To study the impact of heat storage on the coal consumption of power supply of supercritical thermal power units under deep peak shaving, the actual heat storage coefficients of the steam-water process components of the boilers under various operating conditions were calculated based on the operating characteristics of the units at steady state and non-steady state. Based on the actual operating data of the unit, the coal consumption for power supply was calculated, and the influence of actual heat storage coefficients of units on the standard coal consumption of power supply was investigated. Taking a 660 MW supercritical unit as an example. Results show that when the unit load reduces from 100% to 30% at steady state and non-steady state, the metal heat storage coefficient increases by 135.47% and 133.93%, respectively, while the steam and water heat storage coefficients change little, with actual thermal storage coefficients increasing by 80.83% and 80.07% respectively. Moreover, the heat storage coefficients of the unit at non-steady state are greater than those at steady state. The standard coal consumption of power supply at non-steady state increases by 11.76-22.40 g/(kW·h) compared with that at steady state. The actual heat storage coefficient and coal consumption of the unit during the upward peak shaving are greater than that during the downward peak shaving. The unit coal consumption of power supply and the actual heat storage coefficient show a strong correlation, and the former increases with the increment of the actual heat storage coefficient.

To solve the problem that it is difficult to achieve the ultra-low emission of NO_x from circulating fluidized bed boiler under low load condition, method were proposed with adding an upper secondary air layer to realize air staging at first and then optimizing the coal sowing air, lower secondary air and ammonia spray grid structure. After which, above methods were implemented to a 350 MW unit. Numerical simulation were conducted on a single coal drop pipe, secondary air nozzle, urea guns and a full-size boiler by computational fluid dynamics and computational particle fluid dynamics. The sowing uniformity of coal particles and the intensity of air staging are improved by optimizing the margin of coal sowing air. Meanwhile, by adding the secondary air layer and reducing the angle of secondary air nozzle, NO_x emission can be further reduced. Full-size simulation results show that the NO emission mass concentration decreases from 264 mg/m³ to 162 mg/m³, with a decrease of 38.6%, and with the moving up of flame center, the temperature of flue gas at furnace outlet increases. Field retrofit results show that the boiler can achieve NO_x ultra-low emission under 35%-100% boiler load without a negative effect on boiler efficiency, while the urea consumption can be saved by more than 40% due to the increase of flue gas temperature at furnace outlet and the retrofit of urea guns.

A safety assessment methodology was proposed for manufacturing defects in steam turbine components based on crack propagation life under multiple damage mechanisms, and multiple damage mechanisms were introduced for safety assessment of steam turbine components. Calculation methods were presented for the crack propagation life under single damage mechanism (e.g., low-cycle fatigue, high-cycle fatigue, creep, stress corrosion), and the low-cycle fatigue crack propagation life at different stages under combined damage mechanisms. Theoretical models of life damage evaluation were established, and based on the crack propagation calendar life, allowable initial crack size, and allowable initial defect area, a safety assessment methodology for defects was constructed. After which, engineering application examples were provided for defect safety assessment of 1 000 MW-class nuclear steam turbine components, including high-pressure rotors, low-pressure inner casings, and low-pressure rotors under multiple damage mechanisms. Results demonstrate that the rotors and casings of nuclear steam turbine containing defects with an equivalent diameter of 2 mm, which are previously classified as unserviceable under conventional assessment criteria, can safely remain in operation.

Reducing tip leakage losses is crucial for improving the aerodynamic performance of low Reynolds number turbines. By using numerical simulation methods, flat blade tip structure was compared with squealer blade tip, and their flow characteristics within tip clearance were analyzed for Reynolds number (Re) range of 3.5×10⁴-35×10⁴. Results show that when there is a relative motion between blade row and casing, the proportion of blade tip leakage losses to total losses is related to the Reynolds number, with leakage losses increasing as the Reynolds number increases. Within the conventional low Reynolds number range, there exists a second critical Reynolds number Re_cr,2=6×10⁴. When Re<Re_cr,2, the squealer tip not only fails to reduce leakage flow rate but actually intensifies the mixing of the leakage flow with the main flow, increasing the flow losses at the blade tip. When Re>Re_cr,2,the squealer structure can effectively suppress leakage flow rate and improve the turbine stage efficiency. In the calculation conditions of this paper, the maximum efficiency improvement can reach 0.28%.

