Airborne Particulate Matter from the Emissions of a Coal-Fired Power Plant in Central Pennsylvania

Amanda Carey

University of Rochester
Dept. of Earth and Environmental Sciences

OPT407: Electron Microscopy
Spring 2009
Final Project

General Setting
Analysis Methodology
Results and Discussion



The use of coal as a source for power generation has increased as U.S. energy demands continue to grow. Coal combustion residue (CCR), the waste product of coal combustion, concentrates inorganic non-combustibles such as heavy and trace metals. This increases the potential for localized environmental damage if emissions and CCR reach the environment in appreciable amounts. The anthracitic coal region of central Pennsylvania contains 15 independent coal combustion power plants and produces upward of 9 million tons of fly ash each year.

Aside from becoming concentrated within CCR, metals can be emitted into the atmosphere by industrial processes, such as coal combustion, in the form of fine particles. Despite the fact that efficient particulate collection methods are utilized in power plants, considerable amounts of fly ash are emitted into the atmosphere daily as a result of the large amounts of coal burned (Querol et al., 1996). These aerosol particles are capable of being transported long distances before being deposited as wet precipitation or dry deposition, with smaller particle sizes (.1-5m) having the longest atmospheric residence times (Damle et al, 1982). Fly ash particles may impact remote environmental settings great distances from point sources because their high surface areas allow for the enrichment of potentially toxic elements through condensation during the cooling of combustion gases (Klein et al., 1975).

Emphasis has been placed on studying the detrimental effects of particulate matter on human and animal health, including damage to cells and DNA. Particulate size controls the likelihood of incorporation into the body through inhalation. Smaller particles are more able to deeply penetrate into the lungs causing the boundary between respiratory and non-respiratory particles to be roughly set at a 10m size. Additional evidence suggests that solubilized components of fine particulate matter (e.g. metals) are able to permeate through the lung and distribute through the body systematically, perhaps including the cardiovascular and other organ systems. Some ultra-fine particles may be capable of themselves evading the bodys defense systems for removing larger particles from the lungs surface and penetrating through the lung tissues. Therefore, it is suggested that individual health outcomes are not only a factor of size, but of the inherent toxicity of the particulates and co-associated materials (EPA, 2008).

This study aims to examine the size and shape of collected aerosols from the Roaring Creek, Pennsylvania watershed. The composition of particulates was also analyzed in an attempt to identify any size-dependent chemical compositional variations as well as examine particulate sources and potential environmental and human toxicity.

(back to top)


General Setting

The Roaring Creek watershed is located directly adjacent to the northern most outcrop of the central anthracitic coal region of Pennsylvania. The watershed, seen in Figure 1, contains three reservoirs and is the main water source for a radius of 15 miles. The creek flows westward through the Conrad Weiser State Park with its origin approximately 8 miles north-east of the Mount Carmel cogeneration power plant. The cogeneration power plant is a 44MW coal fired plant which opened in 1990 and is located less than two miles from the main reservoir (EPA, 1999).

The power plant has been shown to deposit particulates and fly ash, rich in heavy metals, to the watershed in detectable amounts via ICP-MS analysis (Carey et al., 2008). However this anthropogenic input does not appear to be impacting the drinking water quality beyond the limits set by the EPA due to the regions pH and the high Kd for most metals.

Figure 1: Sampling map from aerosol sampling, July 2008. Blue flags denote sample locations and the red x is the location of the coal-fired power plant (created on TopoQuad using field latitude and longitude values).

Seven aerosol samples were collected within the watershed in July of 2008. All samples were collected downwind of the coal-fired power plant under light wind conditions of 5 mph or less. Using a battery powered mechanical pump, aerosols were collected onto a .45 micron Millipore filter for a time period of approximately four hours. Weather conditions remained warm and dry during particulate collection, with temperatures in the low 80s Fahrenheit. Collected particulate filters were stored in Ziploc bags until SEM analysis.

(back to top)


Analysis Methodology

Scanning electron microscope analysis was done using a Zeiss SUPRA40-VP model SEM at the Institute of Optics at the University of Rochester. Prior to analysis, all filters were mounted on a sample stub using double stick carbon tape and sputter coated with approximately 5nm of gold to ensure proper conductivity. Samples of bulk fly ash and magnetically extracted grains from a sediment core were also prepared for analysis. An Everhart-Thornley secondary electron detector (SE2) was used in addition to a large angle scintillator backscatter electron detector (BSD) for the collection of micrographs. The backscatter detector was especially useful for the detection of grains with metal enrichment because enriched grains with a higher atomic number produced more backscattered electrons and therefore appeared brighter than nearby, non-enriched grains. X-ray microanalysis was conducted using energy-dispersive x-ray spectroscopy (EDS detector) capable of producing a spectrum characteristic to the analyzed sample with associated quantitative results. In addition, the bulk fly ash sample underwent x-ray mapping to examine the two-dimensional distribution of elements within a random grain distribution.

