Water Security in South Asian Cities: A Review of Challenges and Opportunities

Abstract

Achieving water security in South Asian cities will require a realistic and holistic understanding of the challenges that are growing in extent and severity. These challenges include the rapid rise in urban household water demand due to both overall population growth and increasing urbanization rate. Additionally, surface water supply in closed river basins is fully utilized, and there is little opportunity in these regions to increase the extraction of surface water to meet rising demands. Furthermore, groundwater extraction in most regions exceeds natural recharge rates, leading to rapidly falling annual water tables and seasonal depletion in hard rock regions and to gradually declining water tables requiring deeper wells and increased pumping effort in alluvial regions. Additionally, even in cities with abundant water resources, poorer segments of the population often face economic water scarcity and lack the means to access it. Nevertheless, there are important potential engineering opportunities for achieving water security in South Asian cities. Much withdrawn water is lost due to urban water distribution inefficiency, and a range of proven techniques exist to improve distribution. Metering of urban water can lead to structural improvements of management and billing, though the water needs of the poorest city residents must be ensured. Industrial water-use efficiency can be significantly improved in manufacturing and electricity generation. The quantities of wastewater generated in South Asia are large, thus treating and reusing this water for other purposes is a strong lever in enhancing local water security. There is limited potential for rooftop rainwater harvesting and storage, though capture-enhanced groundwater recharge can be important in some areas. Some individual inter-basin transfer projects may prove worthwhile, but very-large-scale projects are unlikely to contribute practically to urban water security. Overall, the water challenges facing South Asian cities are complex, and although no single intervention can definitively solve growing problems, numerous actions can be taken on many fronts to improve water security.

1. Water Security Challenges in South Asian Cities

Among the most vexing problems the world is facing today is water security: rainfall patterns are shifting, alternating between drought and flood; groundwater tables are permanently dropping; and pollution in water bodies is worsening in many countries. The combined effect of climate change, population growth, economic development, urbanization, and inadequate regulation are making it very difficult to pinpoint root causes and solutions alike. These problems are particularly severe in developing regions, no more so than in South Asia, which is home to virtually every major water-related problem. This section explores the specific water security challenges facing urban residents in South Asia, while the following section outlines important potential engineering interventions that could enhance water security in South Asian cities. This review article is adapted, condensed, and updated from an in-depth report on technology interventions for water security in South Asia [1], which was based on comprehensive literature review, interviews with experts, and quantitative impact analysis.

1.1. Urban Water Demand Is Growing Strongly

Urban population in South Asia is projected to almost double by 2050 due to a combination of overall population growth and increasing urbanization (see Figure 1, based on total population data from [2], medium fertility projection, and urban proportion data from [3]). Currently, the urban population in India increases by about 10.5 million individuals per year (calculated based on [2,3]). In Bangladesh and Pakistan, the urban population is increasing by about 1.85 and 2.05 million people per year, respectively. This increase in population is causing a rapidly growing demand for urban water consumption for both household and industrial use.

Although the absolute quantity of water supplied to urban residents is much smaller than that used by agriculture, water use in cities is concentrated with a high level of use in a relatively small geographic area. Figure 2 shows that the density of household water use (in units of m³ of household water use per hectare of urban land area per year) of most major South Asian cities is higher than that of most irrigated agricultural fields (in units of m³ of irrigation water per hectare of crop land per year). For cities, the figure is based on population and land area data from [4], per capita household water use data from [5], and assumed distribution losses of 30%; urban industrial water use is not included; thus, city water use is underestimated. For crops, the figure is based on all-India averages, with total cultivated area and irrigated area from [6], total crop water use from [7], and irrigation water application rate from [8].

Furthermore, each litre of urban water generates more value in social and economic terms than a litre of irrigation water. This is because households require timely water access for essential activities such as drinking, cooking, cleaning, and sanitation and because industrial activities generate relatively high economic returns from water use. For example, urban health and hygiene depends on water for the functioning of modern sewage systems. An adequate and reliable supply of water to cities remains essential for civil living. Within an area, cities tend to appropriate water from agriculture through many different formal and informal mechanisms due to the absolute need for household water coupled with the economic concentration of the urban population [9].

The Bureau of Indian Standards recommends a minimum of 135 litres per capita per day for communities with water-based (i.e., flushed) sanitation [10]. Surveys show that actual water consumption among Indian households varies widely both geographically and socio-economically, with an average of about 92 litres per capita per day (see Figure 3, based on [5]). Aggregated per capita urban water supply statistics mask the inequity between citizens, with the poorest residents barely securing water for subsistence, while the richest residents enjoy abundant water supply (this is discussed in detail in Section 1.5).

In addition to household use, another important demand for urban water is industries, particularly water-intensive industries such as steel, textiles, pulp, and paper. Some industries located in urban areas use water from municipal supply systems. Other industries obtain water from local surface and groundwater sources (see Figure 4, based on survey of Indian industries [11]). In cities with water stress, this industrial water demand competes with household water demand to achieve adequate urban water supply.

