The objective of this project was to demonstrate a novel continuous emissions metals analyzer, the Aerosol Beam Focused Laser Induced Plasma Spectrometer (ABF-LIPS). Continuous emission monitoring systems (CEMS) such as this provide an effective means for monitoring the level of hazardous air pollutants (HAPs) in real time. This potentially facilitates better control of processes and improved pollution control without relying on conservative permit limits, which are based on time-averaged integrated traditional sampling techniques with off-site laboratory determination of HAPS. The demonstration objectives were to obtain data under real-world conditions to complement laboratory data and previous field testing.
The ABF-LIPS uses a pulsed laser beam tightly focused onto an aerosol sample to ignite plasma, which breaks down all compounds to their elemental composition. The elements in the plasma volume are vaporized, resulting in an unstable, excited state. When the atoms return from the plasma-excited state to ground, they release light at element-specific wavelengths that can be observed using time-resolved spectroscopy. The wavelengths of the emission spectra correspond to a particular element, and the amplitudes of the peaks correspond to the mass of that element. The aerosol beam focusing capability improves the detection and sensitivity of traditional laser-induced plasma spectrometry by aerodynamically focusing aerosol particles to a point, increasing the local aerosol concentration and significantly improving the signal-to-noise ratio. ABF-LIPS, because of its portability, can be mounted at an emission source and requires no long sampling line. This virtually eliminates sample loss.
The first test was conducted at the Naval Aviation Depot (NADEP) North Island in San Diego, California. A chromium plating bath exhaust, a nickel plating bath exhaust, and a molten metal (Kirksite) furnace were sampled. The second demonstration test was conducted at Tooele Army Depot (TEAD) in Tooele, Utah, using the Ammunition Equipment Directorate’s Ammunition Peculiar Equipment 1236M2 test furnace, a munitions deactivation test incinerator.
The first test at NADEP produced useable data from ABF-LIPS for only the molten metal furnace source. The chromium and nickel plating bath source tests likely failed due to an incorrect setting of the detector exposure time. ABF-LIPS data from the molten metal furnace, in the form of emission spectra peak heights, did not correlate with the spiking levels of the three test metals (cadmium, chromium, and nickel). Pearson correlation coefficients ranged from negative (anti-correlated) to weakly positive (i.e., random). In contrast, the reference method (RM) results showed generally good correlation with the spiking level. The failure of these tests to produce accurate data, the primary performance objective, means that further system development and testing would be required before the instrument could be permitted for use in pollution control systems.
As a result, modifications were made to the ABF-LIPS instrument and spiking apparatus prior to the subsequent field tests at the TEAD. A munitions deactivation furnace was tested at TEAD in 2005, which produced data from ABF-LIPS that correlated well with the spiking levels. Agreement with the RM results, however, was not within the relative accuracy acceptance criteria of 20% for any of the five test metals; again, the primary performance criterion was not met. The ABF-LIPS reported emission rates were higher than the RM results by an average of 67% for the high spike concentration (per-metal ranges of 50% to 88%) and higher by 73% for the medium spike concentration (per-metal ranges of 17% to 99%). ABF-LIPS generally reported lower values for the low spike concentration. No cadmium or chromium was detected, and nickel was 145% and lead 1000% lower than the RM results. Mercury was within 5% at the low concentration.
The TEAD RM data are suspect. Variances within the group of four runs at each concentration were high. Coefficients of variation for each metal averaged 45% with a per-metal range of 27% to 81%. This suggests that the RM data are inherently flawed and comparisons to the ABF-LIPS results are likely unreliable.
Although validation of ABF-LIPS did not meet the relative accuracy criterion, other performance objectives were met: the instrument was relatively easy to transport and set up, it operated under adverse environmental conditions without needing repairs, zero drift was within performance specification 10 (PS-10) criteria, and the analysis cycle was short (less than 6 minutes).
It is unlikely the current prototype unit could be used in a real-world application at this point. Further development of ABF-LIPS will be required prior to additional validation testing and eventual regulatory acceptance. Previous laboratory testing of the system components yielded better accuracies, suggesting that adaptation of system components to a portable, field-deployable unit used in these tests resulted in compromised data accuracy and that the inherent nature of field testing with higher degree of uncontrolled variables likely led to failed field testing. Some of these variables included the delivery of standards to the stack/flue stream, condensation of water in the instrument optics and stack/flue stream, and alignment issues due to environmental vibrations from mechanical systems and wind.