Extending Load Forecasting from Zonal Aggregates to Individual Nodes for Transmission System Operators

TL;DR

The paper tackles the need for high-spatial-resolution load forecasts for Transmission System Operators by introducing a multi-level, interpretable forecasting framework that extends zonal forecasts to individual buses. It combines a scalable hits-GAM model with global training and group-level pooling, and evaluates hierarchical reconciliation strategies to ensure coherence between utility- and bus-level forecasts. The approach yields strong utility-level accuracy with interpretability and substantially improves bus-level forecasts, while providing error-diagnosis tools that attribute deviations to specific buses. Practically, the system enables operators to adjust forecasts confidently, diagnose errors precisely, and integrate with existing EMS workflows, supporting more reliable operation of power grids with high renewable penetration. Key empirical findings include a 24% reduction in MAPE at the utility level and meaningful reductions in RMSE/MAE for bus-level forecasts, demonstrating the value of high-resolution, interpretable, and scalable forecasting for TSOs.

Abstract

The reliability of local power grid infrastructure is challenged by sustainable energy developments increasing electric load uncertainty. Transmission System Operators (TSOs) need load forecasts of higher spatial resolution, extending current forecasting operations from zonal aggregates to individual nodes. However, nodal loads are less accurate to forecast and require a large number of individual forecasts, which are hard to manage for the human experts assessing risks in the control room's daily operations (operator). In collaboration with a TSO, we design a multi-level system that meets the needs of operators for hourly day-ahead load forecasting. Utilizing a uniquely extensive dataset of zonal and nodal net loads, we experimentally evaluate our system components. First, we develop an interpretable and scalable forecasting model that allows for TSOs to gradually extend zonal operations to include nodal forecasts. Second, we evaluate solutions to address the heterogeneity and volatility of nodal load, subject to a trade-off. Third, our system is manageable with a fully parallelized single-model forecasting workflow. Our results show accuracy and interpretability improvements for zonal forecasts, and substantial improvements for nodal forecasts. In practice, our multi-level forecasting system allows operators to adjust forecasts with unprecedented confidence and accuracy, and to diagnose otherwise opaque errors precisely.

Figures (14)

• Figure 1: Scope of the datasets visualized with a traditional power grid structure.
• Figure 2: Architecture of hits-GAM: Trend, Seasonality and Event effects are computed for a given time and subtracted from the scaled load to yield a semi-stationary series, which is subsequently modeled via an autoregressive neural network. Concurrently, temperature forecasts are scaled and modeled with an independent neural network. Finally, all components are summed up and re-scaled to produce the forecast.
• Figure 3: Forecast origin for training and testing using a rolling window approach.
• Figure 4: Error distribution across utilities for utility-level MAE.
• Figure 5: Average utility-level MAE [MW] by hour of day.