(back to top)


Results and Discussion

A total of 68 particulates from three of the seven aerosol collection locations were individually analyzed in the SEM. Analyzed grain sizes ranged from less than 1m to 25m, with an average grain size of 5.34m. Grain sizes found at the three sampling sites, located .83, 3.42 and 6.27 miles from the power plant, showed minimal correlation with respect to distance from the plant. Previous studies have reported grain size segregation as a function of the distance from the source plant, however sampling ranges included particulate collection sites at distances up to 46.8km. A distance of 30km from the power plant was pin-pointed as a cutoff point for grain sizes of 10m and larger, smaller grains capable of transporting further distances from the plant (Querol et al., 1996). The similarities of grain sizes collected within Roaring Creek suggests that sampling distances were not large enough to fully display the extent of grain size segregation.

Analysis of the fly ash sample included x-ray mapping which was carried out for the aluminosilicate matrix elements normally present in significant quantities in fly ash; Al, Ca, Mg, Na, Si, O, and Fe (Damle et al, 1982). Figure 2 shows each elemental map, where brighter colors indicate higher concentrations of the selected element. White portions of grains seen in the original micrograph, seen by rolling over a phase map, are the result of sample charging. Grain sizes range from less than 1m to approximately 100m. While oxygen is relatively evenly distributed, it can be noted that larger grains have higher magnesium and aluminum-silicate compositions while smaller grains tend to have higher iron content. This is consistent with previous studies that have shown preferential concentration or enrichment of certain trace elements in smaller particle sizes (McElroy, 1982). The enrichment is the result of the higher surface area-to-volume ratio of smaller grains that underwent a heterogeneous condensation of evaporated volatile species (Damle et al, 1982).



Figure 2: Roll over image for initial micrograph, acquired using the SE2 detector. In order from the top left to right: aluminum, calcium, iron, magnesium, oxygen, silicon, sodium.

Studies have looked at this enrichment and found that under simple geometrical considerations, the enrichment concentration of volatile species should be proportional to 1/d^2 where d is the fly ash particle diameter. Conversely, other studies have found concentrations of volatile species in particles smaller than 1m to be independent of size (Damle et al, 1982). Approximately 12 particles were analyzed from the Roaring Creek watershed that were 1m or smaller. While 4 of these grains, such as the grain seen in Figure 3, did show enrichment of metals such as Fe, Cu and Zn, higher concentrations of trace metals were seen in grain sizes ranging from 1 to 10m. Only two grains over the size of 10m showed any trace metal enrichment.

Figure 3: One micron sized grain displaying iron enrichment. Micrograph acquired using the SE2 detector.

Several of the additional submicron sized grains indicated the presence of less volatile species such as Al, Mg and Si. One model used to explain the solid-gas reaction kinetics during combustion indicates reducing conditions near the particles surface with oxidizing conditions slightly further away from the particles surface. This model can explain the presence of these less volatile species through the reduction of refractory species Al2O3, SiO2, and MgO to more volatile reducing species AlO, SiO, and Mg within the reducing zone (Sarofim et al., 1977). It should also be noted that several of the collected grains, including some in the 1m size range, showed quartz, calcite and clay compositions. These grains can be assumed to be the result of air-borne dust and collection of these grains may skew any statistical analysis made on collected grains as they represent natural sources and were not emitted from the power plant. Future aerosol collections will include sampling locations upwind of the power plant to gain a better understanding of these naturally occurring particulates.

Morphologically, fly ash is often identified for its spherical characteristics; the result of combustion and quenching. Several spherical grains were identified, however a larger proportion of grains were highly irregularly shaped. Figure 4 displays a large, iron rich grain as well as a smaller, spherical, iron enriched grain. Spherical, magnetically extracted grains from a sediment core taken within the watershed, such as the one seen in Figure 5, indicates the presence of fly ash in regions of core samples that have been shown through trace metal concentrations and magnetic susceptibility analysis to contain higher concentrations of ferromagnetic material (Carey, 2008). The irregular, amorphic shape of the grain seen in Figure 6 is highly characteristic of the type of grain seen within the power plant emissions. This grain, less than 10m in size, has an enrichment of zinc as well as sulfur.

