Interim Report. Cost Recovery and Tariff Practices for Urban Water Supply and Sanitation in India. Volume II. Detailed Case Study Report

Transcription

1 Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Cost Recovery and Tariff Practices for UWSS Sector in India Interim Report on Cost Recovery and Tariff Practices for Urban Water Supply and Sanitation in India Volume II Detailed Case Study Report Chandigarh Kochi Dehradun Hyderabad Indore Pune Ludhiana Prepared for Water and Sanitation Program South Asia June 2008

2 Summary i

3 Introduction To access funds for water supply and sanitation improvements under JNNURM, service providers are required to levy reasonable user charges so that O&M costs are recovered within five years. This study reviews data from a range of case studies within India and beyond, to identify the challenges associated with cost recovery and suggest viable solutions based on international good practices. The national analysis examines data from 23 Indian cities. Of these, seven were studied in detail while secondary data from 16 further cities were used to supplement and to validate the findings. The secondary data are taken mostly from a study by the Asian Development Bank (ADB)_and the Ministry of Urban Development (MoUD) entitled 2007 Benchmarking and Data Book of Water Utilities in India. Based on this study, guidelines on cost recovery have been prepared for municipal managers and service providers. Cities studied Water and Sanitation Program South Asia (WSP-SA) Study Cities Dehradun, Ludhiana, Chandigarh, Indore, Pune, Hyderabad, Cochin ADB-MoUD-GoI Study Cities Ahmedabad, Amritsar, Bangalore, Bhopal, Chennai, Coimbatore, Jamshedpur, Kolkata, Mathura, Mumbai, Nagpur, Nasik, Rajkot, Varanasi, Vijayawada, Vishakhapatnam Scope of the study The study set out to examine tariff and cost recovery practices associated with both water supply and sewerage services in India. Unfortunately, insufficient data were on sewerage costs and revenue, even in cities with a substantial network, reflecting the fact that this is an underdeveloped and neglected area of service provision. As a result, the bulk of the discussion concerns water supply services, though efforts have been made to consider sewerage wherever possible. Generally, it was difficult to access reliable data from the seven cities studied in detail and, this being a snapshot study, it was not possible to generate data independently. Overview of tariff structures Urban development is a state subject and tariff structures for water and sewerage services are usually set by the state governments or are subject to their approval. Some states (for example, Maharashtra and Gujarat) prescribe minimum or maximum tariffs for various user categories while others define tariff structures for different categories of municipality or even for individual towns. Tables 2 and 3 show that 19 out of the 23 cities operate a mixture of fixed charge and volumetric charges for water consumption. 1

4 Ahmedabad Amritsar Bangalore Bhopal Chennai Chandigarh Coimbatore Table 2: Tariff Structure of 23 Cities Type of Tariff Structure Water Metered Single part category-wise uniform volumetric tariffs6 Single part increasing block tariffs Two part category-wise uniform volumetric tariffs7 Two part increasing block tariffs Water Unmetered Ferrule ARV-based (i) Part of property tax (ii) Separate from property tax (iii) Water tax Tap-based No. of time of supply per day Area of land/plot size or building based Free Water Free water for domestic sector Free water to poor General Water Charge/Tax to all (Metered and Unmetered) ARV-based Sewerage Service No service charge Fixed charge ARV-based (i) Part of property tax (ii) Separate from property tax (iii) Sewerage tax No. of WCs in premises Area of land/plot size or building based As % of water For poor, people who were earlier accommodated in rehabilitation colonies at Rs. 100 per month no mention of Ferrule size 6 A single part category-wise uniform tariffs means that the ULB/agency charges only volumetric tariffs but at different rates for different categories of consumers, such as for domestic, commercial, institutional and industrial. 7 A two part category-wise uniform tariffs means that the ULB/agency levies both fixed charges and volumetric charges. But different rates for different categories of consumers apply, such as for domestic, commercial, institutional, industrial. 2

5 Cochin Dehradun Hyderabad Jamshedpur Kolkata Ludhiana Indore Mathura Cost Recovery and Tariff Practices for UWSS Sector in India Table 2: Tariff Structure of 23 cities (Continued) Type of Tariff Structure Water Metered Single part category-wise uniform volumetric tariffs Single part increasing block tariffs Two part category wise uniform volumetric tariffs $ Two part increasing block tariffs Water Unmetered Ferrule size based ARV-based (i) Part of property tax (ii) Separate from property tax (iii) Water tax $$ Tap-based $$$ Area of land/plot size or building based No. of time of supply per day Free Water Free water for domestic sector # Free water to poor General Water Charge/Tax to all (Metered and Unmetered) ARV-based Sewerage Service No service charge Fixed charge ARV-based (i) Part of property tax (ii) Separate from property tax (iii) Sewerage tax $$ No. of WCs in premises $ Area of land/plot size or building based As % of water consumed $ For nondomestic sector; $$ for all unconnected properties; $$$ where ARV is not (in case of Dehradun) # For all plots below 125 square yards 3

