A Spatial Water Footprint Assessment of Recycled Cotton T-Shirts: Case of Local Impacts in Selected China Provinces

Abstract

In global trade, the manufacture and consumption of cotton textile products are intervening in the hydrological cycle. To address the relationship between the product system and the unsustainable use of local water resources, a spatial water footprint assessment is needed. This study presents a spatial water footprint method that was demonstrated in the case of domestically produced cotton T-shirts in three recycle scenarios. The results showed that the water scarcity footprint of conventional T-shirt, eco T-shirt, recycled T-shirt is 2.45 H₂O eq, 1.74 H₂O eq, 8 × 10⁻² H₂O eq, respectively, and the water-eutrophication footprint of conventional T-shirt, eco T-shirt, recycled T-shirt is 1.18 × 10⁻² PDF·m²·yr⁻¹, 9.47 × 10⁻³ PDF·m²·yr⁻¹, 4.04 × 10⁻³ PDF·m²·yr⁻¹, respectively. Two interesting results have been found. Firstly, the hydrosphere has been affected by manufacture and consumption; thus any choice made by manufacturers and consumers may lead to an impact on the water resource in a region that is far from the location. Secondly, the effect of water footprint reduction will be more apparent in places with severe water stress. The spatial water footprint offers a transparent result of each phase’s potential contribution to the local environment and could make a quantitative comparison between product stages, products, and local impacts. Thus, spatial water footprint will be a critical component in the sustainability management improvement of the supply chain.

1. Introduction

Cotton textile has been shown to intervene in the global water cycle massively through international trade [1]. Pakistan, China, Uzbekistan, and India are the largest exporters of blue water because of the high water consumption in the cotton agriculture stage. In the twenty-five members of the European Union (EU25), 84% of the EU’s cotton-related water footprint lies outside the EU, which heavily depends on the water resources of India, China, and Pakistan [2]. China is the largest cotton-cultivation country, accounting for approximately one-quarter of global cotton-lint production [3]. In China, the textile industry is an important sector, which directly employs more than 8.67 million workers and has developed 35 thousand enterprises [4]. According to the data of the National Bureau of Statistics of China, domestic textile enterprises contributed more than USD 963 billion in revenue and USD 124 billion in export in 2016 [4].

The agricultural sector for the production of food and fiber requires accounts for 86% of global freshwater consumption. Furthermore, the textile industry produces a large volume of water consumption and water pollutant emissions, which is also recognized as one of the priorities for the implementation of cleaner production [5]. Cotton textile products, from the life cycle perspective, compared with other man-made textiles like polyester, have a more complex and multi-tiered chain that consists of three major processes (i.e., cultivation; textile production; use stage). These processes also include many sub-processes, such as planting, irrigation, spinning, dyeing, and washing. Thus, an array of concerns in terms of water has been raised associated with cotton cultivation and textile production, as well as different consumer behaviors, because water is indispensable in those phases. In 2021, Xinjiang’s cotton output was 5.129 million tons, accounting for 89.5% of the national output [6]. The planting area was 2506.1 hectares, accounting for 82.8% of the national planting area. Because the land type of Xinjiang is semidesert, it poses high pressure to the water resource of the planting region [7]. A report prepared by Sphera (Sphera Solutions) revealed that water consumption per tonne of seed cotton production: conventional cotton 1.71 × 10⁶ kg; organic cotton, 1.88 × 10⁶ kg, Better Cotton 1.75 × 10⁶ kg [8]. Liu et al. [7] revealed that 79.1% of the contribution to the water depletion is attributable to cotton cultivation, wherein irrigation water, compound fertilizer, and pesticides are the dominant contributors. In the textile production phase, wet processes (e.g., sizing, desizing, scouring and bleaching, dyeing, finishing) require a large amount of freshwater, and the wastewater emission from those processes is highly polluted. Tremendous artificial chemicals are input during dyeing and finishing, such as dyes, wetting agents, and softeners, are emitted into the nature environment with wastewater flow and have made researchers consider their toxici impact on ecosystems and human beings [9]. Different consumer behaviors (e.g., lifetime span, frequency of use, and washing habits) have been indicated, leading to different environmental results. A significant difference in consumer habits between regions is evident when laundering jeans [10]. U.S. consumers use more water, and Chinese users wash more frequently. This difference depends on culture, habits, technology, and washing machine standards; however, in general, handwashing saves water and is a more environmentally friendly way [11].

