Auriol and Biancini have a nice theoretical take on the subject of regional power pools in the World Bank Economic Review:
Power market integration is analyzed in a two-country model with nationally regulated firms and costly public funds. If the generation costs between the two countries are too similar, negative business stealing outweighs efficiency gains so that, subsequent to integration, welfare decreases in both regions. Integration is welfare enhancing when the cost difference between two regions is large enough. The benefits from export profits increase the total welfare in the exporting country, whereas the importing country benefits from a lower price. In this case, market integration also improves incentives to invest compared to autarky. The investment levels remain inefficient, however, especially for transportation facilities. Free riding reduces incentives to invest in these public-good components of the network, whereas business stealing tends to decrease the capacity to finance new investment.
There is a lot of excitement in the power sector in Africa these days. Eastern Africa will probably be the most power-rich in the next few years: Ethiopia’s 1870MW Gibe III will likely come online later this year (oh, and there is the ongoing Grand Renaissance Dam project that will produce 6000MW [or 2800MW?]); Kenya has had a massive push for geothermal, solar and wind generation; and Kenya, Uganda, and Tanzania have commercially viable deposits of oil and gas that will also come online within the next six years. The region is also very well integrated, although there is need to upgrade power transmission infrastructure. The Bank has an ongoing project in this regard [that is way behind schedule].
For more on this subject see my commentary on energy sector integration in Africa over at the African Development Bank blog.