[Objective] With the continuous advancement of China’s electricity spot market reform, electricity prices experience real-time fluctuations driven by dynamic supply and demand. Concurrently, under the national "dual carbon" strategic goals, deep peak shaving of thermal power units has become normalized. This operational shift causes grid dispatch commands to exhibit intense randomness and high-frequency volatility. Traditional prediction models exhibit insufficient accuracy in predicting coal consumption and suffer from severe transient response lag when thermal power units operate under deep peak shaving conditions. To address these critical issues, this paper proposes a dual-stream fusion prediction model based on conditional variational autoencoder and extreme learning machine (CVAE-ELM), incorporating multi-scale spatiotemporal feature fusion. The primary objective is to provide power plants with a highly accurate and fast-responding tool to determine their marginal costs, thereby supporting precise bidding strategies in the electricity spot market. 

[Methods] The proposed CVAE-ELM model adopts a sophisticated dual-stream architecture to capture complex operational dynamics. First, multi-scale convolution operations and an attention mechanism are integrated into the encoder of the conditional variational autoencoder (CVAE). This configuration successfully achieves the decoupled extraction of strongly coupled operational  parameters and weakly coupled time-delay features. Second, the model establishes a dual-stream framework consisting of a deep feature stream (designated as Path A) and a raw feature stream (designated as Path B). Within this framework, residual channels are strategically utilized to capture high-frequency transient characteristics under sudden load variations. To evaluate the performance of this model, comprehensive validation was conducted using actual measured operational data collected from a real 600 MW thermal power unit under various complex working conditions. 

[Results] The experimental results demonstrate outstanding predictive performance and robust stability across diverse and changing operating conditions. The root mean square error (RMSE) of the coal consumption prediction under variable working conditions is precisely restricted to 11.047 g/(kW·h). Furthermore, the average relative error of the model is strictly controlled within a narrow margin of less than 1%. Compared with conventional single-line or single-stage prediction models, the transient response speed of the proposed dual-stream fusion model is significantly accelerated, showing an improvement of 15 % to 30 %. 

[Conclusion] This study demonstrates that the developed CVAE-ELM model achieves superior prediction accuracy when dealing with highly nonlinear relationships compared to traditional reference models, such as the standard Extreme Learning Machine (ELM) and Support Vector Regression (SVR). Even under highly complex and volatile operational conditions, the model maintains high precision and remarkable stability in coal consumption forecasting. In conclusion, this research provides an essential and precise marginal cost foundation for thermal power plants, enabling them to optimize their competitive bidding strategies within the reforming electricity spot market.

[Objective] The lower-furnace spiral tube coils of a 660 MW ultra -supercritical coal-fired boiler are subjected to high heat flux and strong thermal deviation conditions. Under such conditions, the flow and heat transfer characteristics of the working fluid inside the tu bes directly affect boiler operation safety. Therefore, it is necessary to clarify the flow and heat transfer behavior of supercritical water in inclined internally ribbed tubes under actual operating conditions, especially in the pseudo -critical region wh ere thermophysical properties vary sharply. [Methods] Based on the actual operating conditions of the boiler, a typical inclined straight tube section in the lower-furnace spiral tube coils was selected as the research object. A three-dimensional numerical model of an inclined internally ribbed tube under half-circumferential heating was established. The shear stress transport (SST) k-ω turbulence model was used to numerically investigate the flow and heat transfer characteristics of supercritical water. Buoyancy and thermal acceleration effects were evaluated using dimensionless criteria. The circumferential wall temperature distribution and the radial distributions of turbulent kinetic energy, turbulent viscosity, wall shear stress,and specific heat capacity were analyzed at three typical cross sections along the heated 2 tube length. The influence of load reduction on heat transfer enhancement in the pseudo -critical region was also discussed. [Results] Under boiler maximum continuous rating (BMCR) load condi tions, the calculated buoyancy and thermal acceleration criteria are at low levels, indicating that their effects on flow and heat transfer are weak. The inner wall temperature shows an obvious non -uniform circumferential distribution, which is significant ly affected by the flow structure induced by the spiral inner ribs. The rib -root temperature is generally higher than the rib-top temperature, and the maximum inner wall temperature appears near the intersection of the rib root and the downstream side. In the pseudo-critical region, the heat transfer coefficient increases obviously, indicating heat transfer enhancement. However, turbulent kinetic energy, turbulent viscosity, and wall shear stress do not increase significantly, which suggests that the enhanc ement is not mainly caused by turbulence intensification. Instead, it is dominated by the sharp increase in specific heat capacity in the near -wall region. With decreasing boiler load, heat transfer enhancement in the pseudo -critical region becomes stronge r, mainly because the near-wall specific heat capacity further increases rather than because of enhanced turbulence, buoyancy, or thermal acceleration effects. [Conclusion] This study reveals the flow and enhanced heat transfer mechanisms of supercritical water in inclined internally ribbed tubes under actual operating conditions, and provides a reference for analyzing water-cooled wall heat transfer characteristics and evaluating the safe operation of ultra-supercritical boilers.

