The global community had launched a global target related to access to sustainable safe drinking water sources and had pursued a global monitoring and reporting program on progress made on access to drinking water sources. The universal indicator considered by the global community for this purpose is the proportion of the population with (or without) access to improved drinking water sources. An improved drinking water source is defined as one that is likely to be protected from outside contamination. However, monitoring and reporting progress using this indicator does not reflect the quality of service delivery, in particular the quantity, quality and continuity of drinking water supply. This paper presents the evaluation of the quality of service of the water supply in Lebanon based on additional national-specific indicators that can properly inform the quality of water supply service delivery. A method of evaluation of quality of service index (QSI) is proposed based on a fuzzy inference system. The results of evaluation of the QSI for the four regional water supply authorities in Lebanon for the years 2000, 2005, 2010 and 2014 are then presented and discussed. Finally, the conclusion and further developments are presented.

INTRODUCTION

The development commitments expressed at the United Nations Millennium Summit in 2000 led to the formulation and adoption of the Millennium Development Goals (MDGs) (United Nations 2000). The MDGs formalize eight development goals, whose progress is monitored and reported upon through a series of targets and associated indicators for measuring achievement by the target year of 2015. The MDG No. 7 on ensuring environmental sustainability includes one target related to drinking water supply and sanitation: ‘halve, by 2015, the proportion of people without sustainable access to safe drinking water and basic sanitation’. The water supply part of this target is evaluated by the indicator ‘proportion of population using an improved (or unimproved) drinking water source’ (United Nations 2008).

The Joint Monitoring Program (JMP) for water supply and sanitation, arranged by the World Health Organization (WHO) and the United Nations Children's Fund (UNICEF), serves as the institutionalized mechanisms for monitoring the MDG target on access to improved water source at the global level (WHO & UNICEF 2005). An improved water source is defined by the JMP as one that, by nature of its construction or through active intervention, is likely to be protected from outside contamination, in particular from contamination with fecal matter. Therefore, the proportion of the population using an improved drinking water source is the percentage of the population who use any of the following types of water supply for drinking: piped water into dwelling, plot or yard; public tap/standpipe; borehole/tube well; protected dug well; protected spring; rainwater collection and bottled water (if a secondary available source is also improved). It does not include unprotected well, unprotected spring, water provided by carts with small tanks/drums, tanker truck-provided water or surface water taken directly from rivers, ponds, streams, lakes, dams or irrigation channels. Definitions and detailed description of these facilities can be obtained from JMP (2014).

The above-mentioned indicator is computed as the ratio of the number of people who use an improved drinking water source to the total population, expressed as a percentage. The estimates of proportion using improved drinking water sources originate from data collected by national statistics offices and international survey programs (WHO & UNICEF 2006).

Quality of drinking water service delivery could not be evaluated using the universal water supply indicator (United Nations Economic and Social Commission for Western Asia (UN-ESCWA) 2013a). Actually, this indicator does not reflect the effective drinking water quantity and quality, nor does it consider the continuity of water supply. For example, while the latest JMP report estimates that 100% of the Lebanese population have access to improved drinking water sources, this does not mean that 100% of the total population have regular and effective access to water supply, or that the supplied water is adequate for drinking. This demonstrates the need to develop additional national-specific indicators that should more appropriately inform the delivery of water supply services. These additional indicators should in turn be used to better measure progress on sustainable access to safe drinking water and, therefore, would result in effective monitoring and reporting of progress towards quality of water supply service delivery.

The recent report of the Open Working Group of the United Nations General Assembly on Sustainable Development Goals includes a proposal for a global goal for water: ‘ensure availability and sustainable management of water and sanitation for all’ (United Nations 2014). The proposal incorporates a target to monitor the access to drinking water: ‘target no. 1: by 2030, achieve universal and equitable access to safe and affordable drinking water for all’. No formal indicators or methods of measurement have been set yet by the international community to measure the proposed target.

This study proposes a methodological framework for quantifying and assessing the quality of water supply service for sustainable access to piped water supply in Lebanon. In addition to the universal drinking water supply indicator (‘proportion of population using an improved drinking water source’), three other indicators are considered and the whole are aggregated into an overall quality of service index (QSI) using a fuzzy inference model. The proposed approach is then applied to the drinking water supply service delivery in Lebanon.

QUALITY OF SERVICE INDICATORS

Various agencies and organizations worldwide have developed detailed water supply performance indicators that extensively cover all the aspects of the water supply systems (Ashley & Hopkinson 2002; Alegre et al. 2006; Ocampo Duque 2008; Kanakoudis et al. 2011; Van Den Berg & Danilenko 2011). The International Water Association (IWA) has proposed a database of 170 water supply performance indicators based on 232 variables; these indicators are grouped into six groups of indicators: water resources, personal (staffing), physical, operational, quality of service (customer satisfaction) and economic and financial (Alegre et al. 2006).

