Environmental Geochemistry and Health 1993 15(4) page 215
Distribution and mode of occurrence of selenium in US coals Lynn Coleman, Linda J. Bragg and Robert B. Finkelman US Geological Survey, MS 956, Reston, VA 22092, USA
Abstract
Selenium excess and deficiency have been established as the cause of various health problems in man and animals. Combustion of fossil fuels, especially coal, may be a major source of the anthropogenic introduction of selenium in the environment. Coal is enriched in selenium relative to selenium's concentration in most other rocks and relative to selenium in the Earth's crust. Data from almost 9,000 coal samples have been used to determine the concentration and distribution of selenium in US coals. The geometric mean concentration of selenium in US coal is 1.7 ppm. The highest mean selenium value (geometric mean 4.7 ppm) is in the Texas Region. Atlantic Coast (Virginia and North Carolina) and Alaska coals have the lowest geometric means (0.2 and 0.42 ppm, respectively). All western coal regions have mean selenium concentrations of less than 2.0 ppm. In contrast, all coal basins east of the Rocky Mountains (except for several small basins in Rhode Island, Virginia, and North Carolina) have mean selenium values of 1.9 or greater. Generally, variations in selenium concentration do not correlate with variations in ash yield, pyritic sulphur, or organic sulphur concentrations. This may be the result of multiple sources of selenium; however, in some non-marine basins with restricted sources of selenium, selenium has positive correlations with other coal quality parameters. Selenium occurs in several forms in coal but appears to be chiefly associated with the organic fraction, probably substituting for organic sulphur. Other important forms of selenium in coal are selenium-bearing pyrite, selenium-bearing galena, and lead selenide (clausthalite). Water-soluble and ion-exchangeable selenium also have been reported.
Introduction
Various elements in coal have been used to correlate coal beds and to help decipher depositional e n v i r o n m e n t s . S o m e elements contribute to technological problems such as boiler fouling (sodium, calcium, magnesium), slagging (iron), and corrosion (chlorine, fluorine). A few elements in coal (e.g. gold, silver, uranium) have potential as e c o n o m i c b y - p r o d u c t s , whereas others (e.g. selenium, mercury, cadmium) pose potential threats to the environment. The reader is referred to Swaine ( 1 9 9 0 ) for a d d i t i o n a l i n f o r m a t i o n on the environmental, technologic, geologic, and economic impact of trace elements in coal. The concentration levels, vertical and lateral distribution, and modes of occurrence (chemical forms) of the elements in coal have been the subject of numerous papers and books (Bouska, 1981; Valkovic, 1983; Swaine, 1990; Finkelman (in press)). Few elements in coal (other than sulphur) have attracted as much attention as has selenium. For example, Boon etal. (1987) note that most western states require selenium monitoring in coal overburden. Concern with selenium is due to the well-documented environmental and health-related problems caused by this element, including several
problems caused by selenium derived from coal. These problems are discussed below. A number of data compilations exist for selenium in US coal.Compilations by Swanson et al. (1976) and Zubovic et al. (1979, 1980) were based on information from the USCHEM (US Geochemical Data) data base, part of the US Geological Survey's National Coal Resources Data System (NCRDS). These earlier compilations were based on analyses of 799, 617, and 659 coal samples (Swanson et al., 1976; Zubovic et al., 1979, 1980, respectively). The present study is based on analyses of almost 9,000 coal samples, including all the data from the previous NCRDS compilations. Gluskoter et al. (1977) found geometric means of 3.4 ppm selenium for 23 Appalachian coal samples, 1.3 ppm for 28 western coal samples, and 2.0 ppm for 114 Illinois basin coal samples. E x a m i n i n g the enrichment (relative to the average concentration in t h e Earth's crust) of the trace elements in coal samples, they found selenium had the highest enrichment (26 to 68 times crustal abundances) in all three coal-bearing areas. Environmental and health-related problems caused by selenium