“Sustainability development aims to satisfy the current needs without compromising the future generation’s needs” [12]. The cleaner production concept is adopted and flourishes in the manufacturing sector, which integrates the continuous application of preventive environmental strategies to processes, products, and services aiming to increase efficiency and minimize the risks to people and the environment [13]. Many sustainable strategies are being taken in the textile sector to develop cleaner products, and more and more companies are undertaking initiatives to gain a better understanding of the environmental performance of their products [14]. The sustainability of textiles can involve eco-friendly material selection, environmentally friendly manufacturing processes, green supply chains, and ethical consumers [15]. Selecting raw materials or substances that are sustainable or processed through sustainable process routes is an effective way to reduce consumption and emissions. For example, organic cotton is considered a good alternative to original cotton, which is planted without insects and fertilizer usage [16]. Furthermore, technical improvement also plays an important role in clean production. Various identified methods in the textile industry are ozone bleaching, enzyme pretreatment, micelle dyeing [17], ozone fading [18], and the hydro-jet technology in denim [19]. Systematically literature review has shown that generally, the reuse or recycling of textiles can significantly reduce environmental impact. except in some circumstances wherein very low replacement rates and the long distance of transport will eliminate the benefits [20]. It was shown that recycling or reuse has the potential to cause certain types of environmental impacts mainly in terms of climate change and acidification because energy and electricity are consumed in this process [21,22]. The main obstacle of the textile recycle market is the low quality of the resultant material, which leads to low strength and short fiber length and thereby the low spinnability of the recycled fiber. This mainly depends on the quality of the waste and the processing technology [23].

Water footprint is a method that effectively measures the environmental burden related to water, which was introduced by Hoekstra et al. in 2002. For the earlier studies of water footprint, they simply quantified the sum of the volume of water usage [24,25], and then they incorporated the dilution of water volume to quantify the impacts of pollution [26,27,28]. However, voluminal accounting provides incomplete information without information on local ecosystem conditions. To overcome this limitation, a water stress index (WSI) was proposed by Pfister et al. [29] to access the extent of water scarcity in each region [30]. To better assess water scarcity, the WULACA group and UNEP/SETAC Life Cycle Initiative in a joint effort proposed available water remaining (AWARE), a generic blue-water scarcity indicator, representing the water availability of each area in the watershed [31]. The ISO 14046 standard for water footprint was developed in 2014, lining with using life cycle assessment (LCA) principles [32], which offer guidelines and principles to conduct an LCA-based water footprint assessment that incorporates two dimensions, i.e., water scarcity and water degradation. For many impact categories, the impact of a given elementary flow depends on where the flow occurs [33]. The LCA method is evolving to incorporate spatial characterization factors (CFs) [34,35] to indicate environment heterogeneity, and an over-reaching LCA method that is globally regionalized was developed, such as IMPACT World+ [36] and LC-IMPACT [37].

The LCA-based water footprint, particularly the water-scarcity footprint, has been calculated for a wide range of products, including primary aluminum [38], oil [39], potato [40], and milk [41]. However, the improved regionalized method on water quality impact (e.g., water acidification, water eutrophication) has not been applied to water footprints as they are presently calculated [42,43]. At present, most LCA studies of cotton products do measure water consumption and water pollutant impact, but they remain the simple summation of different types of water volume (i.e., blue, green) and a general potential impact assessment without a spatial difference [11,44,45,46,47]. Moreover, most existing water footprint assessments are conducted with a country, continent WSI, and water degradation footprint is conducted with general CFs, this presents the risk of misleading consumers. Firstly, for large countries with heterogeneous ecosystems like China, country WSI provides little insight into local water-scarcity levels, so it is unable to be used in the domestic products’ water footprint assessment. However, the high resolution of CFs at the level of 0.5° × 0.5° will lead to difficulties because the specific location is always difficult to identify. As such, the CFs and WSI aggregated in the province as resulted in this study are easy to use and without the loss of the characteristic of the local level. Secondly, general CFs for the water-degradation footprint will mislead consumers because of the confusion on whether a product with a low water footprint is better. Actually, it depends on where it occurs. The published literature remains a gap in the comprehensive and spatial water-footprint assessment of cotton textile products. It can only provide limited information about the environmental performance of cotton textile products and has the potential to mislead producers and consumers because of the confusion regarding whether a product with a lower water footprint is better than others. Actually, this depends on where it occurs.

