3.1. Performance of the Practices
Table 5 and
Table 6 summarize the performance and unweighted performance ranking values of the 12 studied water resource management practices. Determination of the performance rankings was assumed to be proportional to water-supply potential or overall systemic energy efficiency. However, the calculation and assumptions on the water-supply potential or overall systemic energy efficiency might be site specific and can be only applied to the case study. For instance, the water-supply potential for residential water conservation program may be greater in populated urban area than that in sub-urban area. Other the other hand, the calculations for wastewater reclamation potential (Equation (2)) and rainwater harvesting (Equation (3)) are for general uses. One thing to note is that the functional unit for water-supply potential and systemic energy efficiency was based on a unit-period of one year (e.g., annual provision of water supply).
Water-supply potential for the studied practices ranged from 1050 to 655,000 thousand m
3 per year for domestic wastewater reclamation and reduction of leakage rate on irrigation canals, respectively. The lowest potential for domestic wastewater reclamation (1050 thousand m
3 per year) was due to limited source of wastewater for reclamation, as the total population in these two counties only accounted for about 8.5% of total population in Taiwan. Similarly, the potential for a water conservation program in residential sector (5026 thousand m
3 per year) was estimated at the lower end among the studied practices, assuming a water saving of 20 L per capita per day. Water savings for water use efficiency measures of showerhead, toilet and clothes washers were reported at approximately 106, 131, 150 L per household per day, respectively [
27].
Water-supply potential for industrial wastewater reclamation (6300 thousand m
3 per year) was confined by their specification of existing measures. Manufactures that located in government-organized industrial area were mandated to improve their water use efficiency, by detecting of abnormal water usages, increasing water reuse in process system or reusing of reclaimed water for secondary use [
33,
34]. The expected water-supply potential was calculated using existing measures or implementation of additional on-site measures by the Formosa Plastics Corporation (FPC) [
9]. A seawater desalination plant at 100,000 m
3 per day was also proposed by the FPC to top-up their water demand and consequently to reduce the industrial withdrawal of potable water.
The water-supply potential for rainwater harvesting was highly depended by weather condition (precipitation) and available area for rainwater catchment, as described in Equation (3). The potential for rainwater harvesting ranged from 3900 to 8058 thousand m
3 per year, considering the availability of suitable area or site for implementation of the system. A study on using harvested rainwater for buildings revealed fair to moderate level of reliability from satisfying about 37% of non-potable water demand by harvested rainwater. The amount of harvested rainwater was estimated based on annual rainfall of around 1000 mm and effective roof area of 1600 m
2 [
39].
The water-supply potential for agriculture-related practices ranked the highest among other practices. This is due to the fact that the major economic activity in Choshui river basin is agriculture, and it shares the highest water demands among other economic sectors, accounting for about 74% of freshwater withdrawals. The highest potential of 655 billion m
3 per year from reducing leakage rate in irrigation canals could be attributed to the long length of irrigation canals (9,355,730 m). The expected water savings from reducing leakage is at 70 m
3 per meter of canal per year [
24]. Likewise, the potential of 225 billion m
3 per year was obtained from promoting drip irrigation or other smart irrigation measures, based on a water saving rate of 2000 m
3 per hectare, compared to conventional irrigation methods [
9,
24]. Reuse of agricultural return water was proportional to agricultural water demand, resulting a water saving estimation of approximately 200 billion m
3 per year at an ultimate recovery rate of 50% [
43].
The performance of systemic energy efficiency for studied water management practices was assumed to be proportional to the water-supply potential. For practices that provide water savings or reduced water demand were expected to have less systemic energy demand as compared to baseline (i.e., less energy requirement for abstraction and treatment of source water), thus, contributed to relatively higher performance ranking (
Table 6). The ranking of water conservation program for agricultural sector was greater than that for residential sector. On the contrary, practices that require additional treatments were expected to have higher systemic energy demand than those in baseline, resulting in relatively lower performance ranking.
