[Objective] As the accelerated transition toward modern power systems with high-penetration renewable energy integration continuously weakens the equivalent rotational inertia of the grid, 1000 MW ultra-supercritical once-through boiler thermal power units face increasingly strict dispatch performance criteria. Due to their inherent thermodynamic lag and strong non-linear characteristics, these megawatt-class units frequently encounter severe command coupling conflicts and destructive energy clashes when executing automatic generation control and primary frequency regulation instructions simultaneously. To fundamentally resolve these operational bottlenecks, this research introduces a model-driven, coordinated optimization control strategy. 

[Methods] First, adopting a hybrid modeling concept, the intricate steam-water cycle of the boiler is spatially lumped to construct a highly efficient 3rd-order affine non-linear dynamic control model. To guarantee wide-range robustness, an objective-adaptive non-dominated sorting genetic algorithm is employed to perform the online calibration of unknown dynamic parameter matrices, specifically targeting pressure conversion coefficients, energy conversion variables, and thermal time constants. Capitalizing on this reduced-order model's internal boundary constraints, state reconstruction technologies are embedded in the control layer. Specifically, the transient reheat steam flow is extracted through a frequency-domain complementary filtering network to effectively eliminate volume-induced phase delays. Concurrently, the actual equivalent coal feed rate is reverse-calculated via a high-order Butterworth band-pass inversion algorithm to successfully strip away coal transportation delays. Furthermore, aligning strictly with the latest grid auxiliary service assessment rules, a comprehensive coordinated control logic strategy matrix is configured. This matrix features active main steam pressure-raising for energy storage at deep peak-shaving loads, adaptive variable-gain compensation algorithms for small frequency deviations, cross-interlocking protection mechanisms to prevent reverse actions during large frequency deviations, and a discrete micro-step decomposition module for continuous automatic generation control ramp commands. 

[Results] Field test data gathered from an industrial unit thoroughly verify the proposed strategy's exceptional efficacy. The reduced-order mathematical model exhibits remarkable tracking fidelity under wide-range load-varying conditions, maintaining an average main steam pressure prediction error of only 3.15%. Following the control logic upgrade, the initial mechanical response delay for primary frequency regulation was compressed to 2.65 seconds. This drove the monthly primary frequency regulation qualification rate from 62.4% to a stable 81.5%, safely surpassing the penalty-exemption baseline. Simultaneously, the comprehensive automatic generation control performance index registered a leap from 0.51 to 1.18, systematically eradicating power oscillations and inverse adjustments traditionally caused by thermodynamic inertia.

 [Conclusion] Balancing computational efficiency and physical fidelity, the proposed scheme establishes a practical engineering paradigm for 1000 MW units to deeply engage in the ancillary service market through pure software logic reconfiguration. It fundamentally breaks through the response bottlenecks of boiler-turbine coordination and significantly strengthens the unit's support capability for next-generation power systems.

[Objective] To improve the smoothness of the meridional flow passage and thereby enhance the flow efficiency of centrifugal compressors, this study develops a quasi-three-dimensional (quasi-3D) aerodynamic design methodology for CO₂ centrifugal compressors, based on a novel meridional passage design algorithm with high-order smooth curvature. The meridional passage is a critical geometric feature of a centrifugal compressor because it directly governs the through-flow velocity distribution, blade loading and internal losses. The design procedure consists of three steps. 

