High-temperature environments can lead to the deterioration of the heat dissipation performance of indirect air cooling towers. Air inflow spray pre-cooling is an effective method to enhance the heat dissipation performance of indirect air cooling towers. Taking a 2×350 MW indirect air cooling unit in northwest China as the research object, a numerical model coupling the spray evaporation with the ventilation and heat dissipation of the indirect air cooling tower is established to study the effect of air inflow spray pre-cooling on the performance of the indirect air cooling tower with different environmental factors. The results show that crosswind can carry the spray downstream, causing the spray to accumulate and benefiting the radiators in the leeward area the most. The performance improvement of the radiators in the windward area decreases with the increasing wind speed, while the radiators in the side area even experience performance degradation at medium to high wind speeds. Additionally, as the wind speed increases, the spray flows out of the annular evaporation zone, resulting in some pre-cooled ambient air failing to enter the radiators and leading to spray waste and reduced effectiveness. The improvement rate of heat dissipation in the indirect air cooling tower after air inflow spray pre-cooling decreases at first and then increases with the increasing wind speed. At an ambient humidity of 40%, the heat dissipation improvement rate decreases from 5.65% at 0 m/s to a minimum of 2.03% at 8 m/s, and then rises to 3.98% at 12 m/s. The effectiveness of air inflow spray pre-cooling weakens with the increasing ambient humidity. Under windless conditions, as the humidity increases from 20% to 80%, the heat dissipation improvement rate of the indirect air cooling tower decreases from 6.4% to 2.4%.

The distribution characteristics of the air flow field inside the natural-draft direct-air-cooling exhaust tower under low-temperature and low-load operating conditions still remain unclear. There is an urgent need to study its variation laws and propose effective measures to ensure exhaust performance and anti-freezing safety. Through the computational fluid dynamics (CFD) numerical simulation, the flow and temperature fields inside the tower at ambient temperatures of –21 ℃ and –30 ℃, and at different wind speeds are analyzed. The results indicate that, based on the symmetrical operation of steam isolation valves for sector switching of the air-cooled condenser, using louvers to regulate airflow in isolated sectors can effectively optimize the internal airflow field, ensure smooth exhaust under low-temperature conditions in winter, and significantly reduce the risk of localized freezing. Field tests verified that this measure can reduce the unit backpressure by approximately 2 kPa and improve the flue gas flow deviation.

As an important parameter reflecting the combustion process, temperature distribution in a furnace is related to the safety, economy and pollutant emission level of the combustion process, which is of great significance for boiler control and the study of the combustion process in the furnace. The radiation imaging method is suitable for reconstruction of furnace temperature field due to its high temporal and spatial resolution and easy implementation on site. An online measurement technology of furnace temperature field based on optical tomography is proposed. A reconstruction algorithm combining deep learning with regularization algorithm is adopted to solve the ill-posed problem in the temperature field reconstruction process. Firstly, a radiation imaging model is established according to the set parameters such as furnace size, medium radiation characteristics, and CCD camera installation position. A large amount of data is obtained through direct problem calculation. Then, the appropriate Tikhonov regularization parameter is found through an automatic optimization algorithm to construct the training data set, and the accuracy and stability of the solution are evaluated. Finally, a deep neural network model is established to predict the optimal regularization parameter and then reconstruct the temperature field. The results show that this furnace temperature field reconstruction algorithm has an error less than 5%, showing good accuracy. After adding the measurement error, the reconstruction error is within 5%, indicating that the method is robust. At the same time, this method has high computational efficiency and meets the requirements of real-time monitoring of temperature fields.

The optimization design of the first domestically produced full-capacity feedwater pump used for No.9 unit of the Huaneng North Power Dalat Power Plant Phase V expansion project (1×1 000 MW) is introduced. The three-dimensional structural model of the feedwater pump is established by using ANSYS Workbench software, and the thermal stress analysis of the pump body and finite-element calculation of the impeller strength are conducted. Moreover, the trial operation of the turbine-driven feedwater pump unit and optimization suggestions are provided. The feedwater pump runs under various load conditions of the unit, ensuring that the feedwater flow and pump outlet pressure meet the operational requirements, with the temperature and vibration indicators of each bearing in the steam pump unit falling within the excellent range. Based on the performance assessment test data of the feedwater pump, the calculated efficiency is 84.32%, which exceeds the guaranteed efficiency value. The successful application of this domestically produced full-capacity steam feedwater pump unit in a 1 000 MW coal-fired unit can provide experience for planned or newly constructed units and has certain reference value.

Abnormal stator core temperatures in generators can lead to serious issues such as aging of insulating materials and winding shorts, thereby affecting the overall performance and lifespan of the generator. This study presents a stator core temperature prediction model for turbo generators based on FFCM-MHDA-iTransformer. It leverages an improved Transformer architecture, namely the inverted Transformer (iTransformer) model, which adopts an inverted time-series encoding approach to address the limitations of the standard Transformer in handling multivariate variable correlations. The model employs fused Fourier convolution mixer (FFCM) to enhance and extract local features from time-series data. Furthermore, the model replaces conventional self-attention with multi-head differential attention (MHDA), effectively reducing attention noise and directing the model’s focus towards critical information. After training and validation, the proposed model demonstrates higher prediction accuracy compared to other mainstream prediction models. It facilitates timely detection of potential faults, preventing shutdowns for maintenance, and holds significant application value for ensuring stable operation of turbo generators. This approach effectively enhances the accuracy and practicality of temperature prediction technology.