In this study, we focused on giving a spatial water-footprint assessment of a recycled cotton T-shirt in China, with a resolution at the province level. In this way, it could be used for identifying the impact of cotton T-shirt production on the local environment, as well as provide a more comprehensive and detailed understanding of the environmental burden of products. While WSI has been widely applied, the water-degradation footprint in the textile industry continues to remain generally assessed without geographic boundaries. With the usage of WSI, the incorporated regionalized CFs in water degradation could result in a comparison between different stages and products of different regions. The environmental performance of three different recycling scenarios during the fiber production phase was compared to evaluate the benefits taken from recycling. Moreover, combined with spatial water-footprint assessment, it could provide a better understanding of how the local conditions influences the results of environmental performance, which could inform the policy-makers and investors on how to enact their strategies.

2. Materials and Methods

Our study mainly analyzes the spatial water footprint of a piece of 100% cotton T-shirt with a weight of 180 g. The water footprint of the cotton T-shirt was estimated from the cultivation to the use stage, and the phases of disposal and transportation were excluded from the system boundary of this analysis, as there is no direct water consumption for these phases. Figure 1 details the type of water included, phases, and locations in the life cycle chain of the T-shirt assessed in this study. Three different recycling scenarios during fiber production were assessed and compared (i.e., no recycling, 30% recycling, and 100% recycling).

2.1. Data Collection

Primary production data was collected by field investigations on a representative mill in China, including the loss rate of each phase, water input, and wastewater emission. In this mill, cotton fiber recycling is a way to recycle the rag cuttings in the cutting process and the lost fiber in the spinning process in the same production chain. Those processes both contribute to a high loss rate between 20% to 30%. Rag cuttings are shredded and then mixed with the collected lost fiber. After that, they are mixed with original fiber in the composition and spinning phases. As mentioned before, recycling will lead to low quality and spinnability. Therefore, mixing with original fiber is always conducted to improve the quality of recycled cotton products.

Statistical data from this enterprise was allocated according to the production of semi-finished products or finished products. The blue water utilized as irrigation in the cotton cultivation phase was obtained from a crop water-footprint study reported by Xuan et al. [48]. Blue water required for cotton growth was calculated in 13 regions of Xinjiang with the Penman equation and CROPWAT software. Results show that 1 kg of seed cotton requires 4.82 m³ water containing green and blue water. Fertilizer application was allocated according to the annual cotton-lint production that was obtained from the Xinjiang Statistical Yearbook [49]. The phosphorus loss rate through runoff and leaching was obtained from field experiments developed by the First China Pollution Source Census (CPSC) [50], wherein 372 field tests were conducted to assess fertilizer runoff and the leaching process for different cropping systems [51]. The phosphorus loss rate of runoff and leaching caused by fertilizer application for cotton cultivation in this study was searched by region: northwest arid and semi-arid plain, flat land, and drylands, and crop system: mono-cropping systems. Tap water consumption and wastewater emission in the use phase were taken from a questionnaire survey of Chinese consumers, conducted by Zhang et al. [11].

2.2. Water-Scarcity Footprint

The water-scarcity footprint is a metric that describes the impact of human activity on water scarcity in a locality [40]. It was revised from the water footprint, which crudely sums more than one form of water consumption (blue, green, and gray water), by incorporating water stress CFs to address the regional nature difference of water scarcity [30]. For example, the impact of water consumed in a region of water abundance is incomparable to that of water consumed where it is scarce. The impact of the water consumption of one piece of T-shirt was denoted as WSF and was assessed as follows:

$$\mathrm{WSF}={{\displaystyle \sum }}_{\mathrm{i}=\mathrm{n}}^{\mathrm{i}=1}{\mathrm{WC}}_{\mathrm{i},\mathrm{r}}\frac{{\mathrm{WSI}}_{\mathrm{r}}}{{\mathrm{WSI}}_{\mathrm{average}}}$$

(1)

where $\mathrm{WSF}$ is the water-scarcity footprint (kg H₂O eq); ${\mathrm{WC}}_{\mathrm{i},\mathrm{r}}$ (water consumption kg H₂O) is the water consumed in phase i in region r. ${\mathrm{WSI}}_{\mathrm{r}}$ is the average water stress index of region r. ${\mathrm{WSI}}_{\mathrm{average}}$ is the global average water stress index (WSI).

