Water Resources Management

, Volume 27, Issue 7, pp 1913–1930 | Cite as

A Methodology for the Breakdown of NRW into Real and Administrative Losses

Article

Abstract

The estimation of Non Revenue Water (NRW) is simple and easy for water suppliers who keep records of the system input volume and the billed authorized consumption. However, the breakdown of NRW into its two main components real and administrative which refers to the unbilled authorized consumption plus apparent losses is not an easy or straight forward task. Methods reported in the literature for the breakdown of NRW into its components are top down approach and bottom up approach. Both approaches suffer from certain limitations and shortcomings that limits their use and reduce our confidence in the results obtained by them. This paper presents a methodology that can be used to draw a line between the real and the administrative losses with an acceptable level of accuracy. This methodology is based on the fact that the administrative losses are delivered to the demand site and consequently reach the wastewater collection system whereas the real losses are lost from the system and consequently do not reach the wastewater collection system. The methodology applies water balance from the water treatment plant outlet till the inlet of the wastewater treatment plant (WWTP). The mass balance approach of the Water Evaluation And Planning (WEAP) system was implemented for this purpose. In this methodology, the breakdown of NRW into its two main components is adjusted iteratively so that the difference between WEAP calculated and measured inflow to the WWTP is minimal. The presented methodology was applied to Amman and Zarqa cities in Jordan which return their wastewater to As Samra WWTP. The results showed that this methodology is capable of dividing NRW water into its two main components with an acceptable level of accuracy.

Keywords

Administrative losses Non revenue water Real losses Water distribution systems Water Evaluation and Planning system 

1 Introduction

Non Revenue Water (NRW) is defined as the difference between the system input volume and the billed authorized consumption, where the system input volume is the annual volume input to the water supply system (Farley et al. 2008). NRW is a worldwide problem in both developed and developing countries, however, the problem is more severe in developing than in developed countries due to the lack of financial resources to maintain the water distribution system, less availability of needed technologies for detecting and locating leaks, lack of qualified and trained personnel, in addition to the low level of public awareness and corruption. The cost of NRW worldwide is estimated at $ 14 billion per year, one third of which occurs in the developing countries, furthermore, leakage from the water distribution systems in the developing countries is estimated at 45 Million Cubic Meter (MCM) per day, in addition, about 30 MCM of water are delivered to the consumers daily but are not billed due to different reasons, in terms of percentage loss, NRW makes about 15 % of the system input volume in developed countries and about 35 % in developing countries (Kingdom et al. 2006).

NRW is a big financial challenge to the water supplier which hinders its development and the sustainability of its services as the water supplier pumps the water from its source, treats it and distributes it to the consumers but is not paid for it. In addition, the presence of leaks also requires larger infrastructure to account for the additional water required to compensate for the lost water (Water Loss Committee Review (WLCR) 2007), which means additional capital cost. Colombo and Karney (2002) investigated the cost of leaky pipes, they found that leaks increase energy cost substantially, as a general rule, they found that the percentage increase in the energy cost appears to be a second order polynomial function of leakage. In addition to the additional capital and energy costs associated with NRW, leaking pipes pose the risk of contaminating the drinking water as a consequence to the low pressure they induce in the pipelines which multiplies the risk of contaminants ingress into the water supply system which poses serious health risks to the consumers.

From water resources management point of view, minimizing NRW is equivalent to making new water resources available, Kingdom et al. (2006) estimated that saving 50 % of the worldwide leakage from the water supply systems is enough to provide water for additional 100 million people. It is important to note that benefits gained from investing in minimizing NRW are comparable to and may overweigh benefits gained from investing in developing new water resources. Furthermore, in countries with high NRW levels in their water supply system such as Jordan, reducing NRW is one of the management options to adapt to climate change.

NRW in Jordan makes big portion of the water supplied to the customers. According to the most recent Ministry of Water and Irrigation (MWI) estimates, NRW in Amman governorate for the year 1998 was about 48 % which dropped to 38 % for the year 2008 due to the maintenance of the water distribution system (MWI 2008). For other governorates such as Zarqa and Mafraq, NRW is even higher. For Jordan, a country classified among the poorest countries in the world in water resources, the reduction of NRW is key to reducing the gap between supply and demand. However, management options for the reduction of NRW are highly dependent on the water utility knowledge of its components; physical or real losses require maintenance of the distribution system while unauthorized consumption and errors in the meters’ reading require other management options. So the breakdown of NRW into its two main components is crucial to its effective management which is a must to improving the financial efficiency of the water utility.

2 Literature Review

NRW is simply the water supplied by the water utility to the consumers but not billed which can be estimated by subtracting the volume of the billed authorized consumption from the system input volume (Farley et al. 2008). According to the International Water Association (IWA) (Hirner and Lambert 2000; Alegre et al. 2000), NRW is divided into unbilled authorized consumption, and water losses, water losses are further divided into apparent losses and real losses, real losses include leakage from transmission lines and distribution mains, leakage and overflow at the utility storage tanks, and leakage from service connections up to the customers’ meters, apparent losses are divided into unauthorized consumption due to illegal connections, customers’ meters inaccuracies and data handling errors. Unbilled authorized consumption can be metered or unmetered, unmetered authorized consumption includes water used for public services such as irrigating public parks, landscaping, street cleaning, frost protection, flushing of water mains and sewers, firefighting etc. (Farley et al. 2008). The literature that deals with detecting, locating and assessing the different components of NRW is extensive. This section summarizes the methods and the technologies reported in the literature for detecting, locating and assessing the different NRW components.

Puust et al. (2010), in their review paper, classified the methods used for the assessment of real losses from water supply systems into two broad categories which are top down approach methods and bottom up approach methods. The top down approach methods are based on water balance where real losses are assessed by estimating all the other components of NRW i.e. unbilled authorized consumption and apparent losses due to the unauthorized use as well as due to the customers’ meters inaccuracies and data handling errors. The System input volume and the billed authorized consumption are obtained from the water utility records, real losses can then be estimated by simple mass balance.

