FRIEND 2002—Regional Hydrology: Bridging lite Gap between Researcli and Practice (Proceedings ofthe Fourth International FRIIIND Conference held at Cape Town. South Africa. March 2002). IAI IS Publ. no. 274. 2002. 329

Water sharing in the Volta basin

MARC ANDREINI Center for Development Research, GLOW A Volta Project, PO Box 645, Tamale,

PAUL VLEK & NICK VAN DE GIESEN Center for Development Research, Bonn University, Waller-Flex-Slrasse 3, D-53113 Bonn, Germany e-mail: [email protected]

Abstract The geography of the Volta basin is briefly described. This is followed by an overview of water use developments in the two main riparian countries, Ghana and . Ghana's water use consists mainly of hydropower generation for the urbanized south whereas in upstream Burkina Faso agricultural water use dominates. Analysis of the water balance shows a large sensitivity of surface water resources with respect to changes in precipitation and thereby to global and regional climate change. Finally, the GLOWA Volta project is introduced which addresses the issue of optimal water use under changing water demand and supply. Key words international basins; ; integrated basin management; GLOWA

INTRODUCTION

No alternative resource can stand in for the critical role that water plays in development. Especially in sub-humid and semiarid regions, water is becoming a critical constraint and is clearly a resource no country can do without. Yet the economic impact of water resources has not enjoyed the painstaking analysis devoted to resources for which alternate options do exist, such as oil. Since the early seventies, there has been an awareness of the economic importance of oil and the relationships among producing and consuming countries have been examined in detail. Now, as the scarcity of water increases, tensions among riparian states sharing international river basins are emerging with increasing frequency, full attention must be focused on the impact of water resources. The countries sharing the basin are no exception. In this paper, we will show the different expectations with respect to the use of water resources that exist in the main riparian countries, Ghana and Burkina Faso. We then sketch the availability and sensitivity of water supply at the (sub)continental level and the extent to which the riparians depend on water flowing into the country. We finish with a brief overview of a new research project that seeks to model present trends in water supply and demand in an integrated fashion. The Volta River basin drains approximately 400 000 km" (Fig. 1). There are six riparian countries sharing the basin: Ghana, Burkina Faso, , Côte d'Ivoire, Benin and Mali. In this paper we will concentrate on Ghana and Burkina Faso. These two countries comprise the bulk of the basin; approximately three quarters of Ghana (42% of the basin) and two thirds of Burkina Faso (43% of the basin) are drained by the Volta River system. The flows from Burkina Faso along the border of 330 Marc Andreini et al.

Fig. 1 Overview of the Volta basin.

Ghana and Côte d'Ivoire through the Bui Gorge joining the that also flows south from Burkina Faso. Further east, the flows from Burkina Faso through Togo then below the western slopes of the Buem ranges before it also joins the other two reaches. Then the river passes through the narrow Akosombo Gorge, site of the dam built to form Volta Lake, before flowing approximately 100 km to the sea.

WATER USE IN THE VOLTA BASIN

Water use in Ghana

Ghana is a country with a long history of mining and today has aspirations to develop a diversified industrial economy. Gold production was an important source of wealth for the African kingdoms located in present-day Ghana and during the colonial period. The original motivation for the proposed construction of a dam on the Volta River was to provide electricity to produce aluminium from locally mined bauxite. Although Ghana was a largely agrarian country at independence, the local leadership was convinced that construction of the "was the most important way of bringing Ghana forward to a more balanced economy" (Government of Ghana, 1961). It was claimed that the dam would supply "electric power from the Volta River for industry and for lighting our towns and villages" (Government of Ghana, 1961). Through several admini­ strations, the basic theme of development through industrial expansion has remained. The original generating capacity of the Akosombo Dam was 512 MW at its completion in 1964. After modifications increasing the capacity of Akosombo to 833 MW and completion of the Kpong Dam, downstream, the combined capacity of the two plants is 1060 MW. The Akosombo and Kpong dams are still Ghana's major Water sharing in the Volta basin source of electricity. Demand for power continues to increase especially within the urban-industrial sector (Table 1). Long-standing plans, still in the pipeline, to increase Ghana's generating capacity by developing another dam site in the Bui Gorge, indicate the continued commitment to hydropower as an engine of growth. There are relatively modest irrigation schemes at both Tono and Kpong. Further irrigation development in the north, although spoken of frequently, has thus far been given low priority nationally. In Ghana, water is most often seen as a source of hydropower.

