- Email: [email protected]
Economic-environmental modeling of point source pollution in Jefferson County, Alabama, USA
Journal of Environmental Management (2002) 65, 85±94 doi:10.1006/jema.2001.0532, available online at http://www.idealibrary.com on
1
Economic-environmental modeling of point source pollution in Jefferson County, Alabama, USA Ellene Kebede*À, Dean F. SchreinerÁ and Gobena HulukaÀ À
Department of Agricultural and Environmental Sciences, Tuskegee University, Tuskegee, Alabama, USA Á Department of Agricultural Economics, Oklahoma State University, Stillwater, Oklahoma, USA Received 9 June 2000; accepted 6 November 2001
This paper uses an integrated economic-environmental model to assess the point source pollution from major industries in Jefferson County, Northern Alabama. Industrial expansion generates employment, income, and tax revenue for the public sector; however, it is also often associated with the discharge of chemical pollutants. Jefferson County is one of the largest industrial counties in Alabama that experienced smog warnings and ambient ozone concentration, 1996±1999. Past studies of chemical discharge from industries have used models to assess the pollution impact of individual plants. This study, however, uses an extended Input±Output (I±O) economic model with pollution emission coef®cients to assess direct and indirect pollutant emission for several major industries in Jefferson County. The major ®ndings of the study are: (a) the principal emission by the selected industries are volatile organic compounds (VOC) and these contribute to the ambient ozone concentration; (b) the direct and indirect emissions are signi®cantly higher than the direct emission by some industries, indicating that an isolated analysis will underestimate the emission by an industry; (c) while low emission coef®cient industries may suggest industry choice they may also emit the most hazardous chemicals. This study is limited by the assumptions made, and the data availability, however it provides a useful analytical tool for direct and cumulative emission estimation and generates insights on the complexity in choice of industries. & 2002 Elsevier Science Ltd. All rights reserved.
Keywords: economic-environment, chemical emission, ozone, volatile organic compounds, cumulative emission.
Introduction Industry expansion is a frequent strategy designed to exploit regional resource potential and enhance economic growth, employment, income, and local government tax revenue. Cities are centers of ®nancial, technical and other services, pools of skilled labor, and location of supplier industries and hence attract more manufacturing activities. Studies (Schmenner, 1982; Rainey and McNamara, 1999) have identi®ed these as factors that in¯uence industrial location decisions. Literature * Corresponding author. Email: [email protected] 0301±4797/02/$ ± see front matter
has shown that once the automobile industry started locating assembly plants in the Southeast during the 1980s and the 1990s it subsequently attracted parts suppliers (Ngandu and Kebede, 1996). Economic growth based on industrial expansion, however, is often accompanied by discharge of chemical pollutants to the air, water, and land that can affect human health and the environment. Pollution emission is a by-product of industrial activities that is not usually considered in the valuation of commodity output and economic growth. The reason is a lack of common method for the inclusion of pollution emission is such analyses. Integrated modeling can be used to assess # 2002 Elsevier Science Ltd. All rights reserved.
86
E. Kebede et al.
economic and cumulative environmental impacts of industrial expansion on a regional scale. Cumulative environmental effects arise from technical and economic linkages between industrial activities as each industry requires other supportive industries to meet input requirements. One of the services of the environment is the ability to assimilate wastes from human and economic activities. However, accumulation of pollutants in a given area can exceed the assimilative capacity of the environment. Smog is a warning sign of the exhaustion of the assimilative capacity of the atmosphere. Jefferson County, in the State of Alabama, USA, especially the city of Birmingham and its surrounding area, has major industrial activities and the highest smog index in the state of Alabama (Alabama Department of Environmental Management, 1999). These industries individually and collectively affect the environment, especially through the creation of smog. In spite of the noted effects of industrial expansion on the environment, local governments make tremendous sacri®ces to attract these industries. Industry choice can be a complicated decision-making process. Industry choices usually are made after consideration of economic and environmental impact. The economic and environmental impact can be measured by total value of output and total emission per dollar value of output, total emission as a ratio of total output. What is often omitted in measurement consideration is the hazardous nature of the emission. The objective of this study is to use a general analytical framework to assess the environmental and economic effects of major manufacturing industries in Jefferson County. The major industrial activities and inter-industry linkages are identi®ed, and we attempt to quantify total air pollutant emissions, and their contribution to the smog level, and also to examine industries, based on the hazardous nature of the emission.
Conceptual framework Economic activities in general take raw materials from the environment, transform them into useful products, and discharge waste back into the environment. Both producers and consumers use environmental services and generate wastes. Wastes generated by consumption and production processes are treated and recycled; however, it is not generally possible to recover all the residuals, and hence there is a ¯ow of residual discharges back into the environment. Economic waste is a function of
the technology used in the production process. The mass of inputs to the intermediate product sector equals product supplied to ®nal demand plus emission to the environment (Perring, 1987; James, 1985). If we represent industrial output as: X
N X
xi ;
i1 . . . N
1
i1
With a given production function: xi f ai , ki , ei
2
Pollution generation function: ri f ai , ki , ei A
N X i1
ai ;
K
N X i1
ki ;
R
N X i1
3
ri ;
T
N X
ti ,
i1
4 where xi is production function of industry i, ri is pollution generation function of industry i, ai is input requirement of industry i, ki is technology used by industry i, ei is an indicator of environmental condition, air, soil, weather, and water quality, and ti is the level of pollution reduction activity of industry i. Pollution emission is assumed, therefore, to be a function of the regular inputs, level of pollution reduction activity, and the environmental condition. Pollution emission increases with output, but may be reduced dependent on the state of the technology and attempts made at pollution abatement. Environmental condition is an important indicator for assessing the technology used and capacity utilization. It can be measured in terms of concentration of residuals at receptor points, air, water, and land. This paper addresses only industrial air emissions. Integrated economic-environmental models incorporate physical effects, production processes, and interactions of economic activities. The approach describes the economic relationship between inputs and outputs of different activities, and relates physical interaction with the environment in terms of resource use and waste emission levels. The models used to explore and jointly determine economic-environmental systems are developed to avoid the free disposal assumption (Lipnowski, 1976; Victor, 1972; Bergh, 1996). The economic-environmental model, based on the extension of input±output models has been suggested and applied in several regional studies (James et al., 1978; Leontief, 1986; Dixon and
Economic-environmental modeling of point source pollution
Hufschmidt, 1986). Isard (1972) employed input± output to depict the interrelationships among ecological variables, the consumption of resources and ¯ow of matter back to the environment in an economic-ecological framework. Input±output provides an opportunity to observe the direct and the indirect effects associated with an increase or a decrease of the ®nal demand of an industry. Inter-industry linkages show that an increase in production increases the demand for inputs, and the related pollution emission by input suppliers. Isolated treatment of industries may have a minor effect on the environment, but the cumulative result from incremental activities through the interactions of industries may be signi®cant. An integrated model allows us to obtain a proxy of the cumulative environmental effects arising from the technical and economic linkages between industries (James et al., 1978). Although beyond