[Methods] First, a supercritical carbon dioxide (S-CO₂) centrifugal compressor is designed and optimized at the one-dimensional (1D) level, through which the key aerodynamic and geometric parameters, such as the specific speed, the flow coefficient and the impeller diameter ratio, are determined. Second, based on the obtained 1D parameters, a meridional flow passage design algorithm with high-order smooth curvature is developed. In this algorithm, the hub and shroud meridional contours are constructed so that the curvature varies continuously from the inlet to the outlet. Compared with the conventional multi-arc meridional contours, which exhibit discontinuous curvature at the arc junctions, the proposed contours are smoother. A discontinuity of curvature causes abrupt changes in the wall curvature and undesirable near-wall flow disturbances, which may promote boundary-layer separation and increase losses; the continuous curvature distribution, in contrast, suppresses such disturbances and yields a more uniform flow and a higher flow efficiency. Third, on the basis of the curvature-smooth meridional passage and the S1/S2 stream surface method, a quasi-3D blade profile design approach for the CO₂ compressor is developed. The S1/S2 method decomposes the three-dimensional flow into blade-to-blade (S1) and through-flow (S2) stream surfaces solved in a coupled manner, enabling an efficient and physically sound design of the blade camber lines and thickness distributions. Applying this methodology yields the impeller design scheme for the recompressor of a 300 MW-class CO2 power generation system with a split-flow recompression cycle. 

[Results] Numerical simulations of the designed compressor show that the total-to-total isentropic efficiencies of the impeller and the diffuser reach 93%~97% or above, fully validating the effectiveness of the proposed method. The stage-to-stage efficiency variation is mainly caused by the decrease in the specific speed of the last stage: as the pressure rises, the CO2 density increases and the volumetric flow rate decreases, which lowers the specific speed and degrades the efficiency of the final stage. 

[Conclusion] The proposed meridional passage and quasi-3D design methods are therefore expected to significantly promote the aerodynamic design research and engineering application of CO2 compressors.

Purpose: Dimensional tolerances in the flow-channel height of metallic interconnects are unavoidable during the manufacturing and assembly of solid oxide cell (SOC) stacks. Such tolerances can lead to non-uniform internal flow distribution and may threaten cell performance and long-term stability. This study aims to evaluate the effects of flow-channel height tolerance on the coupled electrical, flow, and mass-transfer behavior in single-layer and three-layer SOC stacks, and to provide a theoretical basis for rational tolerance specification. 

Methods: A coupled thermal–mass–electrical–chemical multiphysics model was developed for SOC stacks. The model integrates an electrochemical model with momentum, species, and energy conservation equations, allowing analysis of current density, gas-flow distribution, pressure drop, and temperature. For the single-layer stack, four flow-channel height profiles were considered: uniform height, random height, height decreasing toward both sides, and height increasing toward both sides. For the three-layer stack, two comparative cases were further examined: consistent layer heights (DN1) and inconsistent layer heights (DN2). All simulations were performed under constant-voltage operation at 0.8 V. Macroscopic and interfacial current-density distributions, gas distribution, fuel-channel pressure drop, interlayer voltage variation, and the coefficient of variation (CV) of gas flow were systematically compared. 

Results: For the single-layer stack, the difference in macroscopic current density among the four flow-channel height profiles was less than 0.5%, and the maximum relative deviation of the standard deviation of the interfacial current density was only 0.18%. This indicates that the current distribution was insensitive to flow-channel height tolerance within the studied range. However, further analysis of flow and mass transfer revealed hidden non-uniformity. Under all height-deviation scenarios, the fuel-side channel pressure drop increased by approximately 28.6% compared with the uniform case, indicating that the flow field was far more sensitive to height tolerance than the electrical performance. For the three-layer stack, the layer voltages in DN1 were nearly identical, with an interlayer voltage error of only 0.4% and a gas-flow CV of 1.8%, indicating excellent gas-distribution uniformity. In contrast, in DN2, complete hydrogen depletion occurred in the rear region of the top-layer flow channel, concentration polarization intensified, and the voltage decreased continuously; the interlayer voltage error reached 6.7%, and the gas-flow CV was 18.7%, indicating a highly non-uniform gas distribution. These results show that although the current distribution in a single cell is relatively tolerant of flow-channel height deviations, the flow and mass-transfer fields are strongly affected. In multilayer stacks, interlayer height inconsistency can lead to severe local fuel starvation and performance degradation. 