Table 1 Yearly electric power consumption by sector in TWh (Andreini et al., 2000).

1980 1985 1990 1995 1998 Aluminium 3.319 0.853 2.789 2.198 0.927 Mines 0.272 0.257 0.315 0.654 0.713 Urban 1.098 1.048 1.615 2.760 3.093 Rural 0.000 0.000 0.066 0.189 0.254 Export 0.440 0.684 0.761 0.285 0.460

Water use in Burkina Faso

In contrast, Burkina Faso is a land-locked country with only minimal industrial development. Burkina Faso is an agrarian country; 90% of the population is active in agriculture. At present economic development depends on agriculture because there are no alternatives. The staple crops of sorghum, millet, and corn are grown under rainfed conditions. Failing rains often greatly reduce yields, causing the farmers substantial financial losses. Average yields lie between 800 kg and 1000 kg of grain per ha. Farmers need to invest in more labour, improved seed, and chemical inputs to obtain higher yields. Such investments, however, will remain prohibitively risky, without irrigation to lessen the risk of poorly timed water supply. Burkina Faso lacks economically attractive and environmentally acceptable large- scale dam sites. Irrigation schemes have been developed in Bagré and the Vallée du Sourou and there are modest that supply drinking water to Ouagadougou. However, most irrigation development in Burkina Faso takes the form of village-level schemes with imperfect hydraulic control. The total irrigable area in Burkina Faso is estimated to be 160 000 ha (Sally, 1997). The increase in the use of surface water resources over the last 15 years is much larger in Burkina Faso than in Ghana as shown in Table 2. This activity will affect water availability, but the impact is difficult to quantify because of the diffuse nature of the irrigation development. To give perspective to the issue one may remark that the surface area of is 800 000 ha or five

Table 2 construction in northern Ghana and Burkina Faso (van de Giesen et al., 2001).

Burkina Faso: Ghana 1984- 4986: 1999: Increase (%): 1984- 1986: 1999: Increase (%): no. ha no. ha no. ha no. ha no. ha no. ha <5 ha 109 145 377 421 246 191 96 115 113 170 18 48 5-100 ha 72 1704 154 3 444 114 102 18 197 48 534 167 172 >100 ha 5 789 16 23 421 220 2868 2 1184 2 2218 - 87 Total 186 2638 547 27 286 194 934 116 1496 163 2922 95 332 Marc Andreini et al. times the estimated irrigation potential of Burkina Faso. In any case, in Burkina Faso irrigation is seen as the primary non-domestic water use. The tension over water can be characterized as a conflict between rural and urban communities or as a conflict between Burkina Faso and Ghana. In each instance the basic dichotomy may be defined as between those requiring water for the generation of hydropower and those requiring water for irrigation. Already there has been some sparring between the two regarding issues of water use. Irrigated areas in both Ghana and Burkina Faso are modest in size, but the rapid expansion of irrigation in Burkina Faso and the relative stagnation in Ghana indicate different development paths. Even though potential losses from irrigation development in Burkina Faso are small compared to those of the lake, anxiety exists in urban Ghana concerning irrigation and even drinking water development upstream of Lake Volta in general and in Burkina Faso in particular (Gyau-Boakye & Tumbulto, 2000). To help reduce tensions and perhaps to strengthen their claim to the use of the Volta's water, Ghana offered to sell hydropower to Burkina Faso. Ghana wants to charge US$0.09 per unit, but Burkina Faso is ready to pay only US$0.06 per unit (Africa News, 1998). The Government of Burkina Faso maintained that it would be cheaper for them to generate power from thermal plants.

