Abstract: Objective　With the implementation of ultra-low emission standards, the concentration of filterable particulate matter (FPM) from industrial flue gas has been substantially reduced, whereas condensable particulate matter (CPM) is becoming a dominant contributor to stationary source emissions and ambient PM2.5. Conventional CPM measurement relies on manual sampling and laboratory analysis (e. g., EPA Method 202). However, this approach is limited by positive artifacts (gas absorption), wall losses, low time resolution, and heavy workload, failing to meet the demands of refined air quality management. To address these limitations, this study developed an online continuous CPM monitoring and quality-control device for stationary sources based on a controlled condensation-dilution-drying approach, and systematically evaluated its performance. Methods　The device comprises a thermostatic separation unit, a measurement unit, and an online quality control unit. Flue gas is isokinetically sampled and cooled in a hydrophobic condenser to collect CPM as condensate, which is rapidly transferred to minimize contact with soluble gases. An aerosol generator and drying section convert the condensate into dry particles representing total particulate matter (TPM), whose mass is quantified by a β -attenuation sensor. The CPM concentration is then derived by subtracting the FPM concentration obtained from the continuous emission monitoring system (CEMS). The quality control unit ensures reliability through automatic temperature control, multi-stage cleaning, and blank control. The performance of the device was assessed through laboratory calibration with ammonium bisulfate solutions (1–50 mg/m3) and flue-gas simulation experiments. Specifically, a gas-interference test with 5 μmol/mol SO2 was conducted to verify resistance to gas-absorption artifacts, followed by parallel field measurements at typical stationary sources. Results and Discussion　Laboratory calibration results for the device demonstrate a robust linear response (R = 0.953) within the concentration range of 1–50 mg/m3, and the deviation from the true values for all measurement points is within ±5%. In flue-gas simulations, the measurement errors remain within 10%, and a 24-h continuous run yielded a relative standard deviation (RSD) below 1.77%. Repeated measurements over six days at a nominal concentration of 10 mg/m3 showed high stability (mean: 10.05 ± 0.28 mg/m3). Crucially, gas-absorption experiments revealed that SO2-induced positive bias (1.6 mg/m3) in the online system is significantly lower than that of manual Method 202 (5.5 mg/m3), thus confirming the effectiveness of rapid condensate transfer. Field applications at three industrial sources generate continuous 24-h CPM profiles, which effectively capture emission fluctuations and operational changes. The online CPM data show strong agreement with the results of manual measurements for both inorganic-dominated and organic-rich sources (R = 0.965, p < 0.001). The high-resolution data further reveal the diurnal variation of CPM from a biomass boiler and indicate potential links between CPM formation and desulfurization/denitrification processes. Overall, the developed device provides accurate, stable, and composition-independent online CPM monitoring, offering critical data support for ultra-low emission control. Conclusion　In conclusion, this study developed an online monitoring and quality control device for CPM, addressing the limitations of traditional manual sampling and laboratory analysis. By simulating the condensation process and utilizing β -attenuation sensing, the system enables the transition from discontinuous measurement to continuous online monitoring. The integrated quality control measures of the device, especially automatic cleaning and calibration, minimize errors caused by gas absorption and wall losses. Field applications verify that the device provides 24-h continuous data consistent with EPA Method 202, capturing dynamic concentration changes in flue gas. This technology provides robust data support for stationary source supervision and the implementation of ultra-low emission policies.

Key words: condensable particulate matter, flue gas monitoring technology, online monitoring equipment, stationary source, measurement quality control technology