Conclusions: Within the range studied, the flow-channel height tolerance for single cells can be moderately relaxed to reduce manufacturing costs. However, interlayer dimensional consistency must be strictly controlled during stack assembly. The proposed coupled model and findings provide a theoretical basis for formulating reasonable manufacturing tolerances for SOC stacks.

[Objective] Under deep and rapid load changes driven by power dispatch and coal quality fluctuations, pulverizing systems of coal-fired units frequently operate under wide-range variable loads, easily inducing coal blockage and coal interruption faults. These faults are concealed and evolve gradually with hysteretic observable parameters, so traditional fixed-threshold DCS alarms cannot provide accurate early warning, and diagnosis and closed-loop disposal lag behind. To address these problems, an intelligent diagnosis and self-healing control method for coal mill blockage and interruption abnormal conditions is proposed. 

[Methods] For a medium-speed mill positive-pressure direct-fired pulverizing system, a key state parameter model is built on material balance, air-coal transport and motor load characteristics, describing the dynamic coupling among internal material inventories, differential pressure and mill current. A nonlinear state observer with prediction-correction feedback is introduced to estimate the unmeasurable internal inventories online via the measured mill current residual, and model parameters are identified from healthy operating data by the particle swarm optimization algorithm. Diagnostic criteria integrating virtual material states, model residuals and parameter directionality are then constructed based on differential pressure, current, outlet temperature and instantaneous coal feed rate. Four feature quantities are fused through fuzzy quantification into a continuous comprehensive fault index, and a hierarchical warning mechanism comprising early warning and fault warning is established. Self-healing sequential control procedures for the two faults are designed, with maintenance personnel notification and manual intervention links introduced; a cloud-edge coordinated closed-loop diagnosis and self-healing control system is developed, with modules deployed on a hyper-converged server and linked to the DCS via OPC UA and Modbus TCP. 

[Results] Validation on a 4 h independent data set shows that the reconstructed current and differential pressure track DCS measurements well, with RMSE of 1.12 A and 0.04 kPa and mean relative errors of 1.9% and 0.72%. Field application on a 1 000 MW ultra-supercritical unit shows that the system identifies blockage and interruption trends at the early fault stage and achieves self-healing by coordinating coal feed adjustment, hot/cold air regulation and vibration device activation. The blockage and interruption self-healing processes take about 10 min and 5 min; the disturbance to main unit parameters is small, and key mill parameters are smoothly restored to normal ranges. A model degradation monitoring and parameter online updating mechanism based on sliding residual windows further ensures long-term model effectiveness. 

[Conclusion] The proposed method supports the transformation from post-fault alarm to early trend prompting and immediate disposal, providing a reference for intelligent diagnosis, risk suppression and reduced-manpower operation of pulverizing systems under bidirectional disturbances from power dispatch and coal quality.

 [Objective] Carbon emissions from coal-fired power plants vary with combustion organization, flue-gas conditions, and operating-load disturbances. Conventional inventory methods cannot represent these hourly variations in a timely manner, whereas purely data-driven soft sensors may produce estimates that conflict with available process knowledge. This study aims to improve the accuracy and physical consistency of hourly carbon-emission soft sensing and to coordinate the selection of physical regularization and residual-correction parameters under a limited search budget. 

[Methods] The dataset contained 8,760 hourly records from a large coal-fired unit in eastern coastal China, covering 1 January to 31 December 2025. The records underwent temporal-consistency checks, screening of missing and physically invalid values, standardization using training-period statistics, and sliding-window construction. Data from January to October were used for training, November data were used for calibration, and 744 observations from December were reserved for testing. Flue-gas oxygen content, velocity, temperature, humidity, and pressure were used to construct proxy features for oxygen dilution, standard dry flue-gas flow, and combustion intensity. A long short-term memory (LSTM) network extracted current and historical operating information. Flow-direction, oxygen-direction, and temporal-smoothness regularization were incorporated into network training. Ridge regression then learned a mapping of the calibration residuals from 15 physics-related features, forming a physics-guided residual correction LSTM (PGRC-LSTM). An improved heart rate optimizer (IHRO) coupled heart-rate amplitude with fitness feedback, reused known fitness values, and performed a final boundary check in the favorable search interval. IHRO jointly selected one of three pretrained physical-weight models and the ridge regularization coefficient. All compared optimizers used the same candidate set, forward temporal-validation objective, search range, and budget of 90 evaluations. Twenty search seeds were used. Performance was evaluated using mean absolute error (MAE), root mean square error (RMSE), mean absolute percentage error (MAPE), and the coefficient of determination (R2). 

