The enforcement of the policy, justice of punishment, and the fairness of the emission fee depend on the quality of the testing results. Hence, testing companies should provide accredited measurements. The proficiency test is a common way to evaluate the quality of the testing companies. This project develops a simulation stack to enforce proficiency tests and a math model to predict the standard value of the stack. It also improves testing companies' testing ability and can determine the source of bias. The stack is 60 cm in diameter and can provide a testing velocity of 3 – 16 m/s and a mass concentration of 4 -300 mg/Nm3. The test comprises three velocities, two mass concentrations, and one gas composition. The working items include maintenance, calibration, routine tests, model development, and historical data evaluation. The direct weighting method and liquid washing method for sample recovery and the feasibility of the A101 method uncertainty are evaluated in this study, too. The repair, maintenance, and calibration were completed, and the performance of the stack was similar to previous testing parameters. Twenty-three proficiency tests had been completed. The new velocity and mass concentration models were established with new and historical data. The historical score analysis and qualified ratio in each year were also analyzed. The direct weighing method is suggested to avoid the reaction between particles and liquid. The Top-down method includes human error, and the Bottom-up method is capable of analyzing the sources of the bias. From the viewpoint of a method, the Top-down method can be used to evaluate all sources of bias, and the Bottom-down method can be used to evaluate the uncertainty caused by instrument calibration, specification, and resolution.

PM sampling, Stack, Uncertainty