Electricity Prices Research Articles

Retraction Note: A hybrid Elman recurrent neural network, group search optimization, and refined VMD-based framework for multi-step ahead electricity price forecasting

Retraction Note: A hybrid Elman recurrent neural network, group search optimization, and refined VMD-based framework for multi-step ahead electricity price forecasting

Economic Scheduling Strategy for Multi-Energy-Integrated Highway Service Centers Considering Carbon Trading and Critical Peak Pricing Mechanism

This study proposes an optimized economic scheduling strategy for multi-energy-integrated highway service centers (MEIHSCs) within a 24 h operational timeframe. With the imperative of carbon peaking and carbon neutrality, highway areas are increasingly incorporating renewable energy systems, such as photovoltaic arrays, to capitalize on abundant resources along highways. Considering the diverse load demands of new energy vehicles and the mismatch between energy supply and demand on the highway, MEIHSCs must adapt to these trends by establishing integrated networks for electricity, natural gas, and hydrogen refueling. However, there is a lack of coordination between equipment switching and the phases of low electricity prices and peak renewable energy periods. To address this challenge and improve economic efficiency, this study proposes an economic dispatch strategy that combines economic incentives based on carbon trading and critical peak pricing mechanisms. This strategy aims to maximize economic benefits while fully meeting the load demands of new energy vehicles. Case studies indicate that operating costs are reduced by 28.04% compared to strategies without new energy installations, and by 47.85% compared to strategies without optimization. The results demonstrate that this integrated and optimized strategy significantly reduces energy costs and enhances economic benefits in highway service centers.

Benefit-Loss Analysis of Electricity Generation in Indonesia: Life Cycle Assessment and Economic Input-Output Analysis Application

This paper presents a benefit-loss analysis of Indonesia's electricity generation from 2014 to 2018. The loss and benefits are expressed as Disability-Adjusted Life Years (DALY) per capita. In this paper, the benefit is the contribution of value-added triggered by electricity demand to GDP per capita. The value-added is calculated using the economic input-output model. The relationship between GDP per capita with DALY per capita is derived to estimate the reduction in the value of DALY per capita. The loss is the health damage of pollutants produced by electricity generation and is calculated using Life Cycle Assessment. The rise in DALY caused by the pollutants is quantified using Eco-Indicator 99. The results show that the average benefit-loss ratio is 1.257. In Indonesia, the average rate of change of the benefit with respect to electricity consumption per capita is lower than the rate of change of the loss. The analysis also shows a positive relationship between the rate of change of the benefit with the rate of change of electricity price. Promoting renewable energy and improving efficiency are ways to increase the benefit-loss ratio. The results also suggest that increasing the price is a lucrative alternative to increasing the benefit-loss ratio.

Techno-Economic and CO2 Emissions Analysis of the Molten Carbonate Fuel Cell Integration in a DRI Production Plant for the Decarbonization of the Steel Industry

Developing low-carbon systems for the steel industry is increasingly considered necessary for contributing to the targets of CO2 emission reduction requested by modern environmental policies. This work focuses on a process for decarbonizing primary steel production by integrating an MCFC system in a DRI production plant based on the Energiron ZR process configuration. An MCFC system can be used to remove CO2 from the flue gases while producing electricity and reducing the net electric consumption of the DRI plant. The reference scenario, the carbon capture case, and an additional hydrogen-based H-DRI system using high temperature electrolysis are simulated using Aspen Plus to evaluate and compare their energy and environmental performance. The results show that even considering the additional power consumption of the complete carbon capture and separation system, the overall electrical consumption of the carbon capture case is decreased by 78 % and the direct CO2 emissions by up to 95 %. Furthermore, reduced electricity consumption reduces scope 2 emissions, increasing the sustainability of the process in the steel industry. On the other hand, the H-DRI case decreases primary energy consumption by 24 % but significantly increase electricity requirement; thus, it represents a better future solution when cheap low-CO2 electricity is available. The plant configurations are compared economically by retrieving investment and operational costs from the open literature to estimate the levelized cost of DRI and the cost of CO2 avoided. The results show that installing the MCFC and the anode purification system increases the investment costs by 38 % compared to a conventional plant. However, the lower electricity consumption and carbon tax expenditure lead to a comparable final DRI product cost. Contrarily, due to the electrolyzer high investment and operational costs, the additional marginal DRI cost of the H-DRI case is 31 % higher. Finally, the sensitivity analysis on the main variable costs shows that, independently from the electricity and natural gas price, the carbon capture case is economically competitive to the reference scenario, considering a carbon tax of 60 €/tCO2 or higher. At the same time, the H-DRI solution requires low electricity prices to be competitive. In conclusion, even if, in the long term, the objective is to completely replace the use of fossil fuels with renewable energy, in the short term, the implementation of MCFC in the DRI plant can significantly reduce the environmental impact of primary steel production without significant economic penalization.

