Lithium-ion Battery Energy Storage Research Articles

Optimized multi-head self-attention mechanism for SOH prediction in lithium-ion battery energy storage systems within microgrids

Estimating the State of Health (SOH) of lithium batteries is crucial for the safety and stability of the microgrids. Currently, existing data-driven methods struggle to predict across multiple groups of lithium batteries varying in type, capacity, and degree of degradation, as the prediction process requires the collection of extensive operational data for processing and computation. This paper analyzes the characteristics of lithium battery storage units within the microgrids and proposes a novel prediction method based on an improved attention mechanism. In the operational process, we collected the data under constant current (CC) and constant voltage (CV) charging modes to obtain health indicators with high correlation. We proposed a novel optimized multi-head attention method, designated OM-Attention, which employs a multi-head self-attention-based model to compute multiple sets of lithium batteries within a network node in a simultaneous manner. To validate the effectiveness of the OM-Attention method, we randomly selected four lithium batteries from the NASA public dataset for multiple experiments. Compared to SOH prediction methods, our approach exhibits superior performance with reduced Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Mean Square Error (MSE) to 4.16%, 5.02%, and 0.302×10−4, respectively. Additionally, the R2-score achieved by OM-Attention is 97.7%.

Negative Resistor-Based Equivalent Circuit Model of Lithium-Ion Battery Energy Storage System for Grid Inertia Support

Negative Resistor-Based Equivalent Circuit Model of Lithium-Ion Battery Energy Storage System for Grid Inertia Support

Realization of building grade fire resistance in independent fire zone of energy storage system

Abstract Lithium battery energy storage technology is one of the most promising technologies in the field of energy storage. The fire prevention and control of energy storage system usually needs to consider the cooperation of various parts of the energy storage system. The domestic standard requirements for the fire resistance of energy storage system have not been completely established. In this paper, combined with the US standard and related fire resistance experimental standards, the relevant requirements and test methods for the fire insulation performance of battery cabinet of the energy storage system are studied, and a set of structural design methods and test schemes for the 2h fire resistance limit of the energy storage system are formed. The research shows that the fire insulation of the battery cabin has positive significance for reducing the energy consumption of the air conditioning of the energy storage product, improving the battery life, and enhancing the safety performance of the energy storage system. It is of great significance to study the suitable fire resistance limit of energy storage products for improving the land utilization rate of energy storage system and saving the construction cost.

Explosion characteristics of two-phase ejecta from large-capacity lithium iron phosphate batteries

When a thermal runaway accident occurs in a lithium-ion battery energy storage station, the battery emits a large amount of flammable electrolyte vapor and thermal runaway gas, which may cause serious combustion and explosion accidents when they are ignited in a confined space. With the gradual development of large-scale energy storage batteries, the composition and explosive characteristics of thermal runaway products in large-scale lithium iron phosphate batteries for energy storage remain unclear. In this paper, the content and components of the two-phase eruption substances of 340Ah lithium iron phosphate battery were determined through experiments, and the explosion parameters of the two-phase battery eruptions were studied by using the improved and optimized 20L spherical explosion parameter test system, which reveals the explosion law and hazards of the two-phase battery eruptions. Studies have shown that in a two-phase system explosion, EMC can make the two-phase system more explosive and more powerful, and the thermal runaway gas expands its explosion concentration range. The coupling explosion of the two enhanced the sensitivity and explosive power of the two-phase ejecta. Increasing the concentration of any combustible in a two-phase system will cause the explosion intensity parameters of the system to increase. However, when the combustible concentration exceeds the optimal explosion concentration, the explosion intensity parameters will decrease or even no explosion will occur. Both explosion intensity parameters and sensitivity parameters are more sensitive to EMC concentration, and the higher the EMC concentration, the stronger its dominant role in the explosion of the two-phase system. This work can lay the foundation for revealing the disaster-causing mechanism of explosion accidents in lithium-ion battery energy storage power stations, guide the safe design of energy storage systems and the prevention and control of explosion accidents, and provide theoretical and data support for the investigation of explosion accidents in energy storage power stations.

