0325-2834-MTDC, Laboratory Evaluation of Real-Time Smoke Particulate Monitors: Goals and Methods

Several goals were established before starting the tests. One goal was to determine the accuracy of each instrument compared to the FRM sampler. Also, because we had four E-BAM monitors to test, we wanted to determine the approximate precision of each instrument to the others. Another goal was to determine the effect of enabling the particle-size correction feature on the two DataRAM4 monitors. We ran tests with the size correction feature enabled on each instrument, disabled on each instrument, and enabled on one and disabled on the other. This test protocol will allow us to determine a correction algorithm for the DataRAM4 with the size correction feature enabled.

The tests were conducted in the Fire Sciences Laboratory’s large (131,000 cubic foot) combustion chamber. The instruments were placed on a smoke-sampling platform 55 feet above the chamber floor (figure 6 and figure 7). The experiments were conducted at ambient conditions (65 to 80 degrees Fahrenheit at about 30 percent relative humidity). The tests sampled smoke from small burning beds of white pine needles near the chamber floor. Fans mixed the smoke in the chamber. The instruments were close to each other to ensure that the same smoky air was sampled.

Figure 6—The E-BAM monitor was tested in the Rocky Mountain
Fire Sciences Laboratory’s smoke chamber.

Figure 7—The BGI PQ-200 Federal Reference Method sampler (left)
with MetOne Instruments’ E-Sampler (middle) and the two DataRAM4
monitors (right).

The instruments used for the tests included four E-BAM monitors, two DataRAM4 monitors, one E-Sampler, and an FRM sampler (BGI PQ-200). The instruments were configured with their respective PM_2.5 cutoff device. All the instruments were calibrated before the tests. All the real-time instruments were set for an average time of 1 minute and were run continuously throughout the tests. Before testing each day, a self test was conducted on the E-BAM monitors. The E-Sampler and DataRAM monitors were calibrated to zero.

The FRM sampler was programmed to run for different lengths of time, depending on the current smoke particulate concentrations. The goal was to collect at least 100 micrograms of mass on the gravimetric filters for accurate filter weights. The FRM sampler draws about 1 cubic meter of air per hour. So if the particulate concentration in the chamber was about 100 micrograms per cubic meter of air, the FRM sampler was programmed to run for at least 1 hour. After the filters were weighed by the Montana Department of Environmental Quality, their weight was divided by the total volume of air drawn through the filter for the particular time period to obtain the average mass concentration. That value was compared to the average mass concentration estimated by the real-time instruments during the same period.

Results and Discussion

A total of 30 tests were conducted. The average particulate concentrations ranged from 30 to 1,288 micrograms per cubic meter as determined by the FRM sampler. Most of the data is evenly distributed between 30 and 800 micrograms per cubic meter. The 1,288 micrograms per cubic meter test was not included in these results because it would have influenced best-fit-line equations and correlation coefficients. Statistical results were based on least squares-linear regression equations and correlation factors. Table 1 displays the results for all the instruments. Figures 8 to 14 show the results for each instrument when compared to the FRM sampler. Each figure shows the appropriate data points, the slope equation, and the correlation coefficient for the best-fit line. A line representing a one-to-one relationship is also shown.

Table 1—Linear regression results for real-time instruments compared to the BGI PQ 200 Federal Reference Method sampler.
Real-time particulate samplers	Regression slope	Intercept (micrograms per cubic meter)	Correlation coefficient
E-BAM (serial number 4334)	1.21	–23.77	0.9949
E-BAM (serial number 5321)	1.14	–24.15	0.9939
E-BAM (serial number 5324)	1.10	–11.49	0.9946
E-BAM (serial number 1675)	1.09	–13.80	0.9981
E-Sampler	1.13	3.41	0.9628
DataRAM (serial number D034, size correction enabled)	2.87	22.54	0.9912
DataRAM (serial number D172, size correction enabled)	2.57	51.49	0.9712

E-BAM—Three of the E-BAM monitors operated during all the tests. The fourth arrived late and was run for only 12 tests. Figures 8, 9, 10, and 11 show the results for each E-BAM sampler when compared to the FRM sampler results. On average, the E-BAM samplers overestimated the mass concentration by 13.5 percent with the highest (serial number 4334) overestimating the concentration by 21 percent and the lowest (serial number 1675) overestimating the concentration by 9 percent. Correlation coefficients for all the E-BAM samplers were very high, all over 0.9939.

[image] Line chart comparing results for E-Bam 4334 to Federal reference sampler

Figure 8—Results for the E-BAM (serial number 4334) and the
BGI PQ-200 Federal Reference Method sampler.

Line chart comparing results for E-Bam 5321 to Federal reference sampler

Figure 9—Results for the E-BAM (serial number 5321) and the BGI PQ-200
Federal Reference Method sampler.

Test Goals and Methods

Results and Discussion