Development of Method for Assessing Water Footprint Sustainability

Abstract

Large scale production of water-intensive industrial products can intensify water scarcity, resulting in potential unsustainable water use at local and regional scales. This study proposes a methodological framework for assessing the WF sustainability of multiple interdependent products in a system, and one of China’s four major large modern coal chemical industry bases is used as a case study. A Mixed-Unit Input-Output (MUIO) model was applied to calculate the blue water footprint (WF) for 19 major coal-based energy and chemicals in the study area, based on which the WF sustainability of production of the products were assessed using different indicators. Technical coefficient matrix and direct water consumption vector of the products were constructed based a database that were built by field research in the study area. Accounting result indicates that the blue WF of the coal-based products range from 2.5 × 10⁻⁴ m³/kWh for coal-fired power to 55.25 m³/t for Polytetrahydrofuran. The sustainability assessment reveals that the blue WF of all products produced in the study area are sustainable at both product and regional levels, while over half of them have reached the advanced level. However, the blue WF of a few products with large production capacities has just crossed the sustainable thresholds, posing potential threat to the local environment. This paper concludes with a discussion on the choice of blue WF accounting approach, methods to promote WF sustainability of coal-based products, and suggestions for the WF management in general.

1. Introduction

Water scarcity can cause severe socio-economic consequences from local to global scales [1,2]. The World Economic Forum rated water crises as one of the major global risks over the next decade [3]. Energy and chemical industry is one of the largest water consumers; it demonstrates high water sensitivity because each stage of the entire life cycle of its productions needs water (e.g., mining or extraction, processing and conversion) [4]. The International Energy Agency projected a rise of 60% in global water consumption for primary energy production and power generation through 2040 [5]. In China, the energy and chemical uses have dramatically increased in last decades due to rapid economic expansion. Consequently, China released the Energy Production and Consumption Revolution Strategy (2016–2030), which set up a series of targets for 2030 including the share of non-fossil fuel in the energy mix, and the nation’s energy self-sufficiency rate [6]. However, coal-based energy will continue to form the major part of China’s energy mix over the next decade due to the low cost and the abundance of domestic reserves [6].

There are two major problems exist in China’s fossil energy resources endowment. First, the energy structure of China is characterized by “rich coal, meager oil, and little gas”; the proven reserves are comprised of 94% coal, 5% crude oil, and 0.6% natural gas [7]. China now has become the world’s largest and second largest importer of crude oil and natural gas, respectively [8]. The dependence of imported energy and chemicals poses a great threat to China’s energy and chemicals supplies. Second, the distributions of coal and water resources are severely mismatched across the country’s territory. Nearly 70% of coal production is concentrated in the northern and western provinces, that only account for 6.5% of China’s total water resources, making water a significant vulnerability in the country’s energy and chemical supplies [9]. To cope with these problems, China gave great priority to the development of 14 large coal energy bases and four large modern coal chemical industry bases during its 12th and 13th Five-Year periods (2011–2020). To reduce its dependence on foreign petroleum, China also made great efforts to develop technology to convert abundant coal into clean fuels and value-added chemicals [10]. However, producing coal-based fuel and chemicals in these coal-rich water-limited energy bases has been controversial due to the high water-consuming processes. Large-scale water-intensive industrial production within an industrial base potentially threat the local environment. Thus, life cycle assessments related to water scarcity for the arid industrial bases in China is of great importance to achieve environmental sustainability.

Water footprint (WF) can be used as an indicator of environmental sustainability in water use. The WF concept was first introduced in 2002 by [11]; it functions as a multidimensional indicator of freshwater use (i.e., blue WF and green WF) and pollution status (i.e., gray WF) [11,12,13]. The Water Footprint Network (WFN) community considers WF as a volumetric metric and focuses on the consumptive freshwater use. Simultaneously, the Life Cycle Assessment (LCA) community converts WF into an environmental impacted-oriented indicator by a weighting scheme called characterization, which is recommended in ISO document 14046 on WF [14]. Over the past decade, researchers in the two communities have given rise to a debate over so-called better WF accounting approach [15,16,17,18,19,20,21]. Nevertheless, there is no contradiction in fundamental principles of the methods proposed by two sides; information provided by volumetric WF and impacted-oriented WF should be complementary rather than competing [22], and the choice of the two WF accounting depends on the purpose of a study.