[Objective] With the introduction of the "dual carbon" goals, biomass fuels have gained widespread attention due to their abundant supply and zero -carbon-emission characteristics, leading to the rapid development of direct biomass combustion and co -firing technologies in pulverized coal boilers. Unlike the sulfuric acid corrosion in traditional coal -fired boilers, the low -temperature heating surface corrosion of biomass boilers is mainly caused by deliquescent chlorides. It is primarily driven by deliquescent NH 4Cl and CaCl 2, which can absorb moisture from unsaturated vapor, thereby raising the corrosion temperature. Meanwhile, Cl - ions further promote corrosion, increasing the difficulty of material selection for low -temperature heating surfaces in biomass boilers. 

[Methods] In this study, NH 4Cl and CaCl 2 were used as representative deliquescent salts. Corrosion experiments under simulated flue gas conditions were conducted on three commonly used low -temperature heating surface materials—20G, ND steel, and Corten steel. The corrosion rate, micromorphology, and corrosion products were analyzed using the weight loss method combined with SEM, EDS, and XRD. To match th e temperature of the low -temperature heating surfaces in boilers, the experimental temperature ranged from 60 °C to 120 °C. 

[Results]The following conclusions were obtained: at different temperatures, the corrosion resistances of 20G, ND steel and Corten steel showed only minor differences. When loaded with deliquescent salts, the corrosion rate increased as the temperature decreased. Apart from that, for the loaded NH 4Cl, the corrosion 2 rebounded slightly above 100 °C due to the release of HCl from parti al decomposition. At all experimental temperatures, the corrosion rate of NH 4Cl was significantly higher than that of CaCl 2, and the difference between the two increased as the temperature decreased. 

[Conclusion] The deliquescence ability of salts does not fully correspond to their corrosiveness; instead, corrosiveness correlates strongly with the pH of the solution, and salts with lower pH values exhibit stronger corrosion ability. For example, the critical relative humidity of NH 4Cl is much higher than th at of CaCl2, yet the corrosion rate of NH 4Cl was consistently higher than that of CaCl 2 at all experimental temperatures, and the lower the temperature, the larger the difference in their corrosion rates. When the salt with strong deliquescence ability is mixed with the salt with weak deliquescence ability, the deliquescence ability will not be strengthened, but the corrosion ability of the mixture may be stronger than that of either individual salt. For example, NH4Cl was mixed with CaCl2 at 100 ℃. CaCl2 has a stronger deliquescence ability. This enhanced the corrosion ability of NH4Cl.

[Objective] To address the problems of excessively high exhaust gas temperature, exceeding the design value by more than 40 ℃, and low -temperature corrosion of the air preheater in a 1000 MW ultra -supercritical 2 once-reheat coal-fired unit, an optimization strategy for cascade recovery and utilization of flue gas waste heat was proposed to improve unit thermal economy and operational reliability. 

[Method] Based on the Flownex simulation platform and field test data obtained from actual unit operation, a thermodynamic model of the flue gas waste heat recovery system was establish ed and validated. According to the principle of energy cascade utilization, the flue gas temperature was divided into high -, medium-, and low -temperature regions. Four retrofit schemes were developed and comparatively analyzed, including a combined low -temperature economizer and air heater system, a flue gas –air cascade waste heat recovery system, an air preheater bypass cascade flue gas waste heat recovery system, and a multi -stage coupled cascade utilization system for flue gas waste heat. The thermal performance and energy -saving characteristics of different schemes under variable operating conditions were evaluated in terms of exhaust gas temperature reduction, equivalent enthalpy drop of extraction steam, coal-saving effect, and turbine heat consumption rate. 

