Providing water services at scale: how to move from unsustainable assistance to sustainable development?

The past decade has been marked by rapid urbanisation. This historical phenomenon takes various forms. In areas usually considered to be rural, small towns are springing up and the technology usually associated with urban areas is seeping in. This is also happening in the water sector with the construction of thousands of small piped water systems in semi-urban centres of a few thousand inhabitants which supply public tap stands or even households, with private connections on the premises. The past decade has also been marked by the increasing decentralisation of responsibility for water and sanitation services to the local authorities. In West Africa, public service delegation is usually the favoured management choice for these systems, and this has spurred both market development and the emergence of private operators.

Despite being landlocked and predominantly rural, Burkina Faso (population 16.5 million) is no stranger to this trend. Today, 77% of Burkinabe people live in rural areas and half the urban population lives in the capital, Ouagadougou. But the rural environment has changed. Since 1985, the number of villages with fewer than 200 inhabitants has halved. In 1985, two out of three villages had fewer than 1,000 inhabitants whereas by 2006, half boasted populations of more than 1,000. In 2006, 60% of rural dwellers lived in ‘villages’ of more than 2,000 inhabitants, half of which were small towns of 5,000–10,000 people. By 2050, the urban population is predicted to be either equal to the rural population (22–24 million), or to be double their number (33 million in urban areas as against 16 million in rural areas) (Ministère de l'Economie et des Finances; MEF, 2009).

For the past 10 years, the type of water access available to villagers and the inhabitants of small towns in Burkina Faso has changed. Villages with populations of more than 3,500 are now equipped with small pressurised piped water systems which supply water through tap stands and a few on-the-premises connections for private households. These systems supplement the service hitherto provided by hand pumps. Hand pumps remain the only improved water source in villages of fewer than 3,500 inhabitants and are managed by community structures (Water User Associations (WUA)). In small towns, hand pumps will remain managed by a WUA, while town water will be managed by an operator (privately owned or an association). In the Sahel Region of Burkina Faso, all the piped water systems are managed by the same operator, FasoHydro.

In theory, pressurised water supply should offer a better quality service than hand pumps. Tap stands should bring water closer to the people or even to the home (on-the-premises connections) and supply a smaller number of users (120–170 users per tap) than hand pumps (300 users) (Table 1). As the price is linked to the expected level of service, water at a tap stand is much more expensive than at a hand pump. A user must pay the equivalent of 500 FCFA per m³ (1 US$/m³) when collecting water at a tap stand whereas an annual contribution of 250 FCFA per person per year (0.5 US$/p/y) is required to access a hand pump.

Are these piped water systems delivering the expected level of service? Is there a demand for this type of access? Does it actually deliver a better level of service than the hand pumps? Is it sustainable? Should piped systems be built at considerable cost ($125,000 on average per network according to the Water Resources Directorate) in smaller villages (2,000–3,500 inhabitants) or should the installation of pumps continue? In small towns where piped water systems have been constructed, what should be done with the boreholes that have hand pumps? Should they be closed off? Should those that are broken down be repaired?

The rate of piped system breakdowns (23%) (Ministère de l'eau, des aménagements hydrauliques et de l'assainissement; MEAHA, 2014) gives warranted cause for concern. In Burkina Faso, the demand for water is said to be too low and the price of water at tap stands too high. The pricing policy is said to be hindering the construction of piped systems. It is believed that as the critical mass cannot be reached, it would be risky to envisage gradually reducing the price and/or increasing accessibility including to the poorest, by extending networks and increasing the number of private connections.

This study proposes to measure and compare the cost and quality of service delivered at each stand pipe and hand pump in four small towns of Burkina Faso. How much does each type of service cost per user? Is the level of service offered by piped systems actually of better quality? How can the most effective type of access be developed further?

Burkina Faso is one of the poorest countries in the world. In 2013, Gross National Product per capita was 336,000 FCFA (less than $700) and 44% of the population lived below the poverty line, set at 108,454 FCFA per year ($0.62 per person per day). Life expectancy is lower than 55 years and 71% of the population has no education. In terms of human development, the country is classified 161st out of 169 (Institut National de la Statistique et de la Démographie; INSD, 2012).

According to the Joint Monitoring Programme, Burkina has achieved its development objective in terms of access to water: 80% of the population has access to an improved water source, 6% of which is in the form of household connections and 73% through tap stands, hand pumps or protected wells. In terms of service quality, Burkinabes are at the bottom of the ladder: while in Africa 13% of rural populations have a water supply on the premises, this figure is only 2% in Burkina Faso, which is very low, even with regard to the average for West Africa (5%) (African Ministers, Council on Water; AMCOW, 2012).