LARGE-SCALE WATER BALANCES

Having briefly described the diverging views of water consumption in Burkina Faso and Ghana, we now present a water balance as start to thinking about water sharing within the basin. A water balance is the first step in any hydrological analysis and proves especially useful when it comes to a river that drains major parts of the West African subcontinent. Figure 2 shows the rainfall and runoff in the Volta basin for the 60-year period from 1936 to 1996. The strongly varying runoff is on average about 9% of the rainfall. The average annual rainfall for the period from 1936 to 1963, the year the Akosombo Dam was under construction, was 414 km3 with a coefficient of variation of 0.07. River flow averaged 35 kmJ with a coefficient of variation of 0.57. The water balance or partitioning of rainfall into évapotranspiration and runoff in km3 can be modelled as follows using a simple regression:

runoff = 0.53 x (rainfall - 342) (1) where r2 = 0.80.

1936 1946 1956 1966 1976 1986 1996 Year Fig. 2 Yearly water balance of the Volta basin. The Volta dam was completed in 1964. Water sharing in the Volta basin 333

This crude model provides useful insights into the behaviour of the basin. Once a threshold of 342 knr has been reached, more than half of the rainfall runs off. Therefore small changes in yearly rainfall result in large changes in runoff. This sensitivity also implies sensitivity to changes in the character of the surface that influence the division of rainfall into évapotranspiration and river flow. West Africa appears to be warming and population growth continues at a staggering pace (Schâr et al. 2000). The implications of both of these trends need to be understood in a quantitative way if the riparian countries are to share water equitably. As we have seen from the regression, small changes in rainfall result in large changes in runoff. If this area is becoming drier, quantifying these changes will be extremely important to planners in both countries. With population growth rates of approximately 3%, pressure on the water supply will increase. The rural population will intensify their use of lands they already occupy and move into hitherto unoccupied areas. Those unable to survive in the countryside will migrate to the cities further increasing the already growing demand for employment and as a consequence the hydropower required to supply industry. More aggressive collection of fuel wood, increased cropping, and more extensive grazing all change the land surface and hence have a profound effect on the partitioning of the water. The availability of surface water resources is not evenly distributed throughout the basin. Of all the water flowing into Lake Volta, 67% comes from outside Ghana through the Black Volta, White Volta and Oti rivers (Fekete et al, 2000). The Black Volta crosses the border near Bui bringing 8.3 km3 per year or 24% of the total inflow to Lake Volta. The White Volta, entering from Burkina Faso near Pwalagu, has an average yearly runoff of 3.9 km3 (11%). Finally, 11.2 kmJ per year flows past Sabari along the Oti River. Even though the Oti River contributes 32%, the largest part, its basin is actually smaller than that of the White Volta or Black Volta, showing that changes in some parts of the basin may have more important downstream effects than changes in other parts. The simple fact that most of Ghana's water comes from outside the country complicates further the debate about water for electricity vs water for agriculture.

GLOWA VOLTA PROJECT

Land tenure and water rights give rise to emotionally charged tension. Fanned by the media, conflicts can easily escalate to dangerous levels. Therefore, it is desirable for reliable information on water supply and demand to be made available to those peoples sharing the basin. The German Federal Government started an eight-year science programme on Global Change in the Hydrological Cycle (GLOWA) consisting of four projects, one of which studies the impact of global and regional change on the water resources of the Volta basin. The GLOWA Volta Project, "Sustainable Use of Water Resources: Intensified Land Use, Rainfall Variability, and Water Demands in the Volta Basin ", has the stated goal of creating a scientifically sound decision support system (DSS) for the assessment, development and sustainable use of water resources in the basin by means of an integrated model of the basin. There are two challenges facing the project: the relative scarcity of data in West Africa and the large area of the basin. The first is being met by selecting models that 334 Marc Andreini et al. make optimal use of the available data. Remote sensing and geographical information systems will be used to deal with the vastness of the area to be modelled. Scale issues tend to compound the difficulties as each discipline has its own units of observation, analysis, and prediction. To help overcome this problem, a scale of communication has been defined as a grid with cells of 9 km x 9 km. An important advantage of the project is the fact that the individual modellers are aware, from the outset of their work, of the necessity to render their results in a form palatable to the other modellers. The modellers are devising methods of either up- or down-scaling their outputs to conform to this standard within their own fields of experience. Having demonstrated the sensitivity of water supply to environmental changes, it is clear that the DSS should be able to simulate water demand and supply under various scenarios. A multidisciplinary expert panel identified a set of key variables: (a) precipitation, (b) évapotranspiration, (c) agricultural production, (d) land use, (e) income, (f) population, (g) river flow, (h) water use, (i) hydropower, (j) health, (k) technological development, and (1) institutional development. Variables (a)-(i) will be treated as endogenous variables to be solved simultaneously so that the feedback among these variables can be quantified. Health will be an output variable because the feedback mechanisms are too complex to model adequately. Technological and institutional options are treated as input variables over which the DSS will optimize for given objective functions. Ultimately the technological and institutional options selected by policy makers should be chosen from those combinations that are found to yield optimal results. Two broadly defined areas of agreement should be forged between the two partners. They should agree how to care for the uplands of the basin i.e. generate land-use policies that are ecologically sound while providing for the needs of the rural inhabitants. And they will also have to agree how to apportion the water between irrigation and power generation. To prevent conflicts and mitigate water shortages, policy makers can use the DSS to make scientifically sound policy choices on how to share the continent's water resources.