[Results] The simulation results show that the deviation between the calculated and measured exhaust gas temperatures was within 1%, indicating good model accuracy and reliability. With the decrease in exhaust gas temperature, the additional power g eneration from extraction steam replacement and the equivalent enthalpy drop per unit extraction steam increased significantly, resulting in enhanced coal-saving performance and reduced turbine heat consumption rate, while the energy -saving benefit in the low-temperature region gradually diminished. The application of the secondary air duct heater effectively mitigated the low -temperature corrosion problem of the air preheater and simultaneously recovered part of the flue gas waste heat, thereby improving b oiler efficiency. Under 100% turbine heat acceptance (THA) condition, the optimal bypass ratios of Scheme 3 and Scheme 4 were 0.12 and 0.18, respectively. The corresponding coal-saving effects reached 5.464 g/(kW· h) and 6.868 g/(kW· h), while the turbine he at consumption rate decreased by 54.110 kJ/(kW· h) and 71.919 kJ/(kW· h), respectively. Among the four retrofit schemes, Scheme 3 and Scheme 4 exhibited superior comprehensive energy -saving performance, especially under high -load conditions. 

[Conclusion] Scheme 1 and Scheme 2 are more suitable for low -load operating conditions and exhibit stronger adaptability to coal quality fluctuations, with relatively low investment cost but limited energy -saving potential. In contrast, Scheme 3 and Scheme 4 demonstrate r emarkable advantages in reducing coal consumption and turbine heat consumption rate. Scheme 4 is more applicable to units requiring higher energy -saving performance, whereas Scheme 3 achieves a better balance between technical feasibility and economic effi ciency. The proposed Flownex -based cascade flue gas waste heat recovery optimization strategy provides an effective technical pathway for energy-saving retrofit and efficiency improvement of large-scale coal-fired power units.

[Objective] To address the unclear dynamic mechanisms and the lack of research on thermo -electric coupling fault characteristics during the discharge and power generation process of a 350 MW Advanced Adiabatic Compressed Air Energy Storage (AA -CAES) system, this paper c onstructs a dynamic model for the discharge power generation side. This model can reflect the internal dynamic behaviors of key components and grid-connection transient characteristics, aiming to provide a theoretical basis for the dynamic performance evaluation and fault mechanism exploration of large -scale AA -CAES systems. 

[Methods] A dynamic model encompassing air storage tanks, control valves, multi -stage expanders, heaters, and a fourth -order synchronous generator was built based on the Python platform . On this basis, dynamic simulation research was conducted under various operating conditions and disturbance scenarios. The analysis includes the impact of different initial air storage pressures and control methods on the system's dynamic performance, as well as the transient characteristics under heat -side faults, electrical -side short -circuit faults, and thermo -electric coupled complex faults. Through comparative simulations, the response characteristics of system variables such as output power, terminal voltage, valve opening, and control pressure were investigated to reveal the influence patterns of 2 operating conditions and fault combinations on the stability and recovery capability of the power generation side. 

[Results] The simulation results indicat e that the initial air storage pressure and control methods significantly affect the system's dynamic response and output power. A higher air storage pressure leads to a stronger power increase capability after a disturbance and a relatively smoother regulation process. The fault analysis results show that heat -side faults cause the valve opening to continuously increase, reducing the regulation margin. Electrical-side short-circuit faults mainly affect the recovery process of terminal voltage and output po wer, and different fault types cause varying degrees of disturbance. When a heat -side fault is superimposed on a short-circuit fault, the system's recovery process becomes more complex, and the control pressure further increases. 

[Conclusions] The dynamic model of the power generation side constructed in this paper can effectively reflect the transient response characteristics during the discharge power generation process and under grid-connection disturbances. It enables dynamic simulation and transient an alysis under multi -operating conditions and multi -fault scenarios, revealing the influence patterns of initial air storage pressure, control strategies, and thermo-electric coupled faults on system performance. This study can provide a valuable reference for the dynamic analysis, control optimization, and fault mechanism research of the AA -CAES discharge power generation side, while also providing technical support for the safe and stable operation and future engineering applications of large-scale AA-CAES power stations.

[Objective]To reveal the impact mechanism of biomass co -firing on the flue gas temperature distribution along the flow path in a 660 MW W-flame boiler, and to provide a theoretical basis for optimizing the co-firing operation of large -scale W -flame boilers. 