[Results] Under the fixed configurations used for basic-model screening, LSTM achieved the lowest mean MAPE of 6.9833% among the 11 compared models. Incorporating physical regularization reduced the mean MAPE to 6.7923%, and residual correction further reduced it to 5.5528%. The PGRC-LSTM therefore lowered MAPE by 1.4305 and 1.2395 percentage points relative to LSTM and the physics-guided regularized LSTM, respectively. In the conditional direction-consistency analysis, its average combined violation rate was 22.33%, compared with 24.91% for LSTM; this statistic was interpreted together with prediction error rather than as independent proof of engineering improvement. After optimization, IHRO-PGRC-LSTM obtained an MAE of 34,176.5604 kg, an RMSE of 46,830.4620 kg, a MAPE of 4.4631%, and an R2 of 0.9511 on the December test set. All four mean metrics were the best among the models listed in Table 5. The mean MAPE was 1.0897, 0.8342, and 0.3197 percentage points lower than those of unoptimized PGRC-LSTM, Bayesian optimization, and the baseline heart rate optimizer, respectively. Its MAPE standard deviation was higher than that of Bayesian optimization, indicating that the observed advantage primarily concerned mean accuracy rather than the lowest search variability. 

[Conclusion] The proposed method combines weak physical priors, residual correction, and constrained parameter search to improve hourly carbon-emission soft sensing when the required flue-gas inputs are synchronously available. Because online coal quality, total fuel and air flow, and flue-gas carbon dioxide concentration were unavailable, the imposed relationships are directional constraints rather than a strict carbon balance. The findings are limited to the historical unit data and search space examined. Independent-period, cross-season, and cross-unit validation is required before broader deployment. 

[Objective] Excessive horizontal vibration dominated by the once-per-revolution (1X) frequency component was observed in a vertical centrifugal pump during commissioning of a coal-fired power unit. This study aimed to distinguish structural resonance from rotor, electrical, bearing, connection, and operating-condition faults and to evaluate a field reinforcement strategy. 

[Methods] Start-up vibration velocity, spectra, and phase were recorded at three heightwise locations on the motor and supporting cylinders in the horizontal, vertical, and axial directions. Motor no-load measurements were used to assess whether the response depended on hydraulic load, while shutdown external-excitation response tests were used to identify the first-order horizontal natural frequencies of the motor, intermediate cylinder, and support cylinder. Common fault sources were screened using the available online balancing check, connection inspection, operating records, electrical observations, and bearing and electromagnetic frequency features. A frequency-separation margin was calculated to quantify proximity to the 49.95 Hz 1X excitation and was treated as a continuous descriptive quantity rather than a fixed risk or acceptance threshold. The equivalent-stiffness ratio was estimated from the squared natural-frequency ratio only when the same mode was tracked, the boundary conditions were comparable, and the added mass was small. The retained historical records documented the instrument model, measurement positions and directions, operating states, and measured peaks, but did not preserve the complete acquisition configuration; this limitation is stated explicitly. 