The universal water supply indicator falls within the quality of service group of indicators. A variety of water system performance indicators have been elaborated to evaluate a water system's serviceability (Alegre et al. 2006; Kanakoudis et al. 2011; Van Den Berg & Danilenko 2011; Haider et al. 2013; Zeraebruka et al. 2014). IWA had considered 34 indicators related to quality of service aspect, which are clustered into six subgroups as follows: coverage (five indicators), public taps and standpipes (four indicators), pressure and continuity of water supply (eight indicators), quality of supplied water (five indicators), service connection and meter installation and repairs (three indicators) and customer complaints (nine indicators) (Alegre et al. 2006).

The evaluation of the above-mentioned indicators is constrained by many practical problems regarding a lack of data and particularities related to local conditions of water supply systems in each country. An effort has been pursued at the Arab regional level with the launching of the regional initiative for the development of a mechanism to monitor the implementation of the MDGs related to water and sanitation in the Arab region (MDG + initiative), which is an outcome of a series of resolutions adopted by the Arab Ministerial Water Council through a set of indicators that respond to regional concerns for monitoring and reporting on access to water supply and sanitation services in the Arab region (UN-ESCWA 2013a, 2013b). These additional regional-specific indicators are as follows:

  • proportion of population connected to water supply network (network coverage);

  • water consumption;

  • continuity of water supply;

  • water quality.

Aside from the fact that the above-mentioned indicators were regionally proposed and agreed upon, they fully reflect the concerns about the regularity and reliability of the water supply service delivery in Lebanon. Therefore, we will consider these indicators to evaluate an overall quality of service index (QSI) to be applied to the water supply service delivery in Lebanon.

Set forth is a brief description and justification of use of each of the proposed indicators.

Proportion of population connected to water supply network (network coverage)

This indicator is the universal water supply indicator that measures the proportion of the population using piped drinking water on premises (%). This measure is considered according to the MDG water supply target as access to an improved drinking water source.

Water consumption

This indicator measures the proportion of the effective domestic water consumption by person to the national water consumption target (%). The effective domestic water consumption is measured as the actual average amount of water consumed daily by each person using a household connection (liters per person per day). The national water consumption target should be set by water authorities as part of the national water sector strategy (Ministry of Energy and Water (MEW) 2010). In some water supply systems in Lebanon, access to the water supply network does not imply that the connected population have access to sufficient water quantity. Therefore, the evaluation of the quality of service of water supply delivery necessitates the measurement of the average water quantity that is actually consumed by the connected users.

Continuity of supply

This indicator measures the level of service received by consumers ranging between continuous and intermittent supply, it is calculated as the proportion of the annual number of hours of water supply as a percentage of the total annual number of hours (%). In many developing countries, the water supply network is frequently subject to intermittent supply, in this case, the reliability of the service is usually very low and may oblige households to buy supplemental water supply or to equip with in-house water storage facilities, both of which may exert an additional financial burden, and can also have detrimental impacts on water quality. Therefore, the continuity of water supply indicator is a crucial element of the evaluation of the quality of service of water supply delivery.

Water quality

This indicator measures the proportion of the annual volume of the produced and supplied water that has been disinfected at the source as a percentage of the total annual volume of produced and supplied water (%) (UN-ESCWA 2013a). Some water distribution systems in many developing countries are supplied with water directly from sources that may be considered of good quality, or, if installed, the disinfection equipments may not operate due to technical difficulties. Although this indicator does not require measurement of the quality of water at the user end due to the high costs of water testing, it is assumed that disinfection at the water source provides adequate protection for users. Definitely, additional information on water quality can elucidate the safety of the supplied water, but for the time being, weighing the added value of the water testing data at the user end against the associated costs, the proposed indicator serves the purpose with minimal cost.

Table 1 provides measurable variables and methods of calculation of the considered indicators. The data of the measurable variables are obtained from the available records at the Regional Water Authorities (RWAs) and the MEW in Lebanon.

