Circulating fluidized bed (CFB) boilers play a pivotal role in China’s power generation landscape. However, the intricate combustion system within the CFB boiler furnace exhibits strong coupling characteristics, characterized by multiple parameters, variables, nonlinearity, and time-varying dynamics, posing a significant challenge for precise system modeling and prediction. Machine learning (ML), with its robust nonlinear processing capabilities and predictive performance, holds immense promise in the domain of CFB technology. This paper delves into the application of ML techniques in this field, encompassing the prediction of minimum fluidization velocity, emissions forecasting, bed pressure forecasting, bed temperature/thermal efficiency prediction, particle circulation rate prediction, reduced-order models of computational fluid dynamics (CFD) flow fields, and boiler safety control system models. The paper critically evaluates the strengths and limitations of these technologies in various scenarios, providing an insightful perspective on the opportunities and challenges faced by CFB boilers in the era of big data. Emphasizing aspects like model interpretability, enhancing generalization capabilities, improving data quality and diversity, integrating models with conventional methods, and experimental validation are crucial areas worth attention for future advancements.

In the context of “carbon peak, carbon neutral”, how to utilize biomass fuel safely, efficiently and environmentally friendly has become a hot research issue in the industry. The research progress of biomass fuel fluidized bed combustion technology both domestically and internationally was summarized based on the extensive practical engineering experience in fluidized bed combustion, providing a comprehensive overview of progress in biomass fluidized bed combustion power generation technology from an engineering perspective. The analysis compares the characteristics and problems of biomass boilers, with particular emphasis on the research and application status of biomass utilization in circulating fluidized bed. In addition, the combustion characteristics of biomass fuels, NO_x pollutant emission control, chlorine corrosion, and ash deposition and slagging issues caused by alkali metals in fluidized bed combustion were discussed. The reaction mechanisms in each process were described, and the future development directions of related technical issues were forecasted.

Nowadays, circulating fluidized bed (CFB) coal-fired boilers face challenges in the process of deep peak regulation, such as high CO emission concentrations and the lack of theoretical guidance for collaborative emission reduction of multiple pollutants including NO_x and SO₂. Taking a 150 t/h CFB coal-fired boiler as the research object, a model for quickly predicting mass concentrations of CO, NO_x and SO₂ emitted from the furnace is established based on the long short-term memory (LSTM) neural network, the Attention mechanism and the XGBoost algorithm. Moreover, an online emission reduction strategy is proposed by coupling with the particle swarm optimization (PSO) algorithm. 36 298 operational data points from the coal-fired boiler throughout 2023 are selected as training samples. A correlation analysis is conducted between the boiler inspection data and pollutant emission mass concentrations to determine the input parameters for the prediction model. The fitness function and boundary function are determined with the prediction model coupled with the PSO algorithm. Through the calculation of emission reduction optimization model, an online emission reduction optimization strategy for CO, NO_x and SO₂ mass concentrations of CFB boilers in different load ranges is proposed, and the feasibility of the algorithm in practical boiler tuning applications is evaluated.

The accurate prediction of SO₂ and NO_x emission mass concentrations can effectively guide the control of pollutants emissions, which is of great significance for the environmental protection operation of circulating fluidized bed (CFB) units. A 330 MW CFB unit is taken as the research object, and the Pearson coefficient is used to realize the screening of input variables, and the interquartile range (IQR) method is applied to screen the outliers and replace them with the normalization at the same time, to complete the data preprocessing. Subsequently, the features of input variables are extracted by convolutional neural network (CNN), and by entering into the gate-recurrent unit (GRU) the time-series features are processed. The multi-head self-attention (MHA) mechanism is introduced to capture the important relationships between features, and the model output is obtained after training. Finally, the results of the test set are evaluated using the mean absolute error (MAE), mean absolute percentage error (MAPE), and the coefficient of determination (R²). The results show that the model is able to predict the pollutants mass concentration in CFBs more accurately and achieve good prediction results, and the superior performance of the model is proved by the comparison of ablation experiments with the model. The proposed CNN-GRU-MHA model can realize the monitoring and optimization guidance of pollutants emissions CFB units, so that the power plant can adjust the operation parameters in time to ensure that the pollutants emissions meet the standards.

