Hans Eggers - Institute for High Energy Physics, Vienna
Wednesday, January 8, 1997

The following is a brief review of some sections of the draft feasibility study (FS), focusing primarily on the Epupa Main Technical Report (Part B2). The Baynes Technical Report is a simple copy of the Epupa report, with only key words and appropriate sentences being replaced. Hence the focus below will be on the Epupa option.

Unfortunately, the study cannot be reviewed properly as the crucial figures and appendices (for example to Chapters 23 and 24) are missing.
- The Feasibility Study does a poor job of weighing the consequences of the reservoir’s high rates of evaporation, and downplays the project’s impacts on Namibia’s water demand.

- The Study offers an inaccurate analysis of alternatives to the project, especially the Kudu Gas project, now underway. The study misleads the reader to believe that Kudu power will be more costly than Epupa

- The analysis is based on scant hydrological data, and does not take into account changing hydrological conditions. Changes in the FS’s hydrological analysis could make this project an economic burden on the country.

- The study’s Terms of Reference severely limited the scope and precluded a comprehensive review of the nation’s energy and water needs, meaning that the analysis is heavily weighted for a dam on the Cunene and against all other scenarios. This means that Namibia has gotten a report that is useless as a tool to determine its best path toward sustainable development of its energy and water resources.

*) Evaporation losses from the Epupa lake would amount to many times the total urban consumption of Namibia, and many times the amount of water expected to be drawn from the Okavango (20 million cubic metres per year initially). This is because of the enormous lake surface of 380 square km at high water level (Strategic Summary p 2-1).

The evaporation rate is calculated as 20 cubic metres/second (630 million cubic metres/year) in the Study. The suggested evaporation rate appears to be on the low side, and the quoted evaporation rate has not been proven to be accurate. Apparently, the authors did not make independent evaporation measurements, despite the importance of these measurements to the project’s economic viability. Instead, they relied on data taken long ago at locations remote from Epupa (p7-65). Based on guesswork, the authors have increased previous evaporation estimates by 15 percent (p7-70), corresponding to 100 million cubic metres per year of water losses from the dam.

*) The authors conclude that the loss of water at the reservoir is insignificant, since "the external costs of evaporative losses must be set at nil due to a lack of alternative uses for this water" (p3-5 Strategic Summary). This statement must be disputed in the strongest possible terms. Setting the value of a natural resource at nil just because it has not been utilised yet amounts to a complete disregard for any future possibilities and needs. (It would be equivalent to a commercial company valuing its unsold stocks at nil.) Considering that arid Namibia has an ever increasing water shortage problem, the Cunene represents an priceless water reserve. The huge evaporation loss and the desperate need for water in Namibia alone should conclusively eliminate any proposal to build a dam which does nothing but waste this reserve.

*) Quotes such as the following clearly imply that a sensible water usage plan is not in place:

  1. "The water demand for central Namibia is estimated at about 200 million cubic metres for the year 2020 compared to the present supply potential of developed sources of about 72 million cubic metres/a"
  2. "...completion of the final Grootfontein-Okavango river component of this national water supply scheme early in the next century appears to be inevitable", (p10-38) and
  3. "The above findings have been accepted in principle by the Namibian Government and the Department of Water Affairs has drawn up a Water Master Development Map for Namibia..." (p10-38).

The study’s authors accepted results without justification from previous studies such as the Water Master Plan, whose contents must be questioned. Accepting the findings blindly, they state that the lost Cunene water has no value and there is no need for it, except of irrigation in Ovambo (p10-38). In the long term the picture will change and the value of water will increase considerably. The possibility of using the water for the central region has apparently not been factored into the project’s cost-benefit analysis.