Selenium (Se) occupies a dual environmental
216
Selenium in US coals
position, as both deficiency and excess of this element can create health problems (NAS, 1974; Lakin, 1972). Selenium has long been identified as a toxin associated with various animal diseases (Beath et al., 1934; Rosenfeld and Beath, 1964; Beath, 1982). Excessive amounts of selenium have been cited as a cause of periodontal disease and dental caries in humans and as a possible carcinogen (NAS, 1974). An endemic disease in Hubei province China has been attributed to selenium intoxication. This disease is characterised by loss of hair and nails, lesions of the skin, and damage to the nervous system. Morbidity was as high as 50% (Yang et al., 1983). However, Zheng et al. (1992) report finding only 400 cases of mild selenosis symptoms in the Hubei province during the past five decades. Selenium has also been cited as the cause of massive fish kills in Martin Lake, Texas, and at three sites in North Carolina (Shepard, 1987) and as the cause of death of thousands of birds in the Kesterson National Wildlife Refuge in California (Presser and Barnes, 1984). Two e n d e m i c diseases in China, Keshan disease and Kaschin-Beck disease, are attributed to the deficiency of selenium. These diseases can be cured and prevented by adding selenates to table salt (Jiahzhong, 1988). Moreover, the US National A c a d e m y of Science (NAS, 1974) has added selenium to the list of essential nutrients. Selenium in fossil fuels and rocks The combustion of fossil fuels has been cited as a major source of selenium in the environment. A n d r e n etal. (1975) estimated that selenium mobilised by coal combustion is 1.5 to 2.5 times the amount released by the weathering of rocks. The NAS (1974, p.61) report states: "The amount of selenium entering the atmosphere from the burning of ~ossil fuels is approximately six times greater than that derived from mined ores. Preliminary estimates of selenium in rainfall indicate that regions of high fossil fuel use have higher selenium rates of deposition by rain". The endemic disease in China has been attributed to the release of selenium from weathering and combustion of a seleniferous coal (Yang et al., 1983). The fish kills in Martin Lake, Texas, and in North Carolina were caused by the mobilisation of selenium from fossil fuel waste products (Shepard, 1987). Zingaro and Cooper (1974, p. 10) state: "It is of the utmost significance to examine the selenium content of fossil fuels and thereby obtain an indication, not only of the distribution of selenium in these materials but also, of the amount of selenium released to the atmosphere from this source". Special attention has been focused on coal because of its enrichment of selenium: coal contains from 5 to 300 times the amount of selenium as do other rocks (Table 1). This enrichment can be viewed from another perspective: selenium in coal is enriched by a factor of 82 when compared to its concentration in the Earth's crust (US National
Table 1 Selenium concentration in major rock types, (Modified from Kabata-Pendias and Pendias, 1984).
Rock type
Se (ppm)
US coala Soil Shale Argillaceous sediments Sandstones Limestones, dolomites Ultramafic rocks Mafic rocks Intermediate rocks Acid rocks (intrusive) Acid rocks (extrusive) Crude oilb
1.7 0.4 0.6 0.4~).6 0.05-0.08 0.03-0.10 0.02-0.05 0.01-0.05 0.02-0.05 0.01-0.05 0.02-0.05 0.01-1.4
a This paper. b Swaine, 1978.
Committee for Geochemistry, 1980). This is about 10 times the enrichment of arsenic, the next element most highly enriched in coal.*
Sample Collection, Preparation and Analyses Since 1975, the US Geological Survey (USGS) and various state Geological Surveys have collected more than 14,000 samples of coal and associated rock from the United States. Most of these coal samples were collected and prepared according to methods outlined by Swanson and Huffman (1976). Full-bed channel samples, drill core samples, and, when necessary, grab samples were collected. Site documentation consists of locating samples in terms of topographic quadrangle, latitude and longitude, elevation, coal bed name, and formation, as well as measuring coal bed thickness, describing the coal bed, and documenting and estimating the thickness of the overlying and underlying lithologies. The USGS conducted comprehensive chemical analyses for each sample. These included proximate