The Sahel is the poorest region in the country: nearly 80% of the population is classified as poor, in other words double the national average (INSD, 2009). The region has only one urban commune (or district), which contains the regional capital, Dori (17,000 inhabitants), and the regional rate of access to water is the lowest in the country (42.5%). Most villagers do not have access to a hand pump and collect water from non-improved sources, mainly unprotected traditional wells, water holes or dams.

In towns such as Gasséliki (4,200 inhabitants), Mansila (7,400 inhabitants), Seytenga (4,900 inhabitants) and Titabé (2,600 inhabitants), piped water systems have been in service since 2009¹. They are managed by FasoHydro, a company that manages seven small water supply systems in the Sahel Region. In each town, the hand pumps are still managed by a WUA. The piped systems supply from three to nine tap stands, each with two to four taps. Piped water systems were introduced at a time when the existing improved sources were 35–90 metre-deep boreholes equipped with hand pumps. In two small towns (Gasséliki and Mansila), the piped systems complement the service that is provided through hand pumps whereas in the other two small towns (Seytenga and Titabé), there are, on paper, enough hand pumps to cover the entire population (Table 2).

The study of water supply service performance in the small towns of Gasséliki, Mansila, Seytenga and Titabé is part of the action research project called Triple-S (Sustainable Services at Scale) for which the pilot area in Burkina Faso is located in eight communes of the Sahel with a total population of 300,000, in other words a third of the region's population. The main question driving this project concerns the scale at which water service management and regulation enable sustainable and universal access to safe water in rural areas (for more information on the Triple-S initiative ‘Water Services that Last’, go to: http://www.waterservicesthatlast.org/).

To establish a water supply services baseline, our methodology consisted of the following:

• Calculating the life-cycle cost of each type of service (hand pumps or tap stands) on the basis of past expenditure to construct, maintain and rehabilitate the 842 hand pumps and six small piped water supply systems in the eight communes, and to organise and manage 200 hand-pump service areas and six stand pipe/on-premises service areas. In 24 villages (three per commune), the levels of service delivered to the users were determined following a two-stage survey of the entire population (55,000 people).

• Determining the quality of the water services in 20 villages and four small towns, the former solely supplied by hand pumps and the latter supplied by a mixture of hand pumps and tap stands.

• Measuring the compliance of local institutions with national regulations and the degree to which the various organisational and management roles are fulfilled by the WUA, the private operator, the local authorities, repairmen and water point caretakers.

The results presented in this article concern the performance of the water services in the four small towns where water services consist of a mixture of hand pumps and tap stands and where both cost data and service level data have been collected (to see the results on the cost of water services in the eight communes, go to: http://www.waterservicesthatlast.org/countries/initiative_du_burkina_faso).

The life-cycle costs of both types of service are made up of capital expenditure and recurrent expenditure. The latter consists of expenditure incurred to ensure the equipment works properly (maintenance, repair and rehabilitation) and expenditure to support the local authorities and service managers (WUA and private operator) in fulfilling their roles (organisation and management of water services). Table 3 defines the life-cycle cost components of a water service.

The originality of the life-cycle cost methodology lies in taking the support expenditure into account and calculating the unit cost based on the real number of users rather than the number of people hypothetically targeted by the infrastructure². In this way, the demand expressed at each water point and its actual availability is taken into account. By cross-referencing each payment with its source, the amount each contributor (donors, government and users) has to pay for effective access to water services can be better evaluated (Pezon, 2012).

To determine the quality of the service actually delivered by each type of service, we carried out two surveys. For 4 days, an initial ‘water point survey’ was run at all the tap stands and water pumps. The users at each water point were identified and the amount of water taken by each was recorded. All the water points were geographically referenced. Secondly, a ‘user survey’ was carried out in each of the small towns in the form of interviews with all the users who had been identified during the water point survey. The aim was to understand the household structure (how many people consume the water taken by the user at the water point), what the water collected was used for (domestic use or production) and any use of non-improved sources to meet household demand. All the users' homes were geographically referenced. These surveys were carried out twice; once during the dry season (8 months) and once during the rainy season (4 months).

