Quantifying the impact of load forecasting accuracy on congestion management in distribution grids

• Development of an approach for quantifying and simulating load-specific uncertainty within a distribution grid. • Integrating smart meters reduces load uncertainty, leading to substantial cost reductions in congestion management. • Prioritizing the strategic deployment of smart meters optimizes the redispatch of generation units. Digitalization is a global trend in energy systems and beyond. However, it is often unclear what digitalization means exactly in the context of energy systems and how the benefits of digitalization can be quantified. Providing additional information, e.g., through sensors and metering equipment, is one concrete angle where digitalization contributes. This paper provides a framework to quantify the value of such additional information in distribution grids. We analyze to what extent smart meters improve the accuracy of day-ahead load forecasts and quantify the savings in congestion management costs resulting from the improved accuracy. To quantify the cost reduction, we conduct a case study employing a simplified IEEE test system. Historical electricity load data from over 6,000 smart meters was used to improve day-ahead load forecasts. We assessed and compared the forecasting performance to estimates based on standard load profiles with multiple load forecast simulations in the network based on uncertainty parameterizations from forecasts with and without smart meter data. Calculating redispatch cost in the distribution grid, we find that the forecast based on smart meter data reduces key redispatch parameters such as the share of expected voltage violations, the amount of rescheduled generation by more than 90%. These improvements translate into a reduction in congestion management costs by around 97%. Furthermore, we shed light on whether the gains increase linearly with the number of smart meters and available data in the grid. When smart meter shares are increased uniformly throughout the grid, savings are concave, i.e., the first 10% of smart meters reduces congestion management costs by around 20% while the last 10% reduces these costs only marginally. Focusing smart meter installation on the most congested nodes reduces congestion management costs by around 60% with just 10% smart meter coverage, significantly outperforming a uniform rollout. However, savings in congestion management alone are not likely to recover the installation and operation costs of the installed smart meter.