Early Internal Short Circuit Diagnosis for Lithium-Ion Battery Packs Based on Dynamic Time Warping of Incremental Capacity

Timely identification of early internal short circuit faults, commonly referred to as micro short circuits (MSCs), is essential yet poses significant challenges for the safe and reliable operation of lithium-ion battery (LIB) energy storage systems. This paper introduces an innovative diagnostic method for early internal short circuits in LIB packs, utilizing dynamic time warping (DTW) applied to incremental capacity (IC). Initially, the terminal voltages of all cells within the LIB pack are ordered at any moment to determine the median terminal voltage, which is then used to generate the median IC curve. This curve acts as a reference benchmark that represents the condition of healthy cells in the pack. Subsequently, the DTW algorithm is utilized to measure the similarity between each cell’s IC curve and the median IC curve. Cells exhibiting similarity scores that exceed a specified threshold are identified as having MSC faults. Lastly, for the cells diagnosed with MSC conditions, a method for estimating short-circuit resistance (SR) based on variations in maximum charging voltage is devised to quantitatively evaluate the severity and evolution of the MSC. Experimental findings reveal that the proposed method effectively identifies MSC cells in the LIB pack and estimates their SRs without the necessity of a battery model, thereby affirming the method’s validity.

Techno-financial analysis of 100 % renewable electricity for the south west region of the UK by 2050

In the UK, the transition to 100 % Renewable Energy Systems is a critical strategy in addressing climate change, aligning with the UK government's 2050 net-zero objective. This study explores the potential for such a transformation within the South West region of the UK by 2050. EnergyPLAN16.2 software was utilised for scenario modelling and the fmincon algorithm in MATLAB for optimisation analysis. The focus was on two scenarios involving hybrid renewable energy systems incorporating onshore wind, solar, offshore wind, and two energy storage systems which are lithium-ion battery energy storage and hydrogen fuel cell energy storage. The optimisation process entailed two stages: determining renewable energy configurations for cost-efficient coverage of annual electricity demand and computing the minimum energy storage capacity to ensure a constant power supply. The results suggest an optimal configuration of 4,460 MW onshore wind, 9,880 MW solar, 6,982 MW offshore wind, and 2,273,662 MWh of lithium-ion battery energy storage for 100 % hybrid renewable energy systems in the South West region of the UK in 2050. This configuration achieves the lowest annual cost of 3208 M£/year, with a balanced wholesale electricity price ranging from 4.3p/kWh to 6.6p/kWh and a levelized cost of energy of 6.44p/kWh. Furthermore, hydrogen fuel cell storage systems can be another alternative, however, the present study showed it is not economical enough compared to lithium-ion battery storage systems. This research broadens the understanding of hybrid renewable energy systems implementation in large regions, offering valuable insights for sustainable energy planning in the South West region of the UK.

Novel state of charge estimation method of containerized Lithium–Ion battery energy storage system based on deep learning

State of charge (SOC) is a critical indicator for lithium–ion battery energy storage system. However, model-driven SOC estimation is challenging due to the coupling of internal charging and discharging processes, ion diffusion, and chemical reactions in the electrode materials. These factors result in frequent changes in their internal characteristics. Therefore, we devise a novel deep learning SOC estimation method based on the attention mechanism (A) and convolutional neural networks-long short term memory (CNN–LSTM) network model. The CNN layers are employed to extract the critical features of the battery dataset. The LSTM layers are utilized to capture the long-term dependencies in the time series data. The attention mechanism can assign attention weights to different features within the output of the convolution layer. The A–CNN–LSTM method not only amplifies the positive influence of critical knowledge but also provides a global reference. The comparative experiments of three datasets and ablation study are carried out. The root mean square error, mean absolute error, and mean absolute percentage error obtained from the A–CNN–LSTM model decreased by 32.75 %, 45.88 %, and 53.36 %, respectively, compared with the best results obtained from the existing CNN, LSTM, gate recurrent unit (GRU), CNN–LSTM, and CNN–GRU models. The proposed model demonstrates greater stability and consistency than other models, underscoring its efficacy in SOC estimation.

