Growing Renewable Energy (RE) means a greater increase in variability of supply, a relatively newer phenomenon for grids where demand was the usual variable, and supply was tightly controlled, or ‘despatchable’. One cannot control the wind or sun, but one at least needs to predict it well, so that the rest of the grid can plan its output accordingly. This is one of the several key aspects of making RE grid integration cheaper and more scalable. Otherwise, as RE penetration grows, its challenges for the rest of the grid will increase.
Forecasting is just one part of the puzzle – “what next” is the key issue. If you deviate from the forecast, should you be penalised, and by how much? These are key issues. These will link closely to the norms for forecasting, not just in terms of accuracy or time (how far ahead) but even the scale of aggregation. Forecasts per wind turbine or solar panel are expensive and don’t add much value, but aggregated forecasts across a very wide area (like state) also don’t help the grid operator who needs to balance not just generation but also transmission. An ideal balance is usually forecasts aggregated at a transmission pooling station level.