Appreciating Wind Energy’s Probabilistic Nature within the Uncertainty Context of Electric Power System Network Planning Daniel J. Burke 1* and M.J. O’Malley 2 1 Australian Energy Market Operator (AEMO), Melbourne, Victoria, Australia. 2 Electricity Research Centre, University College Dublin, Ireland. * Email – DNLBRK157@gmail.com, Ph - +(61) 396098570 Abstract Electric power system network planning is influenced by the uncertainty in many parameters, such as future customer-demand/fossil-fuel-price parameter projections and new generation plant locations, which can generally be modeled in an approximate or subjective manner at best. Historically recorded wind power data presented here supports the contention that medium-term wind power generation profiles tend to follow more stable probability distributions. While it is widely acknowledged that the addition of wind power to a system makes network investment analysis more difficult in some respects, such data patterns nevertheless evoke an important discussion on how this relatively more probabilistic (i.e. not uncertain) characteristic of wind power may have some positive value for transmission planning and its related decision making procedures. 1. Introduction Many power systems are currently experiencing a significant expansion in wind energy as a source of clean, renewable electricity generation. Wind energy characteristics e.g. its variability and difficulty to predict, its relatively low capacity factor or energy yield, and the fact that it generally uses alternative generation technology, pose many challenges to integrating this resource into electric power systems [1]. Some new decision making tools have been devised to cope with these challenges recently. For short term supply/demand balancing operations purposes the variability and uncertainty has been tackled with increasing reserves [2][3][4][5], improved wind forecasting [6] and the possibility of using stochastic scheduling methodologies [7][8]. Challenges in the infrastructure planning domain have also attracted significant attention both in power generation expansion [9] and electricity transmission network planning. A detailed outline of the industry and research state of the art of wind and transmission planning is given in [10]. Combined regulatory and industry initiatives have made significant progress in some power systems, with the ERCOT “CREZ” project in the USA being one such notable solution [11]. Some academic contributions have proposed advanced statistical analysis [12] combined with investment optimisation models [13] as a means to increase the speed of connection of new wind projects while implicitly minimising the transmission requirement. However, significant investment will still be required in electricity network infrastructure in order to integrate the number of proposed renewable energy projects over the next 10 years [14]. From the point of view of the transmission system planner, this network planning task is most astutely formulated as a ‘decision-making-under-uncertainty’ type of problem [15][16][17]. Precise values for most of the input parameters are generally not available. Model parameters that are a-priori unknowns in the transmission planning problem will generally result in increased costs to system investment. For example, the system might have to be over-dimensioned at the outset when anticipating the risk of extreme parameter values which may not actually result. Or on the other hand, the network capability might end up being insufficient in some respect, if a particularly extreme value of an uncertain parameter is observed outside the range countenanced a-priori in the system investment planning timeframe. Predicting customers’ future electricity demand time series profile and peak value growth levels has always been an approximate process at best (e.g. annual customer energy demand dropped by ~ 7% in 2008 alone in the Republic of Ireland due to unforeseen economic conditions [18]), and conventional plant fossil-fuel and carbon price volatility has been routinely apparent in recent times. For example, the unconventional shale-gas boom in 2013 46th Hawaii International Conference on System Sciences 1530-1605/12 $26.00 © 2012 IEEE DOI 10.1109/HICSS.2013.98 2213 2013 46th Hawaii International Conference on System Sciences 1530-1605/12 $26.00 © 2012 IEEE DOI 10.1109/HICSS.2013.98 2215