Analysing the impact of renewables on Iberian wholesale electricity market prices using machine learning techniques

Cristina Ballester; Dolores Furió; Cristina Ballester; Dolores Furió

doi:10.3934/GF.2024014

Green Finance

2024, Volume 6, Issue 2: 363-382. doi: 10.3934/GF.2024014

Previous Article Next Article

Research article

Analysing the impact of renewables on Iberian wholesale electricity market prices using machine learning techniques

Cristina Ballester ¹,
Dolores Furió ^{2
,
,}

1.
Department of Analytics, Bionline, S.L. c/ Catedrático Agustín Escardino, 9, 46980 Paterna, Spain
2.
Department of Financial Economics, Universitat de València. s/n, Avinguda dels Tarongers 46022 Valencia, Spain

Received: 20 February 2024 Revised: 06 June 2024 Accepted: 18 June 2024 Published: 24 June 2024
JEL Codes: Q41, C69

The integration of renewable generation sources into wholesale electricity markets is expected to reduce day-ahead marginal prices. This effect has been widely evidenced by previous literature and is commonly referred to as the merit order effect. However, the factors influencing the components of final prices, other than the day-ahead market price, have not been subjected to as much study. Nevertheless, they may prove crucial in understanding the dynamics between the interrelated trading segments in the wholesale electricity market. Furthermore, in the context of the energy transition process, the penetration of intermittent renewable energy sources (mainly wind and solar photovoltaic) and the non-storability of electricity at a large scale may result in increased market balancing needs and costs. The objective of this study was to identify the primary drivers of final wholesale electricity prices in the Iberian electricity market, apart from the day-ahead market price, using machine learning techniques. The results indicate that the share of renewable generation in the day-ahead market is a significant factor influencing both the cost of managing technical constraints, which aims to address network capacity issues, and the cost of managing balancing processes and resolving adjustment issues by the TSO. However, both of these costs can be readily accommodated by the market, as they represent a minimal percentage of the final price. These findings are of interest to both practitioners and regulators, as they provide a better understanding of the functioning of the market and have implications for the restructuring of the market towards a more sustainable and competitive electricity system.
- renewable energy,
- network congestions,
- balancing markets,
- machine learning
Citation: Cristina Ballester, Dolores Furió. Analysing the impact of renewables on Iberian wholesale electricity market prices using machine learning techniques[J]. Green Finance, 2024, 6(2): 363-382. doi: 10.3934/GF.2024014

Related Papers:

Abstract

The integration of renewable generation sources into wholesale electricity markets is expected to reduce day-ahead marginal prices. This effect has been widely evidenced by previous literature and is commonly referred to as the merit order effect. However, the factors influencing the components of final prices, other than the day-ahead market price, have not been subjected to as much study. Nevertheless, they may prove crucial in understanding the dynamics between the interrelated trading segments in the wholesale electricity market. Furthermore, in the context of the energy transition process, the penetration of intermittent renewable energy sources (mainly wind and solar photovoltaic) and the non-storability of electricity at a large scale may result in increased market balancing needs and costs. The objective of this study was to identify the primary drivers of final wholesale electricity prices in the Iberian electricity market, apart from the day-ahead market price, using machine learning techniques. The results indicate that the share of renewable generation in the day-ahead market is a significant factor influencing both the cost of managing technical constraints, which aims to address network capacity issues, and the cost of managing balancing processes and resolving adjustment issues by the TSO. However, both of these costs can be readily accommodated by the market, as they represent a minimal percentage of the final price. These findings are of interest to both practitioners and regulators, as they provide a better understanding of the functioning of the market and have implications for the restructuring of the market towards a more sustainable and competitive electricity system.