Previously, the WFN-WF has been adopted in studies focusing on the optimal water resources allocation and productivity of freshwater use [10,23,24], whereas LCA-based WF accounting using input-output (IO) approach has been used in assessing the potential environmental impact of products [4,25,26,27,28]. The IO framework are extensively used to estimate the WF of industrial sectors at global scale (e.g., [29,30]), national or multiregional scale (e.g., [31,32,33,34,35,36]), and basin scale (e.g., [37,38,39,40]), but rarely used to assess the potential impact of the production of interdependent products at local or sub-local scales, because IO tables are compiled only at national or provincial levels due to cost and resource constraints. This study attempts to fill in this research gap by introducing a methodological framework for assessing the WF sustainability of multiple interdependent products in a system. The Mixed-Unit Input-Output (MUIO) model is adopted in the framework to account WF of the products, and three sustainability assessment indicators are then proposed. A large modern coal chemical industry base in Northwest China is used as a case study. Technical coefficient matrix and direct water consumption vector for the products in the study area were constructed based on a database, which was built by our research team through on-site survey and investigation. Since zero liquid regulation has been enacted in China’s major arid industrial bases, the assessment was conduct at product and regional levels with a focus merely on blue WF.

2. Methods

2.1. The Blue WF Assessment Framework

The blue WF assessment of a framework is proposed in this section. The overall methodology of the WF assessment is shown in Figure 1. The first phase of the proposed framework is the scope setting, which determines the major industrial products that are considered in the assessment. This phase also includes the analysis of the interdependence among the products. The functional unit of WF for each product should then be decided in this phase. The second phase is data collection in the region where the products are produced, preliminary analysis of the data can be conducted. In the third phase of the framework, the technical coefficient matrix, and the direct water consumption vector for the are constructed based on the data collected. The coefficient matrix is a p-by-p matrix that describes the interdependence between products within the scope of the study determined in phase one, and the direct water consumption vector is a p by 1 vector contains the direct water coefficients for each product. The accounting model then is built in the fourth phase, followed by WF sustainability assessment of the interdependent products at product scale and regional scale.

2.2. Blue WF Accounting Model

The MUIO model is employed to account the blue WF of each coal-based energy and chemical product. The MUIO model is a top-down framework used for environmental life-cycle analysis; it was first introduced by Hawkins et al. [28]. The MUIO model in the form of WF can be expressed in Equation (1) shown below:

$${\displaystyle \sum }_{n=1}^{P}W{F}_n={\displaystyle \sum }_{n=1}^P{\displaystyle \sum }_{m=1}^P{a}_{nm}W{F}_n+{\displaystyle \sum }_{n=1}^{P}d{w}_n$$

(1)

where WF_n represents the blue WF of the nth product; a_nm is the technical coefficient, which represents the amount of product m directly consumed for producing unit product n; dw_n indicates the direct water use coefficient of product n. P is the total number of major products in the system. a_nm and dw_n are computed using Equation (2):

$$a_{nm}=\frac{{{\displaystyle \sum }}_j{z}_{nm}^j}{{{\displaystyle \sum }}_i{x}_n^i};d{w}_n=\frac{{{\displaystyle \sum }}_i{w}_n^i}{{{\displaystyle \sum }}_i{x}_n^i}$$

(2)

where $z_{nm}^j$ is the total amount of product m flows from the jth enterprise for the production of product n during a time period; $x_n^i$ is the total amount of product n produced in the ith enterprise during the same period; $w_n^i$ is the total freshwater consumed for producing product n in the ith enterprise during the period. Equation (1) can be further expressed in matrix terms as follows:

$$DW=\left(I-A\right)WF$$

(3)

where $WF$ represents the vector containing blue WF of each products; I denotes an identity matrix; $A$and $DW$ represents the technical coefficient matrix and the direct freshwater consumption vector, respectively, which are in the form shown below.