1.2. Surface Water Supply in Closed River Basins Is Fully Utilized

A river basin is considered “closed” when all of the surface water available in an average year is fully allocated, and little or no river water discharges into the ocean [12]. Several major river basins in South Asia are now closed basins, including the Indus in Pakistan, the Krishna, Kaveri, Penna, and Vaigai in southern India, and the Sabarmati and Banas in western India [13,14,15,16]. In these basins, the total annually available surface water supply is fully or nearly fully utilized. There is little or no opportunity to increase extraction of surface water in these regions to meet rising demand from cities and farms.

The Indus River basin, for example, is extensively developed to use surface water in the largest contiguous irrigation system in the world, the Indus Basin Irrigation System (IBIS). Major cities including Karachi are dependent on the Indus for their water supply. Such intense withdrawal and consumption use about 88% of all the water in the Indus River, leaving very little water flowing to the sea (see Figure 5). The figure is based on data from [17,18,19]; river inflow includes the Indus at Kalabagh, Chenab at Marala, Jhelum at Mangla, and minor inflows from Ravi and Sutleg Rivers; and river outflow is the discharge below Kotri barrage. Most of the river water available in an average year is already allocated and used, and no water flows from the Indus River to the ocean in years of below-average rainfall.

The Krishna River basin in peninsular India is similarly developed to capture and use almost all available surface water, constraining water supply expansion to Hyderabad and other urban centres. While average rainfall in the Krishna basin has remained steady over the long term, average discharge to the ocean has decreased to near zero. An increase in reservoir storage capacity has enabled the capture and utilization of most surface water in the Krishna basin. Between 1901 and 1956, an average of 55 km³ of discharge water flowed to the ocean each year; between 1984 and 2008, the average annual discharge was only 15 km³ (see red dashed lines in Figure 6, based on data from [14,20,21]).

In closed basins, while river water is fully used during normal and dryer-than-normal years, some flood water typically reaches the ocean during years of above-average precipitation. Attempting to capture and utilize this “lost” water leads to diminishing returns, requiring significant infrastructure that will only be used irregularly, making 100% utilization of river flow impractical.

During years of less-than-average precipitation, the surface water supply will not meet demand in closed basins. Such years are projected to become more frequent due to precipitation variability resulting from climate change. In the absence of a structured water allocation system, upstream water users will continue to extract river water, while downstream users will be left with inadequate supply.

Among the various uses of freshwater, environmental flows are those required to sustain riverine and estuarine ecosystems and to maintain other natural river processes. In best-practice resource management, a sufficient amount of river water is allotted to environmental flows, just as water is allotted to agriculture, industry, and communities. Environmental flow requirements are complex and varied, involving quantity, quality, and timing of the water flows. Comprehensively defining and maintaining the required environmental flows is challenging in South Asia due to the high hydrological variability, the difficulty and expense of waste treatment scale-up, water resource disputes between states/provinces, and a lack of quantitative data on relationships between river flows and river ecology [22].

As river basins approach closure, maintaining adequate environmental flows becomes more difficult since they reduce the amount of water available for other potential uses. Ensuring adequate environmental flows in rivers has not been a high priority in South Asia. Generally, little respect has been paid to the natural ecology of river basins and the biological, chemical, and physical interactions within the watersheds [22]. In Pakistan, environmental flow requirements of the Indus River are not met during 11 months of the year, while the Ganges River in India faces environmental flow deficits during 8 months of the year [23].

1.3. Groundwater Over-Extraction Causes Seasonal and Long-Term Depletion

Groundwater has long been harnessed by the people of South Asia, using traditional technologies to create wells and extract water. Groundwater use increased rapidly in South Asia after the 1960s, as a key component of Green Revolution agriculture based on borewells and motorized pumps. Groundwater is now a major source of water in the region for agriculture, industry, and households. In locations where high-quality groundwater is abundant, wells can provide more flexible and reliable water supply, compared to surface water supply. Many urban households in South Asia now use groundwater, either from centralized utility supply or from individual household wells.

Groundwater exists in porous underground formations, or aquifers, that are naturally recharged over time by rainwater infiltration into the ground. Some aquifers are very porous and extensive, and other aquifers are less porous and of limited extent. Groundwater recharge depends on the timing and amount of precipitation, land surface characteristics, and subsurface geology. Groundwater is removed from aquifers both by natural drainage and by extraction from wells. When the rate of removal exceeds the rate of recharge, the level of groundwater (i.e., the “water table”) declines.

Water security in South Asia is strongly influenced by the geologic origins of the regions. The Indian tectonic plate is slowly colliding with the Eurasian plate, forming the Himalayan mountain range. Eroded sediment from that range has formed the basins of the Indus, Ganges, and Brahmaputra river systems [24]. This has given rise to two distinct groundwater regimes in South Asia: hard rock and alluvial. While there are also some semi-consolidated formations in South Asia, they are relatively minor in extent and significance compared to hard rock and alluvial formations.

“Hard rock” is a generic term applied to igneous and metamorphic rocks such as granites, basalts, gneisses, and schists. This geologic formation underlays most of peninsular India. Although western Pakistan is part of the Eurasian tectonic plate, it shares similar hydrogeological conditions. In hard rock regions, limited quantities of groundwater are stored in the weathered soil and sub-soil layers (typically tens of meters deep) that overlay the bedrock. The bedrock itself has zero “primary porosity” in the rock itself but has limited “secondary porosity” due to cracks and fissures where groundwater may enter. The storage volume for groundwater in hark rock regions is thus very limited, confined to the shallow soil layer and some deeper cracks, and may deplete seasonally upon heavy pumping.

