Intraday Electricity Prices Research Articles

Multivariate probabilistic forecasting of intraday electricity prices using normalizing flows

Electricity is traded on various markets with different time horizons and regulations. Short-term intraday trading becomes increasingly important due to the higher penetration of renewables. In Germany, the intraday electricity price typically fluctuates around the day-ahead price of the European Power EXchange (EPEX) spot markets in a distinct hourly pattern. This work proposes a probabilistic modeling approach that models the intraday price difference to the day-ahead contracts. The model captures the emerging hourly pattern by considering the four 15min intervals in each day-ahead price interval as a four-dimensional joint probability distribution. The resulting nontrivial, multivariate price difference distribution is learned using a normalizing flow, i.e., a deep generative model that combines conditional multivariate density estimation and probabilistic regression. Furthermore, this work discusses the influence of different external impact factors based on literature insights and impact analysis using explainable artificial intelligence (XAI). The normalizing flow is compared to an informed selection of historical data and probabilistic forecasts using a Gaussian copula and a Gaussian regression model. Among the different models, the normalizing flow identifies the trends with the highest accuracy and has the narrowest prediction intervals. Both the XAI analysis and the empirical experiments highlight that the immediate history of the price difference realization and the increments of the day-ahead price have the most substantial impact on the price difference.

Deep distributional time series models and the probabilistic forecasting of intraday electricity prices

SummaryRecurrent neural networks (RNNs) with rich feature vectors of past values can provide accurate point forecasts for series that exhibit complex serial dependence. We propose two approaches to constructing deep time series probabilistic models based on a variant of RNN called an echo state network (ESN). The first is where the output layer of the ESN has stochastic disturbances and a Bayesian prior for regularization. The second employs the implicit copula of an ESN with Gaussian disturbances, which is a Gaussian copula process on the feature space. Combining this copula process with a nonparametrically estimated marginal distribution produces a distributional time series model. The resulting probabilistic forecasts are deep functions of the feature vector and marginally calibrated. In both approaches, Markov chain Monte Carlo methods are used to estimate the models and compute forecasts. The proposed models are suitable for the complex task of forecasting intraday electricity prices. Using data from the Australian market, we show that our deep time series models provide accurate short‐term probabilistic price forecasts, with the copula model dominating. Moreover, the models provide a flexible framework for incorporating probabilistic forecasts of electricity demand, which increases upper tail forecast accuracy from the copula model significantly.

Machine Learning Models and Intra-Daily Market Information for the Prediction of Italian Electricity Prices

In this paper we assess how intra-day electricity prices can improve the prediction of zonal day-ahead wholesale electricity prices in Italy. We consider linear autoregressive models with exogenous variables (ARX) with and without interactions among predictors, and non-parametric models taken from the machine learning literature. In particular, we implement Random Forests and support vector machines, which should automatically capture the relevant interactions among predictors. Given the large number of predictors, ARX models are also estimated using LASSO regularization, which improves predictions when regressors are many and selects the important variables. In addition to zonal intra-day prices, among the predictors we include also the official demand forecasts and wind generation expectations. Our results show that the prediction performance of the simple ARX model is mostly superior to those of machine learning models. The analysis of the relevance of exogenous variables, using variable importance measures, reveals that intra-day market information successfully contributes to the forecasting performance, although the impact differs among the estimated models.

Trading on short-term path forecasts of intraday electricity prices

We introduce a profitable trading strategy that can support decision-making in continuous intraday markets for electricity. It utilizes a novel forecasting framework, which generates prediction bands from a pool of path forecasts or approximates them using probabilistic price forecasts. The prediction bands then define a time-dependent price level that, when exceeded, indicates a good trading opportunity. Results for the German intraday market show that, in terms of the energy score, our path forecasts beat two well performing literature benchmarks by over 30%. Moreover, the forecasts provide empirical evidence that the increased computational burden induced by generating realistic price paths is offset by higher trading profits. Still, the proposed approximate and bootstrap-based methods offer a reasonable trade-off — they do not require generating path forecasts and yield only slightly lower profits.

