New approach to domestic grey water footprinting: country-scale accounting using statistical methods in Türkiye

Freshwater scarcity has become a global environmental problem. Along with population growth and climate change, the consumption of water resources is expected to increase significantly in the future. Global water demand for all uses is expected to increase between 20 and 30% by 2050 with significant differences across global regions (Baggio et al. 2021). To minimize future pressures on water resources, it is essential to analyze the current state of water resources from a sustainability perspective (Dong et al. 2022).

In 2002, Professor Arjen Y. Hoekstra from the University of Twente, Netherlands, created the water footprinting (WF) approach, and since then, several initiatives have emerged incorporating the WF concept (Hoekstra 2017). The WF includes not only direct freshwater use but also indirect water (needed to produce, grow, or manufacture the items) along the supply. It is also a multidimensional indicator that investigates quantities of water consumption by source and volumes of water-diluting pollutants. The term ‘blue WF’ describes how blue water resources (surface and groundwater) are used for unit production. The use of green water resources is referred to as ‘green WF’. Grey WF is the volume of freshwater required to assimilate the discharged pollutant loads into the surface waters (Hoekstra et al. 2011). In other words, it is the amount of fresh water needed to dilute pollutants (treated or untreated) in order to meet specified ambient water quality standards. The GWF can reflect the degree of water pollution and has been used in recent years to determine the effectiveness of management strategies (Dong et al. 2022).

There are a few studies in the literature that deal with grey water footprinting of anthropogenic activities on a national and global scale. The preceding studies generally focused on a specific sector such as industry, agriculture, etc. (Ayni et al. 2011; Shrestha et al. 2013; Sun et al. 2013; Cao et al. 2014; Denis et al. 2016; Li et al. 2021, 2022; Banerjee et al. 2023). The studies evaluating domestic uses did not investigate country-scale grey water footprints by assessing temporal and spatial differences using high-resolution (spatially and temporally) site-specific data (e.g., applied treatment technology, wastewater production rate, etc.).

In these researches, nitrogen emissions were estimated per year and per country and were generally based on dietary per capita protein consumption, which was abstracted from statistical institutions (Herrebrugh 2018), as well as typical wastewater production rates and water pollutant loads (per capita) taken from the literature. Mekonnen & Hoekstra (2015) studied nitrogen-related water pollution in river basins with a specification of the pollution by economic sector and by crop for the agricultural sector. Zhang et al. (2019) evaluated the GWF characteristics of 31 Chinese provinces. They mostly investigated GWF as a total amount on the basin or country scale instead of giving value on a smaller spatial scale per capita. Mekonnen & Hoekstra (2011) published a report on ‘National Water Footprint Accounts: The Green, Blue, and Grey Water Footprint of Production and Consumption.’ In this study, GWF_domestic showed variation between 150 and 500 m³/ca.year in European countries. Boyacioglu (2018) developed an approach to investigate the GWF of municipalities in the Aegean Region in Türkiye. In that study, GWF_domestic was changeable from one city to another and had a range of 450–1,500 m³/ca.year.

In this study, 81 city municipalities in Türkiye were assessed to account for GWF_domestic. The wastewater production and treatment profiles of the cities were examined, and then the GWF_domestic was calculated. Furthermore, spatial and temporal variations were assessed using statistical methods. This will be one of the first studies assessing municipal grey water footprints and examining differences at a country-scale with high spatial and temporal resolution.

The natural concentration (C_nat) of a water quality variable in a waterbody is the concentration representing pristine condition (before human influences in the catchment). Maximum allowable concentration (C_max) is the criteria for safe levels of exposure to chemicals that can be used to assess the acceptability of receiving media for intended uses (Franke et al. 2013).

In the study, the GWF_domestic was calculated as the amount of water necessary to reduce the total nitrogen concentrations of domestic wastewaters (treated or untreated) released into receiving media (river, sea, etc.) from 81 cities where the total population served by sewage systems was about 72 million inhabitants as of the year 2020 (TUIK 2021). The reason to choose this variable was that discharging large amounts of domestic wastewater drastically increases the reactive nitrogen content in discharged media, which causes severe ecological stress and biodiversity loss.

Data were provided from the series of official waste water surveys at the municipal level published annually or bi-annually by the National Institute of Statistics (TUIK) since 2001.

Pollutant loads were calculated based on the following components for each city individually:

• •

  amount of untreated wastewater

• •

  amount of treated wastewater that was classified according to the applied treatment process as:

• • ○

    physical treatment

  • ○

    biological treatment (secondary)

  • advanced treatment (nutrient removal)

In the calculations, discharged wastewater total N concentration (C) values were referenced from the literature. The total N levels of various types of wastewater are given in Table 1. Load-L was calculated by multiplying the wastewater amount (Q) by these concentration values. The GWF was then estimated by dividing the pollutant load by the difference between C_max and C_nat.

Since the receiving media were rivers for the municipal discharges across the country, in the GWF calculations, threshold values set for rivers were considered. In this scope, C_nat was accepted as 0.38 mg N/L. This was the average value in rivers that was reported by Meybeck (1982). As was proposed by the Canadian Council of Ministers of the Environment-CCME based on the guidelines for the protection of aquatic life, C_max was considered to be 2.96 mg N/L (CCME 2013).