6 Mumbai Nagpur Nasik Pune Rajkot Varanasi Vijayawada Vishakhapatnam Cost Recovery and Tariff Practices for UWSS Sector in India Table 2: Tariff Structure of 23 cities (Continued) Type of Tariff Structure Water Metered Single part category-wise uniform volumetric tariffs Single part increasing block tariffs Two part category wise uniform volumetric tariffs Two part increasing block tariffs Water Unmetered Ferrule size based * ARV-based (i) Part of property tax (ii) Separate from property tax (iii) Water tax Tap-based Area of land/plot size or building based No. of time of supply per day Free Water Free water for domestic sector Free water to poor General Water Charge/Tax to all (Metered and Unmetered) ARV-based ** Sewerage Service No service charge Fixed charge ARV-based (i) Part of property tax (ii) Separate from property tax (iii) Sewerage tax No. of WCs in premises Area of land/plot size or building based As % of water consumed * For poor, domestic independent connection, domestic joint connection, non-domestic **As water benefit tax * For nondomestic sector; ** for all unconnected properties (in case of Pune) Government assistance to service providers State (and sometimes federal) governments offer a range of capital grants, loans, loan guarantees and operating subsidies to public water and sanitation service providers. Some of these are intended to offset exceptional production costs due, for example, to high electricity requirements to pump water or the need to transport bulk water over a long distance. In such cases, as a general rule, subsidies should be designed to meet the gap between own source revenue of the service provider and unavoidably high costs of service provision, so that the consumer does not pay a higher charge for water. They should not, however, be used to compensate for operational inefficiency. In practice, government support to service providers is ad hoc and rarely linked to cost recovery strategies (Table 19). Dehradun, for example, receives sporadic O&M grants from the state government even when revenue is sufficient to cover operational costs. There is therefore little incentive to improve operational efficiency. Indore, too, receives operational subsidies (in the form of salaries and power grants) even though it could potentially recover all its operational costs from revenue, by improving operational efficiency. 4

7 Dehradun ( Ludhiana ( Chandigarh Indore Pune Hyderabad Kochi Cost Recovery and Tariff Practices for UWSS Sector in India Capital grants can be justified in Hyderabad and Indore due to the need to transport water from remote sources and use multi-stage pumping. There are, however, no explicit criteria for providing capital grants and no performance benchmarks are set as the basis for accessing the funds. In some other cities where sourcing water is not exceptionally difficult and the revenue potential is high, it is difficult to make a case for further capital grants. Table 3: State Assistance for Water and Sewerage Services Particulars (Rs. million) Revenue grant for water service Capital grant for water service Total own source revenue from water service Revenue grant for sewerage service Capital grant for sewerage service Total own source revenue from sewerage service Total revenue grant received by ULB/utility Figure not separately for water head As mentioned above No grants, funds service from general account surplus Separate figure not for water head As mentioned above As mentioned above As mentioned above Figure not separately for water head Data gap Data gap No grants; it funds services from general surplus No revenue grants Figure not separately KWA* receives sizeable grant but data not for Kochi As mentioned above Data gap Data gap As mentioned above As mentioned above As mentioned above Data gap As mentioned above No revenue grants As mentioned above Data for city not maintained for both services Total capital grants received by ULB/utility -- Data gap 4292 for both services Total own source revenue of ULB/utility Data gap *KWA= Kerala Water Authority Aspects of Cost Recovery This section reviews data from the 23 cities and identifies the factors contributing to a generally low level of cost recovery. The findings must be treated with caution given that some of the source data were incomplete or unreliable. In particular, costs were often underreported due to poor accounting and recordkeeping and failure to disclose a range of operating subsidies relating to power, salaries, and interest on loans (amongst other things). In Dehradun and Kochi, for example, water supply and sewerage services are managed by parastatals and city units are unable to access data on expenses covered by headquarters. Nevertheless, it is possible to identify some broad trends across the 23 cities. None of the service providers consulted in the course of this research was able to provide information on total capital expenditure, the present value of the existing network or depreciation. It is nevertheless evident that capital costs in the water and sanitation sector are substantial and 8 Water charges including meter rent. 9 Only water cess. 10 Only sewerage cess 11 Only sewerage cess 12 Inclusive of other income, interest etc. 5

8 Cost Recovery and Tariff Practices for UWSS Sector in India their recovery through user charges would cause a substantial revision in tariffs. Recognizing this, JNNURM does not currently require recovery of capital costs. Operating ratio The operating (working) ratio is the simplest and most widely used measure of operational efficiency in an organization. It expresses a service provider s operating expenses as a percentage of revenue, and if the ratio is greater than 1 it indicates that expenses exceed revenues and the operation is losing money. Conversely, if the ratio is less than 1 it indicates an operating surplus. 13 Of the 23 cities studied, the data indicate that 11 have an operating surplus while the others are operating at a loss. The surpluses range from 4 to 56 percent, while losses range from 14 to 373 percent. Cost of service provision O&M costs ranged from Rs. 2 to 22 per kl (Table 4) Table 4: Operation and Maintenance Costs Operation and maintenance cost (Rs./kl) Number of cities Cities Dehradun, Ahmedabad, Coimbatore, Nagpur, Nasik Pune, Jamshedpur, Mathura, Rajkot, Varanasi, Vijayawada Ludhiana, Amritsar, Bhopal, Kolkata, Mumbai Chandigarh, Cochin, Vishakhapatnam Hyderabad, Chennai Bangalore Indore Note: These data do not factor in the depreciation costs of new assets. The wide range of costs is partly attributable to differences in power requirements for bulk supply and distribution. Where water is provided by gravity flow, as in Pune, Nasik, Varanasi and Kolkata, power costs are relatively low, while in cities such as Hyderabad, Bangalore and Indore a considerable amount of pumping is involved. More importantly, however, high costs arise from operational deficiencies such as technical losses, a low customer base, and inefficient billing and collection. The data in Table 5 suggest that, for 19 out of the 23 cities, Rs. 4 per kl would be a reasonable benchmark for current O&M costs. This takes account of the tendency to underestimate real supply costs. An earlier study by WSP-SA (2002) suggests Rs. 15 as a reasonable benchmark, but the current data indicate that this would only be valid for Bangalore and Indore. 13 Moody s calculates the U.S. median figure of operating (working) ratio for water and wastewater enterprise projects to be approximately 0.6. In other words, operating expenses amount to 60% of operating revenue and there is a operative surplus of 40%. 6