Groundwater in hard rock is characterized by limited productivity of individual wells, unpredictable variations in productivity of wells over relatively short distances, and poor water quality in some areas. In hard rock regions, groundwater supply is limited and discontinuous and thus may not be available regardless of the number or depth of borewells. Groundwater over-extraction in hard rock regions leads to rapidly falling water tables and seasonal depletion, imposing hard limits on groundwater extraction rates. In some regions of peninsular India, rising demand for groundwater vastly exceeds annually recharged amounts, leading to rapidly falling water table and many wells that are seasonally or permanently dry. Due to the high density of wells, interference between wells is common in hard rock areas, where neighbouring wells compete for limited groundwater.

In contrast to hard rock regions, alluvial regions have thick porous aquifers that contain large amounts of groundwater. Thick and unconsolidated alluvial formations that are conducive to recharge are found in the Indus–Ganges–Brahmaputra basins in Pakistan, northern India and Bangladesh. The coastal alluvial belt on the eastern coast of India also has relatively high replenishable groundwater resources. Within the alluvial aquifers of the Indus–Ganges–Brahmaputra basin, there is a strong distinction between the relatively arid northern region and the more humid eastern region.

Groundwater over-extraction in the more arid northern region is leading to steadily falling water tables, gradually requiring deeper wells and increased pumping effort. This primarily affects the Punjab region in India and Pakistan, as well as western Rajasthan, northern Gujarat, and Haryana states in India, where water tables in some areas are declining by about 1 m per year [25]. Depending on the water table decline rate and the aquifer thickness, groundwater over-extraction can continue for many years. Alluvial formations in the Indus–Ganges basins are hundreds or even thousands of meters thick [24]. As technologies for deep well drilling are available, the constraining factor is the additional energy needed for pumping as the water table becomes deeper. Although abundant groundwater stores now exist in these alluvial formations, the extraction and use of this stored water for current and near-future requirements raises questions of intergenerational equity and how future generations of South Asians will access water.

Groundwater remains abundant in the eastern Ganges and Brahmaputra regions, which have greater rainfall. In these regions, the phenomenon often described as the Ganges Water Machine operates [26,27]. The more groundwater that is withdrawn before the onset of the annual monsoon, the more soil pore space is created for water to recharge. During recent decades, increased groundwater extraction has increased the overall extent of groundwater recharge in Bangladesh [28]. Thus, in the eastern Ganges and Brahmaputra regions, groundwater from alluvial aquifers does not face the problem of over-extraction and is unlikely to face it in the future.

1.4. Much Water Is Lost due to Urban Water Distribution Inefficiency

Reliable distribution of life-sustaining water to urban residents is a requirement of civil life. To satisfy this need, extensive distribution networks are used to bring water within reach of all citizens. These networks must withstand an array of challenges such as pressure transients, corrosion processes, and contamination sources. Within South Asia, the performance of urban water utilities varies from city to city, but they typically offer poor service delivery, poor maintenance of physical systems, and poor recovery of costs [29]. Many South Asian water utilities face difficulties in accessing and distributing sufficient water supply to meet growing demands, and markets for alternative water sources based on private assets (borewells, trucks, etc.) are growing in many cities.

Municipal utilities in South Asia suffer from high levels of non-revenue water (NRW). NRW is water that has been sourced and prepared for distribution but is lost before it reaches the customer. NRW losses of 50% are not uncommon in South Asian cities (see Figure 7, based on [29,30,31]). Losses can be physical losses (through leaks, also referred to as technical losses) or apparent losses (for example through theft or metering inaccuracies). About three-quarters of NRW is real physical losses of water, and one-quarter is apparent loss [29,30]. From a human development perspective, physical losses are the most critical because they represent potentially available water that is not ultimately used for people’s welfare. Apparent losses are also important, particularly regarding the economic sustainability of a water utility, but they pose less humanitarian concern because they represent water that is actually used by people. In a broader context, physical losses from a water utility perspective are not ultimately lost in a river basin perspective, and leaked water may be accessed later as groundwater or may sustain urban and peri-urban vegetation.

The most commonly used indicator to measure water losses is the percentage of NRW as a share of total water produced. Although this indicator is a useful benchmark, it can be misleading since it does not distinguish between physical and apparent losses, and it is dependent on supply time, pressure, and level of consumption, which vary widely between countries and regions. A better but less-used indicator is the Infrastructure Leakage Index, which considers the total length of the mains as well as private pipes (property boundary to customer meter), the number of service connections, and the average pressure [32]. In general, urban water distribution losses are poorly understood and documented in quantitative terms. In a survey of Indian cities, the Ministry of Urban Development found NRW values averaging 44%, and ranging from less than 20% to greater than 60% (see Figure 8, based on data from [31]; only cities with a data reliability grade of “A” or “B” are shown).

Intermittent water supply (IWS) is an important factor in the poor performance of South Asian water utilities. Water is not available continuously to all households and industries but instead is provided in turn to different urban zones, each for a limited period (see Figure 9, based on data from [33]). There are many drivers of IWS, including physical water supply constraints due to seasonal and population trends, limiting water leakage from damaged pipes, and prioritizing access due to privatization or local governance policy. Galaitsi et al. [34] described three types of IWS. Listed from least disruptive in consumers’ lives to most disruptive, they are predictable intermittency, with water supply that generally occurs on a predictable and anticipated schedule; irregular intermittency, with supply arriving at unknown intervals within short time periods of no more than a few days; and unreliable intermittency, with uncertain delivery time and the risk of insufficient water quantity.