The energy requirements for water sectors at a global level are 0.0002–1.74 kWh/m
3 for surface water supply, 0.37–1.44 kWh/m
3 for groundwater pumping and 2.4–8.5 kWh/m
3 for desalination (using membrane-based technology) [
44]. The energy demand for state of the art seawater reverse osmosis (RO) with energy recovery devices (ERD) was assumed to be 3.1 kWh/m
3 [
37], which is within the range of a modern seawater RO plant at overall energy demand of 3.0 to 4.0 kWh/m
3 [
38]. Fane reviewed developed strategies and technologies for desalination to achieve lower energy demand and reduced carbon footprint [
29]. The study proposed the use of improved membranes for potentially reducing energy demand by 15% to 20%. Rainwater catchment, on the other hand, may only require energy for pumping from its simple design, with median energy intensity from 0.2 to 1.4 kWh/m
3 [
4]. An evaluation study of alternative water sources by Cook et al. reported an energy demand for rainwater system at 0.44 kWh/m
3, which was mainly from pumping from storage tank [
39]. A mean value of 1.0 kWh/m
3 was used as energy demand for rainwater harvesting system in this study.
Wastewater reclamation, either from domestic or industrial, requires a remarkable amount of energy for treatment. Energy demand for wastewater reclamation in this study was 2.14, 1.5, and 1.1 kWh/m
3 for domestic, industrial, and agriculture, respectively. The highest energy demand for domestic wastewater reclamation was due to the use of tertiary treatment for better product water quality for industrial end-uses [
3], while reusing of irrigation return water might only require relatively lower energy for pumping and simple treatment such as filtration. In particular, salinity levels from sodium and chloride salts are undesirable in irrigation water [
45]. The energy demand for industrial wastewater reclamation fell in between the demand for domestic and agriculture, as a result of application of advanced treatment (activated carbon filtration and RO) from cascaded wastewater [
33].
3.2. MCDA Results
The MCDA results showed differentiated ranking for the 12 studied water resource management practices (
Figure 3). A sensitivity analysis was also conducted to evaluate the influence of criteria weights on the MCDA results (
Figure 4).
Figure 4a shows the results from making all the criteria are equally important, while
Figure 4b,c depict the results of ranking water-supply potential and systemic energy efficiency the most important criteria, respectively, considering those two criteria may be prioritized before economic feasibility. These scenarios were set for maintaining the stability in both water and energy systems, which could be attributed to the minimization of time delay associated with the constructions of new infrastructures (e.g., desalination or wastewater reclamation plants) [
46].
Practices for improvement in effective water management such as monitoring and management of groundwater pumping (C3) scored the highest, followed by leakage rate reduction for agricultural irrigation canals (C2). This could be attributed to their relatively higher economic feasibility and water-supply potential. Their high water-supply potential also led to moderate to high systemic energy efficiency, as a result of providing both water and energy savings. Groundwater depletion is an inevitable global challenge from over-withdrawing water from aquifers, which also raised the issues of decreased well yields, increased pumping costs, deteriorated water quality, damaged aquatic ecosystem, and increased land subsidence [
47]. The action of scientific monitoring and assessment of groundwater therefore was applied to ensure a sustainable source of groundwater and to prevent saltwater intrusion in coastal aquifers [
48].
Water conservation programs for residential (A1) and agricultural (A2) sectors were also ranked as high priority in the studied water resource management practices, as a result of their relatively high economic feasibility, especially for the residential sector. Studies on residential water end-uses have found that implementation of water conservation programs showed positive influence on overall water demand and led to significant water savings [
27,
49], but offsetting behavior or rebound effect might occur after a period of time [
50]. The water savings also contributed to significant energy savings during water uses [
27]. The study on determinants of residential water demand in Italy claimed that application of tariff had a negative impact on residential water consumption, which cloud be seen as a relevant driver of domestic water consumption [
49]. Yunlin County, the case study, has a variety of water use patterns. Several residential areas in the county were reported to have the highest water footprints, whereas manufacturing-dominated area such as Mailiao Township showed relative smaller water footprint than in other townships. Industrial development in the township had little influence on the individual use of water. Therefore, attention on water conservation in the region was suggested to be paid to the reuse and recycling of industrial wastewater [
17].