$$A=\left[\begin{array}{ccc}{a}_{11}& \cdots & \begin{array}{ccc}{a}_{1m}& \cdots & a_{1p}\end{array}\\ ⋮& \ddots & \begin{array}{ccc}⋮& \ddots & ⋮\end{array}\\ \begin{array}{c}{a}_{n1}\\ ⋮\\ a_{p1}\end{array}& \begin{array}{c}\cdots \\ \ddots \\ \cdots \end{array}& \begin{array}{ccc}\begin{array}{c}{a}_{nm}\\ ⋮\\ a_{pm}\end{array}& \begin{array}{c}\cdots \\ \ddots \\ \cdots \end{array}& \begin{array}{c}{a}_{np}\\ ⋮\\ a_{np}\end{array}\end{array}\end{array}\right] ;DW=\left(\begin{array}{c}d{w}_1\\ d{w}_2\\ \begin{array}{c}⋮\\ d{w}_n\\ \begin{array}{c}⋮\\ d{w}_p\end{array}\end{array}\end{array}\right)$$

(4)

WF can then be calculated by the MUIO model as Equation (5):

$$WF={\left(I-A\right)}^{-1}DW$$

(5)

where ${\left(I-A\right)}^{-1}$ is called the Leontief inverse matrix. The indirect water use coefficient vector can be calculated by abstracting DW from WF, shown in Equation (6).

$$IDW=WF-DW$$

(6)

where IDW denotes the vectors containing indirect freshwater water for each product.

2.3. Blue WF Sustainability Assessment

The sustainability assessment of WF of a product is twofold: the blue WF assessment against a reference standard, and the WF assessment against the regional water availability. The former indicator given in Equation (7) can be used to evaluate the overall technology level of existing industrial plants in terms of water, energy and material conservation.

$$WSS{I}_p=\frac{W{F}_p}{W{F}_p^{ref}}$$

(7)

where $WSS{I}_p$denotes the WF sustainability index based on any reference standard; $W{F}_p^{ref}$ denotes the blue WF of product p under any reference level of technology. In this study, two references are applied: the blue WF of product p using the general standard ($W{F}_p^{ref-1}$) and using the best practicable technology (BPT) ($W{F}_p^{ref-2}$). The current technology is considered unsustainable if both $WSS{I}_{p-1}$and $WSS{I}_{p-2}$are greater than 1, sustainable if $WSS{I}_{p-1}$is less than 1 while $WSS{I}_{p-2}$is greater than 1, advanced if both WSSI_p−1 and $WSS{I}_{p-2}$are not greater than 1 (

Table 1. Water footprint sustainability assessment methods at product and regional levels.

Assessment Level	Indicator Values	Assessment Results
Product-level	WSSIp−1 > 1, WSSIp−2 > 1	unsustainable
	WSSIp−1 ≤ 1, WSSIp−2 > 1	sustainable
	WSSIp−1 ≤ 1, WSSIp−2 ≤ 1	advanced
Regional-level	WSRIp−1 > 1, WSRIp−2 > 1	unsustainable
	WSRIp−1 ≤ 1, WSRIp−2 > 1	sustainable
	WSRIp−1 ≤ 1, WSRIp−2 ≤ 1	advanced

).

The second indicator given in Equation (8) takes into account both the technology level and the regional water resources endowment. That is, a product produced at the same technology level is considered as sustainable in water abundant region does not necessarily means that it is also sustainable in regions where water resources are scarce.

$$WSR{I}_p=\frac{WSI·W{F}_p}{W{F}_p^{ref}}$$

(8)

where $WSR{I}_p$ denotes the WF sustainability index for product p at regional level. WSI is the regional water stress index. Similarly, the current technology is considered unsustainable at regional level if both $WSS{I}_{p-1}$ and $WSS{I}_{p-2}$ are greater than 1, sustainable if $WSR{I}_{p-1}$ is less than 1 while $WSS{I}_{p-2}$ is greater than 1, advanced if both $WSS{I}_{p-2}$ and $WSS{I}_{p-2}$ are not greater than 1 (Table 1). The WSI index is defined by the ratio of total annual freshwater withdrawals to hydrological availability, it was modified by [41] to differentiates watersheds with strongly regulated flows. Later, Ref. [42] computed $WSI$ for China’s provinces and major river basins using ArcGIS 10.0.