Households develop coping strategies for IWS, such as storing water when it is available for use later. Nevertheless, IWS is emerging as an important obstacle to urban water security for numerous reasons [34,35,36]. A intermittent lack of pressure within the pipes allows contaminated water from outside the system to enter through holes or cracks in the pipes. Pressure fluctuations tend to damage the pipes and connections, leading to water leakage and contamination. Intermittent stagnation allows for growth of biofilms within pipes, leading to microbial contamination. Water held in household tanks can become contaminated by improper storage and access methods. During times when intermittent water is flowing, profligate usage leads to high levels of waste. Water meters malfunction due to pressure surges and air bubbles created by intermittent pressurization. Social conflicts arise due to unequal access to water by different zones of urban areas. Finally, prolonged shut-offs of municipal water require some households to rely on costly private water options.

1.5. Poor City Residents Have Limited Access to Water

Economic water scarcity occurs where the lack of human, institutional, and financial capital limits access to water, even though water is available locally in nature to meet human demands. Some places contain abundant water resources, but some segments of the local populations face water scarcity because they lack the means to access it. While traditionally considered a problem of rural communities, there is also a strong urban component of economic water scarcity in South Asia. Aggregated per capita urban water supply statistics mask the inequity between citizens, with the poorest residents barely securing water for subsistence, while the wealthiest residents enjoy abundant water supply (see Figure 10, based on survey of seven Indian cities [5]).

Poor households apply a “hierarchy of water needs” (similar in concept to Maslow’s hierarchy of needs) and prioritize available water for essential uses [37]. Immediate survival needs of drinking and cooking come first, followed by medium-term maintenance needs of personal washing, clothes washing, house cleaning, and waste disposal, then longer-term sustaining uses for domestic gardens, commercial agriculture, and amenity landscaping. Often living in informal settlements without connections to municipal water supply, poor urban households often must pay elevated retail prices for essential water delivery from private service providers or risk disease by using contaminated local water sources. Private markets for drinking water are expanding in cities, serving communities living in slums as well as wealthier classes. Expensive bottled water is increasingly used in developing countries due to the inadequate quantity and quality of piped water supply. Since 2004, the consumption of bottled water in low- and medium-income countries has increased by 174%, compared to 26% in high-income countries [38].

1.6. Many Cities Are Constrained from Increasing Water Supply

Many cities in South Asia already struggle to provide adequate water supply to their citizens. Current water use in some regions is approaching or already exceeds local renewable supplies of surface and groundwater. Rapidly rising urban water demand from growing cities contributes strongly to this overexploitation. South Asian cities that are dependent on groundwater for municipal supply, such as Dhaka and Chennai, observe falling water tables requiring increased drilling and pumping costs. Cities that depend on surface water, such as Bengaluru and Hyderabad, face intermittent or insufficient supply. Reliable water supply is already challenging in many cities due to growing demand combined with aging and insufficient infrastructure.

Table 1. Characteristics of the ten largest cities of South Asia and their municipal water supplies.

City	Population (Million)	Precipitation (mm per Year)	Primary Sources of Municipal Water	Constraint to Supply Increase
Delhi	26	790	Eastern Yamuna and Upper Ganges Canals	$\mathrm{Medium}$
Mumbai	23	2260	Upper Vaitarna and Middle Vaitarna Dams	$\mathrm{Low}$
Karachi	23	220	Hub Dam, Haleji and Keenjhar Lakes	$\mathrm{High}$
Dhaka	16	2120	Groundwater	$\mathrm{Low}$
Kolkata	15	1800	Hooghly River from Ganges	$\mathrm{Low}$
Lahore	10	630	Tarbela Dam from Indus River; Groundwater	$\mathrm{Medium}$
Bengaluru	10	990	Kaveri and Arkavati Rivers	$\mathrm{High}$
Chennai	10	1390	Kosasthalaiyar River; Groundwater; Desalination	$\mathrm{High}$
Hyderabad TL	8	830	Musi, Krishna, Godavari, Manjeera Rivers	$\mathrm{Medium}$
Ahmedabad	7	750	Narmada Canal; Groundwater	$\mathrm{High}$

summarizes the sources of municipal water for the ten largest cities in South Asia. The table is based on population data from [4], annual average precipitation from [39], and water source data from various literature and interviews. The table also provides estimates of the difficulties of future increases in water supply, based on the authors’ analysis of physical constraints to supply such as river basin closure and groundwater depletion.

Mismatch between urban water supply and demand will impact human development in several ways. Importantly, the health of (poorer) urban populations will suffer due to drinking contaminated water or unsanitary conditions resulting from water scarcity. Furthermore, the economic potential of urban areas will be reduced due to disproportionate spending by municipalities, industries, and households to ensure adequate water supply.

2. Opportunities to Achieve Water Security in South Asian Cities

To counter these serious and growing challenges, the following sections address in detail several engineering opportunities to enhance long-term water security in South Asian cities.

