Minhas, Daud Mustafa; Khalid, Raja Rehan; Frey, Georg : Activation of electrical loads under electricity price uncertainty. The 5th IEEE International Conference on Smart Energy Grid Engineering (SEGE 2017), Oshawa, Canada, August. 2017.

Abstract:

The electricity price is an uncertain and changeable entity, mostly depends on power generating source and consumer’s power demands behavior. Activating consumer’s power demands at low electricity price is a cost minimization problem. The problem arises when all the consumers try to avail specific low price time slot to activate their power demands. It ends up with energy congestion or system destabilization. A better strategy is, to forecast a day ahead price and update it instantly, whenever a new consumer purchases some energy for its next day demands. Therefore, every time a new price is applicable for the upcoming purchasers on day-ahead basis. This strategy may be adopted by an aggregator. An aggregator, which is equipped with renewable energy resources (RES), also imports electricity from the grid. It forecasts and updates a day-ahead fluctuating electricity price for its consumers. So that, all the consumers can avail relatively low price electricity slots based on their energy purchasing decisions. In this paper, a regression base statistical model is adopted to investigate two major problems: 1) accurate forecasting of day-ahead electricity price by an aggregator; 2) activate load demands by the consumers at less electricity price, broadcasted by the aggregator. In our proposed solution, linear regression is used to forecast the electricity price, exploiting intermittent nature of the renewables. Whereas, load activation strategy is proposed by introducing electrical loads with different levels of delay tolerance. Applying boundary condition values using logistic regression, a consumer can activate its loads on different electricity prices. The simulation results have established an effectively forecasted value of electricity price and an accurate activation of consumer’s load demands.

Keywords:

Demand response, Price forecasting, Load management, Regression models