Finally, to assess the overall sustainability of production in a industrial base, the weighted sum WF sustainability index for all products at regional level is computed as Equation (9).

$$WSRI=\frac{{{\displaystyle \sum }}_{p=1}^{P}WSR{I}_p·CA{P}_p·W{F}_p}{{{\displaystyle \sum }}_{p=1}^{P}CA{P}_p·W{F}_p}$$

(9)

where WSRI denotes the overall WF sustainability index for industrial base at regional level, $CA{P}_p$ denotes the total production capacity of product p in the base. Water footprint sustainability indicators and their abbreviations and measurement units are shown in

Table 2. Water footprint sustainability indicators and their abbreviations and measurement units.

Indicators	Abbreviations	Measurement Units
Water footprint	WF	m3/t or m3/kWh
Blue water footprint	Blue WF	m3/t or m3/kWh
Water sustainability index at product level based on general standard	WSSIp−1	unitless
Water sustainability index at product level based on BPT	WSSIp−2	unitless
Water sustainability index at regional level based on general standard	WSRIp−1	unitless
Water sustainability index at regional level based on BPT	WSRIp−2	unitless
Overall sustainability of production in a industrial base	WSRIBase	unitless

.

3. Case Study

3.1. The Ningdong Base

The Ningdong Energy and Chemical Industry Base (Ningdong Base, Ningxia, China) in Ningxia Hui Autonomous Region of China is chosen as a case study. Ningxia is covered in arid and semi-arid climate; the annual average precipitation is 289mm while the annual average evaporation is 1250 mm [43]. Ningdong Base is one of the 14 national major large-scale coal bases and four national major coal-to-chemical industry bases in China, making it an ideal industrial base for the analysis. By the year of 2020, many state-owned and private large enterprises have invested more than 60 coal mining, coal-fired power and advanced coal chemical projects in the base, including the world’s largest single CTL project (

Table 3. Statistics of the coal-based energy and chemicals in the Ningdong Base (2020).

Product	Abbreviation	No. of Enterprises	No. of Projects	Total Capacity (104 ton/a)
Washed coal	COAL	2	13	9000
Coal-fired electricity	ELEC	10	15	15,660 (MW)
Coal-to-liquid	CTL	1	1	400
Coal-to-methanol	CTM	3	4	175
Goal gas to methanol	CGTM	2	2	45
Coke	COKE	2	3	590
Goal gas to olefin	CGTO	1	1	60
Methanol to olefin	MTO	3	5	205
Dimethyl ether	DME	1	1	21
Polyoxymethylene	POM	2	2	11
Ammonia	NH3	2	2	55
Urea	UREA	1	1	70
Calcium carbide	CaC2	1	1	115
Acetylene	ACET	1	1	30
1,4-Butanediol	BDO	1	1	20.8
Polytetrahydrofuran	PTMEG	1	1	9.2
Acetic acid	ACA	1	1	30
Vinyl acetate	VAC	1	1	40
Polyvinyl alcohol	PVA	1	1	10

). The annual total capacities of coal mining, coal-fired power generation, and chemical production have reached 90 million tons, 15,660 MW, and 25 million tons, respectively. The massive production of coal, coal-fired power, and coal-based products in Ningdong has intensified the water stress of the province. It is reported that over 250 million cubic meter of freshwater is consumed annually in the base [43].

3.2. Blue WF Accounting for the Products

Prior to WF accounting for the products, the technical coefficient matrix and direct water use a vector for the main products in an industrial base need be constructed. In this study, we conducted field research to collect first-hand data in the Ningdong Base, during which our research team conducted site visits and interviews in many major ongoing coal mines, coal-fired power stations, and coal-to-chemical projects in Ningdong. Because the number of projects for some products are more than one, the weighted averaged values for unit water, energy and raw material consumption based on the production capacity of the projects were computed to represent the overall technical level of the product in the Base. The technical coefficient matrix and direct water use vector for major products in Ningdong is shown in

Table 4. The technical coefficient matrix and direct water consumption vector for major products in the Ningdong Base.

