Addendum Lake Victoria Report

Daniel Kull
Monday, June 5, 2006

Thanks to an anonymous source, new data on Lake Victoria and its outflow management has become available since the original report. The data consists of observed daily lake levels, associated agreed curve discharges and actual Nalubaale and Kiira Dam releases for September 2004 to the beginning of March 2006, as reported by Eskom Uganda Ltd.

Original 2005 Estimates Satisfactorily Accurate

In the original study it was assumed that during 2005, precipitation was 15% lower than normal, and evaporation was 10% greater than normal. Based on the new data and assuming again 10% above average annual evaporation, the Lake Victoria water balance model indicates precipitation was actually 17% lower than normal in 2005.

The original study computed that the average releases for 2005 for Nalubaale and Kiira Dams were 1114 m³/s. The new data shows actual releases averaged 1180 m³/s, about 6% greater than the calculations. Considering the rough data and estimated parameters used in the original study, this magnitude of error is acceptable.

Negligence of Agreed Curve Greater than Originally Reported

Figure A–1 shows the newly available data of discharge prescribed by the Agreed Curve (based on observed lake levels), and the actual releases from the two dams.

Discharge Rate Lake Victoria
Discharge Rate Lake Victoria
Actual and Agreed Curve–prescribed total releases at Nalubaale and Kiira Dams for September 2004 to early March 2006.
It can be seen that the actual releases were far greater than those prescribed by the Agreed Curve. Indeed, during the period September 2004 to March 2006, the actual releases were 194% of those prescribed by the agreed curve: almost double.

For 2005, the over–release was 191% of the Agreed Curve, which accounted for approximately 0.26 m in additional lake level drop, or 51% of the total (drought and over–release induced) lake drop. The original study yielded an estimate of 0.23 m additional lake level drop and 53% of total drop due to over–releases, again a close estimate when compared to the newly available data. During the full period of data (September 2004 to March 2006), over–releases resulted in an approximate additional 0.4 m of water level drop.

New Release Policy Still Not Adhering to Agreed Curve

The newly available data indicates a New Release Policy was implemented on February 6, 2006, limited to a maximum total discharge of 850 m³/s. Actual average releases for the one month of available data (February 6 – March 5, 2006) were less, around 806 m³/s. In any case, with observed lake levels indicating an average Agreed Curve–prescribed release of 440 m³/s during this period, the New Release Policy is still far above the Agreed Curve.

If 2006 turns out to be an average year in terms of precipitation and evaporation for the Lake Victoria Basin, then the New Release Policy would result in the lake level increasing by about 0.15 m. While this is indeed a positive development, an Agreed Curve outflow would result in about a 0.33 m lake level increase. It is unknown how the outflows have continued since the end of the given data (March 5, 2006).

It is clear that the Agreed Curve is still not being adhered to, and although the New Release Policy is moving in the right direction, it continues to ignore the natural water balance of Lake Victoria. Such choices are often political, and in this case it appears that the sustainability of a regional resource of social, economic and environmental benefits is of secondary consideration.


Daniel Kull
Hydrologic Engineer
Nairobi, Kenya