The viability of the Epupa option is strongly dependent on water inflow; the Baynes option critically so. Any adverse hydrological conditions such as drought and upstream use are therefore a major operational and financial risk for Epupa. Hydrological data and its accuracy are therefore critical in determining the viability of Epupa. Unfortunately, hydrological data shows that water supply is decreasing; furthermore the data used is inaccurate:

*) According to the study, the worst water flows in the 49 year record period occurred in 1992-1994. According to Nampower the Cunene flow reached a new all-time low in 1997. The average Cunene water flow over the past twenty years has decreased. There is evidence of changing climate conditions and that drought periods are increasing in duration and severity all over the world. Under changing hydrological conditions, the future performance of Epupa cannot reliably be tested with past river flow records. The study does not address this issue.

This gives reason to believe that Epupa will perform worse than expected, resulting in a financial burden to the country instead of the forecast benefit. According to the risk analysis (p25-9), the project is already economically marginal with 1992-1994 river flow simulations. The project will not be economically viable (p25-10), should a cost overrun also occur.

*) Only 12 years of reliable daily flow data were available for the Cunene at Ruacana (1961-1972). The Ruacana flow sequence was synthetically extended to span the period 1945-1994, by means of correlation to the Okavango River (Rundu) (p7-1). This approach definitely reflects an uncertainty and could have serious consequences on the economic viability of Epupa/Baynes.

*) Angola has the right to withdraw water from the Cunene upstream, and will increasingly do so. Quote from the FS: "The withdrawal of water from the Cunene will increase from about 125 million cubic metres per year to some 600 - 1100 million cubic metres per year in the year 2035. These figures may also be compared to the estimated average discharge at Ruacana of about 5000 million cubic metres per year." Already, Cunene flow rates are so low as to put the Epupa project in doubt. If roughly thirty percent of water flowing into the lake is lost (through upstream abstraction and evaporation), it becomes hard to believe in the viability of Epupa under any, even optimal, hydrological circumstances.

*) The authors of the FS appear to have complete disregard of the overall water scarcity in Namibia, focusing instead only on the immediate problem at hand. The possibility of using Cunene water for urban consumption rather than for power generation, for example, deserves detailed study. Given that Kudu gas will easily provide for all power needs, and given the sensitivity of the Okavango to water abstraction, Cunene water appears an attractive alternative for urban needs.

Even if such wider considerations exceed the Terms of Reference of the Epupa FS, the government should be taking the overall view. Recent estimates by Earthlife Africa, as published in two open letters, claim that, once such an integrated view of development in Namibia is taken, large financial savings are possible. In addition, all the environmental problems associated with both the Epupa and the Okavango proposals would be eliminated. These claimed huge benefits clearly deserve detailed consideration by the government before any decision regarding Epupa is made.

*) "The total financing requirement of Epupa is calculated as US$753.6 million, including price contingencies and interest during construction" (p23-19). Unit costs of Epupa electricity (excluding price contingencies/interest during construction) is given as 4.4 to 4.9 UScents/kWh (USc/kWh), with and without Gove Dam in operation respectively (p23-13). The study states that "most probably the majority of consumers is willing to pay substantially more" (p23-8) for Epupa power to justify the project, since the price of Epupa power is higher than the current electricity price of 3.7 USc/kWh (p23-6). This assumption defies all logic. It is a truism that the consumer would prefer to pay less, and would by implication prefer cheaper import power or Kudu gas power.

In the light of the above figures, claims in the study that Epupa power would be cheaper than any other alternative are self-contradictory. Apparently Appendix 23.1 contains the financial analysis, but the latter is not included in the version of the feasibility study made available to the public.

*) The FS applies inconsistent criteria for evaluating cost of power. On the one hand,
according to the study, "The financial assessment made indicates that Epupa is financially viable" and "Baynes COULD prove financially viable WITH Gove in operation. A key prerequisite appears, however, to be that Nampower is allowed to raise the tariff to 4.5 USc/kWh in 1997 prices no later than 2005 as assumed in the calculations" (p23-19 Epupa, Baynes Technical).

On the other hand, however, when considering renewable energy, the study writes on p5-27 (Approach to alternative energy): "The review is thus pessimistic about the near to medium term future for large scale renewables in Namibia. This is mainly because the available conventional energy alternatives are cheap by world standards and likely to remain so over, at least the next 10-20 years. Large scale promotion of the use of renewables would therefore consume economic resources which could otherwise be used for other investments."

