Smoke Particulate Monitors: 2006 Update, Page 6

Smoke Particulate Monitors: 2006 Update

Conclusions

Of the five instruments we evaluated, the Met One Instrument E-BAM was the monitor that most closely estimated smoke particulate concentrations when compared to the FRM sampler. It overestimated the smoke particulate concentration by 1 percent. The tests also showed that the ac and dc pump versions of the E-BAM produced similar results. The E-BAM loses some of its portability when the bulky ac pump is used. For some users this loss in portability may be a small inconvenience compared to having to replace the dc pump periodically.

The E-Sampler was also fairly accurate when estimating smoke particulate concentrations. The two E-Samplers overestimated concentrations on average by 13 percent with a difference of just +/- 4 percent between results of the two instruments and the FRM average. Users should consider correcting the estimates from the E-Sampler by 13 percent (multiply the values by 0.89) when measuring smoke particulate.

The DUSTTRAK monitors overestimated the concentration by 217 percent. The difference between results of the three instruments was just +/-2.5 percent. This consistency is the best we have seen in all of our evaluations of these types of instruments. A correction algorithm of 0.32 (multiply the estimated concentration by 0.32) is recommended for those using the DUSTTRAK. The use of the PM_2.5 or PM₁₀ cutoff inlet does not make much difference when estimating smoke particulate concentrations with the DUSTTRAK.

The DataRAM 4 monitors overestimated the smoke particulate concentrations by a large amount, although this was not unexpected based on previous evaluations. On average, the four instruments overestimated concentrations by 144 percent. Previous tests have showed that the DataRAM 4 with the size-correction feature "enabled" overestimated concentrations even more. Previous recommendations suggested a correction factor of 0.37. Based on these new results, we recommend a slightly higher correction factor of 0.39.

On average, the DataRAM 2000 monitors overestimated the mass concentration by 15 percent. This is somewhat of a surprise because previous tests indicated that the DataRAM 2000 normally overestimates the concentration by more than 100 percent. The monitors used in these tests had not been evaluated previously. Based on these results, it may be prudent to develop individual correction algorithms for the DataRAM 2000s.

These tests were conducted using smoke generated by burning dry pine needles. Results may differ if the vegetation type or moisture content is significantly different. Smoke from burning buildings or other sources with different fuels may yield dramatically different comparisons between smoke particulate monitors and the FRM sampler.