9 Dehradun Ludhiana Chandigarh Indore Pune Hyderabad Kochi ( Cost Recovery and Tariff Practices for UWSS Sector in India Table 5: Cost Recovery and Tariff Levels ( ) Service provider s overall operating ratio (on own source revenue) 14 Service provider s overall operating ratio (inclusive of operational grants) No operative grants 1.39 No operative grants Water service operating ratio (on own source revenue) Water service operating ratio 2.39 (inclusive of grants) 0.35 No No grants Grants O&M cost per kl Total cost per kl No operative grants Average own source revenue per kl Average tariff per kl ULBs carry out various functions including water supply and sewerage. They also get revenue from various sources. So the service provider s overall operating ratio means operating ratio of the ULB. 15. Water supply service operating ratio takes into account operative revenue and expenditure of water supply service only. Since some cities are served by water supply parastatals for comparison this ratio is taken into account. 16. Includes income from sewerage also. Since the expenditure for water and sewerage are not separately, separate ratios cannot be worked out. 17. Since the grants for water and sewerage are not separately, separate ratios cannot be determined. This figure includes grants spent on sewerage as well. 18. Inclusive of depreciation. 19. Exclusive of interest and depreciation. 20 Total cost here includes capital investment and loan repayment costs only but does not include depreciation. As water service provider follow different accounting practices separate footnotes have been given regarding their accounting practices 21. Figure provided by Uttarakhand Jal Sansthan (UJS), Dehradun, based on the formula provided by the team. Calculations not, 22 Inclusive of depreciation 23. Inclusive of interest and depreciation, 24. Annual billing (current demand) /annual water production/supply (whichever ). 25. Includes demand from sewerage as well. Separate figures for water supply and sewerage not. 7

10 Mathura Mumbai Nagpur Nasik Rajkot Varanasi Vijayawada Vishakhapatnam Ahmedabad Amritsar Bangalore Bhopal Chennai Coimbatore Jamshedpur Kolkata Cost Recovery and Tariff Practices for UWSS Sector in India Table 5: Cost Recovery and Tariff Levels ( ) 26 (continued) Service provider s overall operating ratio (on own source revenue) Service provider s overall operating ratio (inclusive of operational grants) Water service operating ratio (on own source revenue) Water service operating ratio (inclusive of grants) O&M cost per kl Total cost per kl Average own source revenue per kl Average tariff per kl Table 5: Cost Recovery and Tariff Levels ( ) 27 (continued) Service provider s overall operating ratio (on own source revenue) Service provider s overall operating ratio (inclusive of operational grants) Water service operating ratio (on own source revenue) Water service operating ratio (inclusive of grants) O&M cost per kl Total cost per kl Average own source revenue per kl Average tariff per kl Revenue per kiloliter Actual revenue per kl water in the 23 cities ranges from Rs 0.16 to Rs (Tables 5 and 6). Only in nine cities is this revenue higher than the average O&M cost per kl. Interestingly, data on operating ratios indicated that 11, not nine, cities have an operating profit, highlighting problems with the source data. 26. Since data of 16 cities are collected under different study and format, some of the data are not for these cities. 27. As data of 16 cities are collected under different study and format, some of the data are not for these cities. 8

11 Connection charges Cost Recovery and Tariff Practices for UWSS Sector in India Low official connection charges are an impediment to the expansion of the network in most cities. The fee charged by the 23 service providers for a new house connection is Rs. 2,594 on average while expansion costs are three or four times higher: typically Rs. 7,500 to Rs. 10,000. Only Pune has a comparable connection fee, at Rs. 8,541. Next in line are Vijayawada at Rs. 5,500 and Dehradun at Rs. 4,090. Twelve cities have a connection charge between Rs. 1,000 and 3,000 while the remaining eight charge less than Rs. 1,000. Factors undermining cost recovery The solution commonly proposed for improving cost recovery is to increase tariffs. There are, however, a range of matters to consider when planning cost recovery improvements, including: Operational costs (especially power and staffing); Tariff level; Tariff structure; Consumer base (coverage with registered connections); Non-revenue water or unaccounted for water (waste and leakages, plus illegal connections); Metering; and Billing and collection efficiency. Very often it is a combination of these factors, not tariff levels alone, that explains the inability to recover costs. For example, non-revenue water 28 (on account of technical losses) often stand at 20-30% of production, non-revenue water (on account of non-technical aspects 29 ) at 10 to 20% and collection efficiency at less than 60%. Improving cost recovery depends above all, therefore, on operational efficiency, though some tariff modifications may also be needed. Improving operational performance need not involve major capital investments. It may be possible, for example, to increase the number of registered users or reform billing and collection processes (perhaps with the introduction of incentives) using the resources currently. Service coverage Service coverage is commonly defined in terms of the geographical reach of the distribution network on the number of registered connections as a percentage of total households. While coverage is a useful indicator, it must be borne in mind that the presence of infrastructure does not in itself mean that people are receiving adequate water. A large proportion of poor households may be dependent on limited standpost supplies, while house connections can also be subject to low pressure and intermittent supply. 28 Term non revenue water includes water lost or which has not yielded revenue due to technical and nontechnical reasons. For better clarity of causes, we have segregated it under technical and non-technical categories. 29 Non-technical component of non-revenue water results from water supplied free through standposts or water consumed by illegal connections or in-house water consumption by the local body itself or water consumer by anybody but not billed. 9