In the above two costing estimates, the authors compare projected costs (year 2010) of renewables to those of conventional power options at unit costs of 3.0 USc/kWh and reject renewables because they are more expensive than 3.0 USc/kWh. However, according to their own analysis, there is no conventional energy option available to NAMIBIA at unit costs of 3.0 USc/kWh. Epupa is claimed to be closest at 4.5-4.9 USc/kWh. Therefore it would have been correct to compare the costs of renewables to the costs of Epupa rather than the unattainable ideal of 3.0 USc/kWh. If this had been done, the projected wind power costs of 4.9 USc/kWh, as calculated in the FS, would be competitive to Epupa. In addition, the FS cost estimates of renewable energy projects appear to be unrealistically high in any case (see comments below, under ALTERNATIVES).

The authors claim that they have made a thorough study of alternative supply options (Strategic Summary p 3-7). The following points refute this.

*) The study notes, "The purpose of this feasibility study is to evaluate the development of Epupa/Baynes hydropower project and not to present a full master plan for the electricity sector" (p5-2). It is clearly not in Namibia’s best interests to weigh its energy-development future based on one expensive study that looks single-mindedly at one particular project (Epupa), without a comprehensive master plan as a basis for comparison (see "The Overall Picture" above).

Ideally, the better approach would have been to (1) identify all the available alternatives, (2) to study them to such extent that the most competitive and environmentally acceptable option can be pointed out with confidence and finally, (3) to proceed with this option.

Due to the superficial treatment of alternatives, this study does not appear to facilitate this approach but rather compounds the muddled thinking already in evidence. If the FS ducks the responsibility of making realistic comparisons of ALL options in the electricity sector, it would be the duty of the government to do so before making any further decision on Epupa.

*) The study exhibits a general bias towards Epupa and Baynes in all comparisons with alternatives. Several examples have already been documented above. In another example,
the report states, "Experience with solar and wind on large scale is generally limited, grid connected systems have not been tried out in Namibia. Also, wind data for the country is limited. Therefore, much uncertainty is attached to operational reliability and costs for these alternatives." (p5-4) By contrast, the authors are unrealistically positive about the Baynes project even when it is plain that no solutions have yet been found: "There are a number of complex design issues to be addressed, but these are all within the realm of current technology" (Strategic Summary p2-3).

*) The study concludes that wind and solar energy are not feasible as they are more expensive than Epupa. The FS authors do mention that international grants are available, e.g. by the World Bank, to promote and implement environmentally friendly renewable energy technologies. Yet they fail to include such international grants and the resulting cost reductions in their calculations on renewable energies (p5-9). Therefore, the presented simplified cost comparisons are misleading.

The following comments describe the flawed logic found in the report’s analysis of specific alternatives.

*) The evaluation of the Kudu gas alternative is based on the assumption that gas will be supplied to the gas power station only. In other words, the electricity from a small power station must be sold at very high costs in order to recover the full gas field development costs. Thus, the FS compares a very small 360 MW gas power
station to Epupa because of the similar size. It then proceeds to conclude that Epupa/Baynes is more attractive, because of lower electricity supply costs. (The price of electricity from the 360 MW gas station is given as 5.3 USc/kWh) (p5-11).

This is totally unrealistic. Obviously, the large Kudu gas field will not be developed for a small power station only. In fact, it is common knowledge that an agreement between Shell Namibia, Nampower (a Namibian parastatal!) and Eskom has been signed providing for a 750 MW power station at Oranjemund. This power station, plus planned export of gas and/or power to South Africa, already accounts for a large fraction of the field development costs. Additional gas power plants that might be needed by Namibia, therefore do not have to carry the full Kudu field development costs and will hence come out considerably cheaper than claimed in the FS.

Secondly, the larger the wattage of gas power stations, the lower the unit cost. The FS comparison between Epupa and only a small 360 MW gas power plant is therefore unfair towards gas. By contrast, a larger gas power station (1300 MW), capable of exporting excess power to South Africa, would generate power at 2.85 USc/kWh (p5-11), as compared to 4.4-4.9 USc/kWh for Epupa.