Figure 4: Right, colorized micrograph displays the spherical grain seen in bottom right of left, colorized micrograph. Micrographs acquired using the SE2 detector.


Figure 5: Spherical grain removed magnetically from a sediment core, indicating the anthropogenic addition of fly ash into watershed sediments. X-ray spectra indicates iron enrichment. Micrograph acquired using the SE2 detector.


Figure 6: Small, amorphic grain, displaying metal enrichment. Micrograph acquired using the backscatter detector.


Several grains collected were found to have a purely carbon composition. This is common and has been seen in other studies; resulting from when aromatic hydrocarbons vaporize, burn, and partially release fine carbon particles or soot. Carbon grains may also be the result of unburnt carbon from inefficient combustion. Identified carbon particle sizes ranged from submicron to approximately 40m, sharing identical x-ray spectra such as that seen in Figure 7. The level of accuracy received through grain quantification remains questionable as the filter may have added to the collected carbon peak in several spectra. This effect is likely more prominent in smaller grains as larger grains provided a broader sample area for x-ray collection.


Figure 7: Carbon particles ranged from submicron (left micrograph, BSD) to approximately 40m (right micrograph, SE2), sharing identical x-ray spectra.


Seven grains throughout the watershed were found to contain a portion of sulfur, including the grains seen in Figures 6 and 8. These aerosols may have been directly emitted from the power plant having formed similarly to the majority of other grains seen with metal enrichment. Conversely, formation of sulphate aerosols can occur through the oxidation of SO2 during atmospheric transport and deposition.

Figure 8: Sulfur and Zinc enriched, ~2m sized grain. Micrograph acquried using backscatter detector.

(back to top)



Particulate matter collected within the Roaring Creek watershed indicates a correlation between smaller grain sizes (less than 10m) and trace metal enrichment, though it is difficult to quantify the level of enrichment within the smaller grains. It has been shown through trace metal analysis that the emissions are reaching the watershed in appreciable amounts, however SEM analysis has indicated that a sizeable percentage of the emitted particulates fall in the sub 10m size range. These grains are easily capable of entering the respiratory system and could potentially prove harmful when inhaled. Small sampling distances provide a limitation to a full understanding of grain size segregation as a function of distance from the power plant. Smaller grain sizes are likely traveling outside of the sampling range and further analysis would be needed to understand the full toxological impact of the power plants emissions on the surrounding population.

(back to top)



Carey, A. The Application of Magnetic Susceptibility and Trace Metal Analysis in Determining Temporal Chemical Changes in Regional Watershed Cores. Senior Thesis, University of Rochester. 2008.

Carey, A., Harrold, Z. The Effects of Coal-Fired Power Plant Emissions and Fly Ash
on a Regional Watershed. Journal of Undergraduate Research. University of Rochester, 2008.

Damle, A. S., Ensor, D. S. and Ranade, M. B. (1982) Coal Combustion Aerosol
Formation Mechanisms: A Review', Aerosol Science and Technology, 1:1, 119- 133.

Klein D. H., Andren A. W., Carter J. A., Emery J. F., Feldman C., Fulkerson W., Lyon
W. S., Ogle J. C., Talmi Y., Van hook R. I., Bolton N., Pathways of thirty-seven trace elements through coal-fired power plant. Envir. Sci. Technol., 9, pp. 973-979, 1975.

McElroy, M. W., R. C. Carr, D. S. Ensor, and G. R. Markowski. "Size Distribution of
Fine Particles from Coal Combustion." Science, 205 (1982): pp. 13-19.

Querol, Xavier, Andres Alastuey, Angel Lopez-Soler, Enrique Mantilla, and Felicia
Plana. "Mineral Composition of Atmospheric Particulates Around a Large Coal-Fired Power Station." Atmospheric Environment, 30 (1996): pp. 3557-572.

Sarofim, A. F., Howard, J. B., and Padia, A. S. The Physical Transformation of the
Mineral Matter in Pulverized Coal Under Simulated Combustion Conditions. Combust. Sci. Technol, 16 (1977): pp. 187-204.

U.S. Environmental Protection Agency. Air Toxics Website. Technology Transfer
Network. <, 2007.>

U.S. Environmental Protection Agency. Clean Air Research Multi-Year Plan 2008-
2012. June 2008. <>

(back to top)



I would like to thank Brian McIntyre for his expert guidance, advice and for use of the SEM. I would also like to thank Andreas Liapis for his assistance despite my geology background. Special thanks to my advisor, Robert Poreda, as well as Tom Darrah for assistance in field sampling.

(back to top)

Please enter any comments, criticisms, questions, etc. below.

Your name:

Email address: