The implementation of environmental policies needs precise and accurate measurements. Proficiency testing is essential to improve the quality of the measurements. The simulated emission stack established by Taiwan NIEA can generate stable and controllable flue gas velocity and mass concentration for the stack emission testing laboratories to carry out the proficiency testings. By accumulating testing data, the relationship between stack parameters and flue gas velocity and mass concentration can be used to evaluate the proficiency of the testing laboratories. 22 tests had been completed before the final-term report and all items listed in the contract were finished. The low mass concentration test was conducted in 19 tests. The contents of the test included gas composition, flue gas velocity, and mass concentration. The test data was analyzed by linear regression model and the standard error was used as a reference standard to evaluate the proficiency of the testing laboratories. The velocity ranged from 3 to 15 m/s, and the mass concentration can be lower than the 20 mg/Nm3. The historical data shows the proficiency has been improved and the unqualified rate is 11%. The unqualified rate for velocity and mass concentration tests are 5.4% and 5.56% respectively. The measurement errors were mainly from the inappropriate operation and the procedures of sample recovery, conditioning and weighing. There is no obvious error in the flue gas parameters. Standard scores showed that the velocity tests had a higher deviation than the mass concentration tests. This may be because the velocity model had a higher standard error. The 110 model shows the standard error in velocity model is 0.397 m/s and 17.4 mg/Nm3 in mass concentration model. The R2 for velocity and mass concentration models are 0.996 and 0.982, respectively.

PM sampling, stack, particulate matter