Adaptable scheduling of smart building communities with thermal mapping and demand flexibility

This paper proposes a novel hierarchical optimization framework that couples the balancing problem of a community network with building energy management (BEM) to form an integrated model capable of capturing potential joint-flexibilities of connected buildings while co-optimizing the operation schedules of community-operated assets. The proposed framework comprises a community-level power flow-based model to solve the energy procurement problem of the community and a building-level physics-aware simulation model to estimate a set of day-ahead aggregate load scenarios with potential joint-flexibilities. At the building-level, a least-cost multi-objective optimization provides a sequence of optimal temperature setpoints for all thermal zones that is fed into an accurate Gradient Boosting Machine (GBM) to estimate heating, ventilation, and air conditioning (HVAC) load trajectories while considering human comfort, occupancy patterns, building thermal response, and intraday electricity prices. The community-level model co-optimizes day-ahead schedules of shared distributed energy resources (DERs) and electric vehicle (EV) chargings while directing the building cluster to adapt the joint-flexibilities to balance the community network under different operation conditions. Finally, a test building cluster located in Qatar University is investigated through several case studies. The simulation results reveal the model’s practicality in fully capturing the joint-flexibilities of the building cluster for an adaptable community operation which cuts the operation cost by 21.04%.

Measuring the effect of corrective short-term updates for wind energy forecasts on intraday electricity prices

Transitioning toward a decreased reliance on conventional energy sources subjects electricity markets to greater uncertainties due to the weather dependency of renewable energy sources. Wind energy in Germany has experienced a tremendous expansion, especially over the past two decades. This analysis quantifies how updates in predicted wind energy affect intraday electricity prices. It outlines the merit-order theory and derives hypotheses, testing them by means of exploratory and regression analyses. Wind-speed-based wind energy forecasts are computed to obtain forecasts at two points in time, both before the time of delivery. These two forecasts allow us to measure the effect of the short-term updates between them. The analysis finds that wind energy updates negatively affect intraday prices. The magnitude of the effect differs depending on the position within the merit order approximated by the level of residual load. It is largest for high residual loads and can be explained by the merit-order curve’s steeper shape. In situations of low residual load, the price effect and uncertainty are augmented because of negative prices and larger average forecast errors for wind energy. These results imply that varying predicted residual loads necessitate different risk assessments, for which merit-order-based models can help to anticipate and account for the magnitude of price uncertainties due to (updates for) wind energy forecasts.

The Impact of Renewable Energy Forecasts on Intraday Electricity Prices

In this paper we study the impact of errors in wind and solar power forecasts on intraday electricity prices. We develop a novel econometric model which is based on day-ahead wholesale auction curves data and errors in wind and solar power forecasts. The model shifts day-ahead supply curves to calculate intraday prices. We apply our model to the German EPEX SPOT SE data. Our model outperforms both linear and non-linear benchmarks. Our study allows us to conclude that errors in renewable energy forecasts exert a non-linear impact on intraday prices. We demonstrate that additional wind and solar power capacities induce non-linear changes in the intraday price volatility. Finally, we comment on economical and policy implications of our findings.

Intraday Electricity Pricing of Night Contracts

This paper investigates the intraday electricity pricing of 15-minute contracts in night hours. We tailor a recently introduced econometric model with fundamental impacts, which is successful in describing the pricing of day contracts. Our estimation results show that mean reversion and the positive price effect of neighboring contracts are generic features of the price formation process on the intraday market, independent of the time of day. Intraday auction prices have higher explanatory power for the pricing of night than day contracts, particularly, for the first and last 15-minute contract in a night hour. Intradaily updated forecasts of wind power infeed are the only significant fundamental factors for intraday electricity prices at night. Neither expected conventional capacities nor the slope of the merit order curve contribute to explaining price dynamics. Overall, we conclude that fundamentals lose in importance in night hours and the 15-minute intraday market is rather driven by price information.