and ultimate analyses, determination of sulphur f o r m s , m a j o r - , m i n o r - and t r a c e - e l e m e n t concentrations, free-swelling indices, ash-fusion temperatures and heat content. See Golightly and Simon (1989) for descriptions of the analytical procedures. Selenium concentrations for most samples were determined on a whole-coal basis utilising instrumental neutron activation analysis (INAA). The detection limit of I N A A for selenium is 0.1 ppm, and the precision is estimated to be 5%. Tellurium, geochemically similar to selenium, may have the distinction of being the element most highly enriched in coal. On the basis of a limited number of analyses, Chiou and Manuel (1985) report that tellurium
L. Coleman, L. J. Bragg, and R. B. Finkelman
217
Table 2 Geometric mean and maximum concentrations of selenium in US coal regions and provinces. [No data On Southwestern, Black Hills, Tertiary lake beds, and Bighorn Basin regions.] Geographic areas
Total No. of samples
United States
8,695
304
1.7
3.6
0.02
75
Provinces: Eastern Gulf Interior Northern Great Plains Rocky Mountain Alaska Pacific Coast
4,711 214 705 1,154 1,615 258 38
113 3 11 81 45 43 8
2.5 4.6 2.4 0.66 1.2 0.42 0.76
4.2 5.6 3.1 0.99 1.6 1.1 1.9
0.02 0.50 0.20 0.10 0.10 0.10 0.20
75 16 36 13 13 43 7.3
68
2
2.9
4.4
0.30
16
19 91 1,654 1,921 1,039 10 41 172 3 401 301 324 819 11 40 34 336 60 239 689 89 59
14 2.6 28 29 36 4 1 2 0 11 0 18 63 0 0 1 6 4 1 25 3 3
0.22 3.9 2.9 2.9 1.9 0.20 4.2 4.7 2.3 2.3 2.5 0.63 0.66 1.4 1.6 0.69 1.3 0.89 1.9 0.97 1.9 1.4
0.33 0.03 3.8 3.8 3.3 1.1 5.3 5.6 2.3 2.8 3.5 0.81 1.1 1.5 1.7 1.2 1.4 1.3 2.0 1.4 2.5 1.9
0.40 75 0.20 0.02 0.03 0.10 1.4 0.50 2.1 0.50 0.20 0.10 0.10 0.70 0.66 0.10 0.35 0.20 0.64 0.10 0.31 0.38
1.2
Regions: Pennsylvania anthracites Rhode Isand meta-anthracites Appalachian 4,600 Northern Appalachian Central Appalachian Southern Appalachian Atlantic Coast Mississippi Texas Northern (Interior) Eastern (Interior) Western (Interior) Fort Unoin Powder River North Central Wind River Hams Fork Uinta Southwestern Utah San Juan River Green River Raton Mesa Denver
Total No. qualified samples a
Geometric mean b (ppm)
Arithmetic mean c (ppm)
Min. value (ppm)
Max. value (ppm)
75 18 20 1.8 14 16 2.6 13 36 3.3 13 2.7 3.1 7.9 13 3.9 5.8 9.0 8.9 9.6
a Analytical data designated as "greater than" or "less than". b From Cohen, 1976. c From Connor et al., 1976.
Precision estimates are expressed as percent relative standard deviation for a single determination on a typical coal sample as represented by National Institute of Science and Technology reference coal sample 1632. The major analytical interference is with tantalum; however, concentrations of tantalum in coal are c o m m o n l y only a fraction of the selenium concentration. Therefore, this interference is not considered significant. S e l e n i u m in some o f the western US coal
samples was determined by X-ray fluorescence spectroscopy. Sensitivity and precision o f this procedure were not as good as they were in INAA. Most of the samples analysed for the U S C H E M data base are channel or core samples. Data from bench s a m p l e s w e r e w e i g h t e d to d e r i v e a v a l u e r e p r e s e n t i n g the f u l l b e d . W h e n s t a t i s t i c a l parameters were calculated, 'less than' values for selenium were treated as if they were 0.7 of the minimum detection limit (Cohen, 1959).
218
Selenium in US coals
120 ° 110* 70 °
I00 o
'i
/ -) /
?.i. . / /I
t d
,'
'
90*
~3ppm so"
T....
1.6 l
40°
i
13
~ - ' - " -~1.1 pp-~7-mi ppm \
,
.~'\"i-
"
~I~
l -{~22~2~--~ 2.3 ~pm
;
i
1.9 Pi~ 500 MILES
3°~ ~k
~
I
500 KILOMETRES
EXPLANATION Concentration
Regions 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Rhode Island meta-anthracite Pennsylvania anthracite Atlantic coast Appalachian Northern Eastern Western Southwestern Mississippi Texas Fort Union Powder River Black Hills North Central Tertiary lake beds Bighorn Basin Wind River Hams Fork Uinta Southwestern Utah San Juan River Raton Mesa Denver Green River
2.6 ppm 2.7 ppm 2.3 ppm 2.3 ppm 2.4 ppm
Types of coal in fields ~
Anthracite
~
Low-volatile bituminous . Medium- and highvolatile bituminous
~ 4.4 ppm 4.9 ppm 0.7 ppm 0.8 ppm
Sub-bituminous
i:?~i:.~:~p:~i?~:::i~:~Lignite
1.4 ppm
1.6 ppm 1.1 ppm 1.3 ppm 1.1 ppm 1.9 pprn 2.3 ppm 1.5 ppm 1.1 ppm
Figure 1 Selenium concentrations f o r US coal regions.