Nevertheless, the authors maintain that Epupa is the most attractive option, saying, "This is a difference in present value costs of USD54.7 million in favour of the Epupa option" (p15-3 of the Epupa EIA). As shown above, such conclusions are highly skewed by making the wrong comparisons. This is an important issue and throughout the report the reader is misled that a gas power station would necessarily be more expensive than Epupa.

*) The FS-quoted gas price of 2.0 USD/GJ (corresponding to fuel costs of 1.31 US/kWh and 3.8 USD/MCF) BEFORE tax is substantially higher than previous gas price estimates AFTER tax stated by the World Bank/Pencol Engineering Consultants (1.14 to 1.34 USD/MCF). "The gas price (of 2.0 USD/GJ) used was initially quoted by Shell as a reference price for Kudu gas delivered (at) Oranjemund" (p5-8). Since the gas price has a very strong effect on the electricity price, it is strange that the authors do not elaborate on this issue.

And yet the authors question the higher gas price of 2.0 USD/GJ. "Based on the independent calculations of gas cost made, the assumption of USD 2.0/GJ applied in the study implies an advantage given to CC (gas power) relative to other alternatives which appears unrealistic" (p5-9). Since the authors do not elaborate, it can only be assumed that the reason behind this argument is that gas power would be significantly cheaper than Epupa if a lower gas price is used.

The above facts make clear that the feasibility study tries to present gas prices in as unfavourable light as possible. A more balanced comparison, taking into account the probable scenario of large gas offtake, a large power station, plus export to South Africa, would prove that the FS seriously overestimates costs for gas-generated power.

*) Kudu gas power could make Namibia self-sufficient with respect to power for at least 20 years, at which time advances in other renewables (e.g., solar and wind) will make them commercially competitive. Epupa would NOT ensure self-sufficiency. In fact, the FS appears to state that gas would be needed to COMPLEMENT Epupa power. Quote: "If the Gove dam is not in operation, a total of 360 MW of additional gas turbines installation would be required in order to satisfy the dimensioning criteria" (p24-4).

In addition, if the gas needed to power the gas turbines is to originate from the Kudu gas field, then the field development costs should ALSO have been included in the Epupa financial analysis.

*) With this background, it is wrong to state that "Epupa is the least cost option for supply in Namibia..." (p23-12). Kudu gas power would be a more competitive option than Epupa. Several summaries of alternative energy supply options found in the study did not take the development of Kudu gas into account in determining whether their viability is improved once Kudu is developed (e.g. p5-27).

There is little doubt that, given the already existing plans for a power station at Oranjemund, plus the above considerations, Namibia’s power needs are easily and abundantly solved by using Kudu gas. No amount of obfuscation by the FS can hide the simple fact that the gas option comes out cheaper and safer.

*) Dealing with wind energy, the FS is either biased or technologically out of date, as the following quote shows: "Setting up the measurement study, monitoring wind speeds for 2 to 3 years, designing, erecting and monitoring a pilot wind farm will take at least a decade. Any decision on a major investment programme in wind cannot therefore be made for at least this length of time" (p5-26). This is unrealistic, considering that wind plants in Europe and the US have been installed in very short times, far shorter than a decade. Wind power is a mature technology and could be introduced easily. It is characteristic of the FS bias against alternatives that negative remarks are made towards the time needed for developing a wind power project while hydro plants are known for their long construction times.

Furthermore, the authors present a cost calculation without even mentioning the average windspeed for which the evaluation has been done. Neither, apparently was any use made of available weather station information. The conclusion that under present conditions wind power is not an attractive option for the country thus appears based on unknown and incomplete information.