12 Dehradun Ludhiana Chandigarh Indore Pune Hyderabad Kochi Cost Recovery and Tariff Practices for UWSS Sector in India Table 6: Common Causes of Service Providers Inability to Recover Costs Particulars Network coverage for water supply in terms of area 95% 85% 100% 75% 100% 100% 100% Population coverage for water supply 95% Total domestic connections as % of total households Percentage of total NRW/Unaccounted for Water Technical losses NRW not caused by technical losses Total collection efficiency Recovery against current demand (water) % Recovery against 85% (15% undeclared area) 100% 85% 85% 100% 100% 45% 46% 30 62% 42% 69% 39% 83% 30% 38% 43% 30% 42% 25% 30% 4.30% 14-15% 20% 28% 31 13% 13% 25.70% 27% 37% 16% 66% 22% 22% 32 54% 13% Data Not Data Not 92% 52% 30% 0% 10% 11% arrears Staff per thousand connections Sewerage network coverage 30% 67% 100% % 70% 5% Population coverage for sewerage Sewered households / total households Recovery against current demand (sewerage) Recovery against arrears (sewerage) 30% 100% % 38% Data Not Data Not 30 As of October 2007, there are 156,050 domestic connections for a population of 1.7 million. 31 Includes water lost due to lack of coverage in undeclared areas and captive production by industries. 32 Includes sewerage as well. Separate figures are not for water supply and sewerage. 33 Due to lack of data, the figure has been taken from a recent study by ADB. 34 As of October there are 129,143 domestic sewer connections for a population of 1.7 million. 10

13 Ahmedabad Amritsar Bangalore Bhopal Chennai Coimbatore Jamshedpur Kolkata Mathura Mumbai Nagpur Nasik Rajkot Varanasi Vijayawada Vishakhapatnam Table 6: Common Causes of Service Providers Inability to Recover Costs (continued) Cost Recovery and Tariff Practices for UWSS Sector in India Network coverage for water supply in terms of area Population coverage for water supply 74.5% 75.7% 92.9% 83.4% 89.3% 76.1% 74.4% 79.0% 70% 100% 91.5% 92.6% 98.1% 77.7% 70.5% 49.2% Total domestic connections as % of total 60.86% 71.64% 41.86% 33.28% 26.78% 53.82% 37.08% 29.75% 46.43% 98.4% 43.64% 44.84% 97.95% 37.05% 49.86% 41.55% households 35 Percentage of total NRW/Unaccounted % 45.1% -- 17% 41.3% 12.8% 35.0% % 51.9% 59.5% 23.5% 30.0% 23.8% 14.5% for Water Technical losses NRW other than technical losses % of total supply Total collection efficiency % 69.27% % % % 75.33% 100% 100% % % 79.72% 92.08% 45.33% 64.32% % 86.10% Recovery against current demand (water) % Recovery against arrears Staff per thousand connections DATA ON SEWERAGE NOT AVAILABLE 35 Calculated as domestic connections/potential households. Potential households were calculated as total population/5 on assumption of a household size of five members. 36 Calculated as (annual revenue/annual billings) x

14 Cost Recovery and Tariff Practices for UWSS Sector in India In terms of house connections, Mumbai and Rajkot have the highest coverage at approximately 98 percent, followed by Kochi at 83 percent (Table 6). Most cities, however, have much lower coverage, with 15 in the range percent. When the data in Table 6 are compared with those in Table 5, it becomes clear that cities with lower service coverage also have a have higher operating ratio. The majority of service providers in the 23 cities have ample potential to expand their consumer base and improve their revenue generation. Doing so would involve extending the distribution network while ensuring an adequate bulk supply. It would appear that many of the cities could improve revenues substantially by expanding coverage of house connections to 100 percent, assuming that adequate water is to serve the entire population. For example, revenue in Ludhiana could (in theory) see an increase from Rs. 86 million to Rs. 252 million, while Pune could achieve Rs. 120 million (Table 7). Table 7: Revenue Potential of Increasing the Consumer Base 37 City Ludhiana Existing Status/ Potential 15% undeclared area of city not serviced, potential water demand 62,000 kl Working of Revenue Potential 62,000 kiloliter per day (kld) water x Rs minimum rate x 365 Rs. 86 mn Supplying water to industries, which are presently not served. Potential water demand 60,000 kl per day 60,000 kld water x Rs minimum rate x 365 Rs. 166 mn Pune Only 18,500 slum households paying water charge while 181,500 slum households do not pay any water charge 181,500 x Rs. 360 per year = Rs. 65 mn Rs. 65 mn Around 60,000 potential tax payers not in the water charge database 60,000 x Rs. 900 per year minimum water charge Rs. 54 mn Hyderabad Around 100,000 potential consumers can be added as per estimate of service provider 100,000 new consumers x 240 kl water per year consumption x Rs. 6 per kl water Rs. 144 mn Indore Of potential 342,000 consumers only 161,000 consumers under water charges Chandigarh Of potential 210,000 consumers only 130,000 consumers under water charges 181,000 potential consumers x Rs per year minimum water charge 80,000 potential consumers less 30,000 slum dweller = 50,000 consumer x Rs. 720 per year minimum water charge Rs. 391 mn Rs. 36 mn Cochin Of potential 130,000 consumer currently 108,000 consumers are under water coverage 22,000 potential consumers x Rs. 1,080 per year minimum water charge (240 kl water per year consumption x Rs. 4.5 per kl water) Rs. 23 mn Dehradun Of potential 135,000 consumers only 60,000 consumers under water charge 75,000 potential consumers less 25,000 slum dwellers x Rs. 1,800 per year minimum water and sewerage charge Rs. 90 mn Nonrevenue water NRW refers to water that is produced but not sold to consumers. It arises from technical losses occurring during transmission and distribution; water provided free through standposts or under exemptions (often for the poor); and theft including illegal connections. High levels of NRW are common throughout India. For the service providers assessed in this study, it was estimated at percent of production, though it could not be measured 37 In case of four cities - Pune, Hyderabad, Cochin and Chandigarh extension of network is not required as cities have claimed 100% area coverage by water supply network (Pl refer table 4) while in case of three cities Dehradun, Ludhiana and Indore extension of water supply network will be required for increasing the consumer base. This will involve capital investment but figures were not. 12