With the continuously increasing proportion of installed capacity from new energy sources, higher flexibility demands are imposed on coal-fired power generation units, including supercritical circulating fluidized bed (CFB) boilers. By taking a 350 MW supercritical CFB boiler as the research object, the computational particle fluid dynamics (CPFD) method was employed to simulate the furnace response characteristics under variable load conditions, focusing on parameters such as furnace temperature, near-wall particle concentration, and average heat flux density on heating surfaces. The effects of combustion and circulation interventions on the rate of load change were also explored. The results indicate that, during load ramp-up operation, the response rate of average heat flux density in low load range (30%~50% of the unit rated load) decreases by about 38% compared with that in high load range (above 50% of the rated load). Focusing only on the high load range, the heat flux density responds faster during load ramp-down, with the rate of change about 31% higher than that during ramp-up. Under varying load amplitudes, the particle suspension concentration and convective heat transfer intensity inside the furnace can respond rapidly, while temperature changes lag slightly, indicating that CFB boilers rely more on variations in the heat transfer coefficient for rapid thermal regulation. Through combustion interventions, such as substituting 40% of the original coal feedstock with fine coal particles sized several hundred microns, the change in furnace temperature can be effectively accelerated, with the response rate of average heat flux density during ramp-up in the high load range increasing by about 43%, and by nearly 16% in the low load range. Additionally, implementing circulation interventions, such as adding a certain amount of fine bed material during ramp-up, can rapidly increase the particle suspension concentration in a short time and thus effectively improve the response rate of the heat transfer coefficient on the heating surface. If hot fine material is further added (for example, through a hot circulating ash storage and return system), the response rate of the average heat flux density during ramp-up in the high load range can be improved by approximately 31%, and by about 13% in the low load range. The study elucidates the internal response mechanisms of CFB boilers under variable loading conditions, confirms the feasibility of improving load change rates through circulation and combustion interventions, and provides a reference for further tapping the flexibility potential of supercritical CFB boilers and improving their variable load capability.

In order to evaluate the vibration safety of the central full partition wall of the world’s first lignite-fired 700 MW high-efficiency ultra-supercritical CFB boiler, a three-dimensional non-constant hydrodynamic model of the boiler is constructed, and the distribution of the gas-solid flow field within the furnace chamber is simulated. Furthermore, the pressure distribution and its fluctuation law of the gas-solid flow acting on the surface of the central full diaphragm wall within the furnace are solved. The vibration signals of the furnace wall of a supercritical 350 MW CFB boiler with a similar furnace structure are measured and analyzed spectrally to assess the frequency of pressure fluctuations in the furnace. The dynamic stress on the central diaphragm wall under fluctuating pressure in the furnace is calculated using the pressure fluctuations in the furnace obtained from simulation calculations and experimental tests as the excitation. The results demonstrate that the dynamic stress level of the central full diaphragm wall is below the permissible stress of the material, indicating that the central full diaphragm wall is safe.

With the rapid expansion of new energy power generation capacity in China, the insufficient load regulation capability of coal-fired power plants has become increasingly evident. In order to explore the start-stop peak regulation capability of supercritical circulating fluidized bed (CFB) power units, a 350 MW supercritical CFB unit was taken as the research object, and experimental studies on banked-fire hot standby and rapid start-stop operations were conducted. The experimental results demonstrated that the supercritical CFB unit can rapidly reduce its load to near zero (with an average load change rate of about 10%Pe/min) during bank firing, and then maintain hot standby for 108 minutes. After banked firing, the boiler quickly switched to wet-state operation, with the main steam pressure decreasing rapidly at a rate of 0.13 MPa/min. By reasonably controlling the feedwater flow, the working fluid temperature and wall temperature of the water-cooled walls and water-cooled panels were kept stable. The heat released from the combustion of residual carbon caused the bed temperature to decrease slowly during banked firing, which also provided favorable conditions for re-ignition. During the load lift phase, the unit could be quickly started, with NO_x emission mass concentration reaching an instantaneous peak of 101 mg/m³, while the hourly average was stable below 50 mg/m³. Throughout the entire experimental period, SO₂ emission mass concentration was consistently below 35 mg/m³, and pollutant emissions met the ultra-low-emission requirements. All parameters of the steam turbine and generator remained within normal ranges during the hot standby and startup/shutdown. The rapid decline in main steam pressure and the low superheat of the main steam temperature were the main factors limiting the duration of banked firing in this experiment. The relevant research work provides a reference for the start-stop peak regulation of higher-parameter supercritical and ultra-supercritical CFB units.