2.1. There Are Proven Techniques to Improve Urban Water Distribution

Globally, substantial experience has been accumulated in successfully distributing continuous water supply throughout large cities. Best-practice recommendations have been developed for broad actions to reduce NRW and IWS, including complete metering of production and consumption, improved billing and collection, identification and repair of visible and invisible leaks, and elimination of illegal connections [40,41]. Numerous methods and tools have been developed for managing losses in water distribution systems [42]. Illegal connections can be physically identified and removed, though political and economic factors play large roles in their definitive solutions. Effectively managing physical losses (leakage) in distribution systems requires active control measures, speed and quality of repairs, and effective pressure management.

Metering of water flows (described in detail in Section 2.2) at multiple points throughout a municipal water utility system is important to identify leakage and to manage municipal water flows. There are significant economic considerations in investing in leakage reduction measures in urban water supply systems [43]. Modern flow metering, pressure management, and data capture technologies can quickly identify burst pipes and estimate the gradual accumulation of smaller leaks [44]. Dividing the network into sub-zones such as district metering areas can effectively quantify losses within the region. Similarly, bulk meter zones at production sites or water inlets can help in isolating transmission leaks. Flow meters can only detect the general area of leakage but cannot pinpoint the exact location of a leak. For this, sensors such as ultrasonic noise loggers, leak noise correlators, and ground microphones are manually applied to the pipelines to detect the exact location of the leakage for repair.

Another important tool to reduce urban water loss is pressure management, as the rate of leakage in water distribution networks is a function of the pressure applied by pumps or gravity. There is a direct physical relationship between the water pressure, and both slow leakage rate and the frequency of burst pipes. The most common and cost-effective measure is automatic pressure reducing valves (PRV). PRVs are installed at strategic points in the network to reduce or maintain network pressure at a set level. The valve maintains the pre-set downstream pressure regardless of upstream pressure or flow-rate fluctuations [45]. Other pressure management measures include air relief valves to release negative pressures or air bubbles in a pipeline, variable speed controllers, and break-pressure tanks.

An important technological advance in improving urban water management is supervisory control and data acquisition systems (SCADA). SCADA enables acquiring data from remote devices such as valves, pumps, and sensors and allows for overall system control from the utility office. This provides precise process control to efficiently provide water throughout the network. SCADA host platforms provide functions for graphical displays, alarms, trend analysis, and historical operations data. Integrating data from household water meters, district metering areas, and utility transmission meters, SCADA can allow for central monitoring and detection of water losses, which would help the authorities perform selective maintenance and reduce losses.

Notwithstanding the accumulated global best practices for urban water distribution, South Asian municipal utilities are currently challenged to provide their inhabitants with continuous supply of high-quality water. Private markets for household water are expanding in many cities to satisfy unmet demand. IWS, in particular, is impeding the adoption of improved water utility management practices such as metering and automated control systems. These best-practice approaches require continuous water supply and cannot be applied where pressure is supply-driven rather than demand-driven [36]. In systems that are intermittently pressurized, as is common in South Asia, the analytical water balance approach cannot be used, and current automated control systems are unable to operate with non-continuous water supply. IWS can cause malfunctioning of monitoring equipment and incorrect water flow measurements due to air pockets, vacuums, and repeated drying and wetting. IWS also causes premature wear of the infrastructure, further contributing to losses and intermittency.

Though challenging, increasing the distribution efficiency of urban water is a significant lever for improving water security in South Asia. If Indian cities were to reduce their NRW loss from the current average of 44% down to a target average of 15%, about 4.4 km³ of physical water leakage would be avoided per year, assuming 75% of NRW is from physical losses due to leakage and 25% is from apparent losses due to theft.

2.2. Water Metering Improves System Management and Billing

Metering of water does not directly affect the amount of water used. However, combined with a suitable water pricing structure, metering may act as an economic incentive to adopt water-saving behaviours and technologies. Metering and pricing of water may be an essential prerequisite to adopt other water-saving technologies [46]. Water volume may be measured and billed directly, for example, water used by urban households. In rural contexts, water use may be metered and billed indirectly based on electricity used for groundwater pumping for irrigation. In all cases, the water needs of the poorest residents must be considered and respected.

In addition to metering to enable accurate water billing, metering of water flows at multiple points throughout a municipal water utility system is a necessary component of finding and eliminating leaks and theft to reduce non-revenue water. In general, metering is an essential step to understanding and effectively managing municipal water flows: “you can’t manage what you don’t measure.”

A variety of water meters are available, based on different principles of operation [47]. Most residential and small commercial applications use mechanical positive displacement meters, which use oscillating pistons or disks to measure the volume of water that passes through. Other commonly used water meters measure the velocity of water through a fixed area, from which water volume can be determined. Water velocity is commonly measured with mechanical meters such as jet meters and turbine meters. Non-mechanical methods such as electromagnetic and ultrasonic meters can also be used to measure water velocity. Electromagnetic meters apply a magnetic field to the metering tube and determine the flow velocity in the tube based on electromagnetic induction. Ultrasonic water meters use transducers to send ultrasonic sound waves through the water to determine its velocity, based on either the Doppler effect or the transit time between two fixed points.