References

[1]	Apley D (2018) ALEPlot: Accumulated Local Effects (ALE) Plots and Partial Dependence (PD) Plots. R package version 1.1. Available from: https://CRAN.R-project.org/package = ALEPlot.
[2]	Athey S, Stefan W (2019) Estimating Treatment Effects with Causal Forests: An Application. Obs Stud 5: 37–51, Crossref. https://doi.org/10.1353/obs.2019.0001. doi: 10.1353/obs.2019.0001
[3]	Ballester C, Furió D (2015) Effects of renewables on the stylized facts of electricity prices. Renew Sust Energ Rev 52: 1596–1609. https://doi.org/10.1016/j.rser.2015.07.168 doi: 10.1016/j.rser.2015.07.168
[4]	Breiman L (2001a) Statistical Modeling: The Two Cultures. Stat Sci 16: 199–231. https://doi.org/10.1214/ss/1009213726 doi: 10.1214/ss/1009213726
[5]	Breiman L (2001b) Random Forests. Mach Learn 45: 5–32. https://doi.org/10.1023/A:1010933404324 doi: 10.1023/A:1010933404324
[6]	Carvalho N, Pereira P (2019) The 'merit order effect' of wind and solar power. Volatility and determinants. Renew Sust Energ Rev 102: 54–62. https://doi.org/10.1016/j.rser.2018.11.042
[7]	Chen T, He T, Benesty M et al. (2023). xgboost: Extreme Gradient Boosting. R package version 0.71.1. Available from: https://CRAN.R-project.org/package = xgboost.
[8]	Chuliá H, Furió M, Uribe JM (2019) Volatility spillovers in Energy Markets. Energy J 40: 173–198.
[9]	Credit K, Lehnert MA (2023) A structured comparison of causal machine learning methods to assess heterogeneous treatment effects in spatial data. J Geogr Syst 2023: 1–28. https://doi.org/10.1007/s10109-023-00413-0 doi: 10.1007/s10109-023-00413-0
[10]	Debeer D, Strobl C (2020) Conditional permutation importance revisited. BMC Brief 21: 307. https://doi.org/10.1186/s12859-020-03622-2 doi: 10.1186/s12859-020-03622-2
[11]	Debeer D, Hothorn T, Strobl C (2021) permimp: Conditional Permutation Importance. R package version 1.0–2. Available from: https://urldefense.com/v3/__https://CRAN.R-project.org/package = permimp__; !!D9dNQwwGXtA!Si7ijZba0umEBlLcxRbIt09Pg0rMLgl8N2cQjJsbKN4SiuCSUk6A2q1ERjzqJ2NGI7EXDjeteO31jpDeBlWI8MxbE7fPIvM$.
[12]	De Lagarde MC, Lantz F (2018) How renewable production depresses electricity prices: Evidence from the German market. Energ Policy 117: 263–277. https://doi.org/10.1016/j.enpol.2018.02.048 doi: 10.1016/j.enpol.2018.02.048
[13]	Duras T, Javed F, Mansson K, et al. (2023) Using machine learning to select variables in data envelopment analysis: Simulations and application using electricity distribution data. Energ Econ 120: 106621. https://doi.org/10.1016/j.eneco.2023.106621. doi: 10.1016/j.eneco.2023.106621
[14]	Elamin OA (2023) The causal effect of informal job search on wage and job satisfaction: evidence from Egypt and Jordan using random forest method. Int J Soc Econ 50: 522–536. https://doi.org/10.1108/IJSE-05-2022-0318 doi: 10.1108/IJSE-05-2022-0318
[15]	Furió D, Lucia JJ (2009) Congestion management rules and trading strategies in the Spanish electricity market. Energ Econ 31: 48–60. https://doi.org/10.1016/j.eneco.2008.07.004 doi: 10.1016/j.eneco.2008.07.004
[16]	Gianfreda A, Parisio L, Pelagatti M (2018) A review of balancing costs in Italy before and after RES introduction. Renew Sust Energ Rev 91: 549–563. https://doi.org/10.1016/j.rser.2018.04.009 doi: 10.1016/j.rser.2018.04.009
[17]	Hastie T, Tibshirani R, Friedman J (2001) The Elements of Statistical Learning: Data Mining, Inference, and Prediction. Second Edition. Springer Series in Statistics Ed. Springer
[18]	Holttinen H (2004) The Impact of large scale Wind Power Production on the Nordic Electricity System. Dissertation for the title of Doctor of Science in Technology, Helsinki University of Technology.
[19]	Li S, Pu Z, Cui Z, et al. (2024) Inferring Heterogeneous Treatment Effects of Crashes on Highway Traffic: A Doubly Robust Causal Machine Learning Approach. Transport Res C-Emer 160: 104537. https://doi.org/10.1016/j.trc.2024.104537 doi: 10.1016/j.trc.2024.104537
