Henrik Malmberg  presenterar sitt examensarbete

Detecting shortages of wind power in a small system
(Om att förutsäga brist på vindkraft i ett lokalt system)

Abstract:

In the village of Mwinilunga in north-western Zambia, the supply of
electricity comes from diesel generators, installed in the 1970's. The
supply is unstable and unreliable, due to lack of fuel and spare
parts. It is also expensive to operate and maintain them. Even though
many efforts have been done to connect Mwinilunga to the national
electric grid, it is highly unrealistic that it will be done in the
forseeable future since the distance to the nearest connection point
is too large.

Other alternatives remain though to provide Mwinilunga with a stable
supply of electric energy. One is to install wind mills. The
advantages with wind mills are many, but a great disadvantage is the
need for regulation power. A wind mill cannot be the sole supplier to
an electric system without another power source which could cover
possible shortages when the wind speed is too low. In Mwinilunga, the
existing diesel generators could be used in these cases.

In order to keep the electric system stable, the regulation power
needs to be alerted and ready to produce electricity in the very
moment as the production from the wind mill falls short of the
consumption need.

In this thesis a statistical method have been developed to forecast
and alarm if a shortfall is about to happen. The method use the
predictions of the wind power output as a basis for the alarm.