Technical Coefficient Matrix
Products	COAL	ELEC	CTL	CTM	CGTM	COKE	CGTO	MTO	DME	POM	NH3	UREA	CaC2	ACET	BDO	PTMEG	ACA	VAC	PVA
COAL	0	0.0003	2.230	1.926	0	1.175	2.960	5.10	2.12	2.00	1.61	0	0	0	0	0	0	0	0
ELEC	18.29	0.05	1078.37	363.68	151.95	44.06	848.61	2680.0	80.0	990.0	505.46	78.52	3262.2	206.96	918.49	869.41	91.63	162.15	696.20
CTL	0.0001	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
CTM	0	0	0	0	0	0	0.253	0	0	0	0	0	0	0	1.215	0.057	0.538	0	0.873
CGTM	0	0	0	0	0	0	0.253	0	0	0	0	0	0	0	1.215	0.057	0.538	0	0.873
COKE	0	0	0	0	0	0	0	0	0	0	0	0	0.686	0	0	0	0	0	0
CGTO	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
MTO	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
DME	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
POM	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
NH3	0	0	0	0	0	0	0	0	0	0	0	0.571	0	0	0	0	0	0	0
UREA	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
CaC2	0	0	0	0	0	0	0	0	0	0	0	0	0	3.813	0	0	0	0	0
ACET	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.282	0	0	0.325	0
BDO	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1.490	0	0	0
PTMEG	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
ACA	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.710	0
VAC	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1.819
PVA	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
Direct water consumption vector
Freshwater	0.122	0.00022	9.55	11.13	4.19	1.05	15.91	14.05	7.30	5.76	10.89	2.13	0.81	4.57	18.40	6.74	1.93	2.83	16.53

, in which the unit for electricity consumption is kWh/t (kWh/kWh for power self-consumption) and t/t for material consumption (t/kWh for electricity); the unit for water consumption is m³/t (m³/kWh for electricity).

The blue WF accounting is then conducted using the MUIO model based on the technical coefficient matrix and direct water use vector. The blue WF results are shown in Figure 2. It shows that the blue WF of different coal-based energy and chemicals differ significantly. The standard deviations of the direct water, indirect water, and WF of the products are 5.98, 12.24, and 14.83 m³/t, respectively. In general, the blue WF of products at the downstream are greater than that of the upstream products. The average blue WF of electricity in the study area is 2.51 × 10⁻⁴ m³/kWh. The washed coal in the study area has an average blue WF value of 0.126 m³/t. CaC2 and COKE both have relatively small values of WF. The coal-based chemicals at the downstream such as BDO, PVA, and PTMEG require a large quantity of freshwater along their production chain. For example, the blue WF of PTMEG is 55.25 m³/t, which means over 55 tons of freshwater is consumed for producing one single ton of PTMEG. It is worth mentioning that methanol produced using alternative routes can result in significantly different WF values. The blue WF of methanol based on the CTM route (11.65 m³/t) is over 2.75 times greater than that using the CGTM route (4.23 m³/t). Likewise, the blue WF of olefin based on the MTO route is 15.37 m³/t, while it is 18.12 m³/t using the CGTO route.

The analysis for the structure of WF can help to find the major contributor of the total WF, it was therefore analyzed in the study. Figure 3 presents the distribute proportion of direct and indirect WF for each product. The proportion of indirect WF in the total WF covers a wide range, from 0.9% for CGTM to 87.8% for PTMEG. The standard deviation of the proportion of direct WF is 31.62%. Among the products, the direct water consumption of CTL, CTM, CGTM, COKE, MTO, DME, and NH₃ accounts for more than 90% of the total WF. Most of the above-mentioned products are secondary products synthesized directly from coal. On the contrary, the tertiary and quaternary products from coal have much smaller percentage for direct water consumption. For example, the direct water stands for 12.2 to 36.78% for VAC, PVA, and PTMEG.

