on effluent quality: 0.5% of nitrogen is converted to N₂O, while the CH₄ emission factor is 0.06 kg-CH₄/kg-BOD (National Institute for Environmental Studies 2021). Furthermore, 71% (K, C1, C4) to 73% (C3) of indirect GHG emissions were derived from N₂O emissions. As with direct emission, higher-quality effluent, especially with lower nitrogen concentration, reduced the GHG emissions from the effluent. The direct emissions of C2- and C3-types of septic systems were both 190 kg-CO₂eq/year/house, whereas the highest was 252 kg-CO₂eq/year/house for the standard-structure type (K). Direct GHG emissions as CO₂ equivalent were 11.6 (C1) times to 23.4 (C3) times higher than that from the effluent. Hence, direct emissions were more critical than the effluent for reducing non-CO₂ GHG emissions related to the decomposition of water pollutants.

Tervahauta et al. (2013) reported that the highest primary energy consumption of 914 MJ/cap/year was attained within the centralized sanitation concept in an on-site wastewater treatment process that employed similar treatment methods to the septic systems in Japan. The primary energy consumption via electricity use by septic systems and sludge transportation systems, which corresponded to the system boundary of Tervahauta et al. (2013), was 1,173–1,763 MJ/cap/year. The difference between the two studies seems to be derived from scale, from a community-on-site (100–10,000 people) to a centralized system (≥10,000 people), in contrast to an on-site (one household) system. Garfi et al. (2017) reported that the life-cycle GHG emissions of relatively small conventional WWTPs (1,500 PE, activated sludge system) were approximately 1.3 kg-CO₂eq/m³. This was smaller than the results of this study, i.e., 2.3 (type C4) to 2.8 (type K) kg-CO₂eq/m³, when the system boundary was modified to be consistent with the work of Garfi et al. (2017), excluding indirect emissions from the effluent and septic system disposal. The difference between these values was caused by multiple factors, mainly direct GHG emissions and electricity consumption. For example, in C4-type, the direct GHG emissions were 0.72 kg-CO₂eq/m³, which is 22 times that reported by Garfi et al. (2017) (0.03 kg-CO₂eq/m³). The electricity consumption of the C4-type system, including the sludge treatment process, was 1.72 kWh/m³, whereas it was 1.26 kWh/m³ in a conventional treatment plant. When comparing the two studies, note that the electricity consumption is higher at a smaller scale and that the septic system needs more aeration to circulate wastewater in the treatment tanks.

Lehtoranta et al. (2014) compared the life-cycle GHG emissions and freshwater impact of seven options of on-site wastewater treatments for households in Finland. The life-cycle GHG emissions of the sequencing batch reactor and biofilter were approximately 500–600 kg-CO₂eq/year/household, which was slightly lower than the values reported in this study. The GHG emission hotspots were different from those in this study, i.e., sludge transportation as well as electricity and direct emissions. Moreover, the T-N load of the effluent in septic systems was lower (21 g/d). Similar to this study, the operation, including effluent and sludge treatment, accounted for much of the eutrophication potential in the study by Lehtoranta et al. (2014). Lam et al. (2015) evaluated conventional off-site and on-site wastewater treatment options with source-separated on-site treatment options in China. The conventional on-site treatment employed septic systems that corresponded to the C1-type system in this study. The life-cycle GHG emissions of the conventional option in China were approximately 180 kg-CO₂eq/person/year, which is slightly lower than that of the C1-type system in this study (231 kg-CO₂eq/person/year). The difference mainly arose from the inventory data of sludge treatment; the GHG emissions from other processes in the operation stage were similar in both studies. Regarding other impact categories that were common to the two studies (i.e., acidification and eutrophication), the impact in stages other than the operation stage was relatively low. The contribution of sludge disposal was higher in the report by Lam et al. (2015) in these two impact categories because of the assumptions of nutrient leaching and air emissions related to application to agricultural land. The impact on climate change accounted for 30.3% (C1) to 45.1% (C3) of the total environmental impact. Thus, any decision-making should be based on not only GHG emissions but also multiple environmental impacts, as shown in this study. The C3-type systems had the lowest environmental load owing to phosphorus reduction through iron electrolysis technology. Consequently, when the environmental loads of the nutrient removal and compact-type septic systems were calculated, the C3-type systems were found to have clearly performed better. A new C3-type system with fewer blowers and lower power consumption was developed recently. As this new type of system becomes mainstream, its advantages will become apparent.

These results can be used as the basic information for the future management of the water environment by the local authorities or residents. For example, C3-type systems will be preferable for lakeside areas (Figure 7) because nutrient emission impacts on the water environment are more critical. In contrast, C4-type systems are preferable for areas where nutrient emission is not critical. Thus, the local government can recommend different types of septic system based on the tradeoffs of environmental impact and total cost of introduction. To the best of our knowledge, this is the first study to compare the multiple environmental impacts of septic systems based on on-site measurement of effluents. The results clearly show the variations in the environmental profiles of different septic systems. The environmental load can be expressed as a monetary value using the model adopted in this study. Thus, it can be used to evaluate the government subsidy for introducing new technology for wastewater treatment. For instance, the benefits of the present septic systems, which can process all domestic wastewater, can be compared with those of the remaining Tandoku-shori septic systems, which can process wastewater from toilets only.

This method of study is innovative and challenging, and so the results obtained cannot currently be verified by comparison with other studies. The inventory data were not validated by comparison with other studies because no existing study provides comprehensive inventory data with detailed classification. Hence, more study results should be accumulated. LCA results are based on multiple assumptions. The values for various conditions such as household size, water consumption per capita, and background data on grid power are considered to be representative of the Japanese situation, but may differ from these values when evaluated in a specific region. The characterization factor for ammonia nitrogen biotoxicity is an estimate based on the conditions in Saitama Prefecture, and the factor may differ depending on the region. These points should be taken into account when applying the results of this study to other regions or when comparing them with other studies. Furthermore, the areal environmental impacts such as BOD, which causes poor oxygenation in the water environment and undesirable or discomforting odors, should be considered in future studies.

The water-quality data of effluents were collected, and LCA was performed, to enable a comprehensive and comparative evaluation of the environmental impact of septic systems. A model for evaluating the septic systems using LCA was developed, and the obtained water-quality data and system characteristics data were fed into the model. The results showed that the environmental load of the system with phosphorus removal was the minimum. The mean environmental load from the effluent in the operation stage was 37.6%. Therefore, evaluation of the effluent discharge inventory using LCA, inspections, and water-quality monitoring to demonstrate the proper operation of the septic systems is critical for reducing the environmental load. The operation stage accounts for more than 99% of eutrophication, biological toxicity, and toxic chemicals, which are strongly related to the quality of the effluent. The evaluation of an effluent discharge inventory using LCA is of great significance, even for small-scale wastewater treatment systems, in calculating the comprehensive environment impact.

This may be the first study to compare the multiple environmental impacts of septic systems based on on-site effluent measurement. The results clarify the differences in local and global environmental impacts for each type, and can be used to promote appropriate septic system installation in rural areas.

We are grateful to the septic system manufacturing companies for their assistance in this study. This work was supported by grants-in-aid for septic systems research from the Japan Education Center of Environmental Sanitation, JSPS KAKENHI (grant numbers JP16K16209 and JP18H04153) and the Environment Research and Technology Development Fund (JPMEERF20211005) of the Environmental Restoration and Conservation Agency of Japan.

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

The authors declare there is no conflict.