To investigate the flow field characteristics within tip clearance, a planar cascade experiment system was established. A method was introduced to load ultrasonic atomized water mist particles as tracer particles at the inlet of the upstream guide cascade in measurement area. Through numerical simulations, researches were conducted on the impact of loading tracer particle on flow field, and the characteristics of loading. Flow visualization experiments were carried out by particle image velocimetry (PIV), so as to obtain velocity field images within clearance. Results indicate that the tracer particle loading method can meet the requirements for PIV flow field measurements, with the acquisition of high resolution flow field images. Low-speed regions exist in the upstream and downstream of the tip clearance, while high-speed regions exist in the middle and rear sections within the tip clearance. With the increase of main flow velocity, the leakage flow velocity increases, which is particularly obvious in the downstream low-speed region. Additionally, with the increase of the height from the tip, the internal high-speed region expands.

In order to solve the increasingly serious problem of climate change, the development of carbon-free fuels is exposed to a huge opportunity on the background of global carbon emission reduction. Meanwhile, the development of ammonia energy is an effective measure to achieve carbon peak and carbon neutrality. On a 10 kW (input power) laboratory bench with ammonia-blended methane, experimental researches were conducted on the diffusive combustion of mixed methane and ammonia, while combined with chemical reactor network (CRN), analyses were carried out for the combustion and emission characteristics of ammonia/methane under different input powers, equivalence ratios and blending ratios. Results show that methane has a significant effect on promoting and stabilizing the ammonia combustion, but the NO emission would be dramatically increased with the ammonia heat value fraction of only 9.1%. The strategy of axial air staging can be applied to reduce the NO_x and unburned NH₃ emissions, and the combustion chamber outlet temperature can be increased with a increase of ammonia-blended proportion.

To investigate the impact on offshore wind turbines by waves and wind speed non-uniformities caused by wind shear and tower shadow effect, based on a 5 MW monopile offshore wind turbine provided in Openfast, a dynamic model of offshore wind power systems was established to output the mechanical torque under combined impact of wind non-uniformities and waves. The mechanical torque was further fed into a doubly fed induction generator (DIFG) modelled in Matlab/Simulink for co-simulation. Results show that the frequencies from the impact of the wind non-uniformities and waves appear in all spectra of generator speed, power, stator current and rotor current when considering the two factors. The relative influence of wind-speed non-uniformities and waves on generator speed, power, stator and rotor currents varies with wind conditions, and mutual coupling exists among wind shear, tower shadow and wave effects.

Due to the volatility and uncertainty of wind power output, which significantly impact the stability of the power system, grid frequency and power quality, a wind power fluctuation smoothing strategy based on time-scale adaptive wavelet packet was proposed. Firstly, an appropriate time scale was selected to divide the time domain, adaptive wavelet packet decomposition was performed, and arithmetic averaging was applied to smooth fluctuations that exceed the limits at the boundaries of adjacent time domains. Secondly, a comprehensive analysis of hybrid energy storage characteristics was conducted, and the energy entropy difference was used to analyze the wind power fluctuation power, reasonably distributing the fluctuation power. Finally, considering the state of charge (SOC) and cycle life of energy storage devices, a two-stage fuzzy control was applied to optimize the power output instructions of the flywheel energy storage and lithium iron phosphate batteries. Results show that the proposed strategy can effectively mitigate power fluctuations, more accurately allocate hybrid energy storage components, effectively reduce SOC violations, and improve power quality.

Offshore floating wind turbine platforms rely on the restoring forces provided by the mooring system to reduce the platform motion and maintain position. As the structural integrity of the mooring can be reflected by its mechanical properties, accurate prediction of the mooring system tension is essential for an efficient and reliable health monitoring system. A data-driven approach based on deep learning was developed to predict mooring-line tension. To preserve physical information that was often lost in pure time-series forecasting, complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) combined with sample entropy was applied for primary mode decomposition and reconstruction. Variational mode decomposition (VMD) together with box dimension was employed for secondary decomposition and reconstruction, yielding a high-quality feature set. Attention mechanism was then integrated with a convolutional neural network and a bidirectional long short-term memory network (AM-CNN-BiLSTM) to construct the deep-learning tension-prediction model. Results show that the decomposition-reconstruction strategy effectively extracts informative features and improves prediction accuracy when applied to the feature set. Across diverse sea states and mooring-system configurations, the proposed model achieves precise tension predictions, with the coefficient-of-determination (R²) values exceeding 0.879, confirming its strong generalization capability.