[Results] Before reinforcement, the horizontal vibration velocities at measurement points 1, 2, and 3 were 8.37, 6.43, and 5.92 mm/s, respectively, all above the site start-up control value of 4.5 mm/s. Horizontal response increased with structural height and remained dominated by 1X. During no-load operation, the horizontal velocities at the motor non-drive end, motor drive end, intermediate-cylinder upper flange, and intermediate position were 4.09, 3.16, 3.07, and 2.41 mm/s, respectively, while the recorded spectra remained 1X-dominant. The first-order horizontal natural frequencies of the motor, intermediate cylinder, and support cylinder were 48.74, 48.67, and 47.49 Hz, giving separation margins of only 2.4%, 2.6%, and 4.9%. These mutually consistent observations supported a diagnosis of insufficient lateral support stiffness. An adjustable horizontal support was therefore added at the motor top, and axial stiffening ribs were added to the intermediate cylinder. After reinforcement, the corresponding natural frequencies increased to 79.0, 66.5, and 79.0 Hz, and the separation margins increased to 58.2%, 33.1%, and 58.2%. The estimated relative stiffness multipliers were 2.63, 1.87, and 2.77. The horizontal start-up velocities decreased to 2.36, 1.69, and 1.83 mm/s, representing reductions of 69.1% to 73.7%, and all values were below the site control value. 

[Conclusions] The agreement among direction-dependent start-up response, persistent no-load 1X response, natural-frequency proximity, frequency migration, and post-treatment vibration reduction identifies horizontal structural resonance as the dominant fault mechanism in this pump unit. The combined diagnostic path is useful for similar field cases when speed and boundary conditions are stable and competing fault sources have been examined, but the case-specific margins should not be treated as universal acceptance criteria.

[Background] Inconel 617 alloy exhibits a superior combination of high-temperature strength, corrosion resistance, and oxidation resistance, showing broad application prospects in nuclear pressure vessels, industrial gas turbines, and A-USC coal-fired steam units. The lack of systematic development of welding consumables with excellent applicability, combined with ill-defined processing windows for various welding methods, constitutes one of the critical technical bottlenecks restricting its widespread industrial application.

[Objective] This paper systematically addresses the weldability characteristics of Inconel 617, reviewing the research progress on its primary welding processes, welding consumables, and PWHT, with the aim of providing reference for future in-depth research and engineering applications. 

[Analysis] Due to its high alloying element content, Inconel 617 tends to form a high volume fraction of liquid films during the welding thermal cycle, rendering it highly susceptible to solidification cracking and liquation cracking. Consequently, processing parameters such as welding heat input and cooling rate exert a profound influence on the mechanical properties of the joints. Among arc welding processes, TIG welding produces joint properties superior to other arc processes owing to its low heat input and effective metallurgical shielding; furthermore, multi-pulse NG-TIG welding have expanded its applicability across varying plate thicknesses. MIG welding provides high efficiency and degree of automation, but its joint performance is slightly inferior to TIG due to higher heat input—though CMT welding achieves excellent joint quality. SMAW suffers the most severe loss in impact toughness among arc processes, as insufficient metallurgical shielding by the electrode flux coating leads to high inclusion densities in the weld metal. Compared to conventional arc welding, LBW and LHW offer higher power densities and superior heat input control, generally yielding superior overall joint performance relative to TIG. In diffusion bonding, the core issue of oxide and carbide inclusions remains unresolved, resulting in poor metallurgical bonding and severe degradation of ductility and toughness. Regarding welding consumables, solid wires for Inconel 617 have reached industrial maturity, whereas flux-coated electrodes and diffusion bonding interlayers still require compositional optimization to mitigate oxide inclusion formation during processing. Regarding PWHT, Inconel 617 exhibits an excellent solid-solution strengthening response; thus, its welded joints are generally considered suitable for service in the as-welded condition, though stabilization treatment or full solution annealing may be applied when necessary. 

[Conclusion] Taken together, heat input and metallurgical shielding performance represent the central issues influencing the properties of Inconel 617 joints. NG-TIG welding, CMT welding, and laser-arc hybrid welding yield superior joint performance, serving as the most promising joining processes for Inconel 617 applications.

