The effect of averaging time on compliance of a PM threshold when using continuous emission monitors

After two particulate matter (PM) continuous emission monitors (CEMs) were calibrated at a liquid waste incinerator at Eli Lilly and Companyʹs Clinton, IN site, the continuous data from the two instruments were analyzed to determine how different averaging methods and times would impact the number of hours an arbitrary threshold would be exceeded. The arbitrary threshold chosen, based on this data, was 28 mg/acm. Two methods of developing rolling averages for 1-, 3-, 6-, 12-, and 24-h periods were calculated. The calculation method did not affect the number of hours where the averages exceeded 28 mg/acm for either instrument, but the length of the averaging period did have an impact. As the length of the averaging period increased for the Sigrist, the number of hours over the 28 mg/acm threshold decreased. However, for the ESA, the number of hours over the 28 mg/acm threshold increased until the averaging period reached 3-h and then started to decrease. This was due to the ESA instrument having a series of sharp peaks that caused short-term averages to increase but the delay from the longer-term averages were sufficiently buffered that they did not exceed the 28 mg/acm threshold.