Macro Scale Modelling of Wind Plants in Long Term Planning Studies - A Sri Lankan Case Study

Abstract

Harnessing Non Conventional Renewable Energy (NCRE) sources as small-scale embedded generation is rapidly increasing around the world. Modeling most NCRE based generation becomes as arduous task owing to their extremely volatile nature of resource availability along with present day economics, technical, social and environmental constraints. In this context, Sri Lanka is no exception. According to the energy policy-2006, energy share from NCRE is targeted to reach 10% of total electricity generation by 2015. According to National Renewable Energy Laboratory (NREL) report [1], wind energy potential in Sri Lanka is high compared with other technologies; mini hydro and biomass. Recent commissioning of about 30 MW of wind plants in Kalpitiya peninsula is an indication of investor interest in wind energy development for power generation. Therefore, modeling of wind power plants in long term planning can no longer be simple, such as representation as a lumped equivalent thermal plant with high Forced Outage Rate (FOR). Preparation of a macro scale wind plant model for Sri Lanka has not been undertaken before using the economic optimization tool Wien Automatic System Planning (WASP) package. This gap is addressed in this paper. Two models were initially prepared to be compatible with WASP. An in-depth study using (i) Modified Load Duration Curve method and (ii) Run of River (ROR) type hydro equivalent wind model based on five state probabilistic distributions, were investigated. Compatibility of new models was tested with WASP for dispatching as embedded generators. Considering the model simplicity, requirement of time & effect for sensitivity analyses and modifications, the later approach was concluded as the most appropriate for long term planning studies