Based on the computational particle fluid dynamics (CPFD) numerical simulation method, the study takes the 660 MW supercritical circulating fluidized bed (CFB) boiler in Pingshuo, Shanxi as the research object. A full-loop model of the furnace is established and numerically simulated. On the basis of the parameters of the actual furnace, the material distribution characteristics of six cyclone separators are investigated. By altering the ratio of primary to secondary air, the uniformity of primary air, and the uniformity of secondary air, the effect of operational parameter changes on the gas-solid flow field within the furnace and the material distribution at the cyclone separator inlets is analyzed. The results indicate that the distribution characteristics of particles within the furnace lead to a distribution feature of particle mass flow rate at the cyclone separator inlets, which is “high on both sides and low in the middle”. When the total air volume is constant, a larger ratio of primary to secondary air can reduce the deviation in particle mass flow rate at the cyclone separator inlets. The uniformity of air distribution for both primary and secondary air in the furnace also affects the particle mass flow rate distribution at the cyclone separator inlets. The deviations in particle mass flow rate at the inlets of the six cyclone separators reach their minimum when the wind speed deviation at the middle air distribution plate is 10% and the deviation in the middle secondary air volume is 5%, respectively.

Circulating fluidized bed (CFB) boilers are characterized by high thermal inertia and strong heat storage capacity, enabling banked fire and near-zero output peak shaving. However, there is limited experimental research on banked fire and peak regulation in large-scale CFB units, with a lack of studies on key parameter variations and control strategies during this process. In this study, a banked fire and peak regulation test was conducted on a supercritical 350 MW CFB boiler to investigate the evolution of critical parameters during shutdown and propose optimization strategies for boiler feedwater flow rate and integrated turbine valve position control. These optimizations aim to maximize the peak shaving duration while ensuring operational safety. Experimental results demonstrate that the optimized supercritical CFB unit achieved 85 minutes of shutdown peak regulation with a load of 5~8 MW. Throughout the test, the boiler maintained dry operation, while the main and reheat steam temperatures decreased from 566.0 ℃ and 553.0 ℃ to 482.0 ℃ and 472.0 ℃, with average cooling rates of 0.99 ℃/min and 0.95 ℃/min, respectively. The average bed temperature declined from 875.8 ℃ to 730.9 ℃ at a rate of 1.70 ℃/min. During the test, the maximum exhaust temperature of the high-pressure cylinder reached 380.0 ℃, with the steam temperature at the regulating stage exceeding that of the cylinder inner wall. The wall temperature deviation at the outlet of the boiler water-cooled walls and mid-partition walls gradually decreased, peaking at 97.5 ℃. These findings confirm the feasibility of hour-level shutdown peak regulation in supercritical CFB units and provide a reference for engineering applications of similar units.

With the increasing proportion of new energy connected to the grid, the issue of frequency safety in the power system and mastering the regulation ability of the units have become more important. At present, in power system simulation, thermal power unit models suitable for electromechanical transient and medium-long term dynamics mainly adopt the simplified model of drum boilers and the single reheater turbine model recommended by IEEE. If a similar model is also used for the once-through boiler unit, the simulation results of the main steam pressure will deviate significantly from the actual situation due to the dynamic of thermal storage coefficient and the deviation of control system, which leads to a misjudgement of the unit’s regulation ability. By using thermodynamic modeling methods, a supercritical once-through boiler unit model suitable for multi-time scale dynamic simulation is proposed. By establishing a moving boundary model of the water wall and a dynamic heat flow model of the superheater, the heat storage capacity of the once-through boiler can be reflected more accurately. By incorporating feedwater control and superheat control, the control system is more in line with the actual unit. The high simulation accuracy of the model is verified using power plant operation data. Compared with the existing power simulation models, the simulation accuracy of the main steam pressure has been improved significantly. Therefore, the model can describe the dynamics of supercritical once-through boiler units more accurately in primary and secondary frequency regulation and peak shaving, which is helpful for simulating the frequency process of power systems.

