Rainfall forecasting key to efficient hydropower in Ethiopia

Hydropower is a tremendous potential energy source for many developing countries, but managing water reserves to maximize energy production is a tricky business.  Let too much water out of the reservoir and you may not have enough later.  Let out too little, and you aren’t producing all the energy that you could.

Paul Block, Associate Research Scientist in Hydroclimatology and Water Resources Management, at the Earth Institute’s International Research Institute for Climate and Society, presented his study on this subject at the CWC Seminar Series on April 9.  His talk on ‘Tailoring Seasonal Climate Forecasts for Hydropower Operations in Ethiopia’s Upper Blue Nile Basin’ explained the models he’s working with to help Ethiopia manage their energy generation most efficiently.

Currently, Ethiopia is only using a small fraction of the potential it has for hydro-electric production.  Block is looking at four sites in the Upper Blue Nile Basin that have been identified for possible development, which together would generate about 5,500 megawatts per year.

He has been comparing the costs and benefits of models in which reservoir operators plan their resource management based on monitoring of current streamflow conditions, versus based on climatological forecasts.  To be able predict precipitation accurately would help maximize energy production for local use and also for sale.  When forecasts are right, the benefits of using them are great.

Block has found, however, (Block and Rajagopalan 2007) (Block and Strzepek 2009) that when using forecasts as a guide, if the forecast predicts a wetter than normal year and is wrong, the cost is high – higher than when the forecast predicts a dryer than normal year and is wrong.  If the reservoir operators release too much water downstream, expecting rainfall that doesn’t come, then the following year there might not be enough water to meet production needs.  Having held back too much water in expectation of a dry year that turns out to be wet after all, has a smaller impact.

This presents a problem in using forecasts, as the utility operators may not be able to afford to get it badly wrong even once.  One way of addressing that is to use the forecasts to guide reservoir management only when they are predicting a dryer than normal year, which can increase benefits while reducing risks.

Better yet, continuing research is looking for ways to forecast precipitation on a consistently reliable basis, to allow stability and efficiency in meeting energy production goals.

Follow Columbia Water Center on Twitter: http://twitter.com/columbiawater