Multi-Scale Risk-Informed Comprehensive Assessment Methodology for Lithium-Ion Battery Energy Storage System

Lithium-ion batteries (LIB) are prone to thermal runaway, which can potentially result in serious incidents. These challenges are more prominent in large-scale lithium-ion battery energy storage system (Li-BESS) infrastructures. The conventional risk assessment method has a limited perspective, resulting in inadequately comprehensive evaluation outcomes, which impedes the provision of dependable technical support for the scientific appraisal of intricate large-scale Li-BESS systems. This study presents a novel Li-BESS-oriented multi-scale risk-informed comprehensive assessment framework, realizing the seamless transmission of assessment information across various scales. The findings from a previous smaller-scale analysis serve as inputs for a larger scale. The evaluation process of this method is more scientifically rigorous and yields more comprehensive results compared to assessment technologies just relying on a single perspective. By utilizing the proposed comprehensive assessment methodology, this study utilized the emergency power supply of nuclear power plants (NPPs) as an application scenario, demonstrating the complete implementation process of the framework and conducting a comprehensive assessment of Li-BESS feasibility as an emergency power source for NPPs. Our findings propose a novel paradigm for the comprehensive assessment of Li-BESS, which is expected to serve as a scientific foundation for decision-making and technical guidance in practical applications.

Effect of ambient pressure on the fire characteristics of lithium-ion battery energy storage container

As lithium-ion battery energy storage gains popularity and application at high altitudes, the evolution of fire risk in storage containers remains uncertain. In this study, numerical simulation is employed to investigate the fire characteristics of lithium-ion battery storage container under varying ambient pressures. The findings reveal that the peak heat release rate of fires at normal pressure is significantly higher than at lower pressure. Specifically, the heat release rate at 100 kPa is 9215 kW, exceeding the value at 40 kPa by 42%, which is only 3900 kW. This peak heat release rate also demonstrates a power function relationship with ambient pressure. In addition, fires tend to last longer in lower pressure, where high-temperature areas expand and spread rates increase. Moreover, higher pressures produce elevated peak concentrations of CO and CO2, while smoke spreads faster in lower pressure, despite lower peak smoke concentrations. The study findings can serve as a foundation for assessing the fire hazards and designing fire protection measures for lithium-ion battery storage containers exposed to varying ambient pressures.

Degradation-Aware Derating of Lithium-Ion Battery Energy Storage Systems in the UK Power Market

Degradation-Aware Derating of Lithium-Ion Battery Energy Storage Systems in the UK Power Market

Coordinated control algorithm of hydrogen production-battery based hybrid energy storage system for suppressing fluctuation of PV power

Photovoltaic (PV) power generation has issues of volatility and intermittency. Currently, PV plants are generally equipped with 10% rated capacity lithium-ion (Li) battery energy storage systems in China, who often fail to suppress fluctuation in the output power of PV plants effectively and meet the grid-connected standard. The hybrid energy storage system (HESS) combining with hydrogen production and Li battery system can produce hydrogen by water electrolysis during the peak period of PV power generation, effectively improving PV utilization efficiency, while smoothing PV power fluctuation and improving grid connection electricity quality. Firstly, models of the solid oxide electrolysis cell (SOEC) and alkaline electrolysis cell (AEC) systems for hydrogen production, and Li battery energy storage system are established, and the transient response characteristics of each system are analyzed. Secondly, an adaptive wavelet packet decomposition (AWPD) method for PV power signal decomposition is proposed based on the wavelet packet decomposition (WPD) method. Thirdly, a capacity configuration method for HESS are proposed based on the APWD method. Fourthly, a coordinated control strategy for HESS is proposed with the transient response characteristics of different energy storage systems and the state of charge for Li battery system. Finally, the proposed method is validated through simulation experiments based on the actual power data of the PV plant. The results show that the developed methods can effectively utilize partial PV power generation to produce hydrogen, improve PV utilization, and the combined output power of PV plant and HESS can fulfill the grid-connected standard.

The state of charge predication of lithium-ion battery energy storage system using contrastive learning

The state of charge (SOC) is a critical state quantity that must be determined in real-time for a battery energy storage system (BESS). It is a prerequisite for the operation of a BESS. However, obtaining the precise value of SOC is challenging due to it being a hidden state quantity. Existing neural network models commonly employ an end-to-end prediction paradigm for SOC estimation, which fails to fully exploit the rich information present in the time-series battery data. Unlike most studies available in the literature, we propose a novel SOC prediction method named CLDMM. This method is the first to apply contrastive learning techniques from the image field to the SOC prediction of lithium batteries. The method utilizes data augmentation, a multi-scale encoder, and multi-layer perceptrons to learn latent representations and mix these with raw data proportionally for downstream predictive tasks. The performance of the proposed method is evaluated using the Panasonic NCR18650PF dataset, and ablation study were conducted. Experimental results show that CLDMM outperforms baseline methods, achieving an average mean absolute error (MAE) of 0.64% and an average maximum error (MAX) of 2.66%.

