1. Introduction
Decoupling environmental impacts from human activities is one of the important targets for achieving sustainable development. As underlined by Kanda and Nakagami [
1], the linkage between wealth and environmental impact is complex combining policies related to production (cleaner production, resource productivity) and consumption (service intensity and sufficiency). Though the literature on production and consumption indicators is abundant [
2,
3,
4,
5], the combination of the two approaches seems to be the appanage of product-service systems (PSS). Product service systems can be defined as a mix of tangible products and intangible services designed and combined so that they are jointly capable of fulfilling final customer needs [
6]. The sustainability effect of PSS is hypothesized from an aligned incentive between producers and customers on reducing the need for production compared to traditional business models in which producers’ benefits are directly dependent on the amounts produced [
7]. Though environmental benefits have been observed, they are not automatic [
8] or are difficult to assess because of the intangible part of the PSS [
9]. Besides, the implementation of PSS in the economy is still rare, making it difficult to assess its efficiency for decoupling policies [
10].
Product service systems focus on the service provided to the consumer. In the field of water supply, the main service is the quality of the water supplied. Indeed, the water quality is of core importance as 88% of the 1.8 million annual deaths due to diarrheal diseases are attributable to unsafe water supply, sanitation, and hygiene [
11]. Besides, the absence of a safe water supply can severely impede economic development and the associated reduction of poverty. Consequently, access to drinkable water has been recognized as a core target of the millennium development goals [
12]. The quality standard of drinking water is established by the World Health Organization [
13] and declined at national levels as drinking water norms. The present research focuses on turbidity (T) measured in Nephelometric Turbidity units (NTU) and total organic carbon (TOC) as physical quality parameters. Turbidity standards exist in 85 countries and vary between 0.3 to 25 NTU with a median value of 5 NTU [
14]. In developed countries such as Japan and France, where large municipal supplies are common, the standard is in the lower range and equals 2 NTU [
15,
16]. In practice, however, the water supplied in France has a turbidity of 0.1 NTU [
17] following the WHO recommendations of 0.2 NTU or less for large municipal supplies [
13]. The standard values for TOC are 3 mg/L in Japan [
16] and 2 mg/L in France [
15]. In practice, French water suppliers use the value of 1.3 mg/L [
17].
The difference in quality between the extracted water and the quality standards demands the use of treatment processes. These processes consume energy, materials, and financial resources. The municipal water system indeed accounts for approximately 3–4 percent of direct energy use in the United States and represents as much as 30–40 percent of the total energy used by many local authorities [
18]. Typical energy efficiency values for conventional treatment processes are 0.05–0.15 kWh/m
3 but it reaches as much as 5.5–8.0 kWh/m
3 for desalination processes [
19]. The energy consumption for distribution process is 0.10 kWh/m
3 in Sweden [
20], 0.03–0.58 kWh/m
3 in Germany, and 0.18–0.32 kWh/m
3 in California [
21]. Material consumption is also large combining materials used during the treatment processes (ozone, chlorine, coagulants, sand, etc.) and those related to the treatment plant and distribution infrastructures such as tanks and pipelines.
All the usages of water by households do not, however, require the drinking standards. Usages such as toilet flushing, bath or showers, laundry, cooking, gardening, and house cleaning indeed require lower quality water so that a reduction in the energy and material consumption associated with water treatment could be achieved. Water systems with differentiated water quality targets have been denominated as dual water systems in the literature. Barker et al. [
22] showed the reuse of treated wastewater effluent for non-potable purposes in North Carolina could consume less energy than when using the potable network. Rainwater harvesting potential was investigated in Malaysia [
23] and Vietnam [
24] whereas the deterioration of source water quality in China [
25] made a dual water distribution system both more economical and practical than the conventional system. For developed countries, the investigation of dual water systems is important in the context of aging infrastructure as a possibility for improving the efficiency of the current centralized municipal water system [
26]. For developing countries, the poor quality and the absence of maintenance of the distribution system have often led to water leakages and poor water quality. The delivery of bottled water for potable use and non-potable water by pipelines has been proposed [
27] as a solution but at the cost of the high energy consumption associated with the production of bottled water, 316–370 kWh/m
3 [
28].
Another issue of dual water systems is the determination of the share of the household water consumption associated with non-potable purposes. The amount of water consumed varies indeed among developed countries from 130 L/capita/day in Hamburg, Germany to 350 L/capita/day in Winnipeg, Canada and is associated with different usages [
29]. These differences are due to cultural differences, e.g., washing cars at home instead of at a car-washing facility, and technical differences such as those related to flushing technologies. The potable use of water concerns the cooking activity, which represents 15–25% of the household water consumption in developed countries [
29]. Dual water systems studied in the literature are nonetheless stricter considering bath and showers as part of the potable use. This is understandable as most countries have only bathing regulations related to recreational activities. In this context, it may be safer for health authorities to qualify as non-potable the activities for which the water does not enter into contact with the human body. This safety net has, however, consequences on the interest of developing dual water systems since the water consumption associated with bath accounts for as much as 25–40% of household water consumption [
29]. In Japan, however, due to the ancestral culture of public baths, a bathing regulation exists and recommends a water quality of 2 NTU for turbidity and of 2 mg/L for TOC [
30]. The consequence of considering bathing activities as a non-potable usage on cost, material, and the energy consumption is investigated in the present research. In this regard, our analysis of the literature on dual water systems revealed that material, energy, and cost savings are rarely investigated simultaneously and that dual water systems have so far not been integrated into the broader concept of product-service systems. This may be explained by the technical difficulties of linking water quality targets to cost, material, and energy consumption of treatment processes. Such an integrated model does not exist yet in Japan for example, but recent developments such as the EVALEAU model used in the present study propose process modeling life cycle assessment models to analyze the characteristics of treatment processes based on the delivered water quality [
31,
32]. The EVALEAU model deals with the municipal water supply in France where Euzen and Levi [
33] estimated that 17% and 39% of household water were consumed for cooking and bathing activities, respectively. The application of the Japanese bathing standard on this framework means that as much as 83% of the water supplied to French households could be considered as non-potable.
The present research has several objectives. First, dual water systems are incorporated into the broader concept of product-service system using water quality as the service indicator and the functional hierarchy model. Second, bathing activities are, to our knowledge, investigated for the first time with lower quality standard requirements than drinking standard. Third, contrary to most researches on dual water systems that are focused on the use of reclaimed water [
22,
23] and the comparison of the energy efficiency between dual and conventional systems, the present work assesses simultaneously the material, energy, and cost savings induced by the differentiation of water quality by usage. The integrated model EVALEAU is applied to a conventional treatment system of French municipal water supply as an illustration of how material, cost, and energy can be saved when degrading the water quality requirements. Three steps are considered:
- (1)
The description of the water quality framework, the treatment line, and the important input parameters retained for our case study.
- (2)
The detailed analysis of the material and energy savings obtained by the differentiation of water quality targets.
- (3)
A thorough discussion of the implications of the observed savings for the future of water supply.