Table 1

Measurable variables and methods of calculation of the selected indicators

IndicatorsMeasurable variablesCalculation
Network coverage (NC) CP = no. of population connected to the water supply network, TP = total no. of population  
Water consumption (WC) AC = average water consumption (l/person/day), TC = national target of the average water consumption (l/person/day)  
Continuity of water supply (CS) NH = number of hours of water supply during a year, TH = total hours in a year  
Water quality (WQ) PW = volume of the produced and supplied drinking water per year that has been disinfected at the source (m3/year), TW = total volume of produced and supplied drinking water per year (m3/year)  
IndicatorsMeasurable variablesCalculation
Network coverage (NC) CP = no. of population connected to the water supply network, TP = total no. of population  
Water consumption (WC) AC = average water consumption (l/person/day), TC = national target of the average water consumption (l/person/day)  
Continuity of water supply (CS) NH = number of hours of water supply during a year, TH = total hours in a year  
Water quality (WQ) PW = volume of the produced and supplied drinking water per year that has been disinfected at the source (m3/year), TW = total volume of produced and supplied drinking water per year (m3/year)  

Before presenting the methodological approach to evaluate the QSI in the next section, it is important to mention that each country is particular by its institutional, social, environmental and economic conditions that influence its water supply policies and strategies. Therefore, the identified indicators and their methods of calculation could be approached in different ways according to the specific conditions of each country.

PROPOSED METHOD OF EVALUATION

Various international development organizations and groups of research have sought to define aggregate indicators for the purpose of monitoring and reporting well-being (Bossel 1999; Sadiq 2010; Hák et al. 2012; Mori & Christodoulou 2012). The Organisation for Economic Co-operation and Development, OECD (2008), provides a guide to the construction and use of aggregate indices, and a review of major sustainable development indices/indicators can be found in Mori & Christodoulou (2012). However, several other studies elucidate the difficulties and risks of developing aggregate indices, in view of the uncertainty inherent in the aggregation process and the multidimensional aspects of well-being (Beranger & Verdier-Chouchane 2007; García & Kovacevic 2011; Klugman et al. 2011; Mori & Christodoulou 2012; Uppal & Mudakkar 2013). The commonly used operator to aggregate multiple indicators measuring multidimensional sustainable development aspects is the weighted average operator (WAO), where weights are given by experts to represent the importance of the considered indicator (Bossel 1999; Sadiq 2010; Hak et al. 2012; Mori & Christodoulou 2012). Simplicity is the main advantage of this commonly used aggregation operator. However, the WAOs are suitable only if all the measures of the indicators are additive, i.e., the indicators are independent of each other. If there is synergy or redundancy among the indicators, i.e., measures are non-additive, information that is redundant is not properly accounted for by using WAO, which may lead to a bias in the overall assessment of the aggregated index (Karnib 2014).

There is a need for evaluation of the quality of service of the water supply delivery as an overall performance index, in view of the four above-mentioned indicators. The four indicators considered in this study are not independent of each other. For example, the water consumption is directly linked to the proportion of population connected to water supply network (network coverage) and to the continuity of water supply indicators. In addition, water quality, water consumption and continuity of water supply are strongly associated with health concerns. Moreover, the quality of service could not be scored high if the performance of one of the proposed indicators is low. No WAO can yield to the above-mentioned performance of the aggregation index. Thus, it is necessary to use a well-documented approach that responds to the above-mentioned performance and achieves an acceptable level of accuracy. Among the existing approaches, the fuzzy sets theory, and particularly, fuzzy inference is a suitable approach that responds to our guiding principles mentioned above (Zadeh 1975; Mamdani 1976; Tong Tong 1995). Therefore, the development of our model will be based on the Mamdani-type fuzzy inference system to generate the aggregate QSI from a set of individual indicators.

To develop a fuzzy inference system, three components should be defined: if-then rules, membership functions and fuzzy logic analysis (Mamdani 1976; Tong Tong 1995). These components will be addressed in a sequence.

If-then rules

Conditional rules are the most easily understood by water experts (Karnib 2004; Bagheri et al. 2006). They could be simply applied for the representation of know-how in the system studied. These rules are as follows:

  • if ‘continuity of water supply’ is high, then ‘QSI’ is high;

  • if ‘network coverage’ is high and ‘water consumption’ is very low, then ‘QSI’ is very low;

  • if ‘continuity of water supply’ is very high, then ‘QSI’ is very high;

  • if ‘network coverage’ is very high and ‘continuity of water supply’ is very low, then ‘QSI’ is low.

Two types of uncertainty exist in conditional rules, for example, for the following rule: if ‘water supply network coverage’ is low then ‘QSI’ is low, the first form of uncertainty is imperfect information. To what degree do we believe that ‘water supply network coverage’ is low? The second form of uncertainty relates to the imperfection of the rule. To what degree do we believe that the ‘QSI degree’ should be low given that the ‘Water Supply Network Coverage is low’? In the following two sections, we present the fuzzy set theory and fuzzy inference to elucidate the principles of handling these two types of uncertainty.

Fuzzy sets

The first type of uncertainty can be handled using fuzzy set theory (Zadeh 1975). A fuzzy set A is represented by the following membership function:
formula
1
This function measures numerically the degree to which element x belongs to set A. It takes values between 0 and 1. The membership function is assessed subjectively, with small values representing a low degree of membership and high values representing a high degree of membership. In this study, we have used fuzzy sets of the trapezoidal type; in this mode of representation, the membership function is defined by five parameters (m, n, α, β, h) and two functions L (left) and R (right) (Figure 1).
Figure 1

Fuzzy set of a trapezoidal type.