Model-free detection and quantitative assessment of micro short circuits in lithium-ion battery packs based on incremental capacity and unsupervised clustering

Timely diagnosis of micro short circuit (MSC) faults is crucial for ensuring the safe operation of lithium-ion battery energy storage systems. Existing diagnostic methods face limitations such as high dependency on battery models, difficulty in determining accurate diagnostic thresholds, or low computational efficiency. This work presents a model-free approach for the detection and quantitative assessment of MSCs in lithium-ion battery packs, with incremental capacity (IC) and unsupervised clustering. First, the IC is extracted from charging voltage data to effectively characterize MSC faults in lithium-ion batteries. Next, principal component analysis is used to map the high-dimensional feature space onto a two-dimensional plane to facilitate fault detection and result visualization. Then, an unsupervised clustering algorithm is employed for anomaly detection to identify MSC cells within the battery pack. For the detected MSC cells, a method based on the maximum charging voltage difference between adjacent cycles is designed to estimate the MSC resistance, quantitatively assessing the severity and evolution stage of the MSC. Experimental results show that the accuracy of MSC detection is 99.17 % and the minimum relative error of short-circuit resistance estimation is 1.20 %, which demonstrates the effectiveness and feasibility of the proposed method.

IoT-Enabled energy conservation in residential Buildings: Machine learning models for analyzing annual solar power consumption

This study investigates enhancing the self-consumption of solar panels by integrating thermal and electrical energy storage using lithium-ion batteries in a duplex building over a year. Machine learning algorithms are applied to understand energy usage patterns. An IoT-connected model is introduced to reduce electricity expenses by boosting consumption with solar energy and extending solar panel lifespan through optimization. The model includes lithium-ion batteries for energy storage and a thermal cooling system to regulate panel temperatures. Data collected from sensors monitoring occupancy, temperature, humidity, energy, and thermostats are sent to a cloud-based platform via IoT. Results show peak monthly energy conservation measures of 41.6 kW and 28 kW during summer for the two residences. Winter witnesses a reduction of 13.9 % and 7 % in energy consumption, while autumn sees an 18 % decrease for one residence and stability for the other. Real-time IoT monitoring aids model training, with the Gaussian Probability model emerging as the most accurate. Optimization analysis underscores the thermal cooling system’s efficacy in improving solar panel performance and reducing electricity costs. Integration of machine learning and IoT curtails energy consumption to 22 kWh and 15.7 kWh for the respective residences, offering inhabitants free energy from renewable sources and ensuring swift recovery of solar panel installation costs.

Profile of Round-Trip Efficiency of Lithium-Ion Battery Energy Storage System

Lithium-ion battery energy storage systems (Li-BESSs) are expected to adjust variable renewable energies such as photovoltaic energy. In the viewpoint of the energy management cost, profiles of round-trip efficiency (RTE) are important characteristics of Li-BESSs. The profiles of RTE are related to the characteristics of both LIBs and inverters installed on Li-BESSs. The characteristics of Li-BESSs would be various because of reusing and degradation of lithium-ion batteries (LIBs). And inverters made by a lot of manufacturers would be so on. Accordingly, it could be said that the understanding of the profile of each Li-BESS’s RTE is complex and difficult. Therefore, we studied the time series analysis of operation data and tried to estimate the characteristics of both LIBs and inverters. The important ones of LIB are full-charge capacity (FCC), open-circuit voltage (OCV), and internal resistance (R). In addition, characteristics of OCV and R can be expressed as function profiles with state of charge (SOC) as elements. Similarly, the important one of inverter is the typical efficiency curve. In this study, we proposed the estimation method of the profiles of OCV and R and inverter efficiency curve. On the other hand, FCC was given as assuming that general diagnosis results would be provided. Then, a RTE profile of Li-BESS (including 9.94kWh LIB and 10kW inverter) was tried to establish from the results of it. The profiles of OCV and R were estimated by the method called MGFFD (Modified Gaussian Free-Form Deformation) that we have been established [1]. The novelty of this study could be described as: Development of the estimation method of an efficiency curve of inverter installed on BESS during operation (Fig.1). Quantitatively visualized profile of RTE of Li-BESS by combining results of MGFFD estimation applied for LIBs and inverter (Fig.2). The profile described on Fig.2 is the relationship between SOC, input-output power rate of inverter, and RTE of Li-BESS. To understand Li-BESS’s charge-discharge efficiency of from state at that time to the end of an operation plan, it would be necessary information. This proposed method could clarify various RTE of Li-BESSs, consequently, be useful for its efficient operation by information communication technology.Reference[1] M. Arima, L. Lin, and M. Fukui, "Kalman-Filter-Based Learning of Characteristic Profiles of Lithium-Ion Batteries", Sensors, vol. 22, no. 14, 5156. Figure 1

