Final Paper - Executive Summary
November 2000

Note on Pilot Study

The World Commission Dams (WCD) faced a considerable challenge in meeting the mandate given to it by the Gland reference group at its inception in April 1997. This included, among other things, an assessment of the "development effectiveness" of dams, an issue that the Commission decided to address through thematic studies, the cross-check survey and a series of case studies of dam projects worldwide with a view to learning lessons from these experiences.

The aim of the pilot study of the Orange River Development Project in South Africa, a major initiative of the South African Government beginning in the early 1960's, was to develop the methodology to be used for the 7 WCD focal dam project/river basin case studies being undertaken world-wide. It was designed to test a common approach that sought to inform the Commission on development effectiveness as well as a range of issues associated with the planning, design, appraisal, construction, operation and decommissioning of large dams. The Minister for Water Affairs and Forestry gave his agreement to the pilot study in September 1998.

The ORDP was a major project involving two dams on the Orange River and a diversion tunnel that was planned during the apartheid years in the 1960s, and which has hydropower, irrigation, water supply and flood control functions. The multi-purpose nature of the project provided a good opportunity to test the methodology, and the study was undertaken between October 1998 and March 1999.

In adopting a pilot approach for methodology development, it is recognised that this study addresses real issues and real experiences that can not simply be used as a "testbed" without giving some substantive feedback to the interest groups affected by the ORDP on the findings of the pilot study. This is the main purpose of this report. A secondary purpose was to provide a model for the consultants working on the other case studies world-wide. Where additional work would be required to bring the study up to the level of a full case study, this is identified in the text.

Methodological improvements following the experience of the pilot study

The following parts of the case study methodology were improved or addressed as a result of the study, and these modifications were presented in detail in the "Methodology Guidelines for the Implementation of Phase II of the WCD Case Studies" provided to case study team leaders.

Revision of data sheets and enquiry forms

The two sets of pilot data collection sheets (focal dam and cross-check survey), and the enquiry form were thoroughly tested and the pilot versions were revised to be more flexible and to include only data points of direct relevance to the study. The full data collection sheet for the focal dam was scaled down. Only two responses were received on the stakeholder enquiry form, which was felt to be too long and complex and a revised stakeholder interview process was therefore devised.

The consultation process

Valuable lessons were learned on the composition, structure and agenda of the scoping meetings that were put into practice in other case studies worldwide. Experience from the final stakeholder meeting demonstrated the challenge inherent in soliciting views on complex issues from a wide range of stakeholders. It also showed that due balance must be found between a qualitative and a quantitative approach. Experience from the second stakeholder meeting proved invaluable in the 7 case studies undertaken around the world, and the Secretariat substantially modified its approach as a result of the experience of the Orange River meeting.

The structure and content of the report

The final report has gone through a series of revisions before adopting the proposed structure. In addition a number of gaps were identified of which the following are the most important :

1. Definition of "the project" used as a baseline for defining the predicted benefits needs to be more rigorous.

2. More emphasis should be placed on operations and operating rules and how these have changed through time since project completion.

3. The methodology for the distribution of benefits analysis should include a time dimension. One matrix should be completed for the context at project completion, and another for today's context. The positively and negatively affected groups should be better disaggregated and quantified.

4. Sedimentation/reservoir characteristics should be included in "Project design - lifetime of the reservoir."

5. Strengthen "Context at the time of project planning" for each sector.

6. Hydropower benefit methodology should use real tarifs not simulated benefits.

7. Basin-wide issues and cumulative impacts should be strengthened.

8. Strengthen the section on justification for the project (context at the time) and Criteria and guidelines.

This pilot study therefore greatly contributed to refining and improving the case study methodology, and provided a model to guide the work of the case study teams. In the normal case study process, the final report would have taken into account the detailed comments given by stakeholders through the consultation process, and adressed them aftre the meeting, wherever possible. Given the timing and nature of the pilot study it was not possible to respond to the constructive criticism received, and this should be borne in mind by the reader. The draft report remains substantially the same as the circulation draft, with the exception of Chapter 7 - Lessons learned, that draws in discussions from the final meeting.

Executive Summary

Proposals for irrigation projects in the Orange River Basin were initiated in the 1920s to meet the increased need to utilise its water resources given South Africa's semi-arid climate. These proposals were considered too costly to be implemented through the 1950s, but development of the Orange River Basin became more of a priority after the successive electoral victories of the Nationalist Party in 1948, 1953 and 1958.

The actual decision to plan a comprehensive Orange River Development Project (ORDP), however, was triggered by the objective of the Nationalist Party Government to stem the outflow of capital from South Africa. This flight of capital was partially caused by the increasing domestic resistance and international criticism against the Governments Apartheid policies, particularly following the Sharpville massacre in 1961. The decision to embark on the ORDP was also motivated by the Nationalist Party Government's goal of building a political symbol to Afrikaner Nationalism and a monument to the "grand scheme of apartheid" at a time when South Africa had just withdrawn from the British Commonwealth.