Figure 1

Fuzzy set of a trapezoidal type.

Figure 2

Graphical illustration of the Mamdani method.

Figure 2

Graphical illustration of the Mamdani method.

Table 2 shows the numerical limits of the fuzzy linguistic parameters (very low, low, average, high and very high) related to the different indicators and to the QSI. These numerical limits are determined based on estimations made by Lebanese water supply experts.

Table 2

Representation of the different knowledge elements

IndicatorsVery lowLowAverageHighVery high
Network coverage (NC) (0, 0, 0.025, 0.2, 1) (0.2, 0.225, 0.275, 0.2, 1) (0.2, 0.475, 0.525, 0.2, 1) (0.2, 0.725, 0.775, 0.2, 1) (0.2, 0.975, 1, 0, 1) 
Water consumption (WC) (0, 0, 0.05, 0.15, 1) (0.15, 0.2, 0.3, 0.15, 1) (0.15, 0.45, 0.55, 0.15, 1) (0.15, 0.7, 0.8, 0.15, 1) (0.15, 0.95, 2, 0, 1) 
Continuity of water supply (CS) (0, 0, 0.025, 0.2, 1) (0.2, 0.225, 0.275, 0.2, 1) (0.2, 0.475, 0.525, 0.2, 1) (0.2, 0.725, 0.775, 0.2, 1) (0.2, 0.975, 1, 0, 1) 
Water quality (WQ) (0, 0, 0.025, 0.2, 1) (0.2, 0.225, 0.275, 0.2, 1) (0.2, 0.475, 0.525, 0.2, 1) (0.2, 0.725, 0.775, 0.2, 1) (0.2, 0.975, 1, 0, 1) 
Quality of service index (QSI) (0, 0, 0.05, 0.15, 1) (0.15, 0.2, 0.3, 0.15, 1) (0.15, 0.45, 0.55, 0.15, 1) (0.15, 0.7, 0.8, 0.15, 1) (0.15, 0.95, 1, 0, 1) 
IndicatorsVery lowLowAverageHighVery high
Network coverage (NC) (0, 0, 0.025, 0.2, 1) (0.2, 0.225, 0.275, 0.2, 1) (0.2, 0.475, 0.525, 0.2, 1) (0.2, 0.725, 0.775, 0.2, 1) (0.2, 0.975, 1, 0, 1) 
Water consumption (WC) (0, 0, 0.05, 0.15, 1) (0.15, 0.2, 0.3, 0.15, 1) (0.15, 0.45, 0.55, 0.15, 1) (0.15, 0.7, 0.8, 0.15, 1) (0.15, 0.95, 2, 0, 1) 
Continuity of water supply (CS) (0, 0, 0.025, 0.2, 1) (0.2, 0.225, 0.275, 0.2, 1) (0.2, 0.475, 0.525, 0.2, 1) (0.2, 0.725, 0.775, 0.2, 1) (0.2, 0.975, 1, 0, 1) 
Water quality (WQ) (0, 0, 0.025, 0.2, 1) (0.2, 0.225, 0.275, 0.2, 1) (0.2, 0.475, 0.525, 0.2, 1) (0.2, 0.725, 0.775, 0.2, 1) (0.2, 0.975, 1, 0, 1) 
Quality of service index (QSI) (0, 0, 0.05, 0.15, 1) (0.15, 0.2, 0.3, 0.15, 1) (0.15, 0.45, 0.55, 0.15, 1) (0.15, 0.7, 0.8, 0.15, 1) (0.15, 0.95, 1, 0, 1) 

The trapezoidal fuzzy sets are defined by five parameters (m, n, α, β and h) as shown in Figure 1.

Fuzzy inference

The principle of fuzzy inference is based on the Mamdani method (Mamdani 1976; Tong Tong 1995). We present here an illustration of the method on an example of two input indicators I1 and I2 and one output index O (Figure 2) as follows:
formula
Figure 3

The proposed QSI along with the associated indicators for the four RWAs.

Figure 3

The proposed QSI along with the associated indicators for the four RWAs.

The elements A11, A21, A12 and A22 are the fuzzy qualification (‘very low’, ‘low’, ‘average’, ‘high’ and ‘very high’) of premises I1 and I2 (which are the measured indicators in our case), B1 and B2 are the fuzzy qualification of the output O (which is the QSI in our study), which have, respectively, the following membership functions: μA11, μA21, μA12, μA22, μB1 and μB2 as shown in Figure 2.