Operation optimization in large-scale heat pump systems: A scheduling framework integrating digital twin modelling, demand forecasting, and MILP

The integration of large-scale heat pumps and thermal energy storage can facilitate sector coupling, potentially lowering heating and cooling costs in industries and buildings. This cost reduction can be extended by optimizing the utilization of the available thermal energy storage capacity in accordance to fluctuating electricity prices. Although the literature offers methods for optimizing the operation of these integrated systems, they often overlook the impact of heat pump performance degradation over time, such as from fouling. This oversight can lead to suboptimal system performance and inaccurate operational cost estimates. The present study addresses this gap by introducing a novel operational scheduling framework that aimed to reduce the operational costs of a commercial large-scale heat pump system. The system comprised an open cooling tower, a thermal storage tank and two heat pumps affected by fouling. The framework incorporated a mixed-integer linear programming (MILP) model, thermal demand forecasting, and heat pump performance maps that account for varying fouling levels. These maps were obtained from online calibrated simulation models used as digital twins of the heat pumps. The results demonstrated that the proposed framework enhanced the thermal energy storage utilization in response to variable electricity prices and adjusted the heat pump operation based on the influence of fouling. This resulted in a reduction of operational costs of up to 5% compared to the conventional operation of the system. These savings were observed to vary depending on the forecasting accuracy and the prevailing fouling levels. Overall, this study demonstrates the potential of using the proposed framework for cost reduction in large-scale heat pump systems.

Changes in hourly electricity consumption profiles and price elasticities in Denmark 2019–2022

Changes in consumption profiles affect the need for investments in production and grid capacities. Extending a model with categorical variables for hours of the day, months and types of days to include short-term electricity prices, and applying Danish data for the hourly electricity consumption by aggregated categories of customers, consumption profiles for each of the years from 2019 to 2022 are estimated and it is tested if the profiles change significantly. Further, the short-term price elasticity of hourly electricity demand is estimated. During the first COVID-19 lockdown, residential electricity consumption increased and non-residential consumption decreased and the residential consumption profile on weekdays resembles the weekend profile. However, long-term effects are quite limited. For 2022, the residential consumption profile is significantly different from the profile in 2019. The profile has become flatter, increasing consumption during the night and reducing consumption during the daytime. An increasing number of electric vehicles and individual heat pumps contribute to this change. The short-term price elasticity for total electricity consumption increased in 2022. Especially, the price elasticity for residential customers increased and became significant in the second half of 2022 where electricity prices were high and volatile and customers were aware of prices.

Computational Forecasting of Power Prices Using Artificial Neural Networks

In the restructured power markets, the primary responsibility is setting the price of electricity. Thus, it has become increasingly important to accurately and precisely forecast power prices. An Artificial Neural Network (ANN) that was developed specifically for temporary price prediction in restructured electricity markets is presented in this paper. An input level, two hidden layers, and output layer comprise the four levels of the suggested ANN model, which is a perceptron neural network. Instead of using traditional back propagation for ANN training, using Levenberg-Marquardt retrogression (LMBP) methodology is used to accelerate convergence. The performance and efficacy of the suggested ANN model may be shown by training it on the Ontario power market. MATLAB is used to train the model.