[Objective] High-ratio co-combustion of biomass with coal is an effective strategy for reducing greenhouse gas emissions. However, the release of gas-phase potassium during this process frequently induces operational issues such as ash deposition and slagging. These issues not only drastically reduce heat transfer efficiency and increase maintenance costs but also may lead to unscheduled boiler shutdowns and safety hazards. Therefore, it is of paramount engineering importance to develop reliable online detection methods for gas-phase potassium concentration and systematically investigate the potassium release characteristics and combustion behaviors of biomass-coal blended fuels under different mixing ratios. 

[Methods] In this study, a single-particle combustion experimental setup was established to investigate coal and biomass blends at various ratios using emission spectroscopy. This system achieved the simultaneous detection of flame temperature, gas-phase alkali metal concentration, and thermal radiation energy, thereby facilitating an in-depth analysis of the gas-phase potassium release characteristics of the blended fuels. Bituminous coal and corn stover were mixed using an isocalorific method, based on the heat content of 0.5 mg of bituminous coal as a benchmark. The blended fuels were compressed into structurally stable single pellets under a pressure of 13 MPa using a hydraulic press and subsequently combusted on the custom-built experimental platform. 

[Results]The experimental results demonstrate that the variation characteristics of gas-phase alkali metal concentration and thermal radiation energy can clearly distinguish three distinct combustion stages of the blended fuels: the devolatilization stage, the fixed carbon combustion stage, and the ash stage. Every stage could be distinctly identified through the variation characteristics of gas-phase alkali metal concentration and thermal radiation energy. As the biomass blending ratio increased from 0% to 100%, the peak concentration of gas-phase potassium rose significantly from 2.59 mg/m³ to 16.8 mg/m³, exhibiting a trend consistent with the duration of volatile pyrolysis. Conversely, the char combustion duration shortened by approximately 330 s, the average flame temperature decreased by 237 K, and the thermal radiation energy declined by 31.2%. These findings elucidate the promotional effect of biomass co-combustion on alkali metal release and quantify the significant alterations in fuel combustion characteristics. 

[Conclusion] Biomass cocombustion exerts a pronounced promoting effect on the release of gas-phase potassium and significantly alters the overall combustion characteristics of the fuel, including accelerating fixed carbon burnout, lowering flame temperature, and reducing thermal radiation output. The findings of this research provide crucial experimental data and theoretical support for the optimal design of biomass-coal co-firing boilers, the adjustment of operational parameters, and the development of effective alkali metal-related problem prevention and control strategies in industrial applications.

[Objective] Hot dry rock (HDR) possesses unique advantages of enormous reserves, high operational stability, and excellent long-term sustainability, and has been widely regarded as a critical strategic alternative energy resource by countries worldwide in recent years. Enhanced Geothermal System (EGS) represents the most effective technical means for the efficient development of HDR resources, which consists of two interconnected key processes: the underground thermal energy extraction from geothermal reservoirs and the above-ground energy conversion and power generation. However, most existing studies have investigated the thermal extraction performance of HDR reservoirs and the power generation performance of surface systems independently. This decoupled research paradigm fails to fully capture the mutual interactions and inherent coupling relationships between the two subsystems, making it difficult to accurately characterize the overall operational characteristics and identify the optimal performance improvement directions of EGS. Therefore, this study aims to systematically reveal the coupling mechanism between thermal energy extraction and energy conversion processes in EGS, and to provide a comprehensive theoretical basis for the holistic optimization of EGS systems.

[Methods] In this study, two representative EGS configurations are proposed: the water-based EGS (water-EGS) using water as the heat extraction medium and the carbon dioxide-based EGS (CO2-EGS) using supercritical carbon dioxide (scCO2) as the working fluid. Based on the fundamental theories of reservoir engineering, heat transfer, fluid mechanics, and engineering thermodynamics, integrated numerical optimization models coupling thermal extraction and power generation processes are established for both systems respectively. The developed models fully account for the complex physical processes in underground geothermal reservoirs, including heat conduction, thermal convection, and fluid seepage, as well as the operational characteristics of surface power generation systems, such as the performance of heat exchangers, turbines, and condensers. On this basis, a systematic comparative analysis is conducted through numerical simulations to evaluate the dynamic production capacity at the production wellhead and the dynamic power generation performance of water-EGS and CO2-EGS under identical geological conditions (including reservoir temperature, permeability, and thickness) and meteorological conditions (including ambient temperature and solar radiation intensity).