In practice, if i01 and i02 represent values attributed to I1 and I2, then the characteristics of the output (or conclusions of rules) become μB'1(o) and μB'2(o); they are calculated from the previous membership functions.

By applying the inference method of Mamdani, we can write
formula
2
and
formula
3
The results of the two rules inference (the possibilities that O been B'1 and O been B'2) are calculated by
formula
4
formula
5
The global fuzzy output (B*) is given by the maximum of the previous function as follows:
formula
6
Finally, a crisp value of QSI is calculated using the centroid computation of the resulting global fuzzy output (defuzzification), which returns the center of the functions μ(x). Given the fuzzy functions μ(x), x0 is the centroid value as follows:
formula
7

The essence of the proposed method of aggregation is that if water experts can simulate the required performance of the aggregation index in qualitative linguistic terms that takes into consideration the existing correlations (interactions) among the individual indicators, then fuzzy inference system as described above can be used to implement this strategy successfully. Therefore, the numerical limits of the fuzzy linguistic parameters and the if-then rules presented in this study should be set by the water experts based on the unique country's conditions and its water supply management policies.

RESULTS AND DISCUSSION

The responsibility of water supply delivery in Lebanon is delegated to the following four RWAs:

  • Beirut and Mount Lebanon (BML),

  • North Lebanon,

  • South Lebanon,

  • Bekaa.

Table 3 presents the population (×1000) served by each RWA for the years 2000, 2005, 2010 and 2014 (MEW 2010).

Table 3

The population (×1,000) served by each RWA for the years 2000, 2005, 2010 and 2014

2000200520102014
BML 1,700 1,815 1,900 1,953 
North Lebanon 703 751 820 843 
South Lebanon 608 649 720 740 
Bekaa 451 482 535 550 
Total 3,462 3,696 3,975 4,086 
2000200520102014
BML 1,700 1,815 1,900 1,953 
North Lebanon 703 751 820 843 
South Lebanon 608 649 720 740 
Bekaa 451 482 535 550 
Total 3,462 3,696 3,975 4,086 

To examine the various steps of the proposed methodology, this section describes the evaluation of the QSI to analyze and report the progress made by the four RWAs throughout the years 2000, 2005, 2010 and 2014.

Firstly, we calculate the values of the considered performance indicators, which are presented in the ‘Quality of service indicators’ section. Table 4 presents the data, measurable variables and the resulting values of indicators.

Table 4

Data of the measurable variables and the resulting values of indicators

Measurable variablesIndicators
Water authorityYearCP (no.) (×103)TP (no.) (×103)AC (l/person/day)TC (l/person/day)NH (no.)TH (no.)PW (m3/year) (×103)TW (m3/year) (×103)NC (%)WC (%)CS (%)WQ (%)
BML 2000 1,190 1,700 100 200 1,080 8,640 96,522 86,870 70 50 13 90 
2005 1,361 1,815 120 200 1,440 8,640 149,030 149,030 75 60 17 100 
2010 1,615 1,900 150 200 2,520 8,640 218,025 218,025 85 75 29 100 
2014 1,725 1,953 170 200 2,520 8,640 281,674 281,674 88 85 29 100 
North Lebanon 2000 387 703 95 200 3,600 8,640 26,838 18,787 55 48 42 70 
2005 451 751 100 200 5,400 8,640 32,923 26,338 60 50 63 80 
2010 558 820 140 200 6,480 8,640 59,404 53,463 68 70 75 90 
2014 607 843 140 200 7,200 8,640 68,928 65,482 72 70 83 95 
South Lebanon 2000 362 608 90 200 1,080 8,640 19,820 13,874 60 45 13 70 
2005 442 649 100 200 1,590 8,640 29,333 22,000 68 50 18 75 
2010 624 720 130 200 2,880 8,640 59,218 47,374 87 65 33 80 
2014 656 740 140 200 3,600 8,640 64,465 58,018 89 70 42 90 
Bekaa 2000 203 451 88 200 1,080 8,640 13,041 8,476 45 44 13 65 
2005 280 482 95 200 2,880 8,640 19,418 12,622 58 48 33 65 
2010 334 535 100 200 3,600 8,640 24,382 17,067 62 50 42 70 
2014 375 550 110 200 4,320 8,640 30,113 27,101 68 55 50 90 
Measurable variablesIndicators
Water authorityYearCP (no.) (×103)TP (no.) (×103)AC (l/person/day)TC (l/person/day)NH (no.)TH (no.)PW (m3/year) (×103)TW (m3/year) (×103)NC (%)WC (%)CS (%)WQ (%)
BML 2000 1,190 1,700 100 200 1,080 8,640 96,522 86,870 70 50 13 90 
2005 1,361 1,815 120 200 1,440 8,640 149,030 149,030 75 60 17 100 
2010 1,615 1,900 150 200 2,520 8,640 218,025 218,025 85 75 29 100 
2014 1,725 1,953 170 200 2,520 8,640 281,674 281,674 88 85 29 100 
North Lebanon 2000 387 703 95 200 3,600 8,640 26,838 18,787 55 48 42 70 
2005 451 751 100 200 5,400 8,640 32,923 26,338 60 50 63 80 
2010 558 820 140 200 6,480 8,640 59,404 53,463 68 70 75 90 
2014 607 843 140 200 7,200 8,640 68,928 65,482 72 70 83 95 
South Lebanon 2000 362 608 90 200 1,080 8,640 19,820 13,874 60 45 13 70 
2005 442 649 100 200 1,590 8,640 29,333 22,000 68 50 18 75 
2010 624 720 130 200 2,880 8,640 59,218 47,374 87 65 33 80 
2014 656 740 140 200 3,600 8,640 64,465 58,018 89 70 42 90 
Bekaa 2000 203 451 88 200 1,080 8,640 13,041 8,476 45 44 13 65 
2005 280 482 95 200 2,880 8,640 19,418 12,622 58 48 33 65 
2010 334 535 100 200 3,600 8,640 24,382 17,067 62 50 42 70 
2014 375 550 110 200 4,320 8,640 30,113 27,101 68 55 50 90 