A water stress CF, namely, the WSI, was developed by Pfister et al. [29] based on a withdrawal-to-availability ratio. WSI is available in a global geographical coverage dataset at a spatial resolution of 0.5° × 0.5°. In this study, the average WSI of each region (${\mathrm{WSI}}_{\mathrm{r}}$) was calculated at a provincial level in China, weighted by the area of each province (${\mathrm{Area}}_{\mathrm{i}}$) according to Equation (2), and performed in ArcGis. Because, a national WSI would be less representative and provide little insight into local water-scarcity levels, especially for a large country with heterogeneous ecosystems like China, but a WSI with a high resolution will lead to difficulty, since the exact location is always difficult to identify. The province is the basic administrative unit in China, which is easily identified during the collection of water inventory. The results of WSI are shown in Figure 2.

$${\mathrm{WSI}}_{\mathrm{r}}=\frac{{{\displaystyle \sum }}_{\mathrm{i}=\mathrm{n}}^{\mathrm{i}=1}{\mathrm{WSI}}_{\mathrm{i}}·{\mathrm{Area}}_{\mathrm{i},\mathrm{r}}}{{\mathrm{Area}}_{\mathrm{r}}}$$

(2)

where ${\mathrm{WSI}}_{\mathrm{r}}$ is the WSI of province r, ${\mathrm{WSI}}_{\mathrm{i}}$ is the WSI of location i in province r, ${\mathrm{Area}}_{\mathrm{i}.\mathrm{r}}$ is the area of location i in province r (m²), ${\mathrm{Area}}_{\mathrm{r}}$ is the area of province r (m²), and n is the number of locations in province i.

In this study, green-water consumption is not included because it does not contribute to local water scarcity, as it is only accessible through access to land occupation and does not directly influence the water flows needed by ecosystems and human beings [52].

2.3. Freshwater-Eutrophication Footprint

The freshwater-eutrophication footprint quantifies the water eutrophication potential caused by a piece of cotton T-shirt throughout its life cycle. It was assessed by applying the Impact World+ characterization model [36,53]. The freshwater-eutrophication footprint was denoted as WEF and assessed as follows:

$$\mathrm{WEF}={\mathrm{M}}_{\mathrm{P}\text{-water}}·{\mathrm{CF}}_{\mathrm{p}}$$

(3)

where WEF is water-eutrophication footprint (PDF·m²·yr), and M_p-water is the phosphorus emissions to water that potentially lead to water eutrophication (kg); CF_p (characterization factor) relates the emissions of phosphorus with species diversity (PDF·m²·yr·kg⁻¹ P) to address the eutrophication potential.

CF is defined by FF (fate fact, yr), which represents the P transported fraction from source to receptor and the persistence of P in the receptor [53], and by EF (effect factor, PDF·m²·kg⁻¹ PO₄³⁻ eq ), which reflects the potential decrease in relative species richness due to P in a freshwater system [54] and was calculated as Equation (4) In this study, the CF of each region (CF_p) was calculated at the province-level and weighted by the population of each region, as Equation (5).

$$\mathrm{CF}=\mathrm{FF}\times \mathrm{EF}$$

(4)

$${\mathrm{CF}}_{\mathrm{r}}={{\displaystyle \sum }}_{\mathrm{i}=\mathrm{n}}^{\mathrm{i}=1}\frac{{\mathrm{CF}}_{\mathrm{i},\mathrm{r}}\times {\mathrm{Pop}}_{\mathrm{i},\mathrm{r}}}{{\mathrm{Pop}}_{\mathrm{r}}}$$

(5)

where CF is the characterization factor of water eutrophication, EF is the fate factor (yr), EF is the effect factor (PDF·m²·kg⁻¹ P), CF_r is the CF of province r (PDF·m²·yr·kg⁻¹ P), CF_i,r is the CF of location i in province r (PDF·m²·yr·kg⁻¹ P), Pop_r is the population of location i in province r, and another is the population of province r.

3. Results

Water Inventory Analysis

The consumption water of one cotton T-shirt is 1.35 t H₂O. For each phase’s contribution, shown in Figure 3, the cotton cultivation phase made a substantial contribution to the overall water consumption (79.31%), which was mainly utilized for irrigation during the cotton-growing season. Data of detailed water consumption were shown in

Table 1. Water inventory of one piece of eco T-shirt.

