There are several approaches to this problem. The one chosen, is to start and stop the diesel motor in a way that gives a balanced minimization of the costs: fuel consumption, number of motor starts and power quality. This approach assumes a windturbine of least at the same size as the diesel generator set.
A start/stop strategy based on adaptive prediction of power output from the windturbine as well as of the load is developed. The strategy thus accounts for the stochastic and the instationary nature of the inputs to the systems.
Real life evaluation of this approach would take years, therefore a sumulation model is developed. The model consists of two parts; the physical system and the inputs, wind velocity and load. The latter part is based on stochastic processes. This model makes it possible to do several independent runs which in turn enables an evaluation of the statistical significance of the result. This is considered to be a very important feature.
The developed strategy is compared to a very simple one. Simulation results show no significant differences between the two strategies. This surprising result is discussed in detail.
The simulation model is capable of producing results from which yearly average costs can be calculated. These calculations show that there probably is a large number of small communities around the world, where it would be economically attractive to connect a windturbine to an existing diesel generator set.