Ensemble forecasting for intraday electricity prices: Simulating trajectories

Recent studies concerning the point electricity price forecasting have shown evidence that the hourly German Intraday Continuous Market is weak-form efficient. Therefore, we take a novel, advanced approach to the problem. A probabilistic forecasting of the hourly intraday electricity prices is performed by simulating trajectories in every trading window to receive a realistic ensemble to allow for more efficient intraday trading and redispatch. A generalized additive model is fitted to the price differences with the assumption that they follow a zero-inflated distribution, precisely a mixture of the Dirac and the Student’s t-distributions. Moreover, the mixing term is estimated using a high-dimensional logistic regression with lasso penalty. We model the expected value and volatility of the series using i.a. autoregressive and no-trade effects or load, wind and solar generation forecasts and accounting for the non-linearities in e.g. time to maturity. Both the in-sample characteristics and forecasting performance are analysed using a rolling window forecasting study. Multiple versions of the model are compared to several benchmark models and evaluated using probabilistic forecasting measures and significance tests. The study aims to forecast the price distribution in the German Intraday Continuous Market in the last 3 h of trading, but the approach allows for application to other continuous markets, especially in Europe. The results prove superiority of the mixture model over the benchmarks gaining the most from the modelling of the volatility. They also indicate that the introduction of XBID reduced the market volatility.

Intraday Electricity Pricing of Night Contracts

This paper investigates the intraday electricity pricing of 15-min. contracts in night hours. We tailor a recently introduced econometric model with fundamental impacts, which is successful in describing the pricing of day contracts. Our estimation results show that the mean reversion and the positive price impact of neighboring contracts are generic features of the price formation process on the intraday market, independent of the time of day. Intraday auction prices have higher explanatory power for the pricing of night than day contracts, particularly, for the first and last 15-min. contract in a night hour. Intradaily updated forecasts of wind power infeed are the only significant fundamental factors for intraday electricity prices at night. Neither expected conventional capacities nor the slope of the merit order curve contribute to explaining price dynamics. Overall, we conclude that fundamentals lose in importance in night hours and the 15-min. intraday market is rather driven by price information.

PCA Forecast Averaging—Predicting Day-Ahead and Intraday Electricity Prices

Recently, the development in combining point forecasts of electricity prices obtained with different length of calibration windows have provided an extremely efficient and simple tool for improving predictive accuracy. However, the proposed methods are strongly dependent on expert knowledge and may not be directly transferred from one to another model or market. Hence, we consider a novel extension and propose to use principal component analysis (PCA) to automate the procedure of averaging over a rich pool of predictions. We apply PCA to a panel of over 650 point forecasts obtained for different calibration windows length. The robustness of the approach is evaluated with three different forecasting tasks, i.e., forecasting day-ahead prices, forecasting intraday ID3 prices one day in advance, and finally very short term forecasting of ID3 prices (i.e., six hours before delivery). The empirical results are compared using the Mean Absolute Error measure and Giacomini and White test for conditional predictive ability (CPA). The results indicate that PCA averaging not only yields significantly more accurate forecasts than individual predictions but also outperforms other forecast averaging schemes.

Asymmetric information in the German intraday electricity market

This paper investigates how private and public information about unplanned power plant outages impact intraday electricity prices in Germany. It uses data from the EPEX day-ahead and continuous intraday markets as well as market messages concerning unscheduled power plant non-usabilities from the European Energy Exchange (EEX) transparency platform. The results of an econometric analysis suggest that private and public information about unplanned power plant outages have a significant positive effect on the intraday price. Furthermore, this paper shows that a reduction of the lead time on the intraday market enhances the possibilities of traders reacting to unplanned non-usabilities: an increased impact of private information on the electricity price is observed. The results also confirm an asymmetric impact of private and public information on the intraday price after the lead time reduction on the power exchange. The findings contradict the main objectives of the Regulation on Wholesale Energy Market Integrity and Transparency (REMIT), which stipulates that the possession of private information must not have an impact on electricity prices. • We analyze how private and public information impact intraday electricity prices. • We use messages from the European Energy Exchange (EEX) transparency platform. • The results suggest an asymmetric impact of information about unplanned outages. • The findings contradict the main objectives of REMIT.