The data collected enabled the number of users at the tap stands and hand pumps to be quantified for each season and showed who used them throughout the year (during both seasons). The data also enabled comparisons to be made between service expectations and the actual service delivered in terms of numbers of users and water volume, and the amount of water used by each user per day was quantified. Therefore, we can determine the level of service delivered in terms of quantity for each infrastructure.

The distance to the water point was calculated on the basis of the geographical coordinates of the primary water points where the users were identified and the geographical coordinates of their homes. The distance is therefore as the crow flies.

Finally, the density of users at the water points was calculated on the basis of the number of users counted at each water point during the water point survey and the size of each user's household, as identified during the user surveys.

As regards the quality, we were not able to test compliance with World Health Organization (WHO) standards in situ. For the water provided at tap stands, we used the results available from tests carried out by the operator FasoHydro once a year. There is no test being done on the water provided by hand pumps.

The indicators, Water Quantity, Water Quality, Distance to Water Point and Crowding at the water point are the four parameters used within Burkinabe regulations to qualify access to water. Therefore, we used these indicators and the levels validated by the national sector policy to determine the quality of service in the four small towns³.

A better level of service is expected from piped water systems. Is this indeed the case? Analysis of the service levels in the small towns of Gasséliki, Mansila, Seytenga and Titabé shows that the piped systems do indeed deliver a better level of service than the hand pumps.

All the piped systems deliver a better level of service than hand pumps in terms of water quality. In theory, the quality should be compliant with WHO standards at all improved sources, that is to say both hand pumps and tap stands. At the hand pumps, compliance is checked when the borehole is drilled and again 15 years later when it is rehabilitated. None of the 55 hand pumps in the four small towns – or any of the remaining 449 hand pumps in these communes – has had the water quality tested since it was installed. Nothing guarantees the user that the hand pumps are providing better quality water than the traditional wells that are still widely used in the small towns and villages. As regards the piped water supply systems, the operator is contractually required to test the water quality each year. While this is not frequent enough should something happen to pollute the water, and is less often than the testing carried out in urban and peri-urban areas on the systems managed by the public operator (ONEA), it is without question an improvement on the service offered by hand pumps. The quality testing carried out on the four pipe water systems by the operator since 2009 show that the water is compliant with WHO standards. The tap stands do provide users with safe water.

The greater accessibility of tap stands is also measured in terms of density, and thus the waiting time at the water points. The number of people served per tap stand is well below that at the hand pumps in all seasons. The tap stands serve 120–170 people per tap throughout the year, while in Gasséliki, Mansila and Seytenga 30–60% of the hand pumps serve more than 400 users a day in the dry season. Note that the number of people served by hand pumps in the dry season is 20–50% higher than that observed during the rainy season, which increases the density of users and waiting time at hand pumps. Therefore, the level of service provided to hand pumps users decreases during this season.

In light of the poverty levels in the Sahel, the existence of such a demand could be seen as surprising. For an average consumption level of between 10 and 20 l per person per day, the water from the piped systems is on average 10 times more expensive than that from the hand pumps⁴. However, the quantities consumed by the users are comparable on average, whether the water is from tap stands or hand pumps.

These quantities remain insufficient but it appears that they can only be increased by providing service delivery through on-the-premises connections. By eliminating the distance to be covered to reach the water points, it would probably become possible for a certain number of users to completely forego non-improved sources. The latter remain, in the four towns, the users' main source of water.

Finally, very few users exclusively fetch water from improved sources to meet all their domestic needs throughout the year. Fewer than 800 users go solely to tap stands, the only improved source that supplies water of confirmed quality.

If the service is to be extended to a maximum number of people, it is important to know the respective cost of the infrastructure. What should be done in the four small towns – should the piped system be extended or should the number of hand pumps be increased and the water quality tested regularly? Are piped systems with a capacity of between 1,500 and 4,500 people or between 10,000 and 33,000 m³ really cheaper than an equivalent supply by hand pumps? Again the questions must be answered empirically. How much does it cost to supply water via tap stands in comparison to hand pumps on the basis of the actual number of users and not on the basis of the number of people theoretically targeted by each installation?

To compare the cost of piped water supply to that via hand pumps, we collected data on the expenses incurred for each type of system from the outset and then calculated the life-cycle unit cost on the basis of the number of users actually supplied⁵.

The gap between users and theoretical users significantly alters the hierarchy of unit investment costs. Indeed, on the basis of theoretical users or users per design, the hand pumps are $20–40 cheaper per user than piped water systems, except in Mansila where the piped scheme is large enough (nine tap stands) to achieve economies of scale which lower the investment cost per theoretical user to $30 less than that of a theoretical hand pump user. However, on the basis of the number of actual users, the situation is very different. In Gasséliki and Mansila, tap stands cost $20–30 more per user than hand pumps and $10–40 less in Seytenga and Titabé.