15 Cost Recovery and Tariff Practices for UWSS Sector in India accurately due to a lack of metering in production and consumption, or faults with existing meters. The lack of hard data on NRW complicates the planning of revenue improvements. Table 8 estimates the revenue potential of a reduction in NRW. A reduction to the internationally accepted level of 10 percent should mobilize huge revenue increases. Table 8: Revenue Potential of Reducing Nonrevenue Water 38 City Existing Status/ Potential Working of revenue potential Revenue Potential Ludhiana Reversing exemption of 72,000 households and imposing water and sewerage charge Minimum annual water charge (Rs. 1,260) + sewerage charge (Rs. 1,260) = Rs. 2,520 x 72,000 households Rs. 181 mn Pune Technical loss if reduced from 25% to 10% will save 150,000 kld water 150,000 kld water x Rs. 3 per kl x 365 Rs. 164 mn Hyderabad Technical loss at 15% if brought down to 10% will save 63,100 kl water per day 63,100 kl water savings per day x 365 x Rs. 6 cost of production per kl Rs. 138 mn As per water audit illegal consumption of water is 7% of total water supply (96,320 kl water per day) kl water per day x 365 x Rs. 6 cost of production per kl Rs. 210 mn Indore Technical loss at 14.4% if brought down to 10% will save 8,000 kl water per day 8,000 kl water per day x 365 x Rs. 12 cost of production per kl Rs. 35 mn Chandigarh Technical loss at 25% if brought down to 10% will save 52,627 kl water per day 52,627 kl water per day x 365 x Rs. 4 water charge per kl Rs. 76 mn Unutilized installed capacity is 8% (30,509 kl water per day) + 5% water through standposts (19,068 kl water per day) kl water per day x 365 x Rs. 4 water charge per kl Rs. 70 mn Cochin Technical loss at 30% if brought down to 10% will save 29,814 kl water per day 29,814 kl water per day x 365 x Rs. 2 minimum tariff per kl Rs. 21 mn Dehradun Technical loss at 30% if brought down to 10% will save 21,372 kl water per day 21,372 kl water per day x 365 x Rs average revenue per kl water Rs. 21 mn Collection efficiency A common problem in determining a service provider s collection efficiency is the lack of separate accounts for past arrears and current demand. Recovery of arrears is, however, known to be in the range of percent. Municipal legislation does not have write off clauses for old debts, hence the cumulative figure continues to grow. There are problems with the data provided and some of the cities have reported a collection efficiency of more than 100 percent, which is clearly not possible. Of the cities providing more reliable data, the highest reported levels of collection efficiency are in Jamshedpur (100 percent), Kolkata (100 percent) and Nasik (92 percent). Jamshedpur is, however, a special case in that a private company provides water, while Kolkata has no charges for domestic users. Coimbatore (75 percent), Nagpur (80 percent) and Vishakhapatnam (86 percent) are the next best performers while the lowest efficiency levels are reported in Ludhiana (16 percent) and Kochi (13 percent). Most cities lack user-friendly payment options; Chandigarh, Hyderabad and Pune being exceptions. 38 Reducing non-technical component of non-revenue water will not involve much of capital investment, as it mainly requires improving operational efficiency and governance, while reducing technical loss component of NRW certainly involves additional capital investment but figures were not. Overall 10% technical loss is accepted because reducing loss beyond this level involves huge capital investment and cost outweigh benefits. 13