In the following section of the study, maps depicting the amount of waste water discharges (m³/ca.year), untreated and treated wastewater percentages (%) according to applied technology (biological, physical, and advanced), and GWF_domestic were created using Arc Map 10.3.1.

Moreover, factor analysis was used to group cities by explaining the correlations between the data sets in terms of the underlying factors that cannot be observed directly. In the scope of this analysis: (a) for all the variables, a correlation matrix was generated; (b) factors were extracted from the correlation matrix based on the correlation coefficients of the variables; (c) to maximize the relationship between some of the factors and variables, the factors were rotated (Boyacioglu 2006). The analysis was performed using SPSS-20.0 (2011) for Windows.

Classification of the cities using factor analysis was based on the following data. Amount of:

• wastewater production

• untreated wastewater

• treated wastewater – physically

• treated wastewater – biologically

• treated wastewater – using advanced technology

In addition to spatial differences, temporal variations in GWF_domestic values were analyzed using the non-parametric trend analysis technique. The objective was to determine the significance of a trend using the Mann–Kendall test. It is a non-parametric signed rank test and has been widely used in environmental studies. It examines the sign of the difference between later and earlier data. Each later set of data is compared to all earlier ones, yielding a total of n(n − 1)/2 data pairings, where n is the total number of data. The data set does not have to fit a particular distribution (Hipel et al. 1988; Helsel & Hirsch 2002; Burn et al. 2012).

Descriptive statistics of data covering the amount of total untreated and treated (according to the applied treatment process) wastewater representing 81 cities as of the year 2020 in L/ca.day are presented in Table 2. It was the latest survey result published by the Turkish Statistics Institute (TUIK 2023).

Wastewater production was about 165 L/ca.day on average. On the other hand, it increased to 339 L/ca.day. The mean value for the amount of untreated wastewater across the country was 42.2 L/ca.day. Domestic wastewater was mainly treated using an advanced or biological treatment process.

Based on the characteristics of cities grouped under each factor, it can be concluded that:

• Factor 1 represents cities that treat wastewater by dominantly using advanced treatment technologies.

• Factor 2 reflects cities that primarily treat wastewater by utilizing biological treatment technologies.

• Factor 3 represents cities that do not use treatment processes for a high percentage of wastewater.

• Factor 4 represents cities that treat wastewater predominantly using physical treatment technologies.

A total of 60 cities out of 81 were comprised of the first two factors. The common features of the cities were that they used either advanced or biological treatment processes. Only a few cities treated their wastewater after physical treatment. On the other hand, in the 15 cities that were grouped by Factor 3, a high percent of wastewater was discharged untreated. Only six cities were classified under Factor 4.

The empirical results of the study showed that:

• (a)

  the produced wastewater amount increased up to 125 m³/ca.year in some cities;

• (b)

  wastewater treatment rates were remarkably higher in the western part, and above 70% of the wastewater was discharged to receiving media after the treatment process;

• (c)

  biological and advanced treatment processes were the mostly used technology over the country;

• (d)

  GWF_domestic values fluctuated between 330 and 1,900 m³/ca.year, and the average value was about 750. Hoekstra and Mekonnen conducted research in 2012 and estimated the national average GWF value as about 500 m³/ca.year for Türkiye (Hoekstra & Mekonnen, 2012). Hence, the estimated average value of 750 m³/ca.year in this study was comparatively high. The reason could be that previous research has made estimations using limited data at country scales around the world.

• (e)

  A total of 81 cities were classified into four groups, and dominantly used water treatment technology was the characteristic for this classification. A total of 60 cities out of 81 were comprised by the first two factors, and they used either advanced or biological treatment processes.

• (f)

  while 23 cities showed a decreasing trend in GWF_domestic, only in 9 cities values increased over time.

There are a few studies in the literature that deal with grey water footprinting of anthropogenic activities on a national and global scale. The preceding studies generally focused on a specific sector and none of which investigated country-scale grey water footprints by assessing temporal and spatial differences. This will be one of the first studies assessing municipal grey water footprints at a country-scale with high spatial and temporal resolution. In the scope of the study, a new approach to domestic GWF_domestic was proposed for country-scale accounting using statistical methods. The application of the methodology was demonstrated in Türkiye using data from municipalities in 81 cities, where 72 million inhabitants were served by sewage as of 2020. Data were handled within the series of waste water official surveys at a municipal level published annually or bi-annually by the National Institute of Statistics (TUIK). The wastewater production and treatment profiles of the cities were examined. Then the GWF_domestic was calculated as the amount of water required to reduce the total nitrogen concentrations of domestic wastewater released into receiving media. Furthermore, factor analysis and Mann–Kendal trend analysis tests were performed to classify cities based on GWF_domestic components and also determine the significance of a trend in GWF_domestic. The results of the study investigated spatial and temporal differences across the country. The proposed approach is believed to assist decision-makers in developing pollution prevention measures by analyzing site-specific high-resolution data.

Both authors contributed to the study's conception and design. H. B. mainly contributed to data collection and statistical analysis, and H. B. contributed to data collection and water footprint calculations. Both authors read and approved the final manuscript.

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

The authors declare there is no conflict.