3.3. Validation of the Accounting Model

The WF of the same industrial product based on different production routes can be significantly different. Even with the same production route, the difference in WF can still be remarkable due to different level of water-saving, energy-saving, and material consumption technologies adopted. Thus, validation of the WF accounting model needs be conducted with the same products produced at the same sites. In the current study, the WF accounting model is validated by comparing the blue WF of the coal-based products calculated by the proposed model with the blue WF of the same products produced in Ningdong, which were accounted by a process-based model reported in literature [10]. Figure 4 illustrates the comparison results. It shows that the blue WF of the two models are consistent, with R² of 0.906, Pearson correlation coefficient (PCC) of 0.952, and normalized root mean squared error (NRMSE) of 0.112 m³/t. The results indicate the good performance of the WF accounting model.

3.4. Blue WF Sustainability Assessment

3.4.1. Standards and Norms

To assess the blue WF sustainability of the coal-based energy and chemicals at two different levels, the product WF under referenced level of technology needs be computed. China has issued a serial of national standards of water intake and energy consumption for power generation and major chemical productions [44,45,46,47,48,49,50,51]. In addition, updated standards and norms have recently been issued by the provincial government [52]. The standards of water and coal consumption for power generation differ with installed unit capacity (

Table 5. Standard quantity of water and coal consumption for unit rated capacity.

Cooling System	Freshwater Consumption for Install Unit Capacity (m3/MWh)
	<300 MW	300 MW~500 MW	≥600 MW
Circulating	3.20	2.75	2.40
DC cooling	0.79	0.54	0.46
Air cooling	0.95	0.63	0.53
Level	Coal consumption for unit of 600 MW (tce/MWh)
	Subcritical	Supercritical	Ultra-supercritical
Standard	0.319	0.306	0.293
Advanced	0.313	0.298	0.288

The conversion coefficient to standard coal equivalent is 0.1229 tce/MWh.

). The norms for water and energy consumption for washed coal and other chemicals are shown in

Table 6. Standard quantity of water and coal consumption for coal and chemicals.

Product	Freshwater Consumption (m3/t)		Energy Consumption (tce/t)
	General	Advanced	General	Advanced
COAL	0.34	0.26	0.007	0.003
CTL	10	7	2.5 *	2.2 *
CTM	15	9	2.2 *	1.8 *
CGTM	15	9	1.65 *	1.3 *
COKE	1.4	1.2	0.155	0.127
MTO	15	12	4.5	3.7
DME	11	--	1.225	1.146
POM	24	--	2.8 *	2.1 *
NH3	14	10	4	3
CaC2	1.1	--	3.2 *	3.05 *
ACET	2.1	--	--	--
BDO	23.8	--	1.5	0.95
ACA	3.2	--	0.429	0.3
VAC	8	--	0.565	0.41
PVA	10.9	--	2.75	2.072

* The ones marked with star indicate that the material coal consumption is also included.

. The analysis does not include the chemicals of which national or provincial standards are not issued.

3.4.2. Blue WF under General Standard and the BPT

Prior to the computation of WF_p^ref for the energy and chemical products, one should update the technical coefficient matrix and direct water use vector using the data in the general standards and the BPT, to compute the WF under general standard and the BPT, respectively. In the study area, there are 15 coal-fired power stations in operation, which include 31 installed units. The average capacity of the power stations is 1044 MW, whereas the average capacity of installed units is 505 MW. To compute the WF_p^ref of electricity, the values in the norms corresponding to ≥500 MW for water consumption and 600 MW for coal consumption were adopted. Ideally, the WF_p^ref is computed with referenced water, energy and materials consumption. However, only a few products have referenced raw material consumption. For example, the scope of energy consumption standards for CTL and CTM covers both thermal coal and material coal consumption. It should be noted that the energy consumption standards are given based on coal equivalent (7000 kcal/kg), thus one needs to convert the actual energy consumption into coal equivalent. The WF_p^ref versus actual WF of the coal-based energy and chemicals are shown in Figure 5.