The “double high” characteristics of new power system make its frequency stability face a huge challenge. Energy storage assisted thermal power unit frequency regulation technology has become a key core technology to ensure the stable operation of the new power system. The mainstream form of energy storage used in this technology, lithium battery storage, suffers from short lifespan and poor safety in use. The features of supercapacitor energy storage like high power, long cycle life, and high security, are highly compatible with the energy-storage requirements of frequency regulation of the energy storage assisted thermal power unit, but the supercapacitor’s response to the continuous unidirectional command is poor. Therefore, it is necessary to explore the technical route of hybrid energy storage to achieve complementary advantages of the two types of energy storage. The hybrid energy storage capacity configuration of supercapacitor and lithium battery was studied, the energy storage capacity configuration method based on the actual AGC frequency regulation command was designed, considering the characteristics of power-type energy storage devices and energy-type energy storage devices. Moreover, the frequency regulation performance and economy of three typical capacity configuration schemes were compared, and the optimal scheme was determined. Finally, engineering verification was carried out. The actual operation data show that, the supercapacitor hybrid energy storage system can improve the frequency regulation performance of the thermal power unit by 59.77%, extend the service life of the lithium battery to 3.6 times, and improve the system economy.

With the proposal of “double carbon target”, the pace of low-carbon transformation of power system has been further accelerated. The flue gas waste heat utilization systems have been widely applied in thermal power units. As the core equipment of the flue gas waste heat utilization system, the flue gas cooler has a high failure rate in actual operation due to factors such as limited space in the flue gas layout, high flue dust concentration, and ammonium bisulfate deposition. This not only reduces the recovery of flue gas waste heat, but also increases system resistance, resulting in an increase in fan power consumption and even affecting the unit’s load capacity and environmental emission. On the basis of extensive researches on the operation of more than 250 sets of flue gas coolers in the thermal power industry, typical faults commonly found in flue gas coolers, such as leakage, boiler ash deposition, bottom flue ash deposition, and low-temperature flue gas corrosion, are summarized. The causes of the above typical faults and their coupling relationships are analyzed in depth. Moreover, the targeted preventive measures are proposed to provide guidance for transformation and operation/ maintenance of flue gas coolers.

Butterfly valves are widely used in industrial field, and under certain working conditions, strong unstable flow will occur in the butterfly valve and cause vibration in pipeline system. By taking the connecting pipe of the medium and low pressure cylinder of a 600 MW heating unit as the research object, the mechanism of unstable flow in the butterfly valve and the vibration of the connecting pipe was revealed through the combination of field measurement and steady numerical simulation. Then, based on the flow pattern optimization, a new type of butterfly valve with valve plate and diversion structure was designed, and the unsteady numerical simulation of the maximum vibration condition of the original butterfly valve and the optimized butterfly valve was carried out. The results show that, after adding the flow-guiding structure to the valve plate, most of the main steam flow moves along the middle of the steam inlet pipe of the low-pressure cylinder. This can effectively weaken the exciting force generated by unstable flow and suppress the vibration of the connecting pipe. The new butterfly valve proposed can be applied to suppress the vibration of the pipeline system with small opening of the butterfly valve.

The industrial production and urban residents’ lives have led to a large amount of wastewater and sludge, and the landfilling of sludge has caused severe ecological damage. To facilitate the large-scale disposal of municipal sludge, transform waste into valuable resources, and prepare low-carbon, low-cost thermal storage materials, an idea is innovatively proposed, in which the silicon carbide, boron nitride, and expanded graphite is added as thermal conductivity enhancers to enhance the thermal conductivity of sludge incineration ash/potassium nitrate composite phase change thermal storage materials (50% sludge incineration ash+50% potassium nitrate). The composite phase change thermal storage materials were prepared, and the effects of thermal conductivity enhancers on thermal performance of these materials were investigated. The results indicate that, the expanded graphite is not suitable as a thermal conductivity enhancer for sludge incineration ash/potassium nitrate composite phase change thermal storage materials. The addition of a thermal conductivity enhancer with a mass fraction of 2% is optimal for improving melting latent heat, with boron nitride performing better than silicon carbide. The samples with 2% boron nitride shows the most significant increase in thermal conductivity, rising by 65%, 93%, 117%, and 203% compared with samples SC3 (without thermal conductivity enhancers) at temperatures of 100 ℃ to 400 ℃, respectively. After undergoing 1 000 cycles of heating/cooling, the samples with 2% boron nitride have a latent heat of 35.29 J/g and a thermal storage density of 292.1 J/g, while the samples with 2% silicon carbide have a latent heat of 40.90 J/g and a thermal storage density of 334.9 J/g. The heat transfer rates for the samples with 2% silicon carbide and 2% boron nitride are 0.16 ℃/s and 0.17 ℃/s, respectively. This preliminary evidence demonstrates the feasibility of using silicon carbide and boron nitride as thermal conductivity enhancers for sludge incineration ash/potassium nitrate composite phase change thermal storage materials.