[Method]Taking a 660 MW supercritical W -flame once-through boiler as the research object, sugarcane leaves were selected as the biomass fuel under the 100% 2 THA (Turbine Heat Acceptance) condition. A thermal calculation model was constructed based on the boiler's heating surface structure. Using an equivalent fuel processing method, five co -firing ratios (2%, 5%, 10%, 20%, and 30%) were established based on the lower heating value. The variation characteristics of the flue gas temperature at the outlet of each heat exchanger along the flow path were systematically investigated. 

[Results] The results indicate that biomass co -firing does not alter the basic rule that the flue gas temperature at the heat exchanger outlets decreases progressively along the flow direction. The heat exchangers along the flow path can be divided into three combined zones based on the temperature variation characteristics, and the trends of outlet flue gas temperatures within the same zone are highly consistent. As the co-firing ratio increases from 2% to 30%, the flue gas temperatures at the outlets of the furnace and the platen superheater decrease simultaneously, with reductions of 33 ° C and 16 ° C, respectively, while the theoretical combustion temperature drops by 142 ° C. The outlet flue gas temperatures of the final superheater and final reheater exhibit a consistent trend of "initially decreasing, then increasing, and eventual ly stabilizing," with fluctuation ranges of 6 ° C and 9 ° C, respectively. Conversely, the outlet flue gas temperatures of the low -temperature reheater, low -temperature superheater, economizer, and air preheater all consistently increase. Notably, the outlet flue gas temperature of the air preheater rises from 140 ° C to 158 ° C, marking an increase of 18 ° C. 

[Conclusion] When biomass is co-fired in a 660 MW W-flame boiler, the variation in the outlet flue gas temperature of heat exchangers within the same char acteristic zone exhibits high consistency. Co -firing has the most significant impact on the flue gas temperatures in the furnace and the upper radiant zone. The flue gas temperatures of the heat exchangers in the horizontal flue are jointly regulated by upstream temperatures and heat redistribution. Furthermore, the consistent increase in flue gas temperatures across the heat exchangers in the rear pass leads to the accumulation of flue gas sensible heat at the tail, which may reduce the thermal efficiency of the boiler.

[Objective] Premixed flame stability is essential for the efficient, low -emission, and safe operation of gas turbines, aero-engines, industrial combustors, and distributed energy systems. With the increasing use of lean premixed combustion, hydrogen-enriched fuels, strong swirl stabilization, compact combustor designs, and thermal-boundary coupling, premixed flames have become more sensitive to local stretch, differential diffusion, wall heat loss, flow non -uniformity, and unsteady coherent s tructures. Conventional stability indicators, such as the limiting equivalence ratio and the critical flow velocity, can describe the global operating boundary of a combustor. However, they cannot adequately explain where instability is initiated, which lo cal mechanism dominates, or how local flame weakening develops into blow -off, flashback, or stabilization -mode transition. Therefore, premixed flame stability should be analyzed by linking global stability limits with local flame response and combustor-scale flow-thermal coupling. 

[Progress] This paper reviews the mechanisms and characterization methods of premixed flame stability. The theoretical basis is summarized from the perspectives of velocity matching, the Damkö hler number, flame stretch, flame curvature, Markstein response, Lewis -number effects, Karlovitz-number-based turbulent combustion regimes, and thermal-boundary coupling. These concepts show that flame stability is not governed by a single parameter. Instead, it depends on the balance among local flame propagation, residence time, strain rate, curvature, scalar dissipation, differential diffusion, and heat exchange. Turbulent premixed combustion regime diagrams provide a useful scale background for evaluating turbulence-flame interaction. However, global regime parameters should not be directly used as local extinction or flashback criteria, especially in flame roots, shear layers, recirculation -zone boundaries, and near -wall regions. Experimental methods are also reviewed, including particle image velocimetry, planar laser-induced fluorescence, high-speed chemiluminescence imaging, pressure measurement, wall -temperature measurement, and multimodal synchronized diagnostics. These methods help identify flame -root motion, recirculation -zone dynamics, local intermittency, heat-release oscillations, and near -limit transition pathways. Existing studies indicate that lean blow-off, flashback, and stabilization -mode transition usually do not result from simultaneous fail ure over the whole flame surface. They are more often caused by the accumulation of local events, including reaction weakening, flame -root detachment, local extinction, downstream residual reaction, hot -product recirculation, intermittent re -ignition, and final re -ignition failure. The roles of direct numerical simulation, large eddy simulation, and conjugate heat transfer simulation are further discussed. Direct numerical simulation is suitable for resolving detailed chemistry, molecular transport, scalar dissipation, local extinction, re -ignition, and radical transport in canonical flames. Large eddy simulation is useful for predicting coherent structures, flame -root dynamics, heat-release oscillations, and near-blow-off behavior in practical combustors. Its reliability depends on subgrid combustion models, inlet turbulence, grid resolution, chemical mechanisms, and wall-boundary treatment. Conjugate heat transfer simulation extends stability analysis to solid conduction, wall heat flux, stabilizer temperature, and thermal feedback. It is important for compact combustors and near -wall flame stabilization. 

[Conclusion] Premixed flame stability is controlled by the competition between local reaction -sustaining capability and flow -induced stretch, mixing dissip ation, residence time, differential diffusion, and heat loss. Global stability boundaries should be understood as the macroscopic outcome of local flame response and combustor-scale flow -thermal coupling. Future studies should establish cross -scale relatio nships between local instability indicators and global stability limits. Experimental validation using multimodal diagnostics should be strengthened. Numerical modeling frameworks based on direct numerical simulation, large eddy simulation, and conjugate h eat transfer simulation should also be further improved. These efforts will support the prediction, characterization, and control of premixed flame stability in complex low-emission combustors.

This paper proposes a novel compressed air energy storage (CAES) system that fundamentally integrates gravity to achieve constant -pressure operation. By simultaneously storing thermal energy, pressure potential energy, and gravitational potential energy, this innovative system facilitates highly efficient and stable-pressure energy storage and release performance. To thoroughly understand its performance, the study employs a comprehensive methodology combining mathematical modeling, thermodynamic simulations, and theoretical analysis to investigate the underlying energy storage mechanisms and the specific characteristics of energy and exergy storage. Furthermore, a scaled -down experimental prototype of the gravity -enhanced constant-pressure air reservoir was constructed and tested, successfully verifying its dynamic pressure -stabilizing capabilities under realistic operating conditions. Theoretical analysis yields several critical insights into the system's thermodynamic behavior. It reveals that the system's total energy storage is predominantly in the form of compressed thermal energy, whereas exergy storage is primarily driven by pressure potential energy. Gravitational 2 potential energy accounts for less than 1% in both metrics. Despite this, the system demonstrates exceptional thermodynamic performance, achieving a maximum round -trip efficiency of 68.33%. When compared under identical heat storage temperatures, this novel system consistently outperforms traditional constant-volume CAES systems. Moreover, the constant-pressure feature provides a massive advantage in spatial efficiency. For a given power generation capacity, the required air storage volume is drastically reduced compared to traditional constant-volume systems, with the maximum volume reduction reaching an impressive 91.07%. This significant reduction in reservoir size can substantially lower capital costs and alleviate geographical constraints for large-scale energy storage deployment. Experimental results further validate the system' s operational robustness, demonstrating that the gravity -enhanced constant-pressure reservoir can maintain a highly stable pressure output even when subjected to significant fluctuations in inlet and outlet flow rates. However, the study also identifies a practical engineering challenge: real -world applications must carefully account for pressure losses caused by mechanical friction within the gravity -based pressure -regulating mechanism. Additionally, economic analysis indicates that the novel system posses ses promising potential in terms of cost and revenue. In conclusion, the gravity-based CAES system offers a compelling solution to the inherent inefficiencies of traditional systems. By maintaining constant pressure during both the charging and discharging phases, it effectively prevents compressors and expanders from operating under off-design conditions. Therefore, the operating efficiency of the compressors and expanders can be maintained at the designed level, leading to round -trip system efficiency improvement. This not only enhances overall system efficiency but also dramatically reduces the required air storage volume. Ultimately, this research provides a novel, highly promising technical pathway for the widespread commercial adoption and scaling of compressed air energy storage technologies in future renewable energy grids.

