Review of Chapter 11 of the Feasibility Study for the Epupa Hydropower Scheme, Namibia

Steve Rothert
Thursday, January 8, 1998

Section 11.2.ii states that the simulation "will calculate downstream effects of the" various reservoir discharge scenarios. No further discussion or details are provided regarding this claim. It is not clear how the simulation model could calculate downstream effects, beyond rough estimates of streamflow.

Section 11.2.iv states that "secondary hydropower generation is calculated and sold to the Republic of South Africa. The price for sold energy is assumed to be 75% of the import price." This assumption is used in the simulation model to estimate benefits of the Epupa Project, but it is not clear if South Africa would consider purchasing power from Namibia because the RSA has a surplus of generating capacity.

Section states that the consultants extended the simulations to the year 2050. The Epupa Project would have been operating for approximately 45 years by that time, yet the simulation model does not account for the inevitable rise in operation and maintenance costs the Epupa Project will experience by that time. A study of several hundred North American hydrodams shows that, on average, hydrodam operating costs rise dramatically after approximately 30 years.1 Long-term estimates of project benefits should reflect this fact.

Section discusses the value of energy that is not supplied to the customers due to temporary shutdowns, drought etc. It states ìthe cost for unserved energy is normally in the range of 0.25 - 5.0 USD, depending on the infra-structure and industrial development level of a country." The study explains that one consequence of high unserved energy costs is that unreliable generation plants are economically penalised. The section goes on to describe that because "the Namibian economy is highly dependent on electricity dependent industries", the cost of unserved energy should be set relatively high. Yet the cost is set at 1.5 USD, which is in the lower one-third of the range of costs. The consultant unjustifiably chose a lower unserved energy cost, which would protect the Epupa Project from economic penalties resulting from the degree of unreliability of Epupa’s generation.

Section discusses the hydrological data used in the simulations, arguably one of the most important components in hydrodam modelling. This section is misleading and extremely vague. It does not describe the streamflow values used, nor does it adequately describe the rationale for choosing them. Without knowing the assumptions used for this component, it is impossible to know if the analysis is reasonable or accurate. The study states that "the last 49 years of streamflow data for the Cunene are used." This is either a misstatement or a gross error in the simulation methodology. Chapter 7 of the Feasibility Study, "Hydrology", explains in detail why the Cunene streamflow record is unusable for modelling purposes except for a period of 12 years (1961-72). Streamflow data for all but these 12 years had to be generated, creating a synthesised data set. Thus, the simulations are based not on empirical streamflow data as is suggested in this section, but on synthetic data. This is an important point because such synthesised data sets have been shown routinely to overestimate future streamflow and exaggerate project benefits.2

Section summarises the technical and economic aspects of the project. The section lists evaporation at 1,765 mm per year, which does not agree with the 1,890 mm per year figure listed in Table 7.10.4 on page 7-63 of the Hydrology chapter. The 1,765 mm figure underestimates evaporation by approximately 7%, which is more than 40 million cubic meters per year.

Section discusses the ability of Epupa and Ruacana to eliminate the need to purchase power from RSA. The picture presented here conflicts with the discussion in Chapter 13, "Epupa Reservoir", which states "Simulation runs presented here indicate that retaining ample storage volumes at the end of dry seasons and compensating the energy shortage thus incurred by import, will result in increased loss of water over the spillway and consequently decreased energy production in the long run." (Emphasis added). Thus, while Epupa and Ruacana together might, in theory, be able to almost eliminate the need to import power, it seems clear that the project likely will be operated in such a manner that Namibia will still be required to import significant amounts of energy, particularly during the months at the end of the dry season before streamflows increase.

Footnotes1 T. Wong, "Determining O&M Costs Over the Life of a Hydro Station", in Hydro in the 90’s, Hydro Review World-wide, Kansas City 1994.
2 The greatly overestimated streamflow estimates for Phase 2 of the Lesotho Highlands Water Project represents only one of the most recent examples.

Steve Rothert
International Rivers
PO Box 2427
Gaborone, Botswana

Further Reviews: