Advanced Chemical Measurements of Smoke from DoD-Prescribed Burns
Dr. Tim Johnson | Pacific Northwest National Laboratory
The objective of this project was to use advanced instrumental techniques to study the particulate and vapor-phase chemical composition of the smoke that results from prescribed fires used as a land management tool on Department of Defense (DoD) bases, particularly bases in the southeastern United States. The work was carried out in collaboration and conjunction with SERDP projects RC-1647 (models) and RC-1648 (particulates, bases in the Southwest).
The approach involved developing techniques for measuring biomass burning emission species in both the laboratory and field, and developing infrared (IR) spectroscopy in particular. Using IR spectroscopy and other methods, emission factors (EFs, g of effluent per kg of fuel burned) were developed for dozens of chemical species for several common southeastern fuel types. The major measurement campaigns were laboratory studies at the Missoula Fire Sciences Laboratory (FSL), as well as field campaigns at Marine Corps Base Camp Lejeune, North Carolina; Fort Jackson, South Carolina; Fort Benning, Georgia; and in conjunction with SERDP project RC-1648 at Vandenberg AFB, California, and Fort Huachuca, Arizona. Comparisons and fusions of laboratory and field data were also carried out, using laboratory fuels from the same bases.
This project enabled new technologies and furthered basic science, mostly in the area of IR spectroscopy, a broadband method well suited to biomass burn studies. Advances in hardware, software, and supporting reference data realized a nearly 20 times improvement in sensitivity and now provide quantitative IR spectra for potential detection of approximately 60 new species and actual field quantification of several new species such as nitrous acid, glycolaldehyde, α-/β-pinene and D-limonene. The new reference data also permit calculation of the global warming potential (GWP) of the greenhouse gases by enabling detection of their ambient concentrations and quantifying their ability to absorb IR radiation.
Measurements of fuel consumption (FC) were made on DoD bases. The main product was hundreds of measurements of EFs for a broad suite of both trace gases and particulate species from the fires. The EFs were based on measurements at the FSL laboratory and three different southeastern DoD sites. The chief deliverable for each effort was a comprehensive table of EFs, with the associated tables for the given ecosystem and fire type. The list of species measured in this project is the most extensive smoke characterization achieved to date and is presented in extensive tables of EFs, including broken out by vegetation type, e.g., semiarid shrublands, pine understory, etc. Emission ratios (ERs) to carbon monoxide (CO) were also derived, and these can be used to estimate downwind levels or photochemical changes for certain species by coupling with data from other CO monitors. Multiple measurements of in-plume chemical transformations were also made via airborne sampling. Good progress was made measuring ozone (O3) formation and some progress was made on secondary organic aerosol (SOA) formation. The first measurements of initial black carbon (BC) and BC coating rates were made with non-filter-based techniques for a suite of U.S. prescribed fires. Biomass burning BC gets coated much faster (< 1 h) than BC from other sources such as diesel trucks. This affects climate assessments because the coatings increase the absorption of solar radiation by a lensing effect, but they also increase the BC solubility. Increased solubility means the BC has a greater ability to reduce cloud droplet sizes, which causes atmospheric cooling. The coating also increases the rate of removal in the atmosphere, reducing both the lifetime that the BC can warm the atmosphere and the likelihood of transport to sensitive snow/ice-covered regions.
Many key trace gases (also PM2.5) were measured using nearly identical Fourier Transform Infrared (FTIR)-based systems for the laboratory and field fires. By using the same systems a fairly direct comparison of the lab and field data can be made for a suite of species that includes both organic and inorganic gases, as well as compounds associated with both the flaming and smoldering phases. Comparisons were made knowing that the fire EFs depend on modified combustion efficiency, i.e., flaming vs. smoldering combustion, and ecosystem type. It has been confirmed that studying laboratory biomass fires can significantly increase the understanding of wildland fires, especially when laboratory and field results are carefully combined and compared. Much of the unlofted emissions are produced by smoldering combustion making lab studies of great value at characterizing smoldering-phase emissions. A suggestion for future research is the need to develop methods that allow safe real-time fuel consumption monitoring in the field or at least intermittent samples starting shortly after flaming has died down.
This project improved characterization of air emissions under both flaming and smoldering conditions for volatile organic compounds and gases. The analyses provide a set of EFs for modeling prescribed fire smoke photochemistry and air quality impacts that go far beyond what was previously available. The new set of EFs includes data for hazardous air pollutants and numerous precursors for the formation of O3 and secondary aerosol, all useful for model predictions of the amount of smoke produced by prescribed burns. For example, the South Carolina studies also saw faster secondary O3 formation in smoke plumes that mixed with urban emissions, which suggests avoiding smoke plume interaction with sources of nitrogen dioxide (NO2) such as from urban areas (e.g., rush-hour traffic or power plants). The generation of terpene oxidation products in downwind plumes was also observed for the first time, as well as fast formation of large amounts of gas-phase organic acids post emission. This demonstrates that the plume evolution is strongly influenced by as-yet unidentified species. It now seems unlikely that managers will have to allow for high amounts of SOA when forecasting downwind impacts of their burns.
For a few species some EF values were much higher when measured from ground by fixed-position (open-path IR) methods than when measured by actively locating smoke on the ground or in air. This indicates that active sampling methods could be biased in some cases. Careful co-deployment of the roving sampling equipment with the static open path measurement is needed to ensure that the differences are not instrumental. It is believed that the airborne IR measurements may best represent overall fire emissions in terms of global model input, but in terms of best representing the combustion-generated gases and respirable particles (including many toxins and carcinogens) to which a person on the fireline is exposed, a fixed open-path system, by virtue of its position on the perimeter, may give better assessment of smoke exposure affecting personnel deployed on containment lines. However, ground-based IR samplers are also relevant to estimating exposure for personnel who actively engage in extinguishing point sources, e.g., smoldering stumps.