The initial planning proposal for the ORDP was quickly assembled in the early 1960s by the Department of Water Affairs oriented by the various development policies of the Nationalist Party Government. The ORDP fit with the Government's interventionist policies of 'economic nationalism' and satisfied a number of the Nationalist Party's key constituencies. In particular, the Nationalist Party Government's support for White agricultural interests as well as decentralised industrial development was to be served by the scheme.

The politically motivated ORDP that was subsequently proposed by the Department of Water Affairs was an ambitious initiative which sought to provide substantial benefits for South Africa, the Orange River Basin, and White farmers in particular during the core years of the Apartheid regime in South Africa. The major objectives of the scheme included:

• provision of irrigation for increased agricultural production,
• supply of water for municipal and industrial use,
• generation of hydro-electric power for peaking purposes,
• prevention of damaging floods,
• creation of recreational facilities,
• further population settlement in the basin,
• creation of additional employment opportunities,
• stimulation of water and agricultural based industrialisation and the regional economy.

These objectives were geared towards the broader local, regional and national development goals of meeting the increasing consumption needs and raising the living standards of a rapidly growing South African population.

As a result of the rapidity with which the 1962-63 planning proposal was assembled, many of the baseline surveys were cursorily conducted and few of the planning assumptions were thoroughly examined. While potential sedimentation problems were assessed, soil studies were cursorily completed and investigation of environmental impacts was not conducted. Moreover, the ORDP was a much larger and more complex water resources scheme involving dams than had been previously promoted by the Government of South Africa. Because of its size, complexity and cost, the scheme was planned to be completed over 6 phases. After minimal appraisal, the South African Parliament quickly authorised implementation of only the 1^st phase of the ORDP for R85million in 1962-63.

Information disclosure and the participation of project-affected peoples in the planning and appraisal of the ORDP was not regular practice at the time and was thus limited. Some public commentary on the proposed scheme was received early on from independent scientists and technical experts in South Africa but with little effect on the planning and implementation of the ORDP through the mid-1970s. White farmers to be displaced by the Gariep and Van der Kloof reservoirs were notified but, despite being somewhat organised and even taking the Government to court a number of times, did not have much voice in determining compensation arrangements. Given the Apartheid context, Black and Coloured peoples were neither notified nor compensated as these groups had been almost completely dispossessed of their lands and rights in the Orange River Basin and across South Africa by the time of the sanctioning of the ORDP.

The rapidity with which the ORDP was initially planned and approved, as well as the complexity of the proposed scheme, resulted in the need for further detailed planning with the assistance of private domestic and foreign consultants. More detailed planning and appraisal, as well as the discovery of more economically viable alternatives, resulted in design changes and additions to the ORDP during construction. For example, based on a more in-depth assessment after Parliamentary approval of the ORDP, the national electricity utility (ESKOM) persuaded the Government in 1964-65 to raise the heights of both the Gariep and Van der Kloof dams immediately to meet increasing demand for peaking power in South Africa. Although only R85million was allocated to implement the first phase of the ORDP by Parliament in 1962-63, the design changes proposed by ESKOM and more detailed planning resulted in a significant increase in the cost of the scheme.

In fact, an approximately four-fold increase in the estimated costs of the scheme over the decades is attributable to unplanned design changes, the need for foreign technical assistance with design and construction, the non-inclusion of inflation estimates in original planning projections, and unexpected costs for unplanned mitigation measures. Shortfalls in funding to meet rising costs resulted in further delays of the ORDP. Design changes and shortfalls in funding also led to schedule overruns during construction of the ORDP which in turn resulted in further cost increases. Because design changes and rising costs were considered amendments to the original authorisation given for phase 1 of the ORDP, however, further Parliamentary review was not required for increasing the actual expenditures on the scheme.

Project infrastructure was substantially modified during implementation, and regular design changes meant that some features of the planned scheme were not built as the context evolved and other options were explored. Additions unplanned at the outset, such as Wellbedacht dam, affected the assumptions made and actual construction of the planned components. The later construction of the upstream Wellbedacht dam (which subsequently silted up) for municipal water supply effectively, but unexpectedly, reduced sediment inflows into Gariep reservoir, thus prolonging its life.

The dynamics during planning, appraisal and design identified above clearly shaped the outcomes of the ORDP. As this pilot study shows, the actual achieved benefits of the ORDP are less than that which was predicted for irrigated agriculture, both in terms of area (46-57% of planned) and value of production (27%). On the other hand, outcomes are greater than predicted in the areas of agricultural productivity (166-457%) and the economic value of hydropower generation (107 %). Actual hydropower generated is broadly as predicted (although from an installed capacity over twice as large as that which was originally envisioned). The generation of dynamic system benefits for the grid from hydropower unexpectedly added 13% to the overall value of hydropower generation.

Objectives for flood control and recreation were broadly met, although they were unquantified at the time of project planning. Unanticipated environmental impacts of the scheme include the proliferation of biting blackfly that caused damage to livestock and the recreation industry requiring an annual control programme of around 2 million Rand (equivalent to 330,000 in 1998 U.S. dollars). In addition, the ORDP disrupted flows and driving variables for the natural riverine ecosystem and river mouth (a wetland site of international importance).