Deep learning-based electricity price forecasting: Findings on price predictability and European electricity markets

Deep learning algorithms have transformed day-ahead electricity price predictions in recent years. This phenomenon has implications for price predictability and perceived risks, which in turn could impact investments. This study aims to unravel the impact of deep learning-based price forecasting (DPF) on price predictability and its correlation with price volatility. This has been done by examining 19 bidding zones across 11 European countries from January 2018 to June 2023. This timeframe captures the unique market dynamics influenced by the COVID-19 pandemic and the Global Energy Crisis. This analysis has been enabled by the development of the Day-ahead European Electricity Price Forecasting Kit (Deepforkit), a forecasting tool developed in-house and open-sourced. Notably, Deepforkit offers a running speed at least 5 times faster than a previously suggested benchmark, enabling large-scale and multi-market analyses such as the one conducted in this study. It is quantitatively proven that if DPF is used, price volatility is just moderately correlated to unpredictability, and using it as a proxy for price unpredictability is not reliable. This is attributed to the capacity of DPFs to adjust their predictions in response to evolving market trends and patterns. This finding has an important implication on the investment risk perception and the cost of capital for new investments in the power sector. Moreover, the predictability behavior of prices is compared and contrasted in 3 markets, namely EPEX SPOT, OMIE, and Nord Pool, and across 11 European countries (Belgium, Denmark, Finland, France, Germany, Netherlands, Norway, Portugal, Sweden, Spain, and Switzerland).

Systematic Literature Review of Heuristic-Optimized Microgrids and Energy-Flexible Factories

Decentralized renewable energy generation and consumption through microgrids, coupled with short- and long-term storage systems and enhanced demand flexibility, represent a promising strategy for mitigating grid stress and reducing emissions in the industrial sector. However, transitioning into a sustainable industry often poses challenges in terms of economic feasibility. This review surveys current optimization approaches and simulation functionalities to enhance feasibility. It follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, covering 1066 studies from 2016 to 2023 across three research areas: optimal system sizing of microgrids (OSS), optimization of electrical energy distribution to storage systems and consumers (EED), and energy flexibilization of factories (EF). As a result, 24 filtered sources from these areas were analyzed. Quantitative analysis indicated that evolutionary and swarm-inspired metaheuristics are predominantly applied in OSS, whereas exact linear problem solvers are favored for EED and EF optimization. A range of functionalities is available, and approaches often prioritize individual functionalities, such as load forecasting, dynamic electricity pricing, and statistical representation of energy generation, rather than comprehensively integrating them. Furthermore, no current approach simultaneously integrates optimization and simulation models across all three research areas.

Part-load analysis and preliminary annual simulation of a constant inventory supercritical CO2 power plant for waste heat recovery in cement industry

The present work investigates the part-load performance of a MW-scale sCO2 power plant designed as waste heat recovery unit for an existing cement plant located in Czech Republic, in the framework of the H2020 funded project CO2OLHEAT. The study first presents the selected power plant configuration and then focuses on the evaluation of its part-load operation due to variation of flue gas mass flow rate and temperature. The range of flue gas conditions at the outlet of the upstream process is retrieved from a preliminary statistical analysis of historical trends obtained through the cement plant monitoring. The numerical model developed for this study aims at providing realistic results thanks to the adoption of turbomachinery performance maps provided by the turbomachinery manufacturer of the project. Moreover, heat exchangers have been modelled through a discretized approach which has been validated against manufacturer data, while piping inventory and pressure losses have been assessed through a preliminary sizing that considers the actual distances to be covered in the cement plant. Performance decay is estimated for the whole range of flue gas conditions, reporting the most significant power cycle parameters, and identifying the main causes of efficiency loss. The part-load analysis is carried out considering a constant CO2 inventory, in order to reduce the system complexity and capital cost and simplify plant operation. Results show that the operation entails minor variation of the compressors operative points in the whole range of operating conditions of the cement plant, avoiding the risk of anti-surge bypass activation. Moreover, the plant is able to work close to the nominal thermodynamic cycle efficiency (20.5 %–23.0 %) for most of the year and benefits from part-load operation in terms of overall performance. In the last part of the work, a preliminary techno-economic analysis of the plant is also presented to highlight the potential advantages of sCO2 technology for waste heat recovery applications. The results of the part-load performance of the plant are combined with the flue gases data obtained from the preliminary statistical analysis and the cement plant historical monitoring. An annual electricity production equal to 13′909.7 MWh is obtained, corresponding to 6560 equivalent hours and a system capacity factor of 74.9 %. The investment cost of each CO2OLHEAT plant component is estimated by means of cost correlations obtained from literature and the non-discounted payback time is computed as a function of the electricity selling price. The results show that, even considering electricity prices before 2022, the payback time of the CO2OLHEAT plant is estimated to be lower than 8 years, justifying the industrial interest in the proposed technology.