Phase	Location	Phase	Water Consumption (t H2O)	Total Phosphorus (kg)	COD (kg)
Fiber production	Xinjiang	Cultivation	1.07	0.25 × 10−2
		Ginning	1.19 × 10−3
T-shirt Production	Guangdong	Spinning	1.71 × 10−3
		Knitting	2.20 × 10−3
		Pre-papering	1.21 × 10−3
		Dyeing	1.88 × 10−2		0.25 × 10−1
		Finishing	9.30 × 10−4
		Making-up	7.02 × 10−3
Consumption	Zhejiang	Use (45washes)	0.24	1.1 × 10−3	0.78 × 10−1
Function unit: 1 piece of eco T-shirt (180 g)

. Cotton in Xinjiang is entirely irrigated, which poses high stress to local groundwater and underground water [16]. Compared with other crops, cotton requires more water during its growth stage, 1.7 times that of soybean, 3.5 times that of corn, and 2.2 times that of wheat, according to the global averages [27] (see

Table 2. Comparison of water demand for crops in the cultivation phase.

Crop	Area	Blue Water (t)	Green Water (t)	Total (t)	Source
Seed cotton	Global	1306	2282	3588	Mekonnen and Hoekstra et al., 2011 [28]
soybean	Global	70	2030	2107
corn	Global	81	947	1028
Seed cotton	Brazil	46	2575	2621	Chapagain et al., 2006 [1]
	China	760	1258	2018
	Pakistan	3860	1054	4914
	Turkey	2812	288	3100
	USA	576	1673	2249
	Uzbekistan	4377	83	4460
Functional unit: 1 ton of crop

). This indicates that cotton is a water-intensive crop, although the irrigation water of cotton varies from different plant regions depending on the climate, land type, and farming practices. High irrigation demand occurs in regions where evaporation is high and the effective rainfall is low, as in Turkey, Pakistan, and Uzbekistan, all of which are global main cotton-production countries. This means that effective rainfall in this location will not meet the water demand for cotton growth and may result in low yields. However, the climate of USA, Brazil, and China will be more appropriate for cotton growth, with low water demands of 2249 t, 2621 t, and 2018 t H₂O per tons of seed cotton, respectively; moreover, the proportions of green water account for 74.5%, 98.2%, and 62% of the total water demand, respectively, in these countries [1] (see Table 2). These reference data are the average of 1997–2001 and show an apparent difference between countries. However, the water demands of cotton are also largely different between China’s main cotton production region and the rate of green water varies from 2.3% to nearly 100% [55]. According to this study, the green-water rate in Xinjiang was very low at 6.97%, but the total water demand was 1.92~2.16 t H₂O per tons of seed cotton grown in Xinjiang, depending on the loss rate during ginning, which is near the national average. Xinjiang is an arid or semi-arid area with high evapotranspiration (ETo) and low annual precipitation, making it the largest cotton-production region in China. The increase in cotton production and high-irrigation-water demand have derived flood irrigation and drip irrigation combinations for the majority of croplands in this region [56]. Regardless of the absolute difference in the water demand between different regions, high irrigation demand has been unwillingly seen in countries by governments, cooperatives, and farmers, indicating a high cost per unit of cotton production and water-resources-management input.

Except for the extreme difference in the cultivation phase, the water inventory of the three recycle scenarios is the same. Take eco T-shirt as an example, the dyeing process contributed 75.28% of the water volume of the textile manufacturing process and led to 0.02 kg of COD emission. Apparently, it is the main contributor of the water footprint in the textile production phase, of which the wastewater also contains high organic pollutants from the chemical inputs like dyes and softeners. In this study, it is assumed that the T-shirt was used in summer, washed once every 2 days, and used continuously for 3 months, so the T-shirt was washed 45 times throughout the whole life cycle and 0.24 t water was consumed per wash. The phosphorus and COD were emitted from the detergent.

Conventional, eco, and recycling T-shirt scenarios have been evaluated and compared in terms of their water-scarcity footprint and water eutrophication. The results distributed by the life cycle stages of the T-shirt and locations are shown in Figure 4. The water-scarcity footprint of conventional, eco, and recycling T-shirts was 2.45 t H₂O eq, 1.74 t H₂O eq, and 0.08 t H₂O eq, respectively. The water-eutrophication footprint of conventional, eco, and recycling T-shirts was 1.18·10⁻² PDF·m²·yr⁻¹, 9.47·10⁻³ PDF·m²·yr⁻¹, and 4.04·10⁻³ PDF·m²·yr⁻¹, respectively (see

Table 3. Comparison of three recycling scenarios.