Sources:MEW (2010) and data collected by the author from the RWAs.

Acronyms of indicators and variables are explained in Table 1.

Based on the judgments of Lebanese water supply experts, sets of 35 rules are determined; they are used, along with the measures of indicators, in the proposed fuzzy inference system.

The outputs of the proposed fuzzy inference system are the QSIs for years 2000, 2005, 2010 and 2014. The results are shown in Table 5.

Table 5

Results of the quality of service index (QSI) (%)

Year
Water authority2000200520102014Progress 2000–2014
BML 49.5 50 53.4 55.7 6.2 
North Lebanon 46.7 59.8 67.7 71.2 24.5 
South Lebanon 43.6 48.7 54.6 60 16.4 
Bekaa 36 44 53.8 62.5 26.5 
Year
Water authority2000200520102014Progress 2000–2014
BML 49.5 50 53.4 55.7 6.2 
North Lebanon 46.7 59.8 67.7 71.2 24.5 
South Lebanon 43.6 48.7 54.6 60 16.4 
Bekaa 36 44 53.8 62.5 26.5 

Figure 3 presents the proposed QSI along with the associated indicators for the four RWAs in Lebanon for years 2000, 2005, 2010 and 2014.

Figure 4

QSIs for the four RWAs.

Figure 4

QSIs for the four RWAs.

Figure 4 presents the QSIs for the four RWAs in Lebanon for years 2000, 2005, 2010 and 2014. The progress made in the quality of service since 2000–2014 is shown in Figure 5.

Figure 5

Progress made in the quality of service since 2000.

Figure 5

Progress made in the quality of service since 2000.

The evaluation of the network coverage indicator shows that 18 and 17% of the population had gained access to the water supply network since 2000 in BML and North Lebanon water authorities, respectively; these proportions rise to 23 and 29% for Bekaa and South Lebanon water authorities, respectively. These high proportions of progress are pointers of the high financial investment in the construction of water supply networks in Lebanon since 2000. On the other hand, by taking into account other aspects of the quality of service such as the indicators mentioned in the ‘Quality of service indicators’ section, North Lebanon water authority has the highest QSI score of 71.2% while BML water authority has the lowest QSI score of 55.7% as can be seen from Table 5 and Figure 4.

As shown in Figure 3, BML water authority has recorded a low performance in quality of service despite the high proportion of the population connected to the water supply network. This is due to the low performances in other quality of service aspects such as the water consumption and continuity of water supply. Moreover, as can be seen in Figure 5, the progress made in QSI by Bekaa water authority since 2000 is the highest among the RWAs with a value of 26.5% followed by North Lebanon water authority. These high percentages of progress made by Bekaa and North Lebanon water authorities are due to the significant improvement of all the considered aspects of quality of service. The good performance of Bekaa and North Lebanon water authorities should induce BML and South Lebanon water authorities to exert additional efforts to improve the quality of service of their water supply systems especially the continuity of supply.