Existing irradiance forecasts are mostly intra-hour models built on coarse weather-type labels, ignoring the intraday weather evolution and the impact of cloud type on irradiance. To address these limitations, a day-ahead direct normal irradiance (DNI) forecasting framework that couples complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN), K-means density-based spatial clustering of applications with noise (KDBSCAN) and the Informer model was proposed. Firstly, principal component analysis was employed to compress the dimensions of raw meteorological variables. The derived principal components were then decomposed and reconstructed via CEEMDAN. KDBSCAN was subsequently employed to perform a two-layer clustering, weather-type clustering and intra-weather-type cloud-pattern clustering. Finally, based on the two-layer labels, the day-ahead DNI was forecasted via the Informer network. Results show that by classifying cloud types to consider weather changes throughout the day in the prediction process, the determination coefficient (R²) for predicting irradiance before sunny days reaches 0.98. Compared with prevailing benchmarks, the proposed framework reduces the overall mean absolute error by 32.42 W/m² and root-mean-square error by 31.71 W/m², achieving overall R² of 0.97.

A new biomass hydrogen-power cogeneration system based on the coupling of gasification and pyrolysis was proposed to address the issue of low conversion efficiency of biomass energy. In this system, the syngas produced by biomass gasification and pyrolysis drives a gas turbine to generate electricity, and then the waste heat in exhaust gas of the gas turbine was used to heat feed water into high-temperature steam, which drives the high-pressure and low-pressure steam turbines to generate electricity. The system was simulated using the thermodynamic simulation software Aspen Plus, and the simulation results were subjected to thermodynamic analysis, economic analysis, sensitivity analysis, and environmental analysis. Results show that the net power of the system is 8 061.07 kW, the thermal efficiency is 49.27% and the total exergy efficiency reaches 45.98%. The payback period of the system is 4.80 years and the net present value is estimated to be 630.555 9 million yuan.

A model of the solid oxide electrolysis cell (SOEC) and lithium-ion battery energy storage system was established, and their respective transient response characteristics were analyzed. A feed-forward model predictive control method for the SOEC system was proposed. Then, based on the continuous variational mode decomposition and variational mode decomposition methods, combined with the whale optimization algorithm (WOA), a capacity optimization allocation method for the SOEC-lithium-ion battery hybrid energy storage system was proposed. Finally, the effectiveness of the proposed method was verified through simulation experiments according to the actual output power data of a wind farm with rated installed capacity of 30 MW. Results show that the hybrid energy storage system can effectively utilize part of the wind energy for electrolysis to produce hydrogen, and realize the goals that the combined output power of the wind farm and hybrid energy storage system fulfills the grid-connected standards and the initial investment cost of the hybrid energy storage system is the lowest. At the same time, the SOEC system reaches the set electrolysis efficiency and the lithium-ion battery storage array is in a healthy state.

To study the impact of spiral fins and circular fins on the heat transfer performance of the vertical tube-and-shell phase change energy storage tanks, the fin structure with better heat transfer performance was selected, and physical models of tube-and-shell phase change energy storage tanks with plain tubes, circular fins, and spiral fins were established. Considering natural convection, numerical simulations were conducted for the charging and discharging processes. The temporal variations of temperature field of the phase change materials, solid-liquid interface, average temperature, liquid phase mass fraction, and the temperatures at different monitoring points in various storage tanks were analyzed. The overall thermal performance of different storage tanks was discussed, and the main factors affecting the charging and discharging rate of the storage tanks were compared and analyzed. Results show that the presence or absence of fins has a significant impact on phase change materials with low thermal conductivity. The fin structure can significantly reduce the melting and solidification time of the phase change materials in the energy storage tanks during the heat transfer process. Radially extended circular fins restrict natural convection, while spiral fins, with their continuous structure, exhibit the best heat transfer rate among the three structures, and the heat transfer effect improves with increasing fin height. During the charging process, in the vertical direction, the upper part of the energy storage tank has a higher temperature, and the phase change materials melt first. In the radial direction, the phase change materials near the inner tube melt faster. The opposite occurs during the discharging process, where the phase change materials at the lower part in the vertical direction and the inner side in the radial direction solidify first.