16 Cost Recovery and Tariff Practices for UWSS Sector in India The study found that if collection efficiency was increased to 90 percent of current demand and at least 20 percent of past arrears, service providers could mobilize substantial additional revenue: as much as Rs. 1,650 million in the case of Hyderabad. In short, all 23 cities are failing to realize their revenue potential by a substantial margin -- up to 80 percent in some cases (Table 9). Table 9: Revenue Potential of Improving Collection Efficiency Existing Status/ Potential Working of revenue potential Revenue Potential Ludhiana Total collection efficiency stands less than 20% 90% of current demand 39 of Rs mn = Rs mn + 20% of past arrears of Rs mn = Rs mn making total potential recovery of Rs mn less current recovery of Rs. 170 mn. Rs. 165 mn Pune Current collection efficiency 59%, current demand Rs. 1,333 mn, current collection Rs. 790 million 90% of current demand of Rs. 1,333 mn = Rs. 1,200 mn less current collection of Rs. 790 Rs. 410 mn Hyderabad Current collection efficiency 54% against annual billing/demand of Rs. 4,536 million, current collection of Rs. 2,433 million 90% of current demand of Rs. 4,536 mn = Rs. 4,088 mn less current collection of Rs. 2,433 Rs. 1,650 mn Indore Total collection efficiency 22%, current collection is of Rs. 116 million 90% of current demand of Rs. 440 mn = Rs. 396 mn less current collection of Rs. 116 Rs. 280 mn Chandigarh Current collection efficiency 66%, collection against past arrears 0%, current demand Rs million 90% of current demand of Rs. 485 mn = Rs. 436 mn less current collection of Rs. 402 mn Rs. 34 mn Cochin Total collection efficiency 13%, current collection is of Rs. 61 million, current demand Rs. 465 million Dehradun Total collection efficiency 37%, current demand Rs. 130 mn 90% of current demand of Rs. 465 mn = Rs. 418 mn less current collection of Rs. 61 mn 90% of current demand of Rs. 130 mn = Rs. 117 mn and 20% of past arrears of Rs. 150 mn = Rs. 30 mn making total potential recovery of Rs. 147 mn less current recovery of Rs. 110 mn Rs. 357 mn Rs. 37 mn Staffing levels Most service providers have a variety of human resource issues to address, not least overstaffing and inefficient task allocation. Indore has the highest number of staff per 1,000 connections at 18.7, while Pune has the lowest at just The majority have a ratio of 5 or more. In reviewing these levels, it should be noted that there is a growing trend of outsourcing O&M functions. For example: Rajkot has outsourced pumping operations, mains and lines maintenance, sluice valve operations and leak repairs through service contracts; Ahmedabad has outsourced production, operations and maintenance through management contracts; and Nagpur has outsourced billing and collection through a management contract. Outsourced positions do not appear in service providers staff count and payroll, but are usually hidden in various operating expenditure accounts. This practice distorts comparisons between 39 Current demand in this table means amount billed for the current year consumption and does not include past arrears 14

17 Cost Recovery and Tariff Practices for UWSS Sector in India service providers, and hence the impression given by the staff ratio may not reflect reality on the ground. Outsourcing can potentially enable significant efficiency gains, but its widespread adoption is currently impeded by restrictive labor laws and union practices and a lack of capacity within service providers to manage the precontracting process. Outsourcing should not, in any case, be seen as a solution to the inefficiency problems of service providers, and some cities that have outsourced operational tasks still have high staffing ratios. Bangalore, for example, still has 4.8 staff per 1,000 connections while Chennai has 13.3 despite outsourcing production and distribution through service, management, and build-operate-transfer (BOT) contracts. Workforce management can also remain a challenge even when services are outsourced. Managerial capacities of potential contractors are often overlooked in favor of technical expertise, and service providers do not always make good choices when considering which functions to outsource. Tariff Practices Economic efficiency and resource conservation Economic efficiency has been compromised in the cities by tariffs that do not reflect the average cost of production, let alone the incremental cost of new sources which tends to be higher than that of water currently supplied. Water pricing should discourage wasteful consumption by making consumers aware of the true cost of the water they use. Metering is essential for economic efficiency, and measured tariffs should reflect the incremental cost of developing new water sources. Tariffs can, therefore, encourage resource conservation. Current pricing, billing and collection systems in the cities do not encourage conservation, but it would still be possible to create inbuilt disincentives for wasteful consumption through metering and the introduction of an increasing block tariff. However, metering in India has often proved problematic due to poor maintenance, malfunctioning of the meters (especially when the supply is intermittent), tampering by consumers and irregular meter reading by the service provider. As a result, many households that have meters are still billed on a fixed charge basis. Only seven of the cities have a high level of metering while four (Indore, Bhopal, Mathura and Varanasi) have no meters at all (Table 10). In Chandigarh, tariffs are set so low that even poor households have no incentive to limit their consumption (see Chandigarh City Report). In Kochi too, the slab sizes are poorly designed and discourage water conservation. Moreover, Kochi has a very high meter default rate; 30 percent meters are nonfunctional. In Hyderabad, the tariff structure is efficiently designed but there are high technical losses and the meter default rate is 25 percent. Irrespective of the presence of meters, the fact that water is only intermittently in many cities (sometimes for just one hour per day) means that consumption is limited physically, whatever the pricing system. Fixed Charges The use of fixed charges is widespread but has inherent equity problems in that charges remain fixed irrespective of the number of occupants or households in a property, though limited hours of supply and low pressure may prevent users from consuming large amounts. When fixed charge tariffs are set, the intention usually is to generate revenue roughly equal to that which would accrue if a volumetric tariff was in place. Rarely is this achieved in practice, since fixed tariffs are often linked to factors such as plot size or property tax rates which are a 15