The presented costs of 7.0 USc/kWh appear to be too high (p5-25) compared to already existing projects. For example, according to the "Wind Power Monthly" Journal (January 1995), the wind power purchase prices in INDIA (also a developing country) ranged from 4 to 7 USc/kWh prior to 1995. Lower prices are also quoted by another source (Renewable Energy, Sources for Fuel and Electricity, edited by Thomas B.Johansson and others, 1993, commissioned by the United Nations): At an average wind speed of 8.5 m/s, the electricity cost from wind power in the year 2000 is estimated at between 2.7 USc/kWh to 3.8 USc/kWh, for an interest rate of 6 to 12 percent. Indications are that the southern coastal regions of Namibia have a high wind potential so that these international cost estimates appear to be appropriate.

A telling argument in favour of wind energy is provided by the German GTZ (Gesellschaft fur Technische Zusammenarbeit) which, in collaboration with local Namibian firms, is undertaking a study of a small-scale power-from-wind plant on the coast near Luederitz. While clearly wind is unsuited for base-load applications, its use in small-scale plants has the potential greatly to reduce overall demand.

*) "The electricity consumption in Namibia is dominated by electricity use in households, small-scale manufacturing, etc." (p 5-28), where water heating constitutes a significant part of the household energy consumption.

The cost of heating water by means of solar water heaters instead of grid electricity has been calculated as 8.8 USc/kWh. It is then said that since the CONSUMER pays 6.0 USc/kWh for grid electricity, this option is not an attractive alternative. Why then did the authors argue that the consumer is willing to pay substantially more for (Epupa) electricity (p23-8)? If the electricity price is to be increased substantially as recommended, the price of 8.8 USc/kWh should have been compared to the new proposed price the consumer would have to pay.

*) An important assumption in the calculation leading to 8.8 USc/kWh is that a solar system lasts only 7 years. According to major solar water heater suppliers, these systems work reliably for 15 years, and much longer if maintained properly. The corresponding costs calculated with a lifetime of 15 years results in saved electricity costs of 5.6 USc per kWh (or less if maintained). This would therefore be cheaper than grid power. An added benefit, and one not quantified in the report, is that the installation and maintenance of these systems could be a reliable, ongoing source of employment, dispersed around the nation (Kenya has reaped such benefits from its own development of small solar systems.)

The study of energy conservation measures was included in the terms of reference for the FS. Unfortunately, only a superficial discussion is presented. No thorough investigations into this important option have been conducted. This is the cheapest of all options immediately available, and clear recommendations to the decision-makers up to the level of the strategic summary would have been appropriate. (p5-28)

Imported power is a cheap supply option. Present costs for imported electric power are are about 1.8 USc/kWh only (see p8-3, Epupa Main Report Summary). Presently, an additional large 400kV interconnector between South Africa and Namibia is under construction. The new line is expected to be in operation from 1999 with a transfer capability of about 540 MW. If reinforced, the new interconnector will increase the total transfer capacity to around 900 MW. With the new line, the total supply capacity from imports and existing stations will be sufficient to meet projected electricity demand well beyond 2015 (p8-3).

No unit electricity costs of import power can be found in the study, but a World Bank study calculates transmission electricity costs (400kV line) at 1.36 USc/kWh, charged over 15 years. Adding the electricity costs charged by South Africa (1.8 USc/kWh, p8-3) adds up to 3.26 USc/kWh, i.e. much cheaper than Epupa. (On p23-8 the authors quote an average import tariff of 3.0 USc/kWh they have applied, instead of the current tariff of 1.8 USc/kWh. But even this price would lead to a lower price than Epupa, at 4.36 USc/kWh.)

It is unclear how the authors nevertheless arrive at the conclusion that Epupa is cheaper than import in terms of its present value (p24-11). The study further argues that imports would make Namibia too dependent on South Africa. On the other hand, the study omits to say that future upstream water uses by Angola would pose exactly the same risk and dependency on a neighbour (see "Hydrology" above).

The Gove dam would increase power output from downstream stations, including the Ruacana station. The proposal and the cost of first repairing Gove is not included in appropriate sections. A recommendation to repair Gove and then to upgrade Ruacana power station could not be found anywhere in the study.


by Hans Eggers
Institute for High Energy Physics, Vienna

Further Reviews:

This page created January 22, 1998