[Results] The simulation results demonstrate that the supercritical CO2 power generation system exhibits significantly higher thermal energy utilization efficiency, with its maximum value reaching 2.49 times that of the water-Organic Rankine Cycle (ORC) power generation system. However, under the same reservoir temperature and reinjection mass flow rate conditions, the water-ORC power generation system outperforms the scCO2 system in terms of heat extraction power and net electrical output power. This finding clearly indicates that the two EGS configurations possess distinct technical strengths and exhibit complementary advantages in different performance indicators.

[Conclusion] This study establishes a fully coupled numerical simulation framework covering the entire EGS process from underground heat extraction to surface power generation, which effectively overcomes the limitations of traditional decoupled research approaches. The research findings provide important theoretical basis and technical support for the optimal design, performance enhancement, and efficient development of enhanced geothermal systems. Furthermore, the comprehensive comparative analysis of water-EGS and CO₂ -EGS offers valuable reference for the rational selection of heat extraction media in practical HDR development engineering applications, and lays a solid foundation for subsequent long-term operational performance evaluation and economic feasibility analysis of EGS projects.

[Objective] In the context of China’s “Dual-Carbon” goals and the low-carbon transformation of coalfired power generation, this study investigates the economic performance and carbon reduction characteristics of ammonia co-firing in coal-fired power units. The effects of unit capacity, ammonia co-firing ratio, and fuel prices on the incremental cost of electricity and carbon emission reduction are analyzed to provide theoretical support and data references for low-carbon retrofits of coal-fired power plants and the application of green ammonia. 

[Methods] Coal-fired power units with capacities of 300 MW, 600 MW, and 1 000 MW were selected as the research objects. An incremental cost of electricity model was established based on the concept of the Levelized Cost of Electricity (LCOE). The incremental electricity generation costs under different coal prices, ammonia prices, and ammonia cofiring ratios were calculated. In addition, annual coal consumption and annual CO₂ emission reductions were estimated using unit coal consumption rates, the calorific value relationship between coal and ammonia, and the CO₂ emission factor of standard coal. A comparative analysis of the economic and environmental performance of different unit capacities was then conducted. 

[Results] The results indicate that the incremental cost of electricity is mainly influenced by the ammonia co-firing ratio and ammonia fuel price, and increases approximately linearly with the co-firing ratio. Higher ammonia prices significantly increase power generation costs, whereas higher coal prices reduce the fuel price gap between coal and ammonia, thereby lowering the incremental electricity cost. Compared with the 300 MW and 600 MW units, the 1 000 MW unit exhibits lower incremental electricity costs and superior economic performance under identical operating conditions due to its lower coal consumption rate and higher generation efficiency. In terms of carbon reduction, annual coal consumption decreases substantially and annual CO₂ emission reductions increase significantly with increasing ammonia co-firing ratios. The 1 000 MW unit achieves the largest reduction, reaching 1.0693 million tons of CO₂ annually at a 30% ammonia co-firing ratio. 

[Conclusion] An economic assessment model incorporating four dimensions—coal price, ammonia price, unit capacity, and ammonia co-firing ratio—was established based on the actual coal price index of Caofeidian in April 2026 and the market pricing ranges of green ammonia and gray ammonia. Ammonia co-firing can effectively reduce coal consumption and CO₂ emissions, representing a promising pathway for the low-carbon transformation of coalfired power generation. Large-capacity, high-parameter units demonstrate superior economic and carbon reduction performance and are therefore more suitable for future large-scale ammonia co-firing applications. However, the high cost of ammonia remains a major barrier to commercialization. Future development should focus on low-cost green ammonia production technologies, carbon trading mechanisms, and low-carbon subsidy policies to further enhance the economic viability and engineering application prospects of ammonia co-firing technology.