Improving the heat transfer efficiency between liquid lead-bismuth eutectic (LBE) and supercritical carbon dioxide (S-CO₂) is of great significance for advancing the development of advanced nuclear energy systems. The heat transfer performance of printed circuit heat exchangers (PCHE) with different channel structures (straight shaped, wing-shaped, S-shaped and Z-shaped) is investigated through numerical simulation. The results show that the thermal resistance on cold side of the heat exchanger is significantly higher than that on the hot side, with the average heat transfer coefficient of the hot side in straight-channel PCHE being 26.2 times that of the cold side. Under the condition of a fixed hot-side channel structure, the effects of different cold-side channel structures on PCHE heat transfer performance are explored. The results indicate that, compared with straight channels, the heat transfer of Z-shaped, S-shaped and wing-shaped channels increases by 23.3%, 22.2%, and 10.6%, respectively, while the specific pumping power improves by 1.48 times, 1.68 times, and 1.44 times, respectively. In addition, the dynamic performance of different PCHE designs when the cold-side flow rate increases by 20% is compared, revealing that the straight-channel PCHE has the shortest rebalancing time. The pressure drop loss is more significant than the improvement in heat transfer. These findings provide theoretical guidance for optimizing the design of LBE/S-CO₂ heat exchangers and contribute to enhancing the thermal efficiency of next-generation nuclear energy systems.

A novel power generation system integrating an organic Rankine cycle (ORC) with an air-cooled coal-fired power plant is proposed to achieve cascaded utilization of steam energy and enhance output power capability. Based on the efficient thermoelectric conversion characteristics of the ORC system at low and medium temperatures, the steam expanding to a certain level in the coal-fired unit is extracted to drive the ORC system for higher output power. This coupling scheme can achieve the multiple purposes of increasing output power, recovering exhaust steam waste heat and preventing air cooler icing in winter. Focusing on a 600 MW coal-fired power plant, thermodynamic performance evaluation has been carried out. The results show that, when the extraction flow rate is 160 kg/s, the thermal efficiency of the system at 50% THA, 75% THA and 100% THA loads increases by 3.48, 1.72 and 1.08 percentage points, respectively. The exergy efficiency increases by 3.38, 1.68 and 1.05 percentage points, and the coal consumption rate decreases by 27.72, 12.88 and 7.92 g/(kW·h), respectively. The heat recovery rate reaches 64.57%, 25.64% and 15.40%, respectively. This approach not only improves the performance of existing air-cooled coal-fired power plants, but also reduces coal consumption, and the research results can provide a way to improve the overall performance of power plants.

The ignition, burnout, and slagging performance of Baoqing lignite raw coal and its dried lignite with different moisture contents was experimentally investigated using an ignition furnace and one-dimensional furnace test platform. The results show that, the ignition temperature of Baoqing raw coal is 415 ℃, which is highly prone to ignition. Compared to the influence of moisture on ignition temperature, the effect of fineness on ignition temperature is more significant. At low loads, the water and steam react with the water gas of coke in the early stages of combustion, which has a significant impact on the consumption rate of coke. Due to its high moisture content, raw coal undergoes intense reactions during the initial combustion stage, resulting in a rapid decrease in mass fraction of combustible materials in fly ash and a stronger tendency towards slagging. However, excessive moisture is not conducive to the complete combustion of coal powder in the later stage of combustion. As the fineness of coal powder R₉₀ increases, the burnout rate of coal powder decreases. But overall, the burnout rate of both Baoqing raw coal and dry coal is above 99%, indicating the Baoqing coal is highly flammable, and the effect of oxygen on burnout rate is not significant.