CONCLUSIONS

The paper has highlighted the different water uses in Ghana and Burkina Faso. In Ghana, water is mainly used to generate hydropower to support development of the urban and industrial sector. In Burkina Faso, the development of water resources in rural areas for household use, livestock and irrigation is most important. Under continued development of the water resources of the Volta, water sharing will become Water sharing in the Volta basin 335 more difficult, especially because water supply is very sensitive to small changes in rainfall. A science-based DSS, which is presently being developed by the GLOWA Volta project, will help to properly evaluate the complex set of factors that determine future water demand and availability.

Acknowledgements We gratefully acknowledge financial support from the German Federal Ministry of Education and Research (BMBF) as the main sponsor of the GLOWA project and the Ministry for Schools and Education, Science and Research of Northrhine-Westphalia. The GLOWA-Volta project is de facto a research network and we express our thanks for the scientific inputs from our partners: Remote Sensing Research Group (Bonn University), Institut fur Stadtebau, Bodenordnung und Kulturtechnik (Bonn University), Fraunhofer Institute for Environmental Atmospheric Research, Institute for Tropical Medicine and Hygiene (Heidelberg University), Savanna Agricultural Research Institute (CSIR), Water Research Institute (CSIR), Remote Sensing Application Unit (University of Ghana), Population Impact Project (University of Ghana), Institute for Statistical, Social and Economic Research (University of Ghana), Institut de l'Environnement et des Recherches Agricoles (Burkina Faso).

REFERENCES

Andrcini, M., van de Giesen, N., van Edig, A.. Fosu, M. & Andah, W. (2000) Volta basin water balance. ZEFDiscussion Papers, no. 21, Zentrum fur Entwieklungsforschung, Bonn, Germany. Fekete, B. M.. Vôrosmarty, C. .1. & Grabs, W. (2000) Global, Composite Runoff Fields Based on Observed Rivet- Discharge and Simulated Balances. CD-rom, Univ. of New Hampshire. Durham, New Flampshire, USA and Global Runoff Data Centre, Koblenz, Germany. Government of Ghana (1961) the Volta River Project, Statement by the Government of Ghana, 20lh February 1961. IV. P. no. 1/61, The Ministry of Information. The state Publishing Corporation, Accra-Tema, Ghana. Gyau-Boakye, P. & Tumbulto, J. W. (2000) The Volta Lake and declining rainfall and streamflows in the Volta River basin. Environment, Development andSttslatnabllily 2, 1-10. Sally, H. (1997) Améliorer les performances des périmètres irrigués. In: Les Actes du Séminaire Régional du Projet Management de l'Irrigation au Burkina Faso. I1M1, Colombo. Sri Lanka. Schar, C, Luthi, D., Beyerle, U. & Heise, E. (2000) The soil—precipitalion feedback: a process study with a regional climate model. J. Climate 12, 722-741. van de Giesen, N., Kunstmann, FF, Jung, G., Liebe, .1., Andreini, M. & Vlek, P. L. G. (2001) The GLOWA Volta project: integrated assessment of feedback mechanisms between climate, land use, and hydrology. In: Climatic Change: Implications for the Hydrological Cycle and for Water Management (ed. by M. Bcnislon). 151-170. Advances in Global Change Research 10, Kluwer Academic Publishers, Dordrecht, The Netherlands.