L. Coleman, L. J. Bragg, and R. B. Finkelman
Distribution of Selenium in US Coal Regional distribution The coal fields of the United States can be divided into provinces and regions (Trumbull, 1960). The geometric mean concentration for selenium in each region is indicated in Figure 1. The geometric mean and maximum concentrations for selenium in the coal provinces and regions are given in Table 2. The geometric mean for selenium in almost 9,000 coal samples from the USCHEM data base is 1.7 ppm. The Gulf province has the highest geometric mean (4.6 ppm) with the Texas region having a mean of 4.7 ppm, the highest of all the regions. The Alaska province has the lowest geometric mean (0.42 ppm). The Atlantic Coast region has the lowest geometric mean (0.20) of all the regions. All western coal regions have mean selenium values of less than 2.0 ppm. All the coal regions in the Eastern and Interior provinces have at least 1.9 ppm selenium, except for the Rhode Island Meta-anthracite and Atlantic Coast regions. The highest selenium concentrations in the USCHEM data base are 52 ppm for a bituminous coal sample from West Virginia and 75 ppm for a bench sample from Iowa. The highest concentration of selenium in coal reported in the literature is 94,000 ppm in a 'dirty' coal from China (Yang et al., 1983). Stratigraphic variation of selenium in coal Four geographically separate areas were selected to illustrate inter-bed variations, and one area was selected to illustrate intra-bed variations of selenium concentration with respect to ash yield, pyritic sulphur, and organic sulphur. The five areas were part of previous coal geochemistry studies. The average selenium concentration for each coal bed is derived from 3 to 12 analyses. In each of the five areas (bituminous coals from West Virginia and Virginia, bituminous coals from Georgia and Alabama, bituminous coals from w e s t e r n K e n t u c k y , a n t h r a c i t e coals f r o m Pennsylvania, and sub-bituminous coals from Wyoming), selenium exhibits different characteristics. In the coal samples from Georgia and Alabama, the high selenium concentration corresponded to the high pyrite concentration. However, in the Pennsylvania anthracites, the high selenium concentration corresponded to a high ash yield. Selenium concentration in the western Kentucky coals may correlate positively with ash yield and pyrite concentrations, but in the West Virginia and Virginia samples such a relationship is not apparent. The selenium concentrations in the Wyoming coals correlated strongly with all three parameters tested. The relationship of selenium concentration with various coal parameters is examined more closely in the next section. Inter-bed variations: Within the West Virginia and Virginia area (Coleman etal., 1975), 14 bituminous coal beds, from the Pocahontas No.3
219
t h r o u g h the U p p e r B a n n e r coal bed, were considered. The geometric mean for selenium for each coal bed r a n g e s f r o m less than 1 to approximately 5 ppm. We did not observe a trend in selenium concentration nor a correlation between selenium concentrations and ash yield, pyritic sulphur, or organic sulphur c o n c e n t r a t i o n s (Figure 2a). In coals from Georgia and Alabama (Coleman et al., 1987a, 1987b), 13 bituminous coal beds were examined from the No.1 through the No.10 coal bed. Selenium concentrations ranged from 0.5 to 6 ppm (Figure 2b). With the exception of the No.2 coal bed, selenium concentration decreases almost linearly with stratigraphically younger beds. In the No.2 coal bed, a marked increase in selenium concentration corresponded to an anomalously high pyritic sulphur concentration (approximately 4 wt % pyritic sulphur compared to 0.5 wt % pyritic sulphur for the other coal beds). No relationship is observed between selenium concentration and ash yield, pyritic sulphur, or organic sulphur concentrations in the other beds. S e l e n i u m c o n c e n t r a t i o n in 10 w e s t e r n Kentucky bituminous coal beds, the Dunbar through the Western Kentucky No.14, ranged from 1 to 3 ppm. A rough correlation appears between selenium concentration and ash yield for coals from the Dunbar coal bed through the Kentucky No.12 coal bed, but the correlation between selenium and ash yield is absent in the two younger beds (Figure 2c). Selenium concentration generally tracks pyritic sulphur concentrations, but a trend is not observed between selenium and organic sulphur concentrations. In 10 coal beds, the No.4 (Lykens Valley) through the No. 14 (Orchard) from the Pennsylvania anthracite region (Medlin etal., 1975), the g e o m e t r i c mean for s e l e n i u m ranges f r o m approximately 1 to approximately 7 ppm. For the No.5 coal bed through the No. 14 coal bed, the range of selenium concentrations is narrow - from 1 to 4 ppm. Coal bed No.4 contains more than 7 ppm selenium. This high concentration of selenium c o r r e s p o n d s to a high ash yield (25 wt %; Figure 2d). For the other samples in this suite, selenium concentrations do not correlate with ash yield, p y r i t i c sulphur, or o r g a n i c s u l p h u r concentrations. Intra-bed variations: Two drill cores from the Canyon sub- bituminous coal bed in the Powder River basin, Wyoming, were used to determine intra-bed variation. The coal from each core was more than 60 feet thick and was sampled in intervals ranging from 3 to 17 feet. Each core showed a strong correlation between selenium concentration and ash yield, pyritic sulphur, and organic sulphur concentrations (Figures 2e and 2f). This positive correlation was attributed by Oman et al. (1988) to a single detrital source for the selenium, ash yield, and sulphur.
