BAM 1020-9800 Manual Rev W.docx
Page 59
7.3 Comparison of BAM 1020 Data to Integrated Filter Sampler Data
Each new BAM 1020 has been calibrated against a reference beta gauge whose calibration
is traceable to a gravimetric standard. This calibration information is provided in the
calibration certificate that accompanies each BAM 1020
as K and as μ
sw
. As the BAM 1020
span response is virtually insensitive to the chemical composition of the sampled PM one
should expect excellent agreement between mass density determined by a manual filter-
based sampler and the mass density determined by a collocated BAM 1020.
Most PM reference methods are based on manual, integrated sampling techniques in which
PM is sampled onto pre-weighed filters. Sampled filters are then equilibrated and then re-
weighed. The net weight gain is used along with the volume of air sampled to determine the
mass density of PM in the sampled volume. PM reference methods may differ from one
jurisdiction to another. Furthermore, the BAM 1020 may be operated differently from one
jurisdiction to another. For example, the BAM 1020 may be configured to operate as a US-
EPA designated PM
2.5
Federal Equivalent Method. Or it may be configured to operate in
accordance with EU guidelines for PM
2.5
.
Users may collocate a PM
10
or a PM
2.5
reference sampler with a newly deployed BAM 1020
and collect data on both devices for a period of time in order to demonstrate reasonable
correlation and acceptable levels of multiplicative (slope) and additive (intercept) bias
between the two methods. Performing such a field test is beneficial as it could reveal an
undetected performance or data reporting issue. Common issues could include improper data
logger scaling, incorrect background (BKGD) values, or improper flow calibration due to a
mis-calibrated flow standard. A scatter plot between the reference standard results (plotted
along the x-axis) and the BAM 1020 results (plotted along the x-axis) can reveal these
problems.
In order for such an analysis to be useful however it is necessary to have a suitable number
of data points, an acceptable level of dispersion (range) in the measured values and
acceptable level of correlation (r
2
) in a regression between the reference results and the BAM
1020 measurements.
Below are several additional considerations:
•
Nozzle leaks can lead to poor correlation between the BAM 1020 and the reference
standard.
•
Improper inlet tube insulation or placing the BAM 1020 directly in the path of an air
conditioner vent during operation under hot, humid conditions can lead to poor
correlation with the reference standard and unpredictable levels of multiplicative and
additive bias.
•
The collocated inlets should be at approximately the same height and within several
meters of one another during the comparison test.
•
The start time and the stop time of the filter-based method should correspond to the
hourly BAM 1020 measurement cycles.