18 Ahmedabad Amritsar Bangalore Bhopal Chandigarh Chennai Coimbatore Dehradun Hyderabad Indore Jamshedpur Kochi Cost Recovery and Tariff Practices for UWSS Sector in India poor indicator of household consumption or ability to pay -- especially when property tax assessments have not been updated. Property tax data are prone to manipulation and annual rental value (ARV) assessments are often out of date, with many properties remaining unregistered. In Ludhiana, households with plot sizes of 50 to 125 sq yd consume 30 kl per month on average, yet are exempted from water charges. If these consumers had meters and were charged the volumetric tariff of Rs per kl, they would pay Rs. 105 per month. The cities of Pune and Dehradun use fixed charge water tariffs based on the ARV of properties. Under this arrangement, registered properties in Pune are charged between Rs. 75 to Rs. 208 per month while in Dehradun bills range from Rs to Rs (see City Reports, volume 2 of this publication). More than 40 percent of households in Dehradun are not registered for property tax and this results in substantial revenue losses. Pune Municipal Corporation, however, has an updated property tax register that is relatively free of manipulation in ARV calculations, and has generated enough revenue to create an operating surplus despite a high level of NRW and low collection efficiency. Distortions in the fixed charge tariff structures can be revealed by comparing what customers with unmetered and metered connections pay for the same monthly consumption, in the same city - assuming both systems are in operation. (See Table 10). It is assumed here that a typical household consumes 20 KL per month (130 lpcd) though there are some locations where supply constraints make this level of consumption impossible. The analysis shows that Ludhiana charges more for unmetered connections (Rs per month per connection) than for metered ones (Rs. 76 per month per connection). The same is true in Pune, which charges Rs per month per connection for unmetered connections, Rs. 60 per month for metered ones. Table 10 shows that only in a few cities are charges for metered and unmetered connections broadly similar. The widespread use of unmetered connections is a significant barrier both to resource conservation and the recovery of O&M costs. Table 10: Equivalence between Tariff Rates for Metered and Unmetered Connections Particulars Minimum fixed charge per month (in Rs.) Volumetric charge for 20 kl consumption per month (in Rs.) *Kochi Municipal Corporation No fixed char ge conn ectio ns /tap/ month kl free No volum etric charg e for dome stic consu mers No fixed charg e conne ctions min. month ly charg e Water supplie d through standp osts paid by KMC* minimu m monthl y charge 40 For ½ connection on pro-rata basis. 16

19 Kolkata Ludhiana Mathura Mumbai Nagpur Nasik Pune Rajkot Varanasi Vijayawada Vishakhapat nam Cost Recovery and Tariff Practices for UWSS Sector in India Table 10: Equivalence between Tariff Rates for Metered and Unmetered Connections (continued) Particulars Minimum fixed charge per month (in Rs) Volumetric charge for 20 kl consumption per month (in Rs.) Free for dome stic cons umer s Free for dome stic cons umer s % of ARV 76 No meter ed conn ectio ns 12.5 % of ARV Charges to nondomestic consumers The basis on which industrial consumers should be charged for water is a subject of much international debate. While some argue that they should be charged a high rate due to their ability to pay, others argue for low charges given the economies of scale offered by bulk supply to large consumers, with minimal distribution costs. In India, industrial tariffs are generally higher than domestic ones, sometimes six or seven times higher. In Pune, for example, the domestic metered rate is Rs. 3 per kl against Rs. 21 per kl for nondomestic metered consumption. In Hyderabad, consumption up to 15 kl is at Rs. 6 per kl, while general industries pay Rs. 35 per kl and water-intensive industries Rs. 60 per kl. In the light of these charges, it is not surprising that bulk consumers often switch to alternative water sources if they become. When this happens, public service providers lose substantial revenue and this can affect the recovery of O&M costs for the service as a whole. Affordability GoI currently defines the poverty line as a monthly income per month of Rs. 1,848 per family of five in rural areas and Rs. 2,795 in urban areas. To assess the affordability of current water charges, the cost of subsistence consumption in the range 5 to 20 kl per month can be expressed as a percentage of the monthly budget of a family of five living on the poverty line (Table 11). Of the 23 cities studied, five provide free water to the urban poor (Ludhiana, Ahmedabad, Amritsar, Kolkata and Rajkot). Current charges in Indore, Hyderabad, Bangalore and Vijayawada give rise to household expenditure above the World Health Organisation (WHO) benchmark of 5 percent of monthly income for 20 kl consumption. In the other cities, the burden of water charges is within the notional 5 percent monthly limit. It must be remembered, however, that in many cases poor households cannot access 20 kl per month from the public supply, and are forced instead to purchase additional water at high rates from private vendors. This can raise monthly water bills enormously. 41 Range based on frequency of water supply (once/twice a day). 42 For regular domestic connection while rates for below the poverty line (BPL) households are Rs. 50/month. 17

20 Ahmedabad Amritsar Bangalore Bhopal Chandigarh Chennai Coimbatore Dehradun Cost Recovery and Tariff Practices for UWSS Sector in India Table 11: Charge for Subsistence Level Consumption compared to Monthly Budget for People Living on the Poverty Line Particulars Metered Water Tariffs Free water for poor Free water for poor through public taps MCC is now converting fixed charge connections to poor into metered connections No metering in domestic sector Water tariff for 5 kl consumption Water tariff for 20 kl consumption Water charge for 5 kl as % of monthly budget Water charge for 20 kl as % of monthly budget Unmetered Water Tariffs Fixed charge/month Water charge for unmetered connection as % of monthly budget Sewerage Tariffs Sewerage charge/cess Sewerage charge as % of monthly budget kl free kl free % 1% 0% 0% 0% 19% 3% 1% 5% 2% % 4% 2% 3% 5/WC/month % 1% 43 Based on average of ARV-based slab rates for water charges for registered domestic users. 44 Based on average of ARV-based slab rates for sewerage charges for registered domestic users. 18

