Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Leveraging Asynchronous Cross-border Market Data for Improved Day-Ahead Electricity Price Forecasting in European Markets

Maria Margarida Mascarenhas, Jilles De Blauwe, Mikael Amelin, Hussain Kazmi

TL;DR

The paper investigates whether asynchronously published cross-border day-ahead electricity prices can improve forecasting in European markets with later gate closures. Using a LEAR and a DNN ensemble, it tests price data from DE-LU, AT, and CH to predict BE and SE3 prices, examining calibration windows, recalibration frequency, and interpretability. The results show substantial accuracy gains (up to ~22% MAE improvement for BE and ~9% for SE3) when incorporating cross-border data, with gains most pronounced near GCT and during extremes, albeit with market- and scenario-dependent variability. The study highlights important trade-offs between forecast accuracy, computation time, and reliance on cross-border information, and provides guidance for market participants and regulators on leveraging cross-border data in interconnected European energy markets.

Abstract

Accurate short-term electricity price forecasting is crucial for strategically scheduling demand and generation bids in day-ahead markets. While data-driven techniques have shown considerable prowess in achieving high forecast accuracy in recent years, they rely heavily on the quality of input covariates. In this paper, we investigate whether asynchronously published prices as a result of differing gate closure times (GCTs) in some bidding zones can improve forecasting accuracy in other markets with later GCTs. Using a state-of-the-art ensemble of models, we show significant improvements of 22% and 9% in forecast accuracy in the Belgian (BE) and Swedish bidding zones (SE3) respectively, when including price data from interconnected markets with earlier GCT (Germany-Luxembourg, Austria, and Switzerland). This improvement holds for both general as well as extreme market conditions. Our analysis also yields further important insights: frequent model recalibration is necessary for maximum accuracy but comes at substantial additional computational costs, and using data from more markets does not always lead to better performance - a fact we delve deeper into with interpretability analysis of the forecast models. Overall, these findings provide valuable guidance for market participants and decision-makers aiming to optimize bidding strategies within increasingly interconnected and volatile European energy markets.

Leveraging Asynchronous Cross-border Market Data for Improved Day-Ahead Electricity Price Forecasting in European Markets

TL;DR

The paper investigates whether asynchronously published cross-border day-ahead electricity prices can improve forecasting in European markets with later gate closures. Using a LEAR and a DNN ensemble, it tests price data from DE-LU, AT, and CH to predict BE and SE3 prices, examining calibration windows, recalibration frequency, and interpretability. The results show substantial accuracy gains (up to ~22% MAE improvement for BE and ~9% for SE3) when incorporating cross-border data, with gains most pronounced near GCT and during extremes, albeit with market- and scenario-dependent variability. The study highlights important trade-offs between forecast accuracy, computation time, and reliance on cross-border information, and provides guidance for market participants and regulators on leveraging cross-border data in interconnected European energy markets.

Abstract

Accurate short-term electricity price forecasting is crucial for strategically scheduling demand and generation bids in day-ahead markets. While data-driven techniques have shown considerable prowess in achieving high forecast accuracy in recent years, they rely heavily on the quality of input covariates. In this paper, we investigate whether asynchronously published prices as a result of differing gate closure times (GCTs) in some bidding zones can improve forecasting accuracy in other markets with later GCTs. Using a state-of-the-art ensemble of models, we show significant improvements of 22% and 9% in forecast accuracy in the Belgian (BE) and Swedish bidding zones (SE3) respectively, when including price data from interconnected markets with earlier GCT (Germany-Luxembourg, Austria, and Switzerland). This improvement holds for both general as well as extreme market conditions. Our analysis also yields further important insights: frequent model recalibration is necessary for maximum accuracy but comes at substantial additional computational costs, and using data from more markets does not always lead to better performance - a fact we delve deeper into with interpretability analysis of the forecast models. Overall, these findings provide valuable guidance for market participants and decision-makers aiming to optimize bidding strategies within increasingly interconnected and volatile European energy markets.

Paper Structure

This paper contains 32 sections, 12 equations, 8 figures, 9 tables.

Table of Contents

  1. Introduction
  2. The Electricity Spot Market
  3. Electricity Price Forecasting
  4. Contributions and Paper Structure
  5. Methodology and Data
  6. Forecasting models
  7. The LEAR Model
  8. The DNN Model
  9. Model Configurations
  10. Choice of input Covariates
  11. Choice of calibration Window length
  12. Recalibration Frequency
  13. Ensembling
  14. Model Evaluation
  15. Accuracy
  16. ...and 17 more sections

Figures (8)

  • Figure 1: Gate Closure Times across Selected European Markets (Non-Exhaustive Overview: EPEX SPOT, Nord Pool, OMIE, SEMOpx, EXAA).
  • Figure 2: The results of the one-sided Giacomini-White tests for the predictions of the BE prices, made with different sets of market data.
  • Figure 3: The results of the one-sided Giacomini-White Tests for the predictions of the SE3 prices, using different sets of market data.
  • Figure 4: Evolution of forecasting accuracy (MAE) for BE and SE3 DA market prices as additional market price information becomes available at respective GCT.
  • Figure 5: The average runtime (minutes) to train and generate predictions across different CWs and RFs with the (a) LEAR AT/DE-LU/CH and (b) LEAR DE-LU ensembles models for the Belgian and SE3 zones, respectively
  • ...and 3 more figures