Time Evolution of Hurst Exponent: Czech Wholesale Electricity Market Study

In this paper we analyse a temporal evolution of the Hurst exponent estimated on hourly returns of intraday electricity prices in the Czech Republic in 2017 and 2018. Firstly we used the log-returns with adjustments due to negative values, and secondly we employed the returns based on the area hyperbolic sine transformation. We implemented a sliding window technique in order to estimate the Hurst exponent using the Detrended Fluctuation Analysis method on subsamples with four distinct window sizes. According to the stylised facts of electricity, the spot prices and their corresponding logarithmic returns should be mean-reverting. Since the Czech intraday electricity market remains mostly unexplored, we examined this phenomenon on the intraday rather than on the spot market. Consequently, our analysis showed that the estimated values of Hurst exponent indicate a mean-reverting process for time scales greater than 24 hours and a weakly mean-reverting process for the shorter time scales. There were a few exceptions, though, since our calculations have revealed the presence of a nearly random or even weakly persistent behaviour on the shorter time scales.

Econometric modelling and forecasting of intraday electricity prices

In the following paper, we analyse the ID3-Price in the German Intraday Continuous electricity market using an econometric time series model. A multivariate approach is conducted for hourly and quarter-hourly products separately. We estimate the model using lasso and elastic net techniques and perform an out-of-sample, very short-term forecasting study. The model's performance is compared with benchmark models and is discussed in detail. Forecasting results provide new insights to the German Intraday Continuous electricity market regarding its efficiency and to the ID3-Price behaviour.

Common Unobserved Determinants of Intraday Electricity Prices

This paper employs multilevel factor modelling techniques to unravel systematic unobserved determinants of the intraday and interzonal price curve dynamics for the Pennsylvania-New Jersey-Maryland (PJM) interconnection. These techniques make an explicit separation of global drivers from region-specific common factors, thereby facilitating the identification of the actual sources of co-variability. Our empirical findings confirm the hypothesis that the common unobserved determinants of power prices in the PJM interconnection obey a block structure, some of which affect different segments of our panel. We argue that a multilevel factor approach offers a more systematic and transparent representation of intertemporal and cross-sectional patterns in PJM electricity prices compared to alternative brute-force VARMAX parametrizations and the single-level factor models, which are often put forward in the literature as viable modelling alternatives.

A long short term memory application on the Turkish intraday electricity price forecasting

Purpose- This paper aims to forecast the Turkish intraday electricity prices accurately. It will be the first intraday electricity price forecasting work, which uses Long-Short Term Memory (LSTM) application. Methodology- LSTM method is based on a special kind of neural network, which is capable of learning long-term dependencies. This paper aims to achieve the best forecasts, in terms of Mean Absolute Error (MAE) and Root Mean Square Error (RMSE), by applying the LSTM model with multistep-ahead prediction approach. Findings- LSTM model created in this study performed better with lagged values, electricity consumption and electricity production values. Especially using the lagged values of the prices and the reserve margin gave successful results. Conclusion- The proposed method has improvement in the accuracy of forecasting. Turkish Intraday Electricity Market needs further research with time series methods as well as other neural network models

The effect of wind and solar power forecasts on day-ahead and intraday electricity prices in Germany

This study analyzes the effects of wind and solar power generation forecasts on electricity prices. Converse to the existing empirical literature in this area, we apply a panel data analysis to control for endogeneity due to unobserved heterogeneity. We use a dataset with 24 daily observations of day-ahead and intraday prices from 2010 to 2016, and we apply a fixed effects regression under consideration of robust Driscoll-Kraay standard errors. A noteworthy element of the regression model is the simulation-based design of a variable indicating the power generation technology that is price-determining at a certain point in time. In this context, we differentiate between the fuel types coal, gas, and others, to model the nonlinear price behavior for a varying demand. For 2016, we find price dampening effects of both wind and solar power of approximately 0.6 €/MWh per additional GWh of feed-in. Along with the rapidly increasing shares of wind and solar power of the total power generation during the last years, their price dampening effect has declined since 2013, due to a drop in fuel prices. Another finding is that a reduction in forecasting errors on the power generation from wind and solar, and smoothing of the cyclical demand would lead to a decreased price volatility.