Type	Recycling T-Shirt (100% Recycling)	Eco T-Shirt (30% Recycling)	Conventional T-Shirt (No Recycling)
Water scarcity (t H2O eq)	8.07 × 10−2	1.74	2.45
Water eutrophication (PDF·m2·yr)	4.04 × 10−3	9.47 × 10−3	1.18 × 10−2

).

Apparently, the cotton-cultivation process has the highest water-scarcity footprint (2.37 t H₂O eq) and water-eutrophication footprint (7.76·10⁻³ PDF·m²·yr⁻¹), and the water footprint from the textile production process and use stages are less significant, suggesting that the high water footprint in this phase was the biggest and also typical problem of nature textile products. The most dramatic decrease in water scarcity and water eutrophication was observed when switching from conventional to eco and recycling T-shirts (up to 29% and 97% for WSF and 20% and 66% for WEF, respectively), mostly because of the partial removal of the cultivation phase in the fiber production process. This suggests that the removal of the cultivation phase in this study was the most effective way to reduce the water footprint due to the overwhelming water footprint in this phase. For the textile production and use stage, the contribution of different recycling scenarios was the same. Apparently, the dyeing process is the main origin of the water footprint of the textile production phase (74.37%); the wastewater contains high organic pollutants from the chemical inputs, like dyes and softeners. The use phase was another water-intensive process due to the daily washing.

For the water-scarcity footprint, an apparent difference could be found in that the contribution of each phase to the overall water scarcity was changed when compared to the volumetric water footprint (See Figure 3). The overall water-scarcity footprint of an eco T-shirt is 1.74 t H₂O eq (see

Table 4. Result of water-scarcity and water-eutrophication footprints for one piece of eco T-shirt.

Phase	Location	Phase	Water Consumption (t H2O)	Water Scarcity Footprint (t H2O)	Water-Eutrophication Footprint (PDF·m2·yr−1)
Fiber production	Xinjiang	Cultivation	1.07	1.66	5.43 × 10−3
		Ginning	1.19 × 10−3	1.85 × 10−3
T-shirt Production	Guangdong	Spinning	1.71 × 10−3	1.78 × 10−4
		Knitting	2.20 × 10−3	2.29 × 10−4
		Pre-papering	1.21 × 10−3	1.25 × 10−3
		Dyeing	1.88 × 10−2	1.95 × 10−3	8.64 × 10−5
		Finishing	9.30 × 10−4	9.65 × 10−5
		Making-up	7.02 × 10−3	7.29 × 10−4
Consumption	Zhejiang	Use (45 washes)	0.24	7.63 × 10−2	3.95 × 10−3
Function unit: 1 piece of eco T-shirt (180 g)

). The water-scarcity footprint of the cotton cultivation, production, and consumption phases are 1.66 t H₂O eq, 4.43·10⁻³ t H₂O eq, and 7.64·10⁻² t H₂O eq, respectively, and occur in Xinjiang, Guangdong, and Zhejiang (see Figure 4), respectively. Regardless of the high water consumption in the cultivation phase, the high WSI in Xinjiang highlights the plant phase contribution to water resources impact, leading to an increase from 79.39% to 95.37%.

Lower water stress in Guangdong Province has benefited the manufacturing sector. The WSI of Guangdong province is 0.065, lower than 65% of provinces in China, as shown in Figure 2. The WSI values of the major textile-production industrial regions in China were 0.668 for Jiangsu, 0.194 for Zhejiang, and 0.948 for Hebei. It can be concluded that Guangdong province is an appropriate production location in terms of its sufficient water resource. As such, the potential T-shirt production phase contributions to water scarcity through water consumption are very small. The water-scarcity footprint of the use stage was 7.62·10⁻² t H₂O eq, accounting for 4.38% of the total WSF. Low water stress contributes to decreasing in the use stage from 13.11% to 4.38%.

The WSI in Xinjiang is 1.5 times that of the world average WSI, nearly 15 times that of Guangdong Province related to production and 4.8 times that of Zhejiang Province where the use stage is located. Although, the water usage occurring where WSI is high will pose more pressure on the water resource and lead to a high water-scarcity footprint for a product, which rarely appears in an eco-label or sustainability report of product for companies. On the contrary, manufacturers also could benefit from a high WSI, when water-saving strategies are conducted and effect successfully in this region. High WSI will also highlight the effect of water-saving action and present superiority when compared with the same water-saving measures occurring in other regions with low WSI.