When multiple units are used for combined heating, the distribution of thermoelectric loads among the units significantly affects overall energy consumption. For a thermal power plant where Unit 1 and Unit 3 adopt a dual-mode coupled heat-supply method with zero output of the low-pressure cylinder and steam extraction, and Unit 2 and Unit 4 adopt a triple-mode coupled heat-supply method with high back-pressure, heat pump, and steam extraction, an off-design condition model was established using EBSILON software. The thermoelectric characteristics and energy consumption characteristics were analyzed by adjusting parameters such as main steam flow, zero output steam volume of the low-pressure cylinder, heat supply power of the heat pump, and high-back-pressure heat-supply flow rate. The operational boundaries of electrical and thermal loads and the relationship between coal consumption and thermoelectric load were fitted using the least squares method. Under the fixed boundary conditions for the entire plant’s heating load and power supply load, the optimization of thermoelectric load distribution was achieved using particle swarm optimization. The results indicate that large-capacity high back pressure heat pump units should provide heat load, and small-capacity high back pressure heat pump units should provide electric load. After optimization, the total coal consumption of the whole plant was reduced by 0.6~10.0 t/h, resulting in a degree of optimization of 0.3%~3.9%.

The deep peak shaving and flexible operation of thermal power units increase the risk of crack faults in the rotors of main and auxiliary equipment, posing a serious threat to the safe and stable operation of the units. According to the established vibration equation of the cracked rotor, the main vibration characteristics of the cracked rotor are summarized. On this basis, combined with the field diagnosis experience, a practical method of identifying the cracked rotor through vibration analysis is proposed, with criteria including continuously climbing fundamental-frequency vibration and ineffective rotor dynamic balance, continuous increase of second harmonic vibration, abnormal Bode curve, and so on. Finally, three cases of rotor crack fault identification in the operation of a steam turbine, a generator and a boiler primary air fan are given to illustrate the practical application process and accuracy of this method.

In view of the problems that conventional fly ash carbon content prediction models are prone to fall into local optimal solution traps and have insufficient generalization ability, based on the boiler hot-state multi-condition tests, 28 key characteristic parameters are selected through data collection, processing, Pearson correlation analysis of variables, and importance ranking, the sparrow search algorithm (SSA) is used to determine the optimal hyper-parameters of the random forest (RF) model, and an SSA-RF prediction model is constructed. The model verification results show that the root-mean-square error of the SSA-RF model in the training set and the test set decreases to 0.010 8 and 0.019 1 respectively, and the coefficient of determination R² increases to 0.999 7 and 0.998 1 respectively, demonstrating the excellent prediction accuracy and generalization ability of the model. Furthermore, the ISSA-RF-SSA algorithm is proposed. The SSA is improved by integrating multiple strategies to achieve global extreme value optimization of combustion parameters. Engineering verification shows that after optimization, the carbon content in fly ash decreased from 2.500% to 1.345%, and the prediction error was only 0.003 percentage points, verifying the accuracy of the model. The research results indicate that the ISSA-RF-SSA method improved by multiple strategies significantly enhances the optimization performance of the algorithm, providing a new idea for the combustion optimization of coal-fired units.

To address the issues of reduced combustion efficiency and increased pollution caused by the easy deposition of pulverized coal particles from lower burners of coal-fired boilers in cold ash hoppers, a CFD numerical simulation method is used to comparatively analyze the boiler combustion characteristics under the working conditions before the supplementary lifting air is applied, when the swirl burners near the side walls are deflected by 5° toward the center of the furnace, and after the supplementary lifting air is applied. The results show that after the supplementary lifting air is applied, the lifting effect on the lower pulverized coal airflow is enhanced, the deposition amount of unburned carbon particles is reduced by 30.5%, and the burnout rate is increased to 99.44%. The deflection of the burners makes the flame narrow and elongated, reduces the temperature of the side walls, but increases the CO concentration in the cold ash hopper. The supplementary lifting air reduces the CO concentration by enhancing the O₂ supply at the bottom. After the lifting air is supplemented, the air staging is significantly intensified, the reducing atmosphere in the main combustion zone is enhanced, and the NO mass concentration (standard condition) at the furnace outlet is reduced from 315.3 mg/m³ to 282.1 mg/m³. The retrofit of theburner deflection and lifting air at the bottom of the boiler can effectively regulate the pulverized coal transport path, inhibit particle sedimentation, and reduce pollutant emissions. The research results can provide a theoretical basis and engineering practice guidance for related boiler transformations.

The hydrophobic resin-based solid amine adsorbent was prepared by modifying porous materials with different-molecular-weight polyethyleneimine (PEI), by taking hydrophobic oily macroporous adsorbent resin as carrier. The specific surface area, pore structure, functional group structure and thermogravimetric properties of the resin-based solid amine adsorbent were characterized by N₂ isothermal adsorption-desorption, infrared and thermal analysis. The effects of PEI loading, air humidity (30%~80%), adsorption time and multiple cycles on the adsorption performance of CO₂ were investigated. The results show that, the hydrophobic resin-based solid amine adsorbent has good trapping performance for CO₂ in dry air (air humidity is less than 50%). The SD300 resin-based solid amine adsorbent modified by 30%PEI can reach more than 90% of the total adsorption capacity after one hour adsorption in atmospheric environment. When the molecular weight of the PEI is 1 800, it shows high adsorption capacity and good cycle stability of adsorption and desorption, mainly due to the high pore size and its excellent high temperature resistance.