After a meter is used to measure the amount of water consumption, that information must be conveyed to the water utility to enable billing and collection. Traditionally, water meters were read manually by workers walking house to house. More recently, various electronic means have been developed to automatically convey meter readings to the water utility. Automated meter reading (AMR) can increase accuracy and timeliness of metering [48]. An automated meter can calculate the flow rate at a client connection and can communicate the data remotely to the host platform. AMR technologies are currently expensive but have the potential to be mass manufactured, allowing deployment of efficient smart meters for all end-users to aid in accurate billing and management.

2.3. Efficiency of Industrial Water Use Can Be Significantly Improved

Industrial production in South Asian cities is increasing rapidly. Many industrial processes require water for various purposes such as cooling and washing. About 68% of total industrial water usage in India is for cooling of thermo-electric power plants [49]. A range of methods can be used for power plant cooling, with a large variation in specific water use (see

Table 2. Thermal power plants can use various types of cooling systems, which each have advantages and disadvantages in terms of water withdrawal, water consumption, cost, generation efficiency, and other factors.

Cooling Technology	Advantages	Disadvantages
Once-through	Low evaporative consumption High cooling efficiency Lowest capital cost Mature technology	Highest water withdrawal Ecosystem impacts from withdrawal and discharge Restrictions on hot water discharge
Wet recirculating	Lower water withdrawal than once-through Mature technology	Highest evaporative consumption Lower power plant efficiency than once-through Higher capital cost than once-through
Dry recirculating	Very low water withdrawal No evaporative consumption	Highest capital cost Low power plant efficiency (when weather is hot) Large land area requirement
Hybrid	Lower capital cost than dry cooling Less evaporative consumption than wet recirculating cooling Flexibility in operation	Higher capital cost than wet recirculating Limited technology experience

, adapted from [50,51,52]). In India, 80% of thermal power generation uses freshwater recirculating cooling systems [53] (see Figure 11), which have the highest rate of evaporative freshwater consumption of all cooling system types. Other major industrial water users include steel, textiles, pulp, and paper. For most of these industries, evaporative consumption is relatively low, with approximately 80% of withdrawn water returned as wastewater. The wastewater quality from these industries is typically poor, with diverse organic and inorganic contaminants that vary by industry.

Current water use efficiency in South Asian industries is typically very low, and the adoption of global best practices would significantly reduce industrial water use (see Figure 12, based on data from [49]). For example, coal-based thermal power plants in India have an average specific water consumption of about 4.5 m³/MWh, more than twice that used in power plants in other global regions [49]. We estimate that the adoption of water-conserving measures for power plant cooling in India could reduce water withdrawals by over 20 km³ per year, assuming an adoption in 80% of Indian thermal power plants of water efficiency technologies that reduce specific water use by 50%.

A further example is the steel industry in South Asia, which typically uses much more water to produce a ton of steel than the global average (see Figure 12, based on data from [49]). There are ample opportunities to reduce industrial water use in the steel and other sectors, e.g., dry cooling, dry coke quenching, seawater cooling, and zero liquid discharge [54]. In recent decades, as water constraints have been felt in various parts of the world, much effort has been expended globally to devise industrial processes that conserve water. Implementation of this global best practice in South Asia could significantly reduce future demand for industrial water as well as reduce the quantity and improve the quality of industrial wastewaters. We estimate that adoption of best-practice water conservation measures in the Indian industrial sector (not including power plant cooling) could reduce water withdrawals by almost 10 km³ per year, assuming adoption by 80% of Indian industries of water efficiency technologies that reduce specific water use by 50%.

While increasing industrial water use efficiency will be critical for enhancing overall water security in South Asian cities, the electric power sector faces additional water-related challenges. South Asia increasingly faces constraints to electricity generation linked to water availability [55]. For example, cooling water shortages in summer of 2010 forced the Chandrapur coal-fired power station in Maharashtra to shut down, leading to power outages across the state. A delayed monsoon in India in 2012 raised electricity demand for pumping groundwater for irrigation while reducing hydropower generation, contributing to widespread sustained blackouts. Between 2013 and 2016, fourteen of India’s largest thermal power utility companies experienced disruptions at least once due to water shortages [53].

2.4. There Is Great Potential for Wastewater Reuse and Recycling

Wastewater reuse is becoming more common in water-scarce regions, where wastewater is increasingly seen as a resource and not as waste [56]. Potential levers for enhancing water security in South Asia include the direct reuse of household grey water and the treatment and reuse of wastewater from industries, municipal sewers, and agricultural run-off [57]. Reusing wastewater brings two important benefits: less pollution entering water bodies and less need for freshwater withdrawals. However, in the areas where it is currently practiced, wastewater reuse is typically considered as a temporary solution for acute needs, instead of implemented as a long-term solution to improve water security.

2.4.1. Large-Scale Water Treatment and Reuse

The quantities of wastewater generated in South Asia are enormous, thus reusing this water for other purposes is a major lever for enhancing water security. If 80% of the wastewater collected by urban sewage networks in India was reused, an additional water resource of 18 km³ per year would be obtained. Nevertheless, the large-scale treatment and reuse of water faces several challenges, depending on the intended use of the recycled wastewater and the scale of the treatment facility [58]. Major reuse applications include agriculture (food and non-food crops), industry, and groundwater recharge, for which increasing effluent quality is required, respectively. For agricultural purposes, nutrient removal (or partial nitrogen removal) can be left out of the treatment process, whereas reuse in industrial applications or groundwater recharge requires nutrient and solids removal. Groundwater recharge applications may also require removal of micro-pollutants as well as organic carbon. Treatment steps typically include activated sludge, filtration, and decontamination. In terms of costs, the higher the quality of treated effluent, the higher the total capital and operational costs are. Generally, larger treatment plants have an increased efficiency, which lowers the lifecycle costs and environmental impacts per m³ of treated water [59].