In the design of the steam charging channel of the thermal accumulator, two-phase flow phenomena present numerous challenges. By introducing the concept of biomimetic flow channel design, the uniformity of the three-dimensional pressure field distribution can be improved, thereby enhancing the condensation efficiency of the steam. Biomimetic principles were employed to design the inlet steam channel, and a parametric model was constructed based on computational fluid dynamics methods. The dataset generated from calculations was used to train an artificial neural network surrogate model, which was then further utilized with simulated annealing to search for the optimal solution, yielding the optimal pipe diameter parameters for this type of biomimetic channel. Results show that under the optimal parameter conditions, the pressure non-uniformity at the channel outlet decreases by 10.91%, and the turbulent kinetic energy loss within the channel reduces by 73.36%.The research findings provide important references for the design of steam inlet channels in steam accumulators.

Aiming at the problem that it is difficult to accurately measure the main steam flowrate of a unit, a main steam flowrate measurement model based on KPCA-BiLSTM-GRU was proposed, and a simulation verification was proposed based on historical data of a certain 1 000 MW ultra-supercritical primary reheat generating unit. Firstly, based on the operating mechanism and experience of the unit during actual production, operating parameters related to the main steam flowrate were selected as candidate variables for input into the measurement model. Secondly, KPCA algorithm was used to reduce the dimension of the original candidate input features, thereby avoiding the impact on the prediction accuracy caused by an excessive number of input variables in the model. Finally, the BiLSTM-GRU neural network model was used to further learn the change law of the input data features, achieving the regression prediction of the main steam flowrate. At the same time, Neural network models such as BP, LSTM, BiLSTM and GRU were selected for comparative experiments to verify the prediction performance of the proposed model. The results show that the proposed main steam flowrate model based on KPCA-BiLSTM-GRU can accurately measure the main steam flowrate. Its root mean square error (RMSE) is 25.76 t/h, and the mean absolute percentage error (MAPE) is 1.21%. Compared with other models in the experiment, the KPCA-BiLSTM-GRU main steam flowrate measurement model has a better prediction effect and is more adaptable to the variable load operation conditions of the deep load-peak shaving steam turbine power generation unit.

To improve the efficiency of limestone-gypsum wet flue gas desulfurization in coal-fired power plants while ensuring stable operation, a weighted composite control scheme based on proportional-integral-derivative (PID) and generalized predictive control (GPC) was proposed for maintaining the pH value of the desulfurization tower slurry and the outlet SO₂ mass concentration at ideal levels, thereby achieving a reasonable desulfurization control target. The desulfurization system model parameters, PID parameters, and the weight factors of the PID-GPC composite control were all optimized using an improved logistic sparrow search algorithm (LSSA). Simulation results indicate that the identified model can serve as the prediction model for GPC; the composite control algorithm integrates the advantages of both PID and GPC, enhancing the system's response characteristics; and the composite control optimized by LSSA exhibits improved dynamic performance, significantly enhances anti-interference capability, and effectively tracks the set targets.

For a 1 000 MW class pressurized water reactor (PWR) cogeneration system, a dynamic simulation model was established by using APROS software to investigate the effects of steam extraction from live steam, high-pressure exhaust steam, and reheat steam on electrical power. The dynamic response characteristics under various heating conditions were also simulated. The results show that high-pressure exhaust steam extraction has the least impact on power fluctuations and achieves the best equivalent power efficiency for heat supply. The response rate of electric power is sensitive to step changes in the extraction flow rates from reheat steam and high-pressure exhaust steam. When there is a step change in the extraction flow rate for low-temperature heating, a steady-state deviation in the average coolant temperature may occur due to the presence of a temperature dead zone.

The large-scale integration of renewable energy sources has imposed significant frequency-regulation stress on power systems. As energy-storage systems can effectively assist in frequency regulation, a comprehensive control strategy for hybrid energy-storage systems was proposed to support primary frequency regulations. First, the charging/discharging power and regulation logic of the acting storage were determined according to the state of charge (SOC) of battery and super-capacitor units. Then, variable-K droop and variable-K inertia controls were coordinated, and an auxiliary frequency-regulation strategy based on an S-shaped function was introduced. When the storage SOC was too low (or too high), the auxiliary control can enhance charging (or discharging) efficiency to accelerate SOC recovery. Finally, a simulation model was built in MATLAB/Simulink and verified against actual grid power-load data. Results show that, compared with the no-storage case, the proposed strategy reduces the maximum frequency deviation by 0.044 Hz and shortens the frequency-recovery time by 0.335 s, effectively mitigating frequency deviation, maintaining SOC, reducing battery switching action times, and extending battery cycle life.