220
Selenium in US coals
A U Banner Kenned,/ Red Ash RA BRed Ash R B /
Jewell
Tiller
Distribution and mode of occurrence of selenium in US coals Lynn Coleman, Linda J. Bragg and Robert B. Finkelman US Geological Survey, MS 956, Reston, VA 22092, USA
Abstract
Selenium excess and deficiency have been established as the cause of various health problems in man and animals. Combustion of fossil fuels, especially coal, may be a major source of the anthropogenic introduction of selenium in the environment. Coal is enriched in selenium relative to selenium's concentration in most other rocks and relative to selenium in the Earth's crust. Data from almost 9,000 coal samples have been used to determine the concentration and distribution of selenium in US coals. The geometric mean concentration of selenium in US coal is 1.7 ppm. The highest mean selenium value (geometric mean 4.7 ppm) is in the Texas Region. Atlantic Coast (Virginia and North Carolina) and Alaska coals have the lowest geometric means (0.2 and 0.42 ppm, respectively). All western coal regions have mean selenium concentrations of less than 2.0 ppm. In contrast, all coal basins east of the Rocky Mountains (except for several small basins in Rhode Island, Virginia, and North Carolina) have mean selenium values of 1.9 or greater. Generally, variations in selenium concentration do not correlate with variations in ash yield, pyritic sulphur, or organic sulphur concentrations. This may be the result of multiple sources of selenium; however, in some non-marine basins with restricted sources of selenium, selenium has positive correlations with other coal quality parameters. Selenium occurs in several forms in coal but appears to be chiefly associated with the organic fraction, probably substituting for organic sulphur. Other important forms of selenium in coal are selenium-bearing pyrite, selenium-bearing galena, and lead selenide (clausthalite). Water-soluble and ion-exchangeable selenium also have been reported.
Introduction
Various elements in coal have been used to correlate coal beds and to help decipher depositional e n v i r o n m e n t s . S o m e elements contribute to technological problems such as boiler fouling (sodium, calcium, magnesium), slagging (iron), and corrosion (chlorine, fluorine). A few elements in coal (e.g. gold, silver, uranium) have potential as e c o n o m i c b y - p r o d u c t s , whereas others (e.g. selenium, mercury, cadmium) pose potential threats to the environment. The reader is referred to Swaine ( 1 9 9 0 ) for a d d i t i o n a l i n f o r m a t i o n on the environmental, technologic, geologic, and economic impact of trace elements in coal. The concentration levels, vertical and lateral distribution, and modes of occurrence (chemical forms) of the elements in coal have been the subject of numerous papers and books (Bouska, 1981; Valkovic, 1983; Swaine, 1990; Finkelman (in press)). Few elements in coal (other than sulphur) have attracted as much attention as has selenium. For example, Boon etal. (1987) note that most western states require selenium monitoring in coal overburden. Concern with selenium is due to the well-documented environmental and health-related problems caused by this element, including several
problems caused by selenium derived from coal. These problems are discussed below. A number of data compilations exist for selenium in US coal.Compilations by Swanson et al. (1976) and Zubovic et al. (1979, 1980) were based on information from the USCHEM (US Geochemical Data) data base, part of the US Geological Survey's National Coal Resources Data System (NCRDS). These earlier compilations were based on analyses of 799, 617, and 659 coal samples (Swanson et al., 1976; Zubovic et al., 1979, 1980, respectively). The present study is based on analyses of almost 9,000 coal samples, including all the data from the previous NCRDS compilations. Gluskoter et al. (1977) found geometric means of 3.4 ppm selenium for 23 Appalachian coal samples, 1.3 ppm for 28 western coal samples, and 2.0 ppm for 114 Illinois basin coal samples. E x a m i n i n g the enrichment (relative to the average concentration in t h e Earth's crust) of the trace elements in coal samples, they found selenium had the highest enrichment (26 to 68 times crustal abundances) in all three coal-bearing areas. Environmental and health-related problems caused by selenium