Decision-making solutions based artificial intelligence and hybrid software for optimal sizing and energy management in a smart grid system

This paper describes a decentralised smart grid system containing renewable energies, storage systems and distributed generation with human control and intervention. The importance of each element and the interaction between them leads to think about a decision-making strategy. In fact, the integration of a Photovoltaic Panel (PVP) is used due to its availability and its participation in the carbon emissions reduction. Also, a battery is required to fill a power gap or absorb extra generated energy. Moreover, an optimal sizing is needed to get an efficient system with minimum cost. Also, an energy management strategy (EMS) is essential to ensure the power resources scheduling in order to keep a continuous equilibrium supply-demand of electricity and avoid instabilities in the grid, with guaranteeing a minimum cost of electricity. In the first part, the proposed smart grid optimal sizing is determined under real weather data (solar radiation) of the city of Sousse, Tunisia, using the Hybrid Optimization of Multiple Energy Resources (HOMER) software technique. This approach is chosen thanks to its simplicity, effectiveness, and high precision compared to traditional techniques. In this paper, several configurations (Grid, (Grid-battery), (Grid-PVP), (Grid-PVP-battery)) are studied. The obtained results prove that the (Grid-PVP-battery) system configuration is the most efficient and economical solution. In the second part, a robust energy management strategy (EMS) is proposed for two smart grid configurations (grid-battery, grid-PVP-battery). This strategy is based on Fuzzy Logic Control (FLC) thanks to its non-linear modelling and its ability to make decisions relating to energy management. The primary goal of the suggested (EMS) is to ensure the energy resources scheduling in order to keep a continuous equilibrium among the production and consumption of electricity and avoid instabilities in the grid, with guaranteeing a minimum cost of electricity. As input data, (FLC) used time-varying price electricity (Price (t)) to solve an instant decision problem by choosing, at each instant, the optimal energy source (which provide electricity at the cheapest price possible). The obtained results, carrying out Matlab simulation, prove the efficacy of the proposed strategy, not only, in the energy resources scheduling to meet the load, but also, for the system cost reduction since the PVP has been used as much as possible since it is inexpensive relative to the costs of battery capacity and the grid.

Nonlinear model predictive control for mode‐switching operation of reversible solid oxide cell systems

AbstractSolid oxide cells (SOCs) are a promising dual‐mode technology for the production of hydrogen through high‐temperature water electrolysis, and the generation of power through a fuel cell reaction that consumes hydrogen. Switching between these two modes as the price of electricity fluctuates requires reversible SOC operation and accurate tracking of hydrogen and power production set points. Moreover, a well‐functioning control system is important to avoid cell degradation during mode‐switching operation. In this article, we apply nonlinear model predictive control (NMPC) to an SOC module and supporting equipment and compare NMPC performance to classical proportional‐integral (PI) control strategies, while switching between the modes of hydrogen and power production. While both control methods provide similar performance across various metrics during mode switching, NMPC demonstrates a significant advantage in reducing cell thermal gradients and curvatures (mixed spatial‐temporal partial derivatives), thereby helping to mitigate long‐term degradation.

Behavior-aware assessment of demand response participation in power systems using causal artificial intelligence

Demand response (DR) is pivotal in enhancing the operational efficiency and reliability of power systems. Understanding prosumers’ behavior and their participation in DR programs is essential for utilities to design effective DR programs. Typically, prosumers are not concurrently exposed to multiple electricity price plans or engaged in distinct DR programs, which make the available data limited and introduces complexity in assessing their DR participation. Causal artificial intelligence offers a solution by enabling causal intervention and counterfactual analysis, facilitating the estimation of prosumers’ responses to demand-side programs without the necessity of experimental and control groups. This paper introduces a framework that integrates causal learning approaches for causal intervention and counterfactual analyses of prosumers’ DR participation and consumption behavior. The proposed framework incorporates domain knowledge of the power system to enhance model accuracy and performance. Simulation results demonstrate the effectiveness of the proposed framework with causal learning models in estimating prosumers’ DR participation and consumption behavior under different prices or DR programs. This capability allows utilities to gain a profound understanding of their prosumers, offering valuable insights for designing DR programs. Ultimately, this knowledge contributes to the reliable and efficient operation of power systems.