A comparison between the planned and actual settlement of new White and Coloured families in the command areas nor the predicted and actual displacement/resettlement from the ORDP was not possible due to data inavailability. However, compensation of White farmers seems to have been quite generous although less so for the payments made for land expropriations required for the Van der Kloof than the Gariep reservoir. Official compensation was neither planned nor offered to the Black and Coloured people negatively affected by the ORDP but the numbers of persons in these categories was already quite low given the centuries of dislocation and marginalisation to which they had already been subjected. Employment and regional development goals are difficult to verify but it seems that jobs were saved by the ORDP in the face of growing on-farm rationalisation and intensification nation-wide. The actual effect of the scheme on national agricultural production was not large and national economic growth even less, although data constraints limited analysis in these areas.

Although data limitations preclude firm conclusions, the distributional consequences of the ORDP during the initial period of the 1970s throughout the mid-1980s after its commissioning were not surprising given the anticipated and actual outcomes of the scheme as well as policy context in which it was planned and implemented. The South African population as a whole benefited from the increased agricultural production, added supply of peaking power, and recreational opportunities created although the overall contributions to the national economic value in these sectors and total gross national product from the scheme are likely to have been small or negligable. In addition, South African taxpayers are likely to have borne most of the financial costs of the scheme as cost recovery from users during this early period was relatively low.

Within the Orange River Basin, municipalities and industries most likely benefited from the water supplied by the ORDP but the access of different racial groups to this water supply was likely to have been unequal. Recreation industries within the basin also benefited by the new recreation resources created by the scheme although some business were negatively affected by the increased incidence of biting blackfly. Similarly, blackfly infestation somewhat reduced the large share of overall benefits derived from the increase availability of irrigation by White farmers and a smaller number of Coloured farmers who were either settled or resettled in the command areas of the ORDP. Some additional Coloureds as well as some Blacks who may or may not have been those displaced by the scheme did benefit from the mostly seasonal and casual jobs generated. Finally, while White farmers who were displaced did receive generous compensation, Coloureds and Blacks were not offered payment for their losses by the Government.

On the other hand, these distributional effects of the ORDP did change as a result of the changing operation and replanning of the ORDP in conjunction with the evolution of societal values and Government policies in South Africa over time. The original planning documents envisioned that operations of the ORDP would shift from an early prioritisation of power generation followed by irrigation, municipal water supply and flood control to one in which needs for municipal water supply and irrigation would increasingly take precedence over time. The evolution of operating rules for the ORDP after the time of commissioning in 1971 were not determined in this study and thus further conclusions can not be offered at this time in this area.

However, a reappraisal of the ORDP was undertaken with the Orange River Systems Analysis between 1988 and 1992 which found that potential demand for water in the Orange River Basin would far exceed supply by the year 2020. The need for a comprehensive replanning exercise was thus confirmed. Correspondingly an Orange River Replanning Study, the most comprehensive assessment of the Orange River Basin and ORDP since the original planning efforts between 1960 and 1965, was undertaken between 1994 and 1998. The appraisal also assessed how the construction of the upstream Lesotho Highlands Water Project (LHWP), jointly planned by South Africa and Lesotho beginning in the 1980s, might condition changes in the operation and effects of the ORDP.

The most important shifts in criteria and guidelines that affected the decision-making and effects of the ORDP occurred as a result of the recommendations of three new policy frameworks promulgated during the ORDP implementation period: the 1970 Report of the Commission of Enquiry into Water Matters, the 1986 Report on the Mangement of Water Resources in the Republic of South Africa, and the new water regulations adopted after the transition to democracy in 1994 (especially the 1998 Water Act). These documents highlight the gradual change in South African policies from an emphasis on government subsidisation to full cost-recovery from users, as well as from agricultural production and industrial development to employment creation and provision of basic needs. A shift from technocratic to participatory decision-making and from a technical approach to one that takes human rights, environmental and social issues into consideration, and distributes ORDP benefits across a broader section of South Africa society also occurred.

As a result, trends towards greater public participation, more in-depth financial and economic analysis, as well as more comprehensive examination of social and environmental issues in planning and options-assessment are evident in the decision making processes around the ORDP and LHWP over time. The more recent focus has been on altering the operation and management of the ORDP to improve the functioning of the scheme and to reorient it towards generating outcomes reflective of the new values that have been internalised in South Africa policies after the 1994 transition. In addition, a greater emphasis on basin-wide management, and coordination between the ORDP and LHWP in particular, seems to have been initiated during the 1990s. However, some conflicts have emerged between these new criteria and guidelines such as that between the prioritisation of cost-recovery versus the goal of achieving greater equity in outcomes.

Within this changing context, attention to environmental concerns has also grown continually since project inception in 1962-63 and these concerns have been increasingly incorporated in planning, appraisal and operation of the ORDP. Environmental concerns have only recently been given force of law under the 1998 Water Act. Much effort, time and money has been spent in setting in-stream flow requirements (IFRs). However, the process of implementation appears far less rigorous. In this context, the passage of the 1998 Water Act has provided the legal framework for the formulation of IFRs for the ORDP but the regulations necessary for guiding implementation have yet to be promulgated.