This inversion is due to demand and, more precisely, to the fact that the demand for service per tap stand is commensurate with the installed capacity. The relatively high nature of the unit investment costs for hand pumps is also due to the breakdown rate. At the four sites, 55 hand pumps are installed, of which 10 were not working during at least one of the two seasons. The investment made to build these hand pumps cannot achieve its objective in terms of number of people supplied, which increases the unit investment cost for this type of access. When fully functional and supplying 300 users per hand pump, a service via hand pumps would cost $40–50 less per user than supply via tap stands. However, full functionality is unheard of and the average rate of hand pump non-functionality in Africa (30–35%) is close to double that observed in the Sahel (18%). In contrast, almost all taps were fully functional.

The operation of hand pumps costs from $0.40–1.10 per user per year as against $1.70–3.60 for piped systems⁶. The capital maintenance expenditure for hand pumps varies from 0 to $1.80 per user and per year as against $0.10 or $0.20 for piped systems which, we must recall, date back only to 2009. A provision of between $1 and $1.60 per person is set aside to fund future capital maintenance of the piped systems.

At this stage, it is important to recall that the costs are calculated on the basis of past financial expenditure: these are not the ideal costs but observed costs. For a hand pump to be replaced after 30 years – the estimated lifespan of a hand pump in Burkina Faso – the equivalent of $2.60 should be set aside per person per year (excluding inflation). However, no provision is made by either hand pump managers or the local authorities to renew this infrastructure.

For the two sites where data on hand pump capital maintenance expenditure was available (Gasséliki and Mansila), the gap between the unit costs of the two types of equipment (excluding the provision) is very small: the piped water systems of Gasséliki and Mansila cost respectively $0.40 and $0.70 more per user per year than hand pumps.

However, the unit cost of operating the piped systems decreases as the amount of water distributed increases. It drops from $0.66/m³ per year for 5,000 m³ distributed to $0.43/m³ for 12,000 m³ distributed. In contrast, the unit cost of operating a hand pump increases as the volume distributed increases: it rises from $0.11/m³ for a hand pump supplying 7,800 m³ per year to $0.25/m³ for a hand pump supplying an annual volume of 14,500 m³. Its capital maintenance unit cost increases by 50% under the same circumstances. In Mansila, where there is the largest piped system (4,500 theoretical users), each m³ distributed by the system costs $0.44 per year in operating and capital maintenance costs, as against $0.68 per year for hand pumps.

Services provided by hand pumps are considerably more expensive in terms of support than piped systems. Less than $1 per user per year for a piped system, support expenditure varies between $4.5 and $9.5 per hand pump user per year.

The support received by hand pump managers (WUA) and local authorities to organise water services using hand pumps is much greater than the support that the piped water system manager and the local authorities receive specifically to organise their piped services. Finally, a piped water system costs less per year per user, including provision for capital maintenance, than the per-user cost of hand pump support alone, excluding pump operation and rehabilitation costs.

Overall, in terms of capital investment the cost of piped water supply is comparable to the supply via hand pumps, even in the case of the small system in Titabé whose capacity is limited to 1,500 people. In terms of recurrent expenditure, water supply via a piped system is clearly cheaper per actual user than supply by hand pump (Table 4).

This observation confirms the economic theory that over a certain size it is cheaper to distribute water via a piped system than in a decentralised manner. Networks (whether for water supply, electricity, communication, or transport, etc.) benefit from increasing returns: within the limits of the installed capacity, each additional m³ produced or each additional user supplied costs less than the previous one. Therefore, it is economically justified to adopt a piped system rather than decentralised distribution (via hand pumps) once the average or unit cost of piped distribution is less than the average or unit cost of decentralised distribution. This is the case on all four sites studied.

It is less costly to supply users via tap stands than by hand pumps, while at the same time they deliver a better level of service. This is the case in Gasséliki, Mansila, Seytenga and even in Titabé, where the piped system's capacity is only 1,500 people or 10,000 m³.