3.4.3. Blue WF Sustainability Indicators

The blue WF sustainability assessment indicators at product and regional levels were computed and shown in Figure 6. Result shows that all products manufactured in the Ningdong Base are sustainable at both product and regional levels. At the product level, the technology of 12 products has reached the advanced level, which includes the primary energy of coal and secondary energy of coal-fired electricity. NH₃ and PVA are very close to the advanced level, with WSRI_p−1 values of 1.075 and 1.016, respectively. CTL and MTO have just crossed the sustainable thresholds, with WSRI_p−1 values of 0.921 and 0.928. At the regional level, the technology of most of the products have reached the advanced level in terms of WF, whereas CTL, CTM, and MTO are close to the advanced thresholds. Considering the production capacity of the energy and chemical products, the overall WSRI value of the Ningdong Base is 0.876 based on WF under the BPT. This means that the Ningdong Base is sustainable and advanced in terms of water consumption. It is noteworthy that, methanol produced based on the CGTM route is much smaller than that based on the CTM route. The WSRI_p−1 value of the latter is 2.71 times greater than that of the former.

4. Discussion

4.1. Volumetric WF versus Impacted-Oriented WF

The choice of WF accounting approach depends on the goal of a study conducted. The WFN-WF is more preferable for studies in the water resources management, especially in the efficiency, sustainability, and equability of water resources allocation and the productivity of freshwater use [10,23,24]. LCA-based WF, on the other hand, is more appropriate for assessing the potential environmental impact of different products or alternative production processes at different levels, especially when regional water availability is considered [4,25,26,27]. Our study adopted the LCA-based WF to assess the sustainability of different coal-based energy and chemicals at product and regional levels. Results indicate that, as one of the national modern coal chemical industry bases, the wastewater treatment, recycling, and reuse technologies that have widely adopted in the base have offset the intensified water scarcity that would have been caused by the large-scale industrial production. However, further WF reducing measures should be implemented in some major projects in the industrial base; investment in technical improvement in the process unit with lower water, energy requirements is highly encouraged. For example, the CTL project, which is regarded as the “No. 1 Project ” of Ningxia Province, has relatively large product WF comparing to that of other projects in the base. Likewise, there also exist a gap between the average level of WF of the MTO projects in the base and the advanced level. Furthermore, more frequent use of byproducts of upstream can reduce the potential environmental impact of large-scale production. For example, the use of coke oven gas, a main byproduct of COKE, for the methanol production may enhance the environmental sustainability of an industrial base.

4.2. Improve the Sustainability

As one of the nation’s major modern coal chemical industry bases, Ningdong has launched a comprehensive control system for energy and water conservation and environmental protection. We, thus, expected that the water-saving and energy-saving technologies adopted for production of the products in Ningdong are superior compared to their corresponding national average. Obviously, the results of the WF sustainability analysis is consistent with our expectation. According to the on-site data we collected, the average wastewater reuse rate of production of the major products is 34.18%; it is contributed by 15 reclaimed water treatment plants with a total capacity of 3.52 × 10⁵ m³/d in the study area. In addition, three pit water treatment plants in the base with a total capacity of 1.2 × 10⁵ m³/d are also in operation. A question is whether pit water use should be included in the WF accounting. In our analysis, the use of pit water was excluded considering that (1) pit water is mainly interstitial water collected along with mining and cannot be used as water resources without proper treatment; (2) direct discharge of the metal-rich pit water is a great threat to the environment. In the study area, the product WF would be 17.02% to 104.4% greater if the use of reclaimed water and pit water were substituted with freshwater, which would lead to unsustainable production for majority of the products. For example, the WF of COAL and CTL would be 0.245 m³/t and 20.64 m³/t, respectively, if only freshwater were used. Therefore, besides the efficiency improvement of water, energy, and material use, increasing the ratio of nontraditional water use is also an important method to enhance the WF sustainability. In fact, according to a recent water utilization plan in Ningxia [53], the Province proposes to substantially increase the utilization of nontraditional water resources. The plan clearly stated that, by the year of 2025, the utilization rate of reclaimed water and pit water shall reach 50% and 90%, respectively [53].