Selenium (Se) occupies a dual environmental
216
Selenium in US coals
position, as both deficiency and excess of this element can create health problems (NAS, 1974; Lakin, 1972). Selenium has long been identified as a toxin associated with various animal diseases (Beath et al., 1934; Rosenfeld and Beath, 1964; Beath, 1982). Excessive amounts of selenium have been cited as a cause of periodontal disease and dental caries in humans and as a possible carcinogen (NAS, 1974). An endemic disease in Hubei province China has been attributed to selenium intoxication. This disease is characterised by loss of hair and nails, lesions of the skin, and damage to the nervous system. Morbidity was as high as 50% (Yang et al., 1983). However, Zheng et al. (1992) report finding only 400 cases of mild selenosis symptoms in the Hubei province during the past five decades. Selenium has also been cited as the cause of massive fish kills in Martin Lake, Texas, and at three sites in North Carolina (Shepard, 1987) and as the cause of death of thousands of birds in the Kesterson National Wildlife Refuge in California (Presser and Barnes, 1984). Two e n d e m i c diseases in China, Keshan disease and Kaschin-Beck disease, are attributed to the deficiency of selenium. These diseases can be cured and prevented by adding selenates to table salt (Jiahzhong, 1988). Moreover, the US National A c a d e m y of Science (NAS, 1974) has added selenium to the list of essential nutrients. Selenium in fossil fuels and rocks The combustion of fossil fuels has been cited as a major source of selenium in the environment. A n d r e n etal. (1975) estimated that selenium mobilised by coal combustion is 1.5 to 2.5 times the amount released by the weathering of rocks. The NAS (1974, p.61) report states: "The amount of selenium entering the atmosphere from the burning of ~ossil fuels is approximately six times greater than that derived from mined ores. Preliminary estimates of selenium in rainfall indicate that regions of high fossil fuel use have higher selenium rates of deposition by rain". The endemic disease in China has been attributed to the release of selenium from weathering and combustion of a seleniferous coal (Yang et al., 1983). The fish kills in Martin Lake, Texas, and in North Carolina were caused by the mobilisation of selenium from fossil fuel waste products (Shepard, 1987). Zingaro and Cooper (1974, p. 10) state: "It is of the utmost significance to examine the selenium content of fossil fuels and thereby obtain an indication, not only of the distribution of selenium in these materials but also, of the amount of selenium released to the atmosphere from this source". Special attention has been focused on coal because of its enrichment of selenium: coal contains from 5 to 300 times the amount of selenium as do other rocks (Table 1). This enrichment can be viewed from another perspective: selenium in coal is enriched by a factor of 82 when compared to its concentration in the Earth's crust (US National
Table 1 Selenium concentration in major rock types, (Modified from Kabata-Pendias and Pendias, 1984).
Rock type
Se (ppm)
US coala Soil Shale Argillaceous sediments Sandstones Limestones, dolomites Ultramafic rocks Mafic rocks Intermediate rocks Acid rocks (intrusive) Acid rocks (extrusive) Crude oilb
1.7 0.4 0.6 0.4~).6 0.05-0.08 0.03-0.10 0.02-0.05 0.01-0.05 0.02-0.05 0.01-0.05 0.02-0.05 0.01-1.4
a This paper. b Swaine, 1978.
Committee for Geochemistry, 1980). This is about 10 times the enrichment of arsenic, the next element most highly enriched in coal.*
Sample Collection, Preparation and Analyses Since 1975, the US Geological Survey (USGS) and various state Geological Surveys have collected more than 14,000 samples of coal and associated rock from the United States. Most of these coal samples were collected and prepared according to methods outlined by Swanson and Huffman (1976). Full-bed channel samples, drill core samples, and, when necessary, grab samples were collected. Site documentation consists of locating samples in terms of topographic quadrangle, latitude and longitude, elevation, coal bed name, and formation, as well as measuring coal bed thickness, describing the coal bed, and documenting and estimating the thickness of the overlying and underlying lithologies. The USGS conducted comprehensive chemical analyses for each sample. These included