This type of service is affordable for a large enough demand and the extension of these piped systems should be encouraged rather than rehabilitating hand pumps or working to support WUA. Not only is the overall performance of the 200 WAUs in the eight communes after 6 years of continuous support fairly unsatisfactory, none of those in the four small towns is operational at all. In Gasséliki, Mansila, Seytenga and Titabé an increase in demand (number of users or volume) at the tap stands would lower the unit cost of piped water supply without harming the quality of the service delivered to the end user. An increase in demand at hand pumps would lower the level of service (already very mediocre in the dry season) and would increase its unit cost.

Local inhabitants are not encouraged to fetch their water at the tap stands, rather the opposite. The prices recommended by the Ministry of Water clearly position the piped water systems as a luxury service whereas the water at the hand pumps must be financially accessible to all. The regulations stipulate that each household must pay a minimal charge of 2,500 FCFA per year ($5 per year for an average family of 10 people) to have access to the hand pumps and take the equivalent of 20 l per person per day. At the tap stands users pay according to volume the equivalent of 500 FCFA ($1) per m³.

This explains why the service at the tap stand is more costly for users than that at the hand pumps. The user of a hand pump does not see that support worth from 10 to 20 times his annual contribution is provided to offer him this service.

The price does not play the same role in financing the two types of service. The price of water at the hand pump must be able to cover the upkeep costs (care and maintenance) of this infrastructure. The price of water at tap stands must be able to cover the upkeep costs, operating costs, part of the capital maintenance costs, the support and the profits of the operator.

The fact that every year, expenditure on supporting the organisation and management of hand pump services costs double the actual management of a piped water service is perplexing to say the least. Indeed, there is no logical justification for subsidising a lower quality and more costly service to such an extent. The inclusion of support expenditure in calculating the cost of access to water and being able to document it would help to better understand this phenomenon.

More radical critics of development aid range from those who, like J. Sachs (Sachs, 2005) call for a massive increase in development aid, to those who, like Easterly and Moyo call (Easterly, 2001; Moyo, 2009), on the contrary, for this regime to cease because it is counterproductive to its main objective: the development of countries and populations. Increased aid for hand pump water services would mean water quality could be tested and broken-down hand pumps rehabilitated. Stopping aid could plausibly lead to more frequent interruptions of the hand pump service and to a higher rate of breakdowns. In the Sahel Region, hand pumps could no longer fulfil their current role of replacing the non-improved sources that dry up at the height of the dry season.

Our objective here is not to decide one way or the other, but rather to reiterate that development assistance is meant to enable development and not to perpetuate development assistance and support the associated stakeholders. The financing allocated to hand pump services is not part of a development strategy but one of assistance. It is responding to an immediate situation. But hand pumps are no more accessible than non-improved sources and their water costs. More worrying still, the water may be polluted. It is not rare to find boreholes in Burkina Faso that the villagers call arsenic boreholes. What type of impact in terms of public health can we expect from a so-called improved source where the water quality is not regularly tested and that is a 2 km walk away? What type of impact in terms of income or well-being can we expect from water carried up to 1 km and which requires such investment in time and effort?

We expect better access to water to spur development arising from improved health and more spare time available to take up other activities, whether income generating or not. At best hand pumps enable people to stay in their villages and survive. The aim of development assistance is thus to uphold the right to water providing the water quality is regularly tested. Is it possible to control the quality of the water distributed by 40,000 hand pumps when we know that in France the regular control of 8,000 water catchments is problematic? Should household water treatment at the point of consumption be promoted and supported too?

Piped water distribution promotes development on several levels. Firstly because the water supplied is drinkable and a piped system can be extended and the scale of the service can be gradually expanded, with an increasing impact on development. We know that drinking water may be safe when it is drawn but may not be by the time it is consumed. Studies have even shown that a household connection alone may have a significant impact on health (Howard & Bartram, 2003; Günther & Günther, 2010). Piped water distribution brings the point of supply closer to the point of consumption. Doing this reduces the risk of contamination as well as reducing the time spent on water chores. Furthermore, its average cost diminishes with the increase in quantity of water distributed. Over a certain number of users, the marginal cost of supply is so low that the poorest could be supplied under acceptable economic conditions.

The problem is not so much the abrupt increase or interruption of development assistance than its effectiveness. What are the obstacles to redeploying aid to types of access that bring development and therefore will be sustainable in the long-term?

The key is not to subsidise water but its access. Sound studies have shown that, as much in the field of water as in that of energy, equity requires not that water (or electricity) be subsidised, which only favours those already supplied, but its access (Komives et al., 2005). It is therefore not a matter of subsidising operation (to lower the price) but of extending the service by building new tap stands and connecting village households.