The water-eutrophication footprint is 9.47·10⁻³ PDF·m²·yr for the entire life cycle of an eco T-shirt. The water-eutrophication footprints of cotton cultivation (occurring in Xinjiang), dyeing (occurring in Guangdong), and consumption (occurring in Zhejiang) are 5.43·10⁻³ PDF·m²·yr⁻¹, 8.64·10⁻⁵ PDF·m²·yr⁻¹, and 3.95·10⁻³ PDF·m²·yr⁻¹ (see Figure 5). The cotton-cultivation phase contributes 57.34% of the water-eutrophication footprint (see Figure 6). Pollutant potentially leads to water eutrophication because the excessive nutrients emitted to lakes and rivers from phosphorus-containing fertilizer applied for cotton growth are rarely treated before leaving the soil system through runoff and leaching. The dyeing phase in production accounts for a minor proportion of 0.91% because the dyes and chemicals are put in this process and because a high content of organic mass emissions flows through the wastewater. If effluents are treated before their emission to lakes or rivers under government standards, a 93% water eutrophication-footprint reduction could be achieved in this mill. Another important water eutrophication hotspot was observed in the use phase. The consumer-use phase accounts for 41.75% of the water-eutrophication footprint, 14.10% for COD, and 27.64% for P emissions in water due to detergent input during washing. When looking into the CFs of the water-eutrophication footprint (see

Table 5. WSI and CFs for water eutrophication.

Location	WSI	Water Eutrophication CF for COD	Water Eutrophication CF for P
Xinjiang	0.9318	0.01547	2.1523
Guangdong	0.0625	0.0035	0.4853
Zhejiang	0.1942	0.0172	2.3917
China *	0.48	0.0231	3.2179
Global *	0.602	0.0797	11.0869

* Source: WSI is from Pfister et al., 2009; Water eutrophication CF is from Bulle et al., 2019.

), high CFs were observed in Zhejiang province, followed by Xinjiang and Guangdong. Although being less important in the total water-eutrophication footprint, textile production benefited from the low CFs in Guangdong, suggesting that Guangdong was also suitable for production in terms of water eutrophication impact. When compared to the CF of China and globally, the three locations all have lower CFs of water eutrophication, making the environmental performance of the products much more competitive. As same as the interesting finding of WSI, companies are unwilling to use a high level of CFs in their supply chain, but they also can benefit from reducing the CFs of their present supply chain to show their efforts in cleaner production.

Overall, from the perspective of consumers, the consumption of a cotton T-shirt in Zhejiang province has a water footprint that occurs outside Zhejiang (95.62% for the water-scarcity footprint and 58.25% for the water-eutrophication footprint respectively). The strategy made by the manufacturer also influences the water resource far from its location. Those indicated that producers and consumers are strongly intervening in the water cycle of regions far from their local environment.

4. Discussion

To improve the sustainability of products and give transparent information about the product’s environmental performance, a spatial water footprint was required to make the impact on the location transparent. In this study, a spatial water-footprint method containing the water-scarcity footprint and water-eutrophication footprint was illustrated. The superiority of the spatial water footprint was demonstrated in the case study of the water-footprint reports of cotton T-shirts in three recycling scenarios.

4.1. Advantages of a Spatial Water-Footprint Method

One of the most important features of the spatial water footprint is that it provides a comprehensive and regionalized water-footprint result whose impact is distributed to each occurring location. It quantifies the degree of the intervention in the hydrosphere, caused by manufacture and consumption. In this study, the whole life cycle of a cotton T-shirt has impacts related to water in Xinjiang, Guangdong, and Zhejiang provinces and all impacts to those regions can be identified in the spatial water-footprint results (see Figure 4 and Figure 5). As such, it could provide a transparent, meaningful, and comprehensive basis for the eco labeling of products to make consumers more aware of the impacts of their purchasing decisions, and thereby consumers should take responsibility for the appropriation of the water resources that may be far from the consumers’ location. Those actions could drive the popularity of sustainable products and thereby reversely drive the manufacturing sector’s cleaner production. The local government should be informed as to which extent their water resource has been deprived by the production of products and how severe the production has put stress on water resources in terms of volume and quality. Therefore, a spatial water footprint also can serve as a significant and effective basis for policy-making for local water resource management.