proximate and ultimate analyses, determination of sulphur f o r m s , m a j o r - , m i n o r - and t r a c e - e l e m e n t concentrations, free-swelling indices, ash-fusion temperatures and heat content. See Golightly and Simon (1989) for descriptions of the analytical procedures. Selenium concentrations for most samples were determined on a whole-coal basis utilising instrumental neutron activation analysis (INAA). The detection limit of I N A A for selenium is 0.1 ppm, and the precision is estimated to be 5%. Tellurium, geochemically similar to selenium, may have the distinction of being the element most highly enriched in coal. On the basis of a limited number of analyses, Chiou and Manuel (1985) report that tellurium
L. Coleman, L. J. Bragg, and R. B. Finkelman
217
Table 2 Geometric mean and maximum concentrations of selenium in US coal regions and provinces. [No data On Southwestern, Black Hills, Tertiary lake beds, and Bighorn Basin regions.] Geographic areas
Total No. of samples
United States
8,695
304
1.7
3.6
0.02
75
Provinces: Eastern Gulf Interior Northern Great Plains Rocky Mountain Alaska Pacific Coast
4,711 214 705 1,154 1,615 258 38
113 3 11 81 45 43 8
2.5 4.6 2.4 0.66 1.2 0.42 0.76
4.2 5.6 3.1 0.99 1.6 1.1 1.9
0.02 0.50 0.20 0.10 0.10 0.10 0.20
75 16 36 13 13 43 7.3
68
2
2.9
4.4
0.30
16
19 91 1,654 1,921 1,039 10 41 172 3 401 301 324 819 11 40 34 336 60 239 689 89 59
14 2.6 28 29 36 4 1 2 0 11 0 18 63 0 0 1 6 4 1 25 3 3
0.22 3.9 2.9 2.9 1.9 0.20 4.2 4.7 2.3 2.3 2.5 0.63 0.66 1.4 1.6 0.69 1.3 0.89 1.9 0.97 1.9 1.4
0.33 0.03 3.8 3.8 3.3 1.1 5.3 5.6 2.3 2.8 3.5 0.81 1.1 1.5 1.7 1.2 1.4 1.3 2.0 1.4 2.5 1.9
0.40 75 0.20 0.02 0.03 0.10 1.4 0.50 2.1 0.50 0.20 0.10 0.10 0.70 0.66 0.10 0.35 0.20 0.64 0.10 0.31 0.38
1.2
Regions: Pennsylvania anthracites Rhode Isand meta-anthracites Appalachian 4,600 Northern Appalachian Central Appalachian Southern Appalachian Atlantic Coast Mississippi Texas Northern (Interior) Eastern (Interior) Western (Interior) Fort Unoin Powder River North Central Wind River Hams Fork Uinta Southwestern Utah San Juan River Green River Raton Mesa Denver
Total No. qualified samples a
Geometric mean b (ppm)
Arithmetic mean c (ppm)
Min. value (ppm)
Max. value (ppm)
75 18 20 1.8 14 16 2.6 13 36 3.3 13 2.7 3.1 7.9 13 3.9 5.8 9.0 8.9 9.6
a Analytical data designated as "greater than" or "less than". b From Cohen, 1976. c From Connor et al., 1976.
Precision estimates are expressed as percent relative standard deviation for a single determination on a typical coal sample as represented by National Institute of Science and Technology reference coal sample 1632. The major analytical interference is with tantalum; however, concentrations of tantalum in coal are c o m m o n l y only a fraction of the selenium concentration. Therefore, this interference is not considered significant. S e l e n i u m in some o f the western US coal
samples was determined by X-ray fluorescence spectroscopy. Sensitivity and precision o f this procedure were not as good as they were in INAA. Most of the samples analysed for the U S C H E M data base are channel or core samples. Data from bench s a m p l e s w e r e w e i g h t e d to d e r i v e a v a l u e r e p r e s e n t i n g the f u l l b e d . W h e n s t a t i s t i c a l parameters were calculated, 'less than' values for selenium were treated as if they were 0.7 of the minimum detection limit (Cohen, 1959).
218
Selenium in US coals
120 ° 110* 70 °
I00 o
'i
/ -) /
?.i. . / /I
t d
,'
'
90*
~3ppm so"
T....
1.6 l
40°
i
13
~ - ' - " -~1.1 pp-~7-mi ppm \
,
.~'\"i-
"
~I~
l -{~22~2~--~ 2.3 ~pm
;
i
1.9 Pi~ 500 MILES
3°~ ~k
~
I
500 KILOMETRES
EXPLANATION Concentration
Regions 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Rhode Island meta-anthracite Pennsylvania anthracite Atlantic coast Appalachian Northern Eastern Western Southwestern Mississippi Texas Fort Union Powder River Black Hills North Central Tertiary lake beds Bighorn Basin Wind River Hams Fork Uinta Southwestern Utah San Juan River Raton Mesa Denver Green River
2.6 ppm 2.7 ppm 2.3 ppm 2.3 ppm 2.4 ppm
Types of coal in fields ~
Anthracite
~
Low-volatile bituminous . Medium- and highvolatile bituminous
~ 4.4 ppm 4.9 ppm 0.7 ppm 0.8 ppm
Sub-bituminous
i:?~i:.~:~p:~i?~:::i~:~Lignite
1.4 ppm
1.6 ppm 1.1 ppm 1.3 ppm 1.1 ppm 1.9 pprn 2.3 ppm 1.5 ppm 1.1 ppm
Figure 1 Selenium concentrations f o r US coal regions.