[Objective] Co-firing of biomass and coal is a pivotal technology for the low -carbon transformation of coal-fired power plants, as biomass is a carbon -neutral renewable energy source with abundant reserves that can 2 effectively reduce carbon emissions from power generation. However, biomass fuels generally contain high concentrations of alkali metal elements, especially potassium. During combustion, these alkali metals are released into the gas phase in large quantities, condense, deposit and sinter on the s urface of boiler convection heating surfaces, causing severe fouling and slagging problems. This reduces the heat transfer efficiency of heating surfaces, increases boiler exhaust heat loss and power generation coal consumption. The traditional offline evaluation method of heating surface scaling has a large lag and cannot meet the actual requirements of real -time monitoring and precise prevention and control of industrial sites. Therefore, it is importance to realize the dynamic evaluation and prediction of the fouling characteristics of heating surfaces during the combustion of coal and biomass blended fuels. 

[Methods] In this study, the designed relevant experimental tests were carried out on a 5MW pilot furnace. Optical detection equipment was used to collect real-time data, including flame temperature and gaseous potassium concentration. They were applied to dynamically evaluate the fouling trend of heating surfaces during coal-biomass coupled combustion . the evaluation results were verified by combining the changes in the heat effectiveness coefficient and the actual observed fouling status on the surface of the fouling tubes. 

[Results] The experimental results showed that for every 1% increase in the furnace outlet oxygen concentration, the flame temper ature and gaseous potassium concentration increased by approximately 22.5℃ and 2.0 mg/m³. When the blending ratio of corn stover increased from 0% to 100%, the gaseous potassium concentration reached a maximum of 15.71 mg/m³, respectively. The heat transfer effectiveness coefficient decreased from 0.9 to around 0.5, and the surface fouling degree of the fouled tube significantly intensified. There is a significant correlation between the increase amplitude of gaseous potassium concentration and the attenuation amplitude of the heat effectiveness coefficient. 

[Conclusion] The established dynamic fouling assessment method based on in -situ measurements of flame temperature and gaseous potassium presents favorable prediction accuracy. The obtained experimental f indings can provide fundamental theoretical guidance and database for developing anti-fouling technologies and optimizing operational parameters for high alkali fuel combustion. Eliminating the inherent delay of traditional off -line detection techniques, t his method possesses promising application prospects for on -line fouling monitoring of industrial boilers.

[Objective] Aiming at the phenomena of poor purification efficiency of Heat-Stable Salts (HSS) and high amine loss caused by CO2 loading in lean amine solutions during the ion exchange purification process of flue gas decarbonization amine liquid, this study investigates the influence mechanism of CO2 loading on HSS removal and amine loss. The primary goal is to provide comprehensive theoretical and experimental support for the optimization of the amine liquid purification process using the ion exchange method. 

[Methods] Taking the simulated amine solution containing heat-stable salts (HSS) and specific CO2 loadings as the research object, both static and dynamic ion exchange experiments were conducted to thoroughly explore the specific impacts of CO2 loading on the purification process. In conjunction with comprehensive mass balance calculations, the migration patterns of amine species and the underlying loss mechanisms behind them were systematically analyzed. Furthermore, a mathematical model was constructed to precisely quantify the relationship between the amount of amine loss and the amount of HSS removal. This model was subsequently utilized to compare the loss -reduction performance 2 brought by various optimized processes under different operational conditions. 

[Results] Static ion exchange experiments indicated that an increase in CO2 loading significantly reduces the effectiveness of the resin in removing HSS from the solution. Dynamic ion exchange experiments further revealed that the resin exhibits no significant adsorption selectivity between CO2-loaded anions and HSS anions, leading to competitive adsorption. Following the amine liquid purification process, while the HSS mass fraction was successfully reduced from 5 % to 1.4%, a concurrent amine loss of 2.90% by mass fraction was generated. Detailed analysis showed that 96.72% of this total loss originated from the adsorption loss of carbamate species onto the resin. The constructed model for quantifying the relationship between amine loss and HSS removal was plotted into analytical charts, enabling the objectives of horizontal process comparison and trend prediction. Three optimization processes —namely pre -alkalization transformation, secondary desorption, and internal circulation —were proposed and compared with t he traditional process through model simulation, and the results demonstrated that all three optimized processes could achieve loss reduction compared to the traditional process. 

[Conclusion] High CO2 loading in flue gas decarbonization amine liquid is the core factor restricting the effectiveness of the ion exchange purification method. The competitive adsorption between CO2 loading anions and HSS anions inevitably reduces purification efficiency. Crucially, the adsorption of carbamate from the CO2 loading anions onto the resin is the direct cause of significant amine loss. These findings underscore the necessity of managing CO2 levels prior to ion exchange to ensure process sustainability.