There are many methods to reuse wastewater for other beneficial purposes. Engineered wetlands, for example, provide a natural way to both treat wastewater and allow it to percolate back into the ground to recharge water tables. Kolkata fisherman have been reusing wastewater from the city to feed their aquaculture ponds for decades. Some sewage treatment technologies result in sludge, which can be dried in the sun and then applied on fields or buried, which recycles nutrients and conditions the soil. Furthermore, biogas and heat energy can be recovered from a range of treatment processes. Several countries have begun growing microalgae on sewage to produce biofuels, bio-plastics, bio-chemicals, and even nutrition supplements for humans and animals [57].

After some level of treatment, all components of sewage could be harvested and reused, including the elements in urine, faeces, and water. The most common reuse of wastewater is for agricultural irrigation, though it is necessary to treat the sewage to a hygienically safe level before application, such as exposing to sunlight for long periods. Some farmers in Pakistan apply untreated wastewater on their fields, where freshwater is scarce. However, many farmers suffer from hookworm infections and their fields are overloaded by nutrients, which underscores the need for pre-treatment before application [60].

2.4.2. Grey Water Reuse

Grey water is the term used to describe household wastewater that does not have significant faecal contamination, i.e., from sinks, showers, baths, washing machines, etc. Grey water can be directly reused for various applications that do not require potable water. For example, a household can use wastewater from their bathroom sink to flush their toilet. At a slightly larger scale, apartment buildings can use grey water to irrigate their landscaping.

Grey water has significantly fewer pathogens than sewage water and is typically safe to directly reuse for certain local, immediate small-scale applications. Larger-scale applications of grey water reuse are less suitable for two reasons. First, the quality of grey water can deteriorate rapidly during storage because it is often warm and contains some nutrients and organic matter which allows microorganisms to multiply. Second, as grey water is aggregated from more disparate sources, it becomes more difficult to control and maintain the quality, resulting in grey water with unacceptable levels of chemical or faecal contamination. Large-scale reuse of wastewater therefore requires some level of treatment and is considered in the next section.

Utilizing grey water typically requires separate piping systems for potable water and grey water. It is thus more economical to build in the systems during construction rather than to retrofit existing buildings. On a smaller scale, there is potential for well-designed consumer products that facilitate the reuse of drain water from bathroom sinks to flush toilets without the need for expensive retrofitting. We estimate that if 80% of Indian urban households used grey water to flush their toilets, it would result in a total water withdrawal reduction of about 2.4 km³ per year.

2.5. Potential for Rooftop Rainwater Harvesting Is Limited

Rooftop collection followed by household-level tank storage is an option for urban rainwater capture and storage. It is instructive to compare the amounts of rainwater falling on house rooftops in different South Asian cities, to the amount of water used by the households. Potential rainfall capture volume is determined by the local precipitation and by the size and type of the rooftop. In most parts of South Asia, rooftop capture will likely be, at best, a supplemental source of household water. Figure 13 shows the potential for rainwater capture in three locations of varying rainfall (dry: Karachi, moderate: Bengaluru, wet: Dhaka) for three different house rooftop sizes (small: 20 m², medium: 70 m², large: 200 m²). The figure also shows the estimated annual drinking water requirements (7 m³ per year) and total household water requirements (167 m³ per year) for a household of five people. The figure is based on average annual precipitation data from [39] and average per capita water consumption data from [5]. This analysis shows that household drinking water requirements could be satisfied by rooftop capture in most cities and house sizes, but total household water requirements could only be satisfied by large houses in rainy cities.

Assuming adequate precipitation and capture surface are available, a further challenge is storing sufficient quantities of water for use during later dry periods. Storing a 6-month supply of drinking water for a household of five people would require a reservoir of 3.6 m³ capacity, while storing a 6-month supply of total household water would need 82 m³. Based on current prices of common plastic storage tanks and assuming a 20-year tank lifespan, the lifecycle cost of each m³ of drinking water obtained from the storage would be almost USD 7, and each m³ of household water would cost over USD 12. Rooftop rainwater capture and storage for domestic use will therefore likely continue to play a minor role in South Asian water supply due to limited capture area and expensive storage volume. In regions of high and regular rainfall, however, it may be locally important [61].

In most South Asian cities, the total amount of rain that falls annually within the city boundaries is substantially greater than the total amount of domestic water used by the city’s inhabitants (see Figure 14, based on average per capita water consumption data from [5], average annual precipitation data from [39], and city population and land area from [4]). However, although the gross urban rainwater supply is abundant, inadequate rooftop collection area and costly tank storage pose significant challenges. Other potential options to economically capture and store a significant fraction of the gross rainfall include large-scale storage in reservoirs and aquifers. While appropriate sites are limited for additional large-scale reservoir storage, there are substantial opportunities in some areas to direct rainwater runoff into groundwater recharge zones. Managed aquifer recharge (MAR) is achieved by reducing the fraction of rainwater that runs off the land surface, thus increasing the fraction that infiltrates through the land surface and enters the soil. This is typically implemented through engineered structures that slow the downstream flow of surface run-off water, allowing more of it to infiltrate into the ground. MAR requires three conditions: the availability of uncommitted surface water, the availability of underground storage space, and the demand for groundwater [62].