Another significant advantage is that it enables the comparison between different stages of a product’s life cycle. The use of WSI and spatial CFs that distinguish the ecosystem heterogeneity, providing a normalization process for water footprint assessment, enables the comparison between life cycle stages occurring in different locations. A few companies, such as Nike [57], have adopted the life cycle concept to optimize their suppliers, which inevitably requires a comparison between the environmental performances of cotton textile products from different suppliers. Since they all have a global supply chain and the suppliers are distributed in many regions, the spatial water footprint enables an effective comparison of the environmental performance of the suppliers to optimize their global supply network. For example, in this study, the WSI values of the major textile production industrial regions in China were 0.668 for Jiangsu, 0.194 for Zhejiang, 0.0625 for Guangdong, and 0.948 for Hebei. As such, the supplier in Guangdong is a prior choice. Companies also can benefit from the use of a spatial water-footprint assessment, when showing their efforts in improving the sustainability of their products, especially occurring where the water resource is severely stressed.

4.2. Advantages of Recycling Scenarios

Recycling in fiber production has been found to contribute to a significant water-footprint reduction due to the removal of the cotton-cultivation phase. A piece of a recycling T-shirt (100% recycling) contributes a 96.71% reduction in the water-scarcity footprint and a 65.76% reduction in the water-eutrophication footprint compared with the conventional T-shirt. For the eco T-shirt, it can reduce 29.01% of the water-scarcity and 19.73% of the water-eutrophication footprint. Recycling in this phase could be viewed as highly effective.

To date, the main obstacle for the market for recycled cotton is the lower quality of the resultant products. The quality of recycling could be evaluated from the recycled fiber length, the spinnability of recycled cotton fibers, and the properties of the resultant yarns. It depends on the fabric structure, tightness, and previous finishing treatments. Most recycling from post-consumer textile waste will lead to lower quality because the fibers are reduced by wear and laundry, and also the dyed fabric will lead to a higher ratio of short fiber. In this study, the waste from recycling was from rag cuttings in the cutting process and lost fiber in the spinning process. In this way, the recycled fiber shows a higher quality than that recycled from post-consumer textile waste and avoids extra impacts from collecting, sorting, and transportation, as the collected waste textile is always from different communities. However, there is a concern that energy used for collecting the lost fiber in the spinning process and for cutting and shredding the rag cuttings may increase impacts in other impact categories. A study reported that the impact of the cutting/shredding in the recycling of 1 kg cotton yarn produced 0.214 kg CO₂ eq for global warming potential (GWP), 0.00025 (kg SO₂ eq ) for acidification potential (AP) and 0.00021 kg PO₄³⁻ eq for eutrophication potential (EP) [21]. Compared with the overall impact caused by a piece of a cotton T-shirt, the cutting/shredding only contributes 0.76% for GWP, 0.1% for AP, and 0.24% for EP [11]. Therefore, there almost is no need to concern about the extra impact caused by the recycling process.

5. Conclusions

Cotton textile production and consumption, as introduced in this report, intervene in the water cycle in multiple regions throughout the value chain. In this study, a spatial water-footprint method, for the first time, covered water scarcity and water degradation, which could provide a comprehensive result. Under an absolutely spatial water-footprint assessment, the performance of sustainable strategies was compared, aiming to discover the most effective way to reduce the environmental burden related to water. The result shows that, when different processing stages of a garment product are distributed in different regions, the spatial water-footprint method can effectively calculate the water-footprint value of the garment product. It reveals that recycled cotton fibers should serve as a highly effective way to reduce environmental impact and as a substitute for virgin cotton fibers. Spatial water footprint also showed efficiency in eco-label, policy making, and company’s strategy. First, by making the relationship between the phase, regions, and impacts transparent, all the impacts caused by products to each location at the province-level could be identified in the water-footprint results. Secondly, it could make a meaningful comparison of the water-scarcity footprint and the water-degradation footprint between products, phases, and regions, as the water footprint is normalized according to the local stress and ecosystem features of each region. Therefore, a spatial water footprint could serve as a basis for company decisions to improve the sustainability of their products, optimize their suppliers, and report their product’s superiority. For consumers, they will be more aware of their purchase choice and take initiatives to reduce environmental burden. One priority for further study should be developing spatial CFs for water toxicity as many artificial chemicals have been input and emitted to wastewater, which will result in damage to the local ecosystem and residents. As the same as CFs of water eutrophication, the CFs of water toxicity also have regionalized features. An uncomprehensive category has the potential to leave out some environmental burdens and thus mislead consumers and manufacturers.