The unbilled authorized consumption is system specific which is usually estimated as a percentage of the system input volume. Lambert and Taylor (2010) estimated the unbilled authorized consumption at 0.50 % of the system input volume for New Zealand, whereas the Water Loss Control Committee Review (WLCR) (2007) of the American Water Works Association (AWWA) estimated it at 1.25 % of the system input volume based on the findings of numerous water audits worldwide. After estimating the unbilled authorized consumption, the unmetered authorized consumption is obtained by subtracting the metered authorized consumption from the estimated unbilled authorized consumption.

The different components of the apparent losses are also system specific which can be estimated from the water utility knowledge of the water supply system. The WLCR (2007) of the AWWA estimated the unauthorized consumption at 0.25 % of the system input volume based on the findings of numerous water audits worldwide, whereas Lambert and Taylor (2010) estimated it at 1 % of the system input volume for New Zealand, however, despite the fact that these percentages are based on numerous water audits worldwide, the author argues that the unauthorized consumption is much higher for developing countries due to the lack of public awareness, lack of accurate data and corruption. Mutikanga et al. (2010) estimated the apparent losses for Kampala city water supply system in Uganda at 37 % based on "a proactive approach through investigations of suspicious trends of billing data consumption (zero consumptions, negative consumptions, etc.) and employing illegal use informers, however, as stated by Mutikanga et al. (2010), this proactive approach of individual site inspections fall under the bottom up approach. The customer meter inaccuracies can be estimated by meter tests at different flow rates that represent typical customer water use rates and meter guidance manuals (WLCR 2007; Farley et al. 2008; Mutikanga et al. 2010). Data handling errors can be estimated by exporting and analyzing historic billing data trends (WLCR 2007; Farley et al. 2008; Mutikanga et al. 2010). The drawback of the top down approach is that the unbilled authorized consumption and the unauthorized consumption suffer from the lack of scientifically based and universally accepted methods for their estimation, they rather differ significantly from community to community which makes water utilities use bottom up approach methods for their estimation.

Puust et al. (2010) in their literature review reported two methods for the assessment of real losses by the bottom up approach which are; 24 h Zone Management (HZM) and Minimum Night Flow (MNF). These two methods are applicable to District Metered Areas (DMAs) which are small parts of the water distribution system that typically contain 1,000 to 3,000 customers that receive water through one or more mains (WLCR 2007). The concept of the DMA is that flows and pressures in the DMA are of sufficient scale that it can be analyzed to distinguish normal flow from leakage rates (WLCR 2007). The drawback of this method is that it is a resource intensive method as it requires the isolation of small zones in the water supply system by valves and the installation of meters on the mains that feed the isolated zones, in addition to monitoring the flows and the pressures in the isolated zones (WLCR 2007). Furthermore, this method requires an experienced operator who can recognize normal flows and pressures from those caused by leakage, in addition, this procedure should be implemented for all the DMAs where leaks are suspected to occur which is lengthy, especially for large water supply systems. Moreover, the HZM method assesses leaks occurring within the isolated DMA without providing any information about the location of these leaks. The assessment of real losses by the MNF requires measuring the MNF into a DMA and subtracting the legitimate night flow from the measured MNF, legitimate night flow is community specific which can be estimated by knowing the activities that are running during the time of the measurements which are supposed to be very limited (WLCR 2007; Farley et al. 2008; Puust et al. 2010). Leakage determined by the MNF method has to be adjusted for pressure variations over the day as leakage is pressure dependent, May (1994) proposed to use a method called Fixed And Variable Discharges (FAVAD) which is based on relating leakage rate to the pressure head. The disadvantage of the MNF method is that the MNF should be measured in the wee hours of the morning, in addition the MNF procedure has to be implemented for every DMA in the network where leakage is suspected to occur. Similar to the HZM method, the MNF method estimates the leakage within the DMA without providing any information about the location of the leak.

Other methods reported in the literature for detecting, locating and quantifying leaks from water supply systems are based on monitoring and detecting the hydraulic changes that take place in the water supply system due to these leaks i.e. changes in flows and pressures (Pudar and Liggett 1992). Others analyzed transient pressure history at a single section in the pipeline to locate and quantify leaks (Brunone and Ferrante 2001; 2004; Misiunas et al. 2005). Kapelan et al. (2003, 2004) formulated the leak detection as a constrained optimization problem and utilized inverse transient analysis for its solution. Haghighi and Ramos (2012) utilized inverse transient analysis for leak detection and calibration of the friction coefficient. Mpesha et al. (2002) developed a method for leak detection in pipeline systems based on the fact that the frequency response diagram of a pipeline with leak has resonant pressure amplitude peaks that are lower than the pressure amplitude peak of a pipeline with no leak. Kim (2005) developed an algorithm for leak detection by the impulse response method. Covas et al. (2005) applied the Standing Wave Difference Method (SWDM) used in electrical engineering for cable fault location to detect leaks in pipeline systems. Beck at al. (2005) presented a new technique to detect pipeline feature and leakage by cross correlation analysis of reflected waves. Mounce and Machell (2006) applied Artificial Neural Network for hydraulic data i.e. flow and pressure for burst detection in water distribution systems. Giustolisi et al. (2008) integrated a pressure driven demand and leakage algorithm into a steady state hydraulic simulation model for the purpose of determining leakage. Aksela et al. (2009) presented a method based on Self Organizing Map (SOM) for leak detection. Other researchers utilized the characteristics of acoustic and vibration signals to detect and quantify leaks in pipes (Gao et al. 2004; Muggleton and Brennan 2004; Gao et al. 2005; Lee et al. 2005; Gao et al. 2006,). Buchberger and Nadimpali (2004) introduced a method for estimating leaks in small residential zones by statistical analysis of flow readings.