Water conservation program or irrigation return water reuse (B7) in the agriculture sector, on the contrary, were expected to provide a great potential in water savings, as the agriculture accounted for approximately 60–90% of consumptive water use worldwide. It was also directly linked to the sustainability goals in agriculture water use. Efforts have been made to initiate programs that focused on agriculture water conservation or withdrawal restrictions. Clean technologies in agriculture aiming at reducing resource inputs (water, energy, and other constitutes), producing renewable energy or protecting the environment were studied and prioritized for decision making [
26]. Field implementation of improved automated irrigation system based on crop and site characteristics along with fuzzy decision support approach was used with remarkable water savings [
51]. An investigation on options and difficulties to improve water-efficient practices in irrigation showed that the farmers may lack adequate knowledge or requires strong incentives to make extra efforts to improve water efficiency level [
52]. An optimization study of land and water resource allocation for irrigation revealed that, although net irrigation return was high for higher deviations in existing cropping pattern, 20% of deviation was suggested to be the best alternative for satisfying the socio-economic requirements [
53].
Several studies have reported on the implementation of alternative water sources, such as rainwater or desalinated seawater, for regional or local use. Notable examples included harvesting runoff from roofs for toilet flushing and landscape irrigation [
39]. In a study of public acceptance of alternative water sources from nine locations, desalinated seawater was preferred for most of the water uses, while harvested rainwater was second preferred for cleaning, toilet flushing and garden watering [
53]. Rainwater harvesting (B4, B5, and B6) is an easy-to-adopt water management measure and has shown great potential to improve water security during extreme weather events. Therefore, it was recommended in this study to collect local available rainwater as an alternative water source. Yet, studies are still discussing the challenges of its practical use, including contamination, general lack of design criteria [
41] as well as on instability relative to climate change. The performance of rainwater harvesting systems was expected to be reduced by 2–14% under the influence of climate change, and the system was more affected in the dry season than in the wet season [
40].
The least preferable water resource management practices fall into the category of wastewater reclamation, either from domestic (B1) or industrial (B2) sources, mainly due to their relatively lower systemic energy efficiency. Wastewater reclamation processes and desalination all required intensive use of energy, particularly for the ones with pressure pumps or advanced treatments [
2]. Water uses and reuses in industries varied greatly and could be industry-specific. This implied the necessity of proper management and selection of water recycling schemes, in order to fully satisfy the water quality standards at user ends [
34]. Studies also showed that adoption of water resource management practices in industries might not be a pure concern for environmental sustainability or business strategy development, economic opportunity turned out to be a key challenge for promoting the practices [
33]. All of these indicated that manufacturers reside in this study region might be obligated to invest adaptation measures by their own, as the water rights of the public and agricultural use were prioritized over industrial use during dry seasons [
9].
Our results demonstrated the importance of considering water–energy nexus in decision making. The results from sensitivity analysis showed completely different trends for practices with relatively high water-supply potential but required additional energy use (e.g., B1, B2, B3, and B7), as shown in
Figure 4b,c. Additionally, those practices may become less favorable considering their potentially large costs than the conventional water sources, human health risk, and public perception [
54]. These findings are similar to the arguments in Ghisi et al. that selection of water management strategies should be based on indicators other than only the potential for water savings [
55]. This is also the reason that the MCDA played an important role in this study to aggregate various criteria with different units of measurement for better decision making. The results from this study further support the observations made by Chen et al. [
9], in which future studies on planning and implementing of water-related adaptation pathways require dynamic monitoring and modification in responding to future climate and social-economic changes.