A Study of the Impact of Load Forecasting Errors on Trading and Balancing in a Microgrid

How to efficiently integrate and exploit flexible demand and distributed renewable generation in the Smart Grid is a question of high interest reflecting the societal need for energy efficiency and urge to lower carbon emissions. This question also provides strong motivation for honing the concept of a microgrid as an atomic cell of future active distribution networks as well as the smallest market entity. The microgrid locally balances demand and supply in a far more efficient way than in current practice. Complementary, the microgrid market offers basis for creating energy prices that stimulate investments in renewable generation, storage, and demand response. As similar to the current energy markets, the microgrid market can accommodate trading of the short term and the long term energy products, balancing energy and capacity. The foundation of trading is forecasting and optimization, and consequently each party involved in microgrid trading must be able to forecast its demand, supply, or flexibility. But how precise these forecasts must be? This paper focuses on the impact of forecasting errors on the economic effects of trading and balancing in the microgrid: while lower forecasting accuracy induces greater differences between forecasted and real consumption/generation, and hence higher need for balancing energy, higher forecasting precision may increase the cost of the system. We present results of an agent-based simulation study of a microgrid with a simple market integrating local suppliers and customers with flexible loads, renewable energy sources and storage capacity. These actors buy energy, sell demand reduction, and sell energy produced by their wind turbines and solar panels, or stored in their battery. The Microgrid System Operator (MSO) operates the local market and balances demand and supply. In our model MSO operates a dedicated storage and interacts with the global grid markets and performs clearing and settlement of balancing energy costs. In t- e presented study we compare the economic results of trading and balancing for different values of forecasting errors in scenarios characterized with different supply levels in the microgrid.