With the purpose to compare the calculated QSI based on the fuzzy inference method and when using the WAO approach, the QSI scores for the year 2014 are calculated by the WAO and the results are presented in Table 6. The weights of NC, WC, CS and WQ indicators have been set by the Lebanese water supply experts, and they are proportional to 1, 1, 2 and 2, respectively. North Lebanon has ranked first since it has the highest QSI score by using both the proposed fuzzy inference method and the WAO method. BML has ranked last since it has the lowest QSI score by using the fuzzy inference method. However, BML has ranked second by using the WAO method. The water supply experts consider that Bekaa and South Lebanon should get better QSI scores compared to BML and they are able to comment about their judgment as follows: BML is excellent at water quality and it is good at network coverage and water consumption, but it is very bad at continuity of supply. This is rather critical, since water is averagely supplied only 49 hours per week instead of 168 hours. These weekly supply hours are usually provided over only 3 days per week, which means no water supply for the four remaining days of the week. This obliges the subscribers to rely on private water tanker vendors where, in addition to the financial burden, the quality of the vended water is very poor, which may put the health of the customers at high risk. Therefore, water experts judge that BML should receive a low score. These results confirm that no weighted arithmetic mean operator can yield to the above-mentioned ranking result. In fact, Lebanese water supply experts judge that the dissatisfaction of only one indicator leads to a low QSI score. This condition could be realized only by using the fuzzy inference method, which enables control of the aggregation process by identifying appropriate inference rules. In line with this, the scores of the QSI calculated by the fuzzy inference method presented in the study are consistent with experts' judgments.

Table 6

Results of QSI (%) for 2014 calculated by using the weighted average operator

Water authority
BML 72.1 
North Lebanon 83.0 
South Lebanon 70.4 
Bekaa 67.1 
Water authority
BML 72.1 
North Lebanon 83.0 
South Lebanon 70.4 
Bekaa 67.1 

Lastly, in order to analyze the robustness of the QSI ranks, a sensitivity analysis to variations in fuzzy linguistic parameters has been performed for the QSI scores in 2014 using 25 random sets of the fuzzy linguistic parameters varying within the interval of ±20% of the original set values expressed by the experts. The sensitivity analysis has shown a complete agreement in ranking along with the original QSI scores for the four RWAs. Nevertheless, the use of the proposed approach to evaluate and monitor the QSI across countries necessitates performing additional thorough robustness/sensitivity analysis to analyze the variations of the QSI scores and ranks compared to reference QSI results.

CONCLUSION AND FURTHER DEVELOPMENT

This study proposes a methodological framework for quantifying the quality of water supply service delivery in Lebanon. In addition to the universal drinking water supply indicator (proportion of population using an improved drinking water source), three other indicators that reflect the quality of service received by water consumers are considered. A fuzzy inference model was proposed to aggregate the indicators into an overall QSI. The proposed approach is then applied to the water supply service delivery in Lebanon.

The established QSI aims to improve monitoring and reporting on access to drinking water supply services in Lebanon and to provide reliable information and analysis regarding the quality of water supply service delivery in the country. The proposed approach could also be used to evaluate and monitor the progress of the QSI across groups of countries; in this respect, it is necessary to perform sufficient coordination between countries to identify optimal common fuzzy linguistic parameters and inference rules.

This method allows a new opportunity in evaluating the access to water supply as a function of numerous indicators by taking into account inter-linkages across indicators and fuzziness inherent in the available data. It is expected that this national research will help to strengthen the capacity of the water supply authorities in Lebanon in monitoring the progress made on water supply service delivery at the national level through an appropriate approach. It is also expected that these additional indicators can help to inform the regional and global debates concerning the formulation of sustainable development indicators related to the water supply sector during the preparation of a post-2015 development framework that considers access to reliable and clean water services a universal human right (United Nations 2010; UN Water 2014).

In this initial stage, the considered indicators are those that have been agreed upon within the Arab region, and they focus on piped water supply services. The extension of this approach will be to consider additional quality of service indicators including those applied to all improved water source options (such as on-site improved water sources in rural areas). This aspect is still under development at our university for promoting access to sustainable safe drinking water sources in developing countries.