2.6. There Is Limited Potential for Inter-Basin Water Transference

Transferring water from river basins with abundant water resources to other basins that are water scarce is an often-discussed large-scale intervention to enhance urban water security. Inter-basin water transference is not a new concept and has been implemented in South Asia since at least the 17th century. Examples of existing inter-basin transfers in South Asia include:

• The Kurnool Cudappah Canal, transferring water from the Krishna River basin to the Penna basin since 1870;
• The Periyar Project from the Periyar River in Kerala to the Vaigai basin in Tamil Nadu, commissioned in 1895;
• The Triple Canal Project in Pakistan, constructed during 1907–1915, linking the Jhelum, Chenab and Ravi rivers, transferring Jhelum and Chenab water to the Ravi river;
• The Indira Gandhi Canal, built in the 1960s, linking the Ravi River, the Beas River and the Sutlej River to irrigate the Thar Desert;
• The Parambikulam Aliyar project, initiated in 1958, transferring water from the Chalakudy River basin to the Bharathappuzha and Kaveri basins;
• The Telugu Ganga project, completed in 2004, bringing Krishna River water to the city of Chennai;
• The Pattiseema scheme, beginning operation in 2015, linking the Krishna and Godavari rivers.

The proposed Indian River Inter-link project continues this tradition but on a much larger scale than previous efforts. The project aims to connect northern Indian rivers with southern Indian rivers through a network of canals, dams, and pumping stations. Given the long timeframe for completion, the significant opportunity costs of the substantial resources required, the social cost to displaced households, and the high ecological disturbance, there is significant uncertainty about the net benefits of the project. Long-distance physical transfer of water requires capital-intensive fixed grey infrastructure with little flexibility to adapt to future changes (e.g., climate change). In contrast, a more flexible and accessible approach may be found in distributed investments in end-use efficiency measures such as irrigation and industrial modernization, and system integration measures such as wastewater treatment and reuse. Furthermore, permanently diverting water away from regions that currently have “excess” water denies those regions the opportunity to develop local productive uses for the water. Although some additional inter-basin transfer projects are likely to prove worthwhile individually, the proposed Indian River Inter-link project as a whole is unlikely to contribute significantly to Indian urban water security.

Other forms of inter-basin water transfer include proposals for towing of icebergs from polar regions to provide fresh water to coastal cities. Eventual small-scale implementation of this idea in some wealthier and dryer regions of the world will not be surprising, though iceberg transfer is unlikely to play a significant role in South Asia’s water future.

Large-scale, long-distance humanitarian water transfer is an increasing phenomenon. In 2016, the Indian government made emergency water transfers by train to Latur, Maharashtra, to supplement local water sources depleted by drought and overuse. With the extremely expensive per litre cost of delivered water, the need for such interventions should be avoided if at all possible by appropriate management of local water supply and demand.

3. Conclusions: Achieving Water Security in South Asian Cities

The water security challenges facing South Asian cities are complex, involving issues of water quantity and water quality across a range of temporal and spatial scales. Urban household water demand is rising quickly; surface water supply in closed river basins is fully utilized; groundwater extraction in most regions exceeds natural recharge rates; and even in cities with abundant water resources, the poorest residents often lack the means to access it. These challenges results from a confluence of factors, including demographic, climatic, behavioural, economic, and political concerns. The challenges are not unique to South Asia, and many other regions are experiencing similar issues.

No single intervention can definitively solve water security problems in South Asian cities, but numerous actions can be taken on many fronts that collectively can significantly enhance urban water security. Although there are serious constraints to accessing adequate amounts of high-quality water to meet growing demands, there are important opportunities for high-impact engineering interventions that can eliminate some inefficiencies of current water use.

A range of proven cost-effective techniques exist to improve urban water distribution and to reduce physical and apparent losses. Municipal water distribution networks can employ modern monitoring and management methods, including pressure management and leak detection and repair, to overcome problems of intermittent water supply. Metering of urban water can enable structural improvements of management and billing, though the water needs of the poorest city residents must be ensured. Industrial water-use efficiency can be significantly improved in the manufacturing and electricity generation sectors.

The quantities of wastewater generated in South Asian cities are large, thus treating and reusing this water for other purposes is a strong lever to enhance local water security. While there is limited potential for rooftop rainwater harvesting and storage, the capture-enhanced recharge of groundwater may be important in some areas. Some individual inter-basin transfer projects may prove worthwhile, but very-large-scale projects are unlikely to contribute practically to urban water security.

The causes of South Asian urban water security problems are multi-faceted and so must be the solutions. Several common-sense things must happen to improve the systemic efficiency of water use in South Asia by implementing current global best practices in multiple sectors. Technology advancement can play an important role in addressing some water security challenges but must be accompanied by thoughtful and nuanced policies. Importantly, South Asians must recognize that water is not an infinite resource as many had previously assumed.