The improved integrated Exponential Smoothing based CNN-LSTM algorithm to forecast the day ahead electricity price

The deregulation of electricity market has led to the development of the short-term electricity market. The power generators and consumers can sell and purchase the electricity in the day ahead terms. The market clearing electricity price varies throughout the day due to the increase in the consumers bidding for electricity. Forecasting of the electricity in the day ahead market is of significance for appropriate bidding. To predict the electricity price the modified method of Exponential Smoothing-CNN-LSTM is proposed based on the time series method of Exponential Smoothing and Deep Learning methods of Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM). The dataset used for assessment of the forecasting algorithms is collected from the day ahead electricity market at the Indian Energy Exchange (IEX). The forecasting results of the Exponential Smoothing-CNN-LSTM method evaluated in terms of Mean Absolute Error (MAE) as 0.11, Root Mean Squared Error (RMSE) as 0.17 and Mean Absolute Percentage Error (MAPE) as 1.53 % indicates improved performance. The proposed algorithm can be used to forecast the time series in other domains as finance, retail, healthcare, manufacturing.•The method of Exponential Smoothing-CNN-LSTM is proposed for forecasting the electricity price a day ahead for accurate bidding for the short-term electricity market participants.•The forecasting results indicate the better performance of the proposed method than the existing techniques of Exponential Smoothing, LSTM and CNN-LSTM due to the advantages of the Exponential Smoothing to extract the levels and seasonality and with the CNN-LSTM methods ability to model the complex spatial and temporal dependencies in the time series.

Finding opportunity in economic power dispatch: Saving fuels without impacting retail electricity prices in fuel-producing countries

In many oil- and natural gas-producing countries, fuels are allocated to the electricity sector at administratively set prices below international market prices for domestic policy reasons. This article shows that dispatching power units based on the opportunity costs of domestically used fuels can yield significant efficiencies, while end-user electricity prices can continue to reflect administratively set fuel prices. In addition, opportunity cost dispatch can bring environmental benefits when it results in switching the priority of oil and gas units in the merit order and facilitating investments in renewable energy. Therefore, this approach can support long-term net-zero emission targets. It can also resolve cross-border electricity trading issues faced by fuel-rich countries, which are reluctant to base their electricity exports on domestic fuel prices.

Evaluating synthetic fuel production: A case study on the influence of electricity and CO2 price variations

To combat climate change, we need to reduce emissions from the transport sector. Synthetic fuels are a long-term solution for aviation, maritime and heavy machinery. Large-scale use requires cost-effectiveness, efficient production and resilience to price changes. In this case study, we simultaneously optimize the cell voltage of the solid oxide electrolysis cell, the heat exchanger network and the heat supply of a PtL-plant. PtL-efficiency and production costs are used as objectives to generate multiple Pareto fronts for future price scenarios. The results show that the sensitivity to price changes has different impacts on design and operating parameters, which can lead to unattractive solution domains in the Pareto front. Currently, synthetic fuels can be produced at 1.83–2.36 €/kg. In the best case, at 1.42–1.97 €/kg and 3.88–4.28 €/kg in the worst case. This paper supports decision-makers in planning PtL-plants to ensure sustainable synthetic fuel availability on a global scale.