21 Hyderabad Indore Jamshedpur Kochi Kolkata Ludhiana Mathura Mumbai Water Pricing and Cost Recovery Practices in India Table 11: Charge for Subsistence Level Consumption compared to Monthly Budget for People Living on the Poverty Line (continued) Particulars Metered Water Tariffs No metered connections. in slums No metered connections in poor households No water charge for domestic users No metered connections In slums No metered connections. Water tariff for 5 kl consumption Water tariff for 20 kl consumption Water charge for 5 kl as % of monthly budget Water charge for 20 kl as % of monthly budget Unmetered Water Tariffs Fixed charges per month Water charge for as % of monthly budget Sewerage Tariff % 0% 0% 8% 2% 2% No unmetered connections 35% of water supply charges % 4% Through standposts paid by municipality Exempt up to 125 sq yards No charge No charge Exempt up to 125 sq yards 12.5% of ARV minimum monthly charge service charge minimum monthly charge service charge. 19

22 Nagpur Nasik Pune Rajkot Varanasi Vijayawada Vishakhapatnam Water Pricing and Cost Recovery Practices in India Table 11: Charge for Subsistence Level Consumption compared to Monthly Budget for People Living on the Poverty Line (continued) Particulars Metered Water Tariffs Poor provided group connections No metered connections in slums Free water for poor through public taps Water tariff for 5 kl consumption Water tariff for 20 kl consumption Water charge for 5 kl as % of monthly budget Water charge for 20 kl as % of monthly budget Unmetered Water Tariffs Fixed charges per month Water charge for as % of monthly budget Sewerage Tariff % 0% 4% 1% 4% 1% 9% 3% No slab roof -25; Slab roof- 50 No slab roof -0%; Slab roof- 1.7% -- 4% of ARV % 2% 3% In summary, in all cities except Hyderabad, metered and unmetered water charges for consumption up to 20 kl per month lie well within the WHO affordability threshold of 5 percent. Arguably, the subsidies for poor consumers are over-generous (and benefit many nonpoor consumers, as discussed above) to the extent that it undermines cost recovery efforts. The data in Table 11 suggest that there is substantial unexploited cost recovery potential, and revenue could be improved substantially while still meeting the needs of the poor. While consumption tariffs may be affordable for many poor consumers, connection charges can be prohibitively expensive. Very few of the 23 cities currently offer assistance with connection charges. Vijayawada and Hyderabad are exceptions and allow payment in installments, a good practice that other service providers could also follow. Simplicity, acceptability and practicality In addition to meeting cost recovery objectives, the ideal tariff should be: Relatively simple to calculate and easy for the user to understand; 20

23 Water Pricing and Cost Recovery Practices in India Free from public criticism and acceptable to political leaders; and Applicable in the local context. Both Hyderabad and Chandigarh have adopted a Citizen s Charter to enable clear communication between the service provider and users on matters relating to water supply and sewerage services. An active complaints redressal system is also in place in each city; in Chandigarh this operates via ward committees, in line with the 74 th Constitutional Amendment Act. Pune s tariff structure is simple and user-friendly, with online self-calculation for property-based charges and bill payment possible through banks and kiosks in a variety of accessible locations For slum dwellers, too, the charges are easy to understand given their fixed charge basis. Indore recently produced a tariff structure valid until 2010 will enable the municipality to make long-term plans for revenue improvement. The tariff also has a simple, user-friendly design. Dehradun switched to a charge based on property tax following public criticism of faulty meters and meter reading. Pro-poor approaches to cost recovery Since water is a basic human need, all service providers have a pro-poor dimension to service delivery. However, this pro-poor water service delivery is rarely provided efficiently and effectively. Pro-poor tariffs and exemptions Increasing Block Tariffs (IBTs) are used by some service providers and are pro-poor in the sense that a subsistence level of consumption (the first block) is charged at a subsidized volumetric rate, with a higher rate applied to consumption beyond that level. The subsidy is poorly targeted, however, since it applies to all users of the service. Wherever water is supplied at a cost per kl, that is, below the unit cost of production, the consumer is effectively being subsidized, and this often applies to both domestic and commercial users in many cases. Additionally there are, of course, many nonpoor users who pay nothing because they have an illegal connection or their property does not appear on municipal registers. A number of cities offer complete exemptions from user charges to poor customers. Dehradun, for example, has a fixed charge water tax with an exemption for people whose property tax rate is assessed at less than Rs However, anyone (rich or poor) who buys a house connection becomes subject to a basic monthly charge of Rs. 75. There is, therefore, a disincentive to taking a house connection, especially for those on low incomes. Chandigarh charges a fixed charge of Rs. 100 per month per dwelling unit for poor households with a house connection, but average monthly bills for non-poor households with meters are only Rs. 79 for consumption of 30 KL water. In other words, the poor pay more than better off households based on typical consumption. In Pune, the tariff for poor households is Rs. 30 per month while other domestic consumers typically pay Rs Interestingly, poor households enjoy not only a lower tariff but a better level of service than others since poor settlements are mostly in low lying areas were water pressure is higher and the supply operates for more hours per day. Ludhiana is an unusual case. Here, the Government of Punjab operates a complete exemption from water and sewerage charges for plots of the area of 125 square yards (sq yd) or less. Since most poor households typically occupy plots no more than sq yd in area, this exemption benefits not merely the poor but also a wide range of better-off consumers. Most poor families do 49 Average holding of poor people/slum dwellers is considered between 25 to 40 sq m. 21