Probabilistic Dynamic Line Rating with Line Graph Convolutional LSTM

TL;DR

The paper tackles underutilization and risk in power systems caused by static line ratings by introducing a network-wide probabilistic DLR forecasting framework. It develops LGCLSTM, which integrates line-graph convolutional networks with LSTM to capture extended spatial and temporal dependencies and produces quantile forecasts for DLR. Case studies on the Texas 123-bus system show LGCLSTM outperforms baselines in forecast reliability and sharpness and yields lower operational costs and renewable curtailment compared with SLR, across day-ahead scheduling and real-time redispatch. The approach enables risk-aware, cost-efficient grid operation under increasing renewable variability, with potential for integration into more advanced market and security-constrained planning.

Abstract

Dynamic line rating (DLR) is an effective approach to enhancing the utilization of existing transmission line infrastructure by adapting line ratings according to real-time weather conditions. Accurate DLR forecasts are essential for grid operators to effectively schedule generation, manage transmission congestion, and lower operating costs. As renewable generation becomes increasingly variable and weather-dependent, accurate DLR forecasts are also crucial for improving renewable utilization and reducing curtailment during congested periods. Deterministic forecasts, however, often inadequately represent actual line capacities under uncertain weather conditions, leading to operational risks and costly real-time adjustments. To overcome these limitations, we propose a novel network-wide probabilistic DLR forecasting model that leverages both spatial and temporal information, significantly reducing the operational risks and inefficiencies inherent in deterministic methods. Case studies on a synthetic Texas 123-bus system demonstrate that the proposed method not only enhances grid reliability by effectively capturing true DLR values, but also substantially reduces operational costs.

Figures (8)

• Figure 1: Example of probabilistic and deterministic DLR forecasting. Deterministic forecasting inevitable contains errors, while probabilistic forecasts provide a range of possible line ratings.
• Figure 2: Overall framework for probabilistic DLR forecasting.
• Figure 3: An example of reliability and sharpness. Low reliability ensures the prediction intervals consistently capture the true rating, while low sharpness keeps those intervals narrow enough to enable precise capacity utilization in grid operations.
• Figure 4: Heat maps of the average QS for each line in 2021 across TX-123BT. The gray dotted arrow points the line #109 where the highest difference in QS between QRF to LGCLSTM is observed.
• Figure 5: Probabilistic DLR forecasting for line #109.