Optimization of wind farm energy supply system costs through machine learning

At present, the energy crisis is becoming increasingly severe, and changing the energy structure and developing clean energy have gradually become the consensus of the world. Establishing a clean energy power supply system, using clean energy for power generation, and investing in real power supply are currently unresolved challenges. Based the current high cost of new energy power supply systems, this project proposes an optimization plan based on the wind power supply system, targeting the capacity of lithium battery energy storage equipment, and constructs an optimization model for wind power supply in electric vehicle systems. In addition, in order to use lithium-ion battery devices more efficiently and ensure the evaluation of the efficiency and battery life of lithium-ion battery energy storage devices, we predicted the cyclic aging phenomenon of lithium-ion battery energy storage devices through machine learning technology. This prediction can provide suggestions and references for the use of this lithium battery.

Predictive State-of-Health (SOH) of lithium-ion batteries throughout their lifecycle based on LSTM neural networks

With the rapid development of new energy sources, the demand for energy storage systems in the power industry has surged. While lithium-ion battery energy storage systems have gained widespread market acceptance due to their performance advantages, concerns regarding battery safety and durability have also escalated. Therefore, accurately predicting the health status of batteries has become a crucial aspect in assessing their safety and durability. State of Health (SOH), as a vital indicator of battery health status, has become a focal point for many researchers aiming to enhance the accuracy of its prediction. Today, owing to the rapid advancements in machine learning and neural networks, data-driven approaches have gradually emerged as one of the primary methods for SOH prediction. This paper proposes a method based on Long Short-Term Memory (LSTM) neural networks, which combines theoretical analysis with neural network models, utilizing authentic battery aging data from NASA to predict the health status of lithium-ion batteries throughout their lifecycle. Experimental results demonstrate that the proposed method exhibits high accuracy.

Predictive State-of-Health (SOH) of lithium-ion batteries throughout their lifecycle based on LSTM neural networks

With the rapid development of new energy sources, the demand for energy storage systems in the power industry has surged. While lithium-ion battery energy storage systems have gained widespread market acceptance due to their performance advantages, concerns regarding battery safety and durability have also escalated. Therefore, accurately predicting the health status of batteries has become a crucial aspect in assessing their safety and durability. State of Health (SOH), as a vital indicator of battery health status, has become a focal point for many researchers aiming to enhance the accuracy of its prediction. Today, owing to the rapid advancements in machine learning and neural networks, data-driven approaches have gradually emerged as one of the primary methods for SOH prediction. This paper proposes a method based on Long Short-Term Memory (LSTM) neural networks, which combines theoretical analysis with neural network models, utilizing authentic battery aging data from NASA to predict the health status of lithium-ion batteries throughout their lifecycle. Experimental results demonstrate that the proposed method exhibits high accuracy.

High-precision state of charge estimation of electric vehicle lithium-ion battery energy storage system based on multi-scale optimized time-varying bounded smoothing variable structure filtering algorithm

High-precision state of charge estimation of electric vehicle lithium-ion battery energy storage system based on multi-scale optimized time-varying bounded smoothing variable structure filtering algorithm

Demonstration Project: 1.86 MWH Battery Energy Storage System and 540 KVA Inverter Integration

This research focused on the implementation of state-of-the-art system integration, involving a three-phase 540 KVA bidirectional inverter and a lithium-ion battery energy storage system with a capacity of 1.86 MWh, at the Florida Solar Energy Center (FSEC). The system was firstly put into use at the FSEC for load shifting. The second step was to add photovoltaic (PV) panels for peak shaving. The third step was to take the FSEC completely off-grid. The goal of this research was to provide a case study by implementing a power management algorithm, estimating the PV panel size, and presenting the test results and an analysis. Based on the power management algorithm, load shifting was achieved by using 180 kWh energy storage. To acquire peak shaving and the uniformity of power consumption, an 80 kW PV farm was sufficient. The optimal PV panel size was 500 kW for the off-grid scenario.