REFERENCES

Alegre
H.
Melo Baptista
J.
Cabrera
E.
Jr
Cubillo
F.
Duarte
P.
Hirner
W.
Merkel
W.
Parena
R.
2006
Performance Indicators for Water Supply Services
.
IWA Publishing
,
London
.
Bagheri
A.
Asgary
A.
Levy
J.
Rafeian
M.
2006
A performance index for assessing urban water systems: a fuzzy inference approach
.
Journal – American Water Works Association
98
(
11
),
84
92
.
Beranger
V.
Verdier-Chouchane
A.
2007
Multidimensional measures of well-being: standard of living and quality of life across countries
.
World Development Journal
35
(
7
),
1259
1276
.
Bossel
H.
1999
Indicators for Sustainable Development: Theory, Method, Applications
.
International Institute for Sustainable Development (IISD)
,
Winnipeg, Canada
.
García
C.
Kovacevic
M.
2011
Uncertainty and Sensitivity Analysis of the Human Development Index. Human Development Research Paper 2010/47
.
UNDP–HDRO
,
New York. Available from http://hdr.undp.org/sites/default/files/hdrp_2010_47.pdf (accessed 30 September 2014)
.
Haider
H.
Sadiq
R.
Tesfamariam
S.
2013
Performance indicators for small- and medium-sized water supply systems: a review
.
Environmental Reviews
21
,
1
40
.
Hák
T.
Janoušková
S.
Abdallah
S.
Seaford
C.
Mahony
S.
2012
Review Report on Beyond GDP Indicators: Categorisation, Intensions and Impacts. BRAINPOOL deliverable 1.1
.
JMP
2014
Improved and Unimproved Water and Sanitation Facilities
. .
Kanakoudis
V.
Tsitsifli
S.
Samaras
P.
Zouboulis
A.
Demetriou
G.
2011
Developing appropriate performance indicators for urban water distribution systems evaluation at Mediterranean countries
.
Water Utility Journal
1
,
31
40
.
Karnib
A.
2014
A methodological approach for quantitative assessment of the effective wastewater management: Lebanon as case study
.
Environmental Processes Journal
1
(
4
),
483
495
.
DOI
[PubMed]
Klugman
J.
Rodríguez
F.
Choi
H. J.
2011
The HDI 2010: New Controversies, Old Critiques. Human Development Research Paper No. 2011/01
,
UNDP
,
New York. Available from http://hdr.undp.org/sites/default/files/hdrp_2011_01.pdf (accessed 15 September 2014).
Mamdani
E. H.
1976
Advances in the linguistic synthesis of fuzzy controllers
.
International Journal of Man–Machine Studies
8
,
669
678
.
MEW
2010
National Water Sector Strategy
.
Ministry of Energy and Water
,
Beirut, Lebanon
.
Mori
K.
Christodoulou
A.
2012
Review of sustainability indices and indicators: towards a new city sustainability index (CSI)
.
Environmental Impact Assessment Review
32
(
1
),
94
106
.
Ocampo Duque
W. A.
2008
On the Development of Decision-Making Systems Based on Fuzzy Models to Assess Water Quality in Rivers
.
PhD Thesis
.
Universitat Rovira i Virgili, Tarragona, Spain
.
OECD
2008
Handbook on Constructing Composite Indicators – Methodology and User Guide
.
Organisation for Economic Co-operation and Development (OECD)
,
Paris
.
Tong Tong
J. R.
1995
La Logique Floue
.
Editions Hermes
,
Paris
.
UN-ESCWA (United Nations Economic and Social Commission for Western Asia)
2013a
Report: Tenth Session of the Committee on Water Resources – Beirut, 20–22 March 2013
.
UN-ESCWA
,
Beirut, Lebanon
. E/ESCWA/SDPD/2013/IG.1/8/Report.
UN-ESCWA
2013b
Water Development Report 5 – Issues in Sustainable Water Resources Management and Water Services
.
UN-ESCWA
,
Beirut, Lebanon
. E/ESCWA/SDPD/2013/4.
United Nations
2000
United Nations Millennium Declaration
.
General Assembly, United Nations
,
New York
,
A/RES/55/2
.
United Nations
2008
Report of the Secretary-General on the Indicators for Monitoring the Millennium Development Goals, Economic and Social Council
.
United Nations
,
New York
,
E/CN.3/2008/29
.
United Nations
2010
General Assembly Adopts Resolution Recognizing Access to Clean Water, Sanitation as Human Right
.
United Nations
,
New York
,
GA/10967
.
United Nations
2014
Report of the Open Working Group of the General Assembly on Sustainable Development Goals
. .
UN Water
2014
A Post-2015 Global Goal for Water: Synthesis of Key Findings and Recommendations from UN-Water
.
UN-Water
,
Geneva, Switzerland
.
Uppal
J. Y.
Mudakkar
S. R.
2013
Human development and economic uncertainties: exploring another dimension of development
.
Lahore Journal of Economics
18
,
305
334
.
Van Den Berg
C.
Danilenko
A.
2011
The IBNET Water Supply and Sanitation Performance Blue Book: The International Benchmarking Network for Water and Sanitation Utilities Data book
,
Water and Sanitation Program
,
World Bank
,
Washington, USA
.
WHO & UNICEF
2005
Water for Life: Making it Happen
.
WHO Press
,
Geneva, Switzerland
.
WHO & UNICEF
2006
Core Questions on Drinking-Water and Sanitation for Household Surveys
.
WHO Press
,
Geneva, Switzerland
.
Zeraebruka
K.
Mayabib
A.
Gathenyac
J.
Tsiged
Z.
2014
Assessment of level and quality of water supply service delivery for development of decision support tools: case study Asmara water supply
.
International Journal of Sciences: Basic and Applied Research
14
(
1
),
93
107
.