Common Unobserved Determinants of Intraday Electricity Prices

Electricity prices are known to have special statistical properties, where long-memory effects, varying-level periodicities, cross-dependencies and other exogenous influences often work together to form a complex data-generating process. The need for accurately representing all these features typically leads us to complex parametrizations, which are prone to numerical instabilities, overfitting and poor transparency. A natural counterpart to an overparameterized model is accomplished by the use of factor decomposition techniques to simplify the modeling process by implementing a more compact factor structure for the cross-section of electricity prices. In this paper, we exploit multi-level factor modelling techniques to unravel systematic unobserved determinants of the intraday and interzonal price curve dynamics. These techniques make an explicit separation of global drivers from region-specific common factors, thereby facilitating the identification of the actual sources of co-variability. Our findings on the Pennsylvania-New Jersey-Maryland (PJM) interconnection confirm the hypothesis that the common unobserved determinants of power prices in this interconnection obey a block structure, some of which affect different segments of our panel. We argue that a multi-level factor approach offers a more systematic and transparent representation of intertemporal and cross-sectional patterns in PJM electricity prices compared to alternative brute-force VARMAX parametrizations and the ``flat factor models, which are often put forward in the literature as viable modelling alternatives.

Short-term dynamics of day-ahead and intraday electricity prices

PurposeSince the liberalization of electricity markets in the European Union, prices are subject to market dynamics. Hence, understanding the short-term drivers of electricity prices is of major interest to electricity companies and policymakers. Accordingly, this paper aims to study movements of prices in the combined German and Austrian electricity market.Design/methodology/approachThis paper estimates an autoregressive model with exogenous variables (ARX) in a two-step procedure. In the first step, both time series, which inherently feature seasonality, are de-seasonalized, and in the second step, the influence of all model variables on the two dependent variables, i.e. the day-ahead and intraday European Power Energy Exchange prices, is measured.FindingsThe results reveal that the short-term market is largely driven by seasonality, consumer demand and short-term feed-ins from renewable energy sources. As a contribution to the existing body of literature, this paper specifically compares the price movements in day-ahead and intraday markets. In intraday markets, the influences of renewable energies are much stronger than in day-ahead markets, i.e. by 24.12 per cent for wind and 116.82 per cent for solar infeeds.Originality/valueKnowledge on the price setting mechanism in the intraday market is particularly scarce. This paper contributes to existing research on this topic by deriving drivers in the intraday market and then contrasting them to the day-ahead market. A more thorough understanding is especially crucial for all stakeholders, who can use this knowledge to optimize their bidding strategies. Furthermore, the findings suggest policy implications for a more stable and efficient electricity market.

Modeling and forecasting realized volatility in German–Austrian continuous intraday electricity prices

AbstractThis paper uses high‐frequency continuous intraday electricity price data from the EPEX market to estimate and forecast realized volatility. Three different jump tests are used to break down the variation into jump and continuous components using quadratic variation theory. Several heterogeneous autoregressive models are then estimated for the logarithmic and standard deviation transformations. Generalized autoregressive conditional heteroskedasticity (GARCH) structures are included in the error terms of the models when evidence of conditional heteroskedasticity is found. Model selection is based on various out‐of‐sample criteria. Results show that decomposition of realized volatility is important for forecasting and that the decision whether to include GARCH‐type innovations might depend on the transformation selected. Finally, results are sensitive to the jump test used in the case of the standard deviation transformation.