Water loss reduction which is the ultimate goal for water utilities beyond detection and assessment has received considerable attention from researchers recently. Karadirek et al. (2012) utilized hydraulic modeling to determine the optimum location for a pressure reducing valve to reduce leakage from Konyaalti water distribution network in Antalya, Turkey. Araujo et al. (2006) utilized a hydraulic simulation model, EPANET and two other operational models to manage pressure levels in water distribution systems for the purpose of leak reduction. Mutikanga et al. (2011) developed an integrated multi-criteria decision-aiding framework for strategic planning of water loss management to prioritize options for water loss reduction in Kampala city.

NRW estimation is an easy task as it is simply the difference between system input volume and billed authorized consumption (Farley et al. 2008), however, the breakdown of NRW into its two main components; real and administrative losses which refer to the summation of the unbilled authorized consumption and the apparent losses is a complicated task. As mentioned earlier, the top down approach suffers from the lack of scientifically based and universally accepted method for the estimation of the unbilled authorized consumption and the unauthorized consumption. On the other hand, the bottom up approach, i.e., The HZM and the MNF require intensive filed measurements, listening equipment, isolating certain zones of the water supply system, skilled personnel in addition to working in the wee hours of the morning. On top of that, these procedures have to be repeated for each zone in the water supply system where real losses are suspected to occur. Other hydraulically based and acoustic methods require detailed knowledge of the physical system, modeling experience, and continuous measurements of the pressures and the flows in the water supply system; in addition, there is no evidence that these methods are applicable or have been applied to large water supply systems, in fact, most of these methods were tested on laboratory scale on a single pipe or on a too small piping system.

The method presented in this paper seeks to breakdown NRW into its two main components; real and administrative without the need for the determination of its different components, nor the need for detailed knowledge of the water supply system, nor the need for pressure and/or flow measurements, nor the need for extensive field work, nor the need to isolate certain zones of the water supply system. The method presented here is a top down approach method based on mass balance that starts at the water treatment plant meter and ends at the inlet of the wastewater treatment plant, it makes use of the fact that administrative losses reach the wastewater collection system, whereas real losses do not, it is based on iteratively adjusting the real/administrative loss proportion of the NRW so that the difference between the measured and the calculated wastewater inflow to the wastewater treatment plant (WWTP) is minimal.

3 The Water Evaluation and Planning (WEAP) System

WEAP which is developed by the Stockholm Environmental Institute (1999) is a water balancing and allocation model. WEAP allocates the available water resources to the competing demands so that coverage at all the demand sites that compete for a certain resource is equal. Coverage is defined in WEAP as water delivered to a demand site divided by the supply requirement for that demand site, supply requirement is demand plus administrative losses. Other criteria considered in the WEAP allocation algorithm are user defined criteria and the physical characteristics of the distribution network. The user defined criteria are demand priority and supply preference. When and where several demands compete for one resource, the user ranks the demands in the order of their priorities so that WEAP allocates water to the demand site of the highest priority first i.e. priority one. On the opposite side, when and where a demand site receives water from different sources, the user ranks these sources in terms of their preferences which is known in WEAP as supply preference, in this case WEAP delivers the water from the resource of preference one first and so on. The physical characteristics of the distribution network are capacity of the storage tanks, storage capacity of dams, and the carrying capacity of the transmission lines and distribution mains. Despite the allocation criteria, WEAP allows the user to specify water supplied form a resource to a demand site which is dealt with as a constraint in the allocation algorithm.

WEAP expresses a basin as a network of resources and demand sites connected by transmission lines. Resources include rivers, ground water and other resources such as desalinated water. Demands include domestic demands, agricultural demands, industrial demands and environmental demands. A WEAP model of a basin may include other elements such as reservoirs, utility storage tanks, wastewater treatment plants, water treatment plants and return flow lines which transfer the wastewater generated at a demand site to the wastewater treatment plant. Furthermore, WEAP simulates the urban water cycle in a basin from the source to the reuse site, that is; supply to the demand site, losses at the demand site which refer to the administrative losses and within the water supply system which refer to the real losses, wastewater outflow from a demand site to the wastewater collection system, treatment at the wastewater treatment plant, outflow from the wastewater treatment plant and then delivery from the wastewater treatment plant to the reuse site. The approach presented in this paper uses that part of the urban water cycle from the resource up to the inlet of the wastewater treatment plant. In allocating water to the different demand sites, the mass balance is conserved for each WEAP element in the urban water cycle i.e. demand site, resource, transmission line, wastewater treatment plant, etc. which is the basis for the methodology presented here. WEAP also does catchment modeling; however, catchment modeling is out of the scope of this paper.

4 Methodology

The methodology presented in this paper for the breakdown of NRW into real and administrative losses is based on the WEAP mass balance at the demand sites, mass balance at the distribution network, and mass balance at the sewer lines that transport the generated wastewater from the demand site to the WWTP. The distribution network includes: the transmission lines, the distribution mains, the utility storage tanks and the service connections up to the customers' meters. The methodology makes use of the fact that the administrative part of NRW is delivered to the demand site and consequently reaches the wastewater collection system whereas, the real losses part of NRW does not reach the demand site and consequently does not reach the wastewater collection system. Based on this fact the wastewater volume generated at a demand site depends on NRW proportions i.e. administrative and real losses. The higher the administrative proportion of the NRW, the higher the wastewater volume generated out of a demand site and vice versa. The method presented in this paper adjusts the breakdown of NRW into real and administrative losses iteratively so that the difference between measured and calculated inflow to the WWTP is minimal. The WEAP mass balance approach implemented in this paper is described below.