Operational optimization of demand management strategy for hybrid power system based on battery capacity fading characteristics

In the hybrid power system (HPS), taking battery capacity fading characteristics and demand management strategy (DMS) into consideration, an operational optimization method is developed in this work. From the energy utilization and economic perspectives, the minimum total renewable energy curtailment rate (TCR), minimum total outsourced electricity (TOE) from the grid and minimum total operating cost (TOC) are respectively taken as the optimization objectives. The corresponding mathematical models of the proposed method are established, which feature mixed integer nonlinear programming (MINLP) ones. A case study of a solar-wind-biomass-battery HPS is adopted to verify and illustrate the application and effectiveness of this proposed method. Results show that the optimal operating schemes and the DMSs of the HPS based on battery capacity fading characteristics can be determined by minimizing the TCR, TOE and TOC. The target values of TOE and TOC of the HPS with adopting DMS are 15,995.3 kWh and 18,112.1 $, which are about 70 % these of the HPS without DMS. Furthermore, the relationships between battery capacity fading characteristics and DMS are comprehensively investigated in this operational optimization of the HPS. Finally, the effects of different battery types, outsourced electricity price changes on the operation of the HPS are analyzed and discussed. This work provides a valuable method for the operational optimization of the HPS.

A multi‐objective interval optimization approach to expansion planning of active distribution system with distributed internet data centers and renewable energy resources

AbstractWith the development of the digital economy, the power demand for data centers (DCs) is rising rapidly, which presents a challenge to the economic and low‐carbon operation of the future distribution system. To this end, this paper fully considers the multiple flexibility of DC and its impact on the active distribution network, and establishes a collaborative planning model of DC and active distribution network. Differing from most existing studies that apply robust optimization or stochastic optimization for uncertainty characterization, this study employs a novel interval optimization approach to capture the inherent uncertainties within the system (including the renewable energy source (RES) generation, electricity price, electrical loads, emissions factor and workloads). Subsequently, the planning model is reformulated as the interval multi‐objective optimization problem (IMOP) to minimize economic cost and carbon emission. On this basis, instead of using a conventional deterministic‐conversion approach, an interval multi‐objective optimization evolutionary algorithm based on decomposition (IMOEA/D) is proposed to solve the proposed IMOP, which is able to fully preserve the uncertainty inherent in interval‐typed information and allow to obtain an interval‐formed Pareto front for risk‐averse decision‐making. Finally, an IEEE 33‐node active distribution network is utilized for simulation and analysis to confirm the efficacy of the proposed approach.

Interior-point policy optimization based multi-agent deep reinforcement learning method for secure home energy management under various uncertainties

Improving the efficiency of home energy management (HEM) is of great significance in reducing resource waste and prompting renewable energy consumption. With the development of advanced HEM systems and internet-of-things technology, more residents are willing to participate in the electricity market through a demand response mechanism, which can not only reduce the electricity bill but also stabilize the operation of the main grid through peak shaving and valley filling. However, the uncertainties from household appliance parameters, user behaviors, penetrated renewable energy, and electricity prices render the HEM problem a non-trivial task. Although previous deep reinforcement learning (DRL) methods have no requirement for system dynamics due to their model-free and data-driven merits, the unrestricted action space may violate the physical constraints of various devices, and few works have concentrated on this area before. To solve the above challenges, this paper proposes a novel interior-point policy optimization-based multi-agent DRL algorithm to optimize the home energy scheduling procedure while guaranteeing safety during its operation. Besides, a time-series prediction model based on a non-stationary Transformer neural network is proposed to predict the future trend of solar generation and electricity price, which can provide the agents with abundant information to make better decisions. The superior performance of our proposed predictive-control coupled method is demonstrated through extensive numerical experiments, which achieves more precise prediction results, near-zero constraint violations, and better trade-offs between cost and safety than several benchmark methods.

Residential demand response online optimization based on multi-agent deep reinforcement learning

Demand side management in the context of smart grids exploits the value stream of load flexibility, promoting stable and economic operation of the power grid. To address multi-uncertainties such as device characteristics, distributed renewable generation, and residential electricity demand, this paper proposes a model-free framework based on multi-agent deep reinforcement learning to achieve efficient online optimization of demand response strategies. Firstly, the demand side management is formulated as a partially observable Markov game. Then, a multi-agent soft actor-critic algorithm is proposed, combining a temporal electricity price feature extraction technology to improve learning efficiency. Reward correction mechanism is designed to avoid new charging and discharging peaks in the demand response. Finally, case studies demonstrate the effectiveness of the proposed method in reducing the electricity cost and aggregate peak demand of residents. The proposed method reduces the daily average cost by 21.40%, 8.09%, and 5.88% compared to MAPPO, MADDPG, and SUMADRL, and exhibits the highest training efficiency and scalability against the benchmarks.