[20]	Liaw A, Wiener M (2002) Classification and Regression by randomForest. R News 2: 18–22.
[21]	McConnell D, Hearps P, Eales D, et al. (2013) Retrospective modelling of the merit-order effect on wholesale electricity prices from distributed photovoltaic generation in the Australian National Electricity Market. Energ Policy 58: 17–27. https://doi.org/10.1016/j.enpol.2013.01.052 doi: 10.1016/j.enpol.2013.01.052
[22]	Mizuguchi T, Sawamura S (2023) Machine learning-based causal models for predicting the response of individual patients to dexamethasone treatment as prophylactic antiemetic. Sci Rep 13: 1–10. https://doi.org/10.1038/s41598-023-34505-0 doi: 10.1038/s41598-023-34505-0
[23]	Prasanna A, Holzhauer S, Krebs F (2019) Overview of machine learning and data-driven methods in agent-based modeling of energy markets. In: David, K., Geihs, K., Lange, M. & Stumme, G. (Hrsg.), INFORMATIK 2019: 50 Jahre Gesellschaft für Informatik – Informatik für Gesellschaft. Bonn: Gesellschaft fürInformatike.V. (S. 571–584). https://doi.org/10.18420/inf2019_73
[24]	Qays O, Buswig Y, Hossain L, et al. (2020) Active Charge Balancing Strategy Using the State of Charge Estimation Technique for a PV-Battery Hybrid System. Energies 13: 3434. https://doi.org/10.3390/en13133434 doi: 10.3390/en13133434
[25]	Quigley DT, Che Y, Yasar M, et al. (2023) Cover Crop Adoption and Climate Risks: An Application of Causal Random Forests. 2023 Annual Meeting, July 23–25, Agricultural and Applied Economics Association, Washington D.C.
[26]	Robette N (2022) moreparty: A Toolbox for Conditional Inference Trees and Random Forests. R package version. Available from: https://urldefense.com/v3/__https://CRAN.R-project.org/package = moreparty__; !!D9dNQwwGXtA!Si7ijZba0umEBlLcxRbIt09Pg0rMLgl8N2cQjJsbKN4SiuCSUk6A2q1ERjzqJ2NGI7EXDjeteO31jpDeBlWI8MxbJKa66QA$.
[27]	Saénz de Miera G, Del Río P, Vizcaíno I (2008) Analysing the impact of renewable electricity support schemes on power prices: The case of wind electricity in Spain. Energ Policy 36: 3345–3359. https://doi.org/10.1016/j.enpol.2008.04.022 doi: 10.1016/j.enpol.2008.04.022
[28]	Schnürch S, Wagner A (2019) Machine Learning on EPEX Order Books: Insights and Forecasts. arXiv preprint. https://doi.org/10.48550/arXiv.1906.06248
[29]	Sensfuβ F, Ragwitz M, Genoese M (2008) The merit-order effect: A detailed analysis of the price effect of renewable electricity generation on spot market prices in Germany. Energy Policy 36: 3086–3094. https://doi.org/10.1016/j.enpol.2008.03.035 doi: 10.1016/j.enpol.2008.03.035
[30]	Strobl C, Boulesteix AL, Kneib T, et al. (2008) Conditional variable importance for random forests. BMC Bioinformatics 9: 307. http://doi.org/10.1186/1471-2105-9-307 doi: 10.1186/1471-2105-9-307
[31]	Therneau T, Atkinson B (2018) rpart: Recursive Partitioning and Regression Trees. R package version 4.1–1. Available from: https://CRAN.R-project.org/package = rpart.
[32]	Tschora L, Erwan P, Plantevit M, et al. (2022) Electricity price forecasting on the day-ahead market using machine learning. Appl Energ 313: 118752. https://doi.org/10.1016/j.apenergy.2022.118752 doi: 10.1016/j.apenergy.2022.118752
[33]	Würzburg K, Labandeira X, Linares P (2013) Renewable generation and electricity prices: Taking stock and new evidence for Germany and Austria. Energ Econ 40: S159–S171. https://doi.org/10.1016/j.eneco.2013.09.011 doi: 10.1016/j.eneco.2013.09.011
[34]	Xu X, Ye T, Gao J, et al. (2024) The effect of green, supply chain factors in predicting China's stock price crash risk: evidence from random forest model. Environ Dev Sustain 2024: 1–24. https://doi.org/10.1007/s10668-023-04300-y doi: 10.1007/s10668-023-04300-y
[35]	Zhang Y, Li H, Gang R (2022) Estimating heterogeneous treatment effects in road safety analysis using generalized random forests. Accident Anal Prev 165: 106507. https://doi.org/10.1016/j.aap.2021.106507 doi: 10.1016/j.aap.2021.106507

Reader Comments

Your name:*

Email:*
© 2024 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)