Mass balance around the water supply system which starts at the water treatment plant outlet and ends at the consumers’ meters is shown schematically in Fig. 1a which can be expressed mathematically by the following equation:
Fig. 1

Schematic representation of mass balance for (a) distribution network, (b) demand site, (c) sewer line

$$ \mathrm{Inflow}\ \mathrm{to}\ \mathrm{the}\ \mathrm{demand}\ \mathrm{site} = \mathrm{System}\ \mathrm{input}\ \mathrm{volume}\text{--}\ \mathrm{Real}\ \mathrm{losses} $$
(1)
Mass balance around a demand site is shown schematically in Fig. 1b which can be expressed mathematically by the following equation:
$$ \mathrm{WW}\ \mathrm{outflow}=\mathrm{Inflow}\ \mathrm{to}\ \mathrm{the}\ \mathrm{demand}\ \mathrm{site}\text{--} \mathrm{Outdoor}\ \mathrm{use} $$
(2)

Where outdoor use refers to the water used at the demand site but does not reach the wastewater collection system such as water used for gardening and car washing,

Mass balance around the sewer line is shown schematically in Fig. 1c which can be expressed mathematically by the following equation:
$$ \mathrm{Inflow}\ \mathrm{to}\ \mathrm{the}\ \mathrm{WWTP}=\mathrm{WW}\ \mathrm{outflow}+\mathrm{Infiltration}/\mathrm{Inflow}\text{--} \mathrm{Ex}-\mathrm{filtration} $$
(3)

Where infiltration is water that enters to the wastewater collection system through the sewer lines and the manholes surfaces and through the joints and breaks in the sewer lines (Metcalf and Eddy 1991), and inflow is water that enters the sewer lines through open or broken manholes, cellar and foundation drains, roof leaders, etc. (Metcalf and Eddy 1991), and ex-filtration is wastewater that flows out of the sewer lines to the surroundings due to defected sewer lines and weak joints in the sewer system (Amick et al. 2000),

Summation of Eqs. 1, 2, and 3 yields the following equation:
$$ \mathrm{Inflow}\ \mathrm{to}\ \mathrm{the}\ \mathrm{WWTP}=\mathrm{System}\ \mathrm{input}\ \mathrm{volume}\text{--} \mathrm{Real}\ \mathrm{losses}\text{--} \mathrm{Outdoor}\ \mathrm{use}+\mathrm{Infiltration}/\mathrm{Inflow}\text{--} \mathrm{Ex}-\mathrm{filtration} $$
(4)
In equation 4, the system input volume is known from the water utility records, The breakdown of NRW into real and administrative losses is determined by model calibration, outdoor use is estimated as explained in the next section, ex-filtration depends on the sewer condition, joint type, pipe material, age, etc. (Amick et al. 2000), three methods are reported by Amick et al. (2000) for the estimation of ex-filtration which are: estimates based on direct measurements; estimates based on Darcy’s law and related theory and estimates based on drinking water - wastewater balance, however, no study to estimate ex-filtration for the study area has been made, so, for the purpose of this study ex-filtration is assumed at 5 %, the accuracy of this assumption is discussed in terms of the reasonability of the results obtained. Infiltration/inflow is neglected. However, the difference between calculated and measured inflows to the WWTP during winter as a result of infiltration/inflow is discussed. NRW is calculated from Eq. 5 as the difference between system input volume and billed authorized consumption which are obtained from the water utility records.
$$ \mathrm{NRW}=\mathrm{System}\ \mathrm{input}\ \mathrm{volume}\text{--} \mathrm{billed}\ \mathrm{authorized}\ \mathrm{consumption} $$
(5)

4.1 Outdoor Use

Outdoor use is that part of the water delivered to the demand site but does not reach the wastewater collection system such as water used for gardening and car washing. Outdoor use depends on temperature, prevailing type of dwellings i.e. individual houses versus apartments, dwellings density, and water availability. Outdoor use varies widely from community to community depending on these variables, however, there is no explicit mathematical model that models outdoor use in terms of these variables, rather, different methods are reported in the literature for its estimation which are: the summer winter method, the average month method, the minimum month method, the hydrologic region method, and the perspective city method (Gleick et al. 2003). The summer winter method is based on the assumption that the difference between water use during winter months and summer months is approximately the outdoor use. In the average month method, indoor use is assumed to be the average of the lowest three months’ use. The outdoor use for each month is then calculated as the difference between each month’s use and the calculated indoor use. The minimum month method assumes that the minimum month’s use is the indoor use. The difference between each month’s use and the minimum month’s use is the outdoor use for that month. The other two methods; the hydrologic region method and the perspective city method are based on typical values by region or by city, typical values are available for certain regions in the United States. The minimum month method was used in this study for the estimation of the outdoor use for Amman and Zarqa cities as it is the most commonly used method for outdoor use estimation. (Gleick et al. 2003) reported that skeel and Lucas (1998) found that 95 % of the increase in the summer uses in Seattle is due to outdoor use. In Jordan, other factors contribute to the difference between summer and winter uses such as population increase as a result of Jordanians working outside returning during summer to spend their vacations, in addition to tourists whose number increase significantly during summer. For these reasons, outdoor use calculated by the minimum month method is adjusted by multiplying it by 0.75. The validity of this assumption is tested by the reasonability of the obtained results as will be seen in the results and discussion section.

5 The Study Area

The study area is Amman Zarqa Basin which is the most developed and heavily populated basin in Jordan. The two largest cities in Jordan are within Amman Zarqa basin which are; Amman, the capital of Jordan and Zarqa, the second largest city in Jordan. Amman population was about 2.6 million for the year 2008 whereas Zarqa population for the same year was about 0.95 million, based on the settlements connectivity to the wastewater collection system, it is estimated by the author that about 80 % of the wastewater generated in Amman and about 93 % of the wastewater generated in Zarqa returns to As Samra WWTP.