4.3. Enhance the Life-Cycle Thinking for Water Management

In the recent years, China’s central and local authorities have issued serious of regulations and plans to improve the conservation and utilization of water resources in water scarcity basins [53,54]. A very recent plan released by five China’s ministries set clear goals for the establishment of rigid restraint system of water resources, promoting water conservation, utilization of unconventional water resources, etc. Meanwhile, the Ministry of Water Resources of China lately claimed that a national water quota system has been basically established. This system covers water quotas for 105 products, including 70 industrial products. However, all the hard efforts that have been made were solely in direct water utilization, neglecting the importance of indirect water use along the supply chains. Consequently, the analysis on water consumption can be incomplete. In the case study, the WF of all energy and chemical products produced are sustainable, which would lead to inconsistent results if only direct water consumption were considered. For example, the direct water consumption for PVA in Ningdong is 16.53 m³/t, which exceeds its general standard of 8 m³/t and will be determined as unsustainable. This inconsistency is due to the fact that the relatively-high efficient use of energy and materials in the PVA production process, as well as the efficient water use of the upstream production compensate the inefficient freshwater consumption in the PVA production process. These results further address the necessity of life-cycle thinking (LCT) for water resources management, which seeks to identify water use improvement opportunities at all stages across the life cycle. The LCT can provide a comprehensive approach in support of the overall reduction of environmental impacts in water resources utilization.

4.4. Establishment of National WF Benchmarks

The purpose of WF sustainability studies generally can be grouped into two categories: (1) assessing whether the WF of a product unnecessarily contributes to global, national, regional or local WF of humanity and (2) assessing whether the WF contributes to specific hotspots [55]. The interest of our study lies in the former one, which compares each separate product WF with a benchmark for that product. However, although we have established WF benchmarks, they were merely based on norms of water and coal consumption for power generation, and the standards of water and energy consumption for washed coal and other chemicals. The standards of materials consumption for most of the products were not considered simply because such standards do not exist. The limitation of the current study reveals the necessity of establishment of standardized national or regional WF benchmarks in China, especially for the major products in the water-intensive coal-to-chemical industry. The WF benchmark values can be used as instruments to evaluate the advancement of specific production technology, set criteria for newly invested production capacity, as well as formulate WF reduction targets. Reference [56] introduced a set of global WF benchmarks for over 120 agricultural products, but the literature lacks studies on WF benchmark setting for industrial products. In addition, the result of the current study also suggests that the regionalization and industrial technology consideration are necessary for the establishment of such benchmarks. That is, the level of regional or local water scarcity, and the choice of different industrial routes should also be fully considered in future the analysis.

5. Conclusions

This study attempts to fill in a current research gap by introducing a methodological framework for assessing the blue WF sustainability of multiple interdependent products in a system. The Mixed-Unit Input-Output (MUIO) model is adopted in the framework to account WF of the products, and three sustainability assessment indicators are then proposed. A large modern coal chemical industry base in Northwest China, in which 19 major coal-based energy and chemical products are produced is used as a case study. Technical coefficient matrix and direct water consumption vector for the MUIO model were constructed based on first-hand data collected by on-site field research in the study area, after which WF accounting and sustainability assessment were conducted at product and regional levels. The conclusions drawn from the proposed framework, as well as from the results and discussion of the real-world case are as follows: (1) although the top-down approach is usually applied to investigate the interdependent among industry sectors in terms of water consumption, our method has generalized it to calculate the blue WF of multiple interdependent products. The validation results indicate the good performance of the model. (2) Instead of using the IO tables that are directly compiled at national or provincial levels for regional and global scale WF analysis, the proposed method usually requires on-site data collection and computations, based on which the technical coefficient matrix and direct water consumption vector of the products are constructed. (3) In the case study, the blue WF of the coal-based products differ significantly. The standard deviation of the blue WF of the products in the study area is 14.83 m³/t, to which the indirect water contributes much more than direct water. (4) Generally, lowering the indirect water use is the key to WF reduction for the downstream products whereas lowering direct water use is more important for the upstream product WF reduction. (5) Although the blue WF of all products manufactured in the study area are sustainable at both product and regional levels, further WF reducing measure should be implemented for several major products such as CTL and MTO. (6) To enhance the blue WF sustainability, the ratio of nontraditional water resources to total water use should also be further increased. (7) The LCT should be adopted to provide a comprehensive approach in support of the overall reduction of environmental impacts in water resources utilization in China’s arid coal bases. (8) It is suggested to establish standardized national or regional WF benchmarks for the major products in the water-intensive coal-to-chemical industry.