To improve the accuracy and reliability of fault warning for coal mills, a spatio-temporal graph attention network-based fault warning method was proposed. The maximal information coefficient and Top-K nearest neighbours were used to obtain the adjacency matrix adaptively. The original list data were reconstructed into sequence graph data. The spatial and temporal features of the graph data were extracted using the graph attention network and bidirectional gate recurrent unit in turn, and the next time data were predicted. In the offline stage, the overall warning threshold and each component threshold were calculated by the exponential weighted moving average method. In the online stage, a warning signal was issued when the total amount of prediction residuals exceeded the limit, meanwhile the heat map of the exceedance score of each component was plotted. Results show that taking the operation data of the medium-speed coal mill in a cogeneration unit as an example, the proposed method can accurately warn the potential anomalies of the equipment and effectively explain the reasons for the warning, which is better than relevant comparison methods.

A dual-tower monoethanolamine (MEA) solution CO₂ absorption scheme was proposed to solve the problems of high regeneration energy consumption and large equipment size of existing chemical absorption carbon capture systems, which was unfavorable for skid-mounted design. The scheme combined inter-tower gas cooling, rich solution split-stream, and heat pump process to reduce tower height and regeneration energy consumption while enhancing capture efficiency. Through interactive simulation using Aspen Plus and Matlab software, the particle swarm optimization algorithm was applied for global optimization of dual-tower coupling parameters, including lean solution loading, inter-tower gas cooling temperature, lean solution split ratio, and cold rich solution split ratio. Results show that the optimal operating conditions obtained after optimization are 0.291 mol/mol of lean liquid load, 40 ℃ of flue gas cooling temperature between towers, 0.577 of lean liquid diversion ratio, and 0.2 of cold-rich liquid diversion ratio. Coupling various energy-saving processes can significantly reduce the regeneration energy consumption. Under the optimal operating conditions, CO₂ capture rate can reach 93.27%, and the comprehensive optimized regeneration energy consumption can be reduced to 2.924 GJ/t, saving 30.87% energy compared with traditional processes.

To cope with the increasing green power curtailments in energy internet, a cogeneration shared heat storage system using molten salt heat storage to recover green power curtailments was proposed. Two system configurations were designed, namely, using molten salt heat storage to gain main steam (S1) or industrial steam (S2). Simulation models, exergy analysis models, and techno-economic evaluation models were established. Besides, evaluation indexes, including round-trip exergy efficiency, daily profit increment, and dynamic payback period, were proposed to analyze the economics and sensitivity of two systems. Results show that the round-trip exergy efficiencies of S1 and S2 are 54.34% and 43.36%, respectively. The cogeneration shared heat storage system improves the overall technical economy, and the daily profit growth rates of using S1 and S2 on the basis of the original system are 1.94% and 2.44%, respectively.

The "power determined by heat" operation mode constrains the deep peak regulation capacity of combined heat and power units. The molten salt heat storage technique can enhance the operation flexibility of power unit, and achieve the decoupling of heat and power to a certain extent. Taking a 660 MW combined heat and power unit coupled with a molten salt heat storage system that can supply heat as the research object, a calculation method for the safe heating operation boundary was proposed. Based on this safety domain, the changes in decoupling and peak-shaving capabilities were calculated and analyzed, and the energy utilization efficiency and exergy efficiency during the heat storage and release processes were also analyzed. Results show that the proposed boundary calculation method is accurate and can determine the constraints on the fixed values of power and heat when the coupled system supplies heat. The 80 MW·h molten salt heat storage system can provide heating steam at 260 ℃ and 2 MPa with a maximum mass flow rate of 65.92 t/h, and the maximum increase in deep peak-shaving capacity is 7.56% of the rated load. The energy utilization efficiency increases with the increase in the mass flow rate of the unit's extraction steam for heating and the decrease in power generation, while the exergy efficiency increases with the increase in both the mass flow rate of extraction steam for heating and the power generation. When the heat storage power reaches its maximum value (36 MW), the unit achieves the highest energy utilization efficiency and exergy efficiency within the operation domain, which are 79.78% and 46.04%, respectively. When the mass flow rate of steam generated during the molten salt heat release process is the maximum, the unit's energy utilization efficiency and exergy efficiency within the operation domain reach the highest levels, being 80.84% and 46.33%, respectively. In terms of the distribution of heating mass flow rate, giving priority to utilizing the unit's extraction steam heating capacity and then supplementing with the molten salt heat storage system results in higher energy utilization efficiency and exergy efficiency.