6 Input Data to WEAP

Inputs to domestic demand sites are population, per capita annual water use in cubic meter, monthly variation, outdoor use, loss at the demand site which refers to administrative loss, and demand priority. Real losses are input at the transmission lines that transports the water from the source to the demand site. Monthly variation defines how the annual water demand/use occurs over the different months of the year. Monthly variation is input as a percentage of the annual water demand/use. Actual monthly water supplied to each of Amman and Zarqa was obtained from the MWI records for the years 2000 till 2006 and input to WEAP. Actual supplied water is used to calculate the per capita water use for the period covered by this study. Monthly variation was calculated based on the actual monthly supply. NRW for both Amman and Zarqa were calculated from the MWI records of the system input volume and the billed authorized consumption. The percentage return flow from Amman and Zarqa which is the bases for calculating the wastewater outflow from these two cities was calculated based on the settlements connectivity to the wastewater collection system.

6.1 Consistency of the Input Data

As explained in the previous section, the input data needed for the breakdown of NRW into administrative and real losses in Amman and Zarqa cities by the proposed methodology were obtained from the MWI records. However, the data were obtained from different sources at MWI and were processed by different MWI employees at different sections which means that data handling errors might have been introduced as the data were processed and transferred among the different sections and among the different employees. So it is important to check the consistency of the input data as an indicator of its quality to enhance our confidence in the results obtained by the method presented here. Data consistency is checked by comparing the system input volume to Amman and Zarqa cities as calculated by WEAP using the input data obtained from the MWI to the actual system input volume from the MWI records. System input volume to a demand site is calculated by WEAP using the following equation:
$$ \mathrm{System}\ \mathrm{input}\ \mathrm{volume} = \frac{\mathrm{Inflow}\ \mathrm{to}\ \mathrm{a}\ \mathrm{demand}\ \mathrm{site}}{{1\text{-}\%\ \mathrm{real}\ \mathrm{loss}}} $$
(6)

In equation 6, system input volume is function of inflow to the demand site and real loss. Inflow to the demand site is in turn function of per capita water use, population, monthly variation, and administrative loss. The per capita water use and the monthly variation were calculated based on actual system input volume and population. The real and the administrative losses as explained before are calibration parameters which are related to the actual system input volume through NRW, so if all these inputs are consistent, then, the calculated system input volume by Eq. 6 should be in good agreement with the actual system input volume from the MWI records.

7 Model Calibration

As mentioned before, the administrative part of NRW reaches the demand site and consequently, reaches the wastewater collection system, whereas the real loss part of NRW is lost from the water supply system and consequently does not reach the wastewater collection system. The higher the real loss is, the lower the wastewater volume out of a demand site and vice versa. On this basis, calibration was made by iteratively adjusting the breakdown of NRW into administrative and real losses so that the difference between the MWI measured and the WEAP calculated inflow to As Samra WWTP is minimal. Following is a description of the calibration process:
  1. a.

    Assume an initial breakdown of NRW into real and administrative losses, (NRW was initially assumed 50 % real loss and 50 % administrative loss),

     
  2. b.

    Run the WEAP model,

     
  3. c.

    Obtain the inflow to As Samra WWTP from the WEAP output,

     
  4. d.

    Compare the WEAP calculated wastewater inflow to As Samra obtained in step c with the MWI measured inflow,

     
  5. e.

    If the WEAP calculated inflow to As Samra is higher than the MWI measured inflow, then the real loss proportion of NRW is decreased for the next iteration and vice versa,

     
  6. f.

    Adjust the real loss proportion of NRW based on step e and go to step b,

     
  7. g.

    Repeat steps b to f until the difference between the WEAP calculated and the MWI measured inflow to As Samra WWTP is minimal,

     

It is important to note that once the real loss part of NRW water is adjusted, WEAP adjusts the administrative part as they are directly related. To avoid a lengthy iterative calibration process, the increment by which the real loss proportion should be adjusted can be calculated by establishing a relationship between the reduction in the wastewater volume per 1 % of real loss increase.

8 Results and Discussion

8.1 Outdoor Use

Figure 2 shows that the outdoor use for the year 2005, ranged from zero to 14.4 % for Amman, and from zero to 13.1 % for Zarqa. The outdoor use in Amman was higher than that in Zarqa due to the fact that there are more individual houses in Amman than in Zarqa. In addition, the housing density in Zarqa is higher than that in Amman. Al Washali (2010), estimated outdoor use for Sana’ city in Yemen at 4 %. In California, the United States, Gleick et al. (2003) mentioned that outdoor use is estimated at 50 % of the residential water use statewide and at about 70 % of the residential water use in some areas of the state. These figures show how widely outdoor use can differ from community to community as a result of the variation in the factors mentioned earlier which are temperature, prevailing type of dwellings i.e. individual houses versus apartments, dwellings density, and water availability. Of course outdoor use in hot regions is higher than that in cold or temperate regions. In addition, where the prevailing type of dwellings is individual houses, outdoor use is higher than where the prevailing type of dwellings is apartments due to the additional number of gardens that needs to be irrigated in the case of individual houses. Moreover, the higher the housing density, the lower the outdoor use due to the smaller open land areas that need to be irrigated. Water availability also affects outdoor use; the more the water availability is, the more the outdoor use.
Fig. 2