L. Coleman, L. J. Bragg, and R. B. Finkelman
Distribution of Selenium in US Coal Regional distribution The coal fields of the United States can be divided into provinces and regions (Trumbull, 1960). The geometric mean concentration for selenium in each region is indicated in Figure 1. The geometric mean and maximum concentrations for selenium in the coal provinces and regions are given in Table 2. The geometric mean for selenium in almost 9,000 coal samples from the USCHEM data base is 1.7 ppm. The Gulf province has the highest geometric mean (4.6 ppm) with the Texas region having a mean of 4.7 ppm, the highest of all the regions. The Alaska province has the lowest geometric mean (0.42 ppm). The Atlantic Coast region has the lowest geometric mean (0.20) of all the regions. All western coal regions have mean selenium values of less than 2.0 ppm. All the coal regions in the Eastern and Interior provinces have at least 1.9 ppm selenium, except for the Rhode Island Meta-anthracite and Atlantic Coast regions. The highest selenium concentrations in the USCHEM data base are 52 ppm for a bituminous coal sample from West Virginia and 75 ppm for a bench sample from Iowa. The highest concentration of selenium in coal reported in the literature is 94,000 ppm in a 'dirty' coal from China (Yang et al., 1983). Stratigraphic variation of selenium in coal Four geographically separate areas were selected to illustrate inter-bed variations, and one area was selected to illustrate intra-bed variations of selenium concentration with respect to ash yield, pyritic sulphur, and organic sulphur. The five areas were part of previous coal geochemistry studies. The average selenium concentration for each coal bed is derived from 3 to 12 analyses. In each of the five areas (bituminous coals from West Virginia and Virginia, bituminous coals from Georgia and Alabama, bituminous coals from w e s t e r n K e n t u c k y , a n t h r a c i t e coals f r o m Pennsylvania, and sub-bituminous coals from Wyoming), selenium exhibits different characteristics. In the coal samples from Georgia and Alabama, the high selenium concentration corresponded to the high pyrite concentration. However, in the Pennsylvania anthracites, the high selenium concentration corresponded to a high ash yield. Selenium concentration in the western Kentucky coals may correlate positively with ash yield and pyrite concentrations, but in the West Virginia and Virginia samples such a relationship is not apparent. The selenium concentrations in the Wyoming coals correlated strongly with all three parameters tested. The relationship of selenium concentration with various coal parameters is examined more closely in the next section. Inter-bed variations: Within the West Virginia and Virginia area (Coleman etal., 1975), 14 bituminous coal beds, from the Pocahontas No.3
219
t h r o u g h the U p p e r B a n n e r coal bed, were considered. The geometric mean for selenium for each coal bed r a n g e s f r o m less than 1 to approximately 5 ppm. We did not observe a trend in selenium concentration nor a correlation between selenium concentrations and ash yield, pyritic sulphur, or organic sulphur c o n c e n t r a t i o n s (Figure 2a). In coals from Georgia and Alabama (Coleman et al., 1987a, 1987b), 13 bituminous coal beds were examined from the No.1 through the No.10 coal bed. Selenium concentrations ranged from 0.5 to 6 ppm (Figure 2b). With the exception of the No.2 coal bed, selenium concentration decreases almost linearly with stratigraphically younger beds. In the No.2 coal bed, a marked increase in selenium concentration corresponded to an anomalously high pyritic sulphur concentration (approximately 4 wt % pyritic sulphur compared to 0.5 wt % pyritic sulphur for the other coal beds). No relationship is observed between selenium concentration and ash yield, pyritic sulphur, or organic sulphur concentrations in the other beds. S e l e n i u m c o n c e n t r a t i o n in 10 w e s t e r n Kentucky bituminous coal beds, the Dunbar through the Western Kentucky No.14, ranged from 1 to 3 ppm. A rough correlation appears between selenium concentration and ash yield for coals from the Dunbar coal bed through the Kentucky No.12 coal bed, but the correlation between selenium and ash yield is absent in the two younger beds (Figure 2c). Selenium concentration generally tracks pyritic sulphur concentrations, but a trend is not observed between selenium and organic sulphur concentrations. In 10 coal beds, the No.4 (Lykens Valley) through the No. 14 (Orchard) from the Pennsylvania anthracite region (Medlin etal., 1975), the g e o m e t r i c mean for s e l e n i u m ranges f r o m approximately 1 to approximately 7 ppm. For the No.5 coal bed through the No. 14 coal bed, the range of selenium concentrations is narrow - from 1 to 4 ppm. Coal bed No.4 contains more than 7 ppm selenium. This high concentration of selenium c o r r e s p o n d s to a high ash yield (25 wt %; Figure 2d). For the other samples in this suite, selenium concentrations do not correlate with ash yield, p y r i t i c sulphur, or o r g a n i c s u l p h u r concentrations. Intra-bed variations: Two drill cores from the Canyon sub- bituminous coal bed in the Powder River basin, Wyoming, were used to determine intra-bed variation. The coal from each core was more than 60 feet thick and was sampled in intervals ranging from 3 to 17 feet. Each core showed a strong correlation between selenium concentration and ash yield, pyritic sulphur, and organic sulphur concentrations (Figures 2e and 2f). This positive correlation was attributed by Oman et al. (1988) to a single detrital source for the selenium, ash yield, and sulphur.
220
Selenium in US coals
A U Banner Kenned,/ Red Ash RA BRed Ash R B /
Jewell
Tiller