Monthly outdoor use for Amman and Zarqa cities for the year 2005

8.2 Consistency of the Input Data

Figure 3 shows the WEAP calculated system input volume to Amman and Zarqa cities against the measured one for the years 2001 till 2006. The slope of the straight line and the y-intercept indicate the level of agreement between the calculated and the measured system input volumes. A slope of one and an intercept of zero indicate a perfect match between calculated and measured system input volumes. However, this is never the case, the closer the slope to one and the closer the y-intercept to zero, the better the agreement between the calculated and the measured system input volumes. Figure 3 shows that in general there is good agreement between the WEAP calculated and the measured system input volumes to Amman and Zarqa as indicated by slopes close to one and by the small y-intercepts which means that the input data to the WEAP model are consistent and reliable; however, differences are noticed from year to year. These differences are due to two reasons; the first of which is data handling errors and the second is neglecting the system input volume to small settlements in the basin which are supplied from the same resource as Amman and Zarqa cities. Neglecting the system input volume to these small settlements resulted in the measured system volume to Amman and Zarqa cities being always larger than the calculated ones.
Fig. 3

WEAP calculated system input volume against measured one in MCM for Amman and Zarqa for the years 2000 till 2006

8.3 Measured Versus Calculated Inflow to as Samra WWTP

Figure 4 shows good agreement between the WEAP calculated and the MWI measured inflows to As Samra WWTP before calibration especially for summer months where no Infiltration/Inflow exists. During winter months, there exists some big differences between the WEAP calculated and the MWI measured inflows to As Samra WWTP which is due to Infiltration/Inflow to the sewer lines. Examples for that are January and February of 2002 and February of 2005 which are both associated with high rainfall. Figure 4 shows that calibrating the model by adjusting the breakdown of NRW into physical and administrative losses improved the agreement between the MWI measured and the WEAP calculated inflows significantly.
Fig. 4

WEAP calculated versus measured inflow to As Samra WWTP

8.4 Real Losses

Table 1 shows a decrease in the percentage real loss in Amman between 2001 and 2006. However, a slight increase in the percentage administrative loss for the same period with noticeable variation from year to year is observed. The reduction in the real loss is attributed to the maintenance program of the water supply system implemented by the MWI. Table 1 indicates that the reduction in the NRW in Amman between 2001 and 2006 is due to the reduction in the real loss as no reduction in the administrative loss was observed. Instead, it has increased slightly from year to year. Table 1 also shows an increase in the administrative loss proportion between 2001 and 2006 which can be attributed to the reduction in the real loss. In 2001, the administrative loss made about 50 % of NRW and in 2006 the administrative loss made about 58 % of NRW. According to Kingdom et al. (2006), NRW level in Amman is higher than the average or typical NRW in the developing countries which is estimated at 35 %. However, the real loss proportion of NRW is higher in the developing countries than the administrative loss proportion which is estimated at 60 %, whereas in Amman, the administrative loss proportion of NRW is higher than the real loss proportion as given in Table 1. Table 2 shows a decrease in the percentage real loss between 2001 and 2006 in Zarqa, whereas the percentage administrative loss has increased for the same period with slight variation from year to year. Table 2 shows that the reduction in NRW between 2001 and 2006 in Zarqa is due to the reduction in the real loss which can be attributed to the maintenance program of the water supply system. Similar to Amman, Table 2 shows an increase in the administrative loss proportion due to the reduction in the real loss for the period between 2001 and 2006. Table 2 shows that the NRW level in Zarqa is considerably higher than the NRW level in Amman. In addition the NRW level in Zarqa is considerably higher than that reported by Kingdom et al. (2006) for developing countries, which is estimated at 35 %. In terms of administrative and real losses proportions of NRW, the administrative proportion of NRW for the Zarqa city is higher than the real loss proportion, contrary to the values reported by Kingdom et al. (2006) for the developing countries which show that the real loss makes the larger proportion of NRW in developing countries which is estimated at 60 %.
Table 1

Estimated real and administrative losses for Amman assuming that outdoor use is 75 % of that calculated by the minimum month method

Year

NRW (% of system input volume)

Admin. Loss (% of system input volume)

Real Loss (% of system input volume)

Admin. Loss (proportion of NRW %)

Real loss (proportion of NRW %)

2001

49.1

24.6

24.5

50.0

50.0

2002

47.3

26.6

20.7

56.3

43.7

2003

48.5

28.7

19.9

59.1

40.9

2004

45.5

25.9

19.5

57.0

43.0

2005

45.2

25.6

19.6

56.7

43.4

2006

42.5

24.8

17.7

58.4

41.6

Table 2

Estimated real and administrative losses for Zarqa assuming that outdoor use is 75 % of that calculated by the minimum month method

Year

NRW (% of system input volume)

Admin. Loss (% of system input volume)

Real Loss (% of system input volume)

Admin. Loss (proportion of NRW %)

Real loss (proportion of NRW %)

2001

54.8

27.4

27.4

50.0

50.0

2002

55.8

30.9

24.9

55.3

44.7

2003

51.5

30.1

21.4

58.5

41.5

2004

51.8

29.1

22.7

56.2

43.8

2005

52.1

29.1

23.1

55.8

44.3

2006

50.6

28.9

21.7

57.1

42.9

9 Error Analyses

It is important to note that errors can be transmitted to the estimated breakdown of NRW into real and administrative losses made by the use of the methodology presented here as a result of the inevitable errors in the input data and the inaccuracies in the assumptions made. The quality of the input data was tested by checking its consistency which was found consistent. This means that the errors introduced to the estimated real loss as a result of the errors in the input data are expected to be minor. Other sources of error to the estimated breakdown of NRW into real and administrative losses can be related to the inaccuracies in the estimated outdoor use and to the inaccuracies in the assumed ex-filtration rate at 5 %.

Outdoor use was estimated at 75 % of that calculated by the minimum month method. If the actual outdoor use is higher than 75 % of that calculated by the minimum month method, then the estimated real loss will be lower, whereas if the assumed outdoor use is less than 75 % of that calculated by the minimum month method, then the estimated real loss will be higher. The results obtained by the assumed ex-filtration and the outdoor use looks consistent with the fact that the estimated real loss decreased consistently between 2001 and 2006 (Tables 1, 2) which is in harmony with the water supply system maintenance program implemented by the MWI. Tables 3 and 4 show the results obtained by assuming that the outdoor use is 50 % of that calculated by the minimum month method while keeping the ex-filtration at 5 % for Amman and Zarqa respectively. Tables 3 and 4 show a slight reduction in the real loss in both Amman and Zarqa between the years 2001 and 2006 with noticeable increase in some years which is inconsistent with the maintenance program of the water supply system which means that the assumption that the outdoor use is 75 % of that calculated by the minimum month method and that ex-filtration is 5 % is more reasonable and produces results consistent with the maintenance program implemented by the MWI.
Table 3

Estimated real and administrative losses for Amman assuming outdoor use is 50 % of that calculated by the minimum month method

Year

NRW (% of system input volume)

Admin. loss (% of system input volume)

Real loss (proportion of NRW %)

Admin. loss (proportion of NRW %)

2000

24.0

24.2

49.8

50.2

2001

27.2

21.9

55.5

44.5

2002

22.4

24.9

47.3

52.7

2003

23.2

25.3

47.9

52.1

2004

21.6

23.8

47.6

52.4

2005

22.1

23.1

48.8

51.2

2006

20.4

22.1

48.0

52.0

Table 4

Estimated real and administrative losses for Zarqa assuming outdoor use is 50 % of that calculated by the minimum month method

Year

NRW (% of system input volume)

Admin. loss (% of system input volume)

Real loss (proportion of NRW %)

Admin. loss (proportion of NRW %)

2000

27.3

27.5

49.8

50.2

2001

30.1

24.7

54.9

45.1

2002

26.6

29.2

47.7

52.3

2003

24.7

26.8

48.0

52.0

2004

24.8

27.0

47.9

52.1

2005

25.5

26.6

49.0

51.0

2006

24.4

26.2

48.3

51.7

10 Conclusions

A methodology for the breakdown of NRW into its two main components real and administrative has been developed and applied to two cities in Jordan; Amman and Zarqa. The methodology is based on the fact that the administrative loss part of NRW is delivered to the demand site and reaches the wastewater collection system whereas the real loss part of NRW is lost from the system and consequently does not reach the wastewater collection system. In addition, the methodology utilizes mass balance starting at the outlet of the water treatment plant and ending at the inlet of the WWTP. The application of the methodology to Amman and Zarqa cities in Jordan showed that the proposed methodology is capable of breaking down NRW into real and administrative losses with reasonable accuracy.

The advantages of the methodology presented in this paper as compared to the top down approach method is that it does not require the estimation of the different components of the NRW especially those with no universally accepted methods for their estimation which are the unbilled authorized consumption and the unauthorized consumption. As compared to the bottom up approach, the method does not require intensive filed work, nor working in the wee hours of the morning. As compared to the hydraulically based and acoustic methods, the method presented here does not require detailed knowledge of the physical system, nor modeling experience, nor continuous flow and pressure measurements, however the hydraulically based and acoustic methods are capable of locating the leak whereas the method presented here, the top down approach methods and the bottom up approach methods are incapable of locating the leak. The methodology presented here can be used to verify and/or complement the bottom up approach and the top down approach as it is completely independent of these two methods. In addition, the methodology presented here can be implemented at the DMA level where the wastewater outflow from a DMA is measured.

Uncertainties in some of the inputs such as Infiltration/Inflow, ex-filtration and outdoor use which are estimated rather than measured transfers into errors in the breakdown of NRW into real and administrative losses. It is important to note that these uncertainties are inherent to the input data not the model itself, which means that the impact of these uncertainties can be reduced by carefully estimating them. The output of the model can be verified by the reasonability of the results based on the user’s knowledge of the water supply system.

References

  1. Aksela K, Aksela M, Vahala R (2009) Leakage detection in a real distribution network using SOM. Urban Water J 6(4):279–289CrossRefGoogle Scholar
  2. Alegre H, Hirner W, Baptista J, Parena R (2000) Performance indicators for water supply services, manual of best practice series. IWA Publishing, LondonGoogle Scholar
  3. AL-Washali T (2010) Non-revenue water management in Sana’a water distribution system. master’s thesis, University of JordanGoogle Scholar
  4. Amick R, Burgess E, Camp D (2000) Exfiltration in sewer systems. U.S. Environmental Protection Agency publication, CincinnatiGoogle Scholar
  5. Araujo L, Ramos H, Coelho S (2006) Pressure control for leakage minimisation in water distribution systems management. Water Resour Manag 20(1):133–149CrossRefGoogle Scholar
  6. Beck S, Curren M, Sims N, Stanway R (2005) Pipeline network features and leak detection by cross-correlation analysis of reflected waves. J Hydraul Eng 131(8):715–723CrossRefGoogle Scholar
  7. Brunone B, Ferrante M (2001) Detecting leaks in pressurized pipes by means of transients. J Hydraul Res 39:539–547CrossRefGoogle Scholar
  8. Brunone B, Ferrante M (2004) Pressure waves as a tool for leak detection in closed conduits. Urban Water J 1(2):145–155CrossRefGoogle Scholar
  9. Buchberger S, Nadimpalli G (2004) Leak estimation in water distribution systems by statistical analysis of flow readings. J Water Resour Plan Manag 130(4):321–329CrossRefGoogle Scholar
  10. Colombo A, Karney B (2002) Energy and costs of leaky pipes: towards comprehensive picture. J Water Resour Plan Manag 128(6):441–450CrossRefGoogle Scholar
  11. Covas D, Ramos H, de Almeida A (2005) Standing wave difference method for leak detection in pipeline systems. J Hydraul Eng 131(12):1106–1116CrossRefGoogle Scholar
  12. Farley M, Wyeth G, Bin Md. Ghazali Z, Istandar A, Singh S (2008) The Manager’s Non-Revenue Water Handbook: A Guide to Understanding Water LossesGoogle Scholar
  13. Gao Y, Brennan M, Joseph P, Muggleton J, Hunaidi O (2004) A model of the correlation function of leak noise in buried plastic pipes. J Sound Vib 277:133–148CrossRefGoogle Scholar
  14. Gao Y, Brennan M, Joseph P, Muggleton J, Hunaidi O (2005) On the selection of acoustic/vibration sensors for leak detection in plastic water pipes. J Sound Vib 283:927–941CrossRefGoogle Scholar
  15. Gao Y, Brennan M, Joseph P (2006) A comparison of time delay estimators for the detection of leak noise signals in plastic water distribution pipes. J Sound Vib 292:552–570CrossRefGoogle Scholar
  16. Giustolisi O, Savic D, Kapelan Z (2008) Pressure-driven demand and leakage simulation for water distribution networks. J Hydraul Eng 134(5):626–635CrossRefGoogle Scholar
  17. Gleick P, Haasz D, Henges-Jeck C, Srinivasan V, Wolff G, Cushing K, Mann A (2003) Waste not, want not: The potential for urban water conservation in California, a report of the Pacific institute for studies in development, environment, and security, Oakland, CaliforniaGoogle Scholar
  18. Haghighi A, Ramos H (2012) Detection of leakage freshwater and friction factor calibration in drinking networks using central force optimization. Water Resour Manag 26(8):2347–2363CrossRefGoogle Scholar
  19. Hirner W, Lambert A (2000) Losses from water supply systems: Standard terminology and recommended performance measures. An IWA task force publication:1–13Google Scholar
  20. Kapelan Z, Savic D, Walters G (2003) A hybrid inverse transient model for leakage detection and roughness calibration in pipe networks. J Hydraul Res 41(5):481–492CrossRefGoogle Scholar
  21. Kapelan Z, Savic D, Walters G (2004) Incorporation of prior information on parameters in inverse transient analysis for leak detection and roughness calibration. Urban Water J 1(2):129–143CrossRefGoogle Scholar
  22. Karadirek I, Kara S, Yilmaz G, Muhammetoglu A, Muhammetoglu H (2012) Implementation of hydraulic modeling for water-loss reduction through pressure management. Water Resour Manag 26(9):2555–2568CrossRefGoogle Scholar
  23. Kim S (2005) Extensive development of leak detection algorithm by impulse response method. J Hydraul Eng 131(3):201–208CrossRefGoogle Scholar
  24. Kingdom B, Liemberger R, Marin P (2006) The challenge of reducing Non-Revenue Water (NRW) in developing countries how the private sector can help: A look at performance-based service contracting, water supply and sanitation sector board discussion paper series paper no.8. The world bank groupGoogle Scholar
  25. Lambert A and Taylor R (2010) Water loss guidelines. A guideline for the New Zealand water and wastes association: water New Zealand, New ZealandGoogle Scholar
  26. Lee P, Vıtkovsky J, Lambert M, Simpson A, Liggett J (2005) Leak location using the pattern of the frequency response diagram in pipelines: a numerical study. J Sound Vib 284:1051–1073CrossRefGoogle Scholar
  27. May J (1994) Pressure dependent leakage. World water and environmental engineering, October 1994Google Scholar
  28. Metcalf and Eddy (1991) Wastewater engineering: treatment, disposal, reuse, McGraw-Hill, 3rd ed., SingaporeGoogle Scholar
  29. Ministry of Water and Irrigation (MWI) (2008) Water for life: Jordan’s water strategy 2008–2020Google Scholar
  30. Misiunas D, Vítkovský J, Olsson G, Simpson A, Lambert M (2005) Pipeline break detection using pressure transient monitoring. J Water Resour Plan Manag 131(4):316–325CrossRefGoogle Scholar
  31. Mounce S, Machell J (2006) Burst detection using hydraulic data from water distribution systems with artificial neural networks. Urban Water J 3(1):21–31CrossRefGoogle Scholar
  32. Mpesha W, Chaudhry M, Gassman S (2002) Leak detection in pipes by frequency response method using a step excitation. J Hydraul Res 40(1):55–62CrossRefGoogle Scholar
  33. Muggleton J, Brennan M (2004) Leak noise propagation and attenuation in submerged plastic water pipes. J Sound Vib 278:527–537CrossRefGoogle Scholar
  34. Mutikanga H, Sharma S, and Vairavamoorthy K (2010) Assessment of apparent losses in urban water systems. Water and Environment Journal 25(2011): 327–335Google Scholar
  35. Mutikanga H, Sharma S, Vairavamoorthy K (2011) Multi-criteria decision analysis: A strategic planning tool for water loss management. Water Resour Manag 25(14):3947–3969CrossRefGoogle Scholar
  36. Pudar R, Liggett J (1992) Leaks in pipe networks. J Hydraul Eng 118(7):1031–1046CrossRefGoogle Scholar
  37. Puust R, Kapelan Z, Savic D, Koppel T (2010) A review of methods for leakage management in pipe networks. Urban Water J 7(1):25–45CrossRefGoogle Scholar
  38. Skeel T and Lucas N (1998) Seattle water’s outdoor use study. http://www.ci.seattle.wa.us/util/RESCONS/papers/p_tsnl.HTM
  39. Stockholm Environment Institute (SEI) (1999) WEAP: Water Evaluation and Planning system. Tellus Institute, BostonGoogle Scholar
  40. Water Loss Committee Review (2007) Water audits and loss control programs, AWWA M36 Publication RewriteGoogle Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2013

Authors and Affiliations

  1. 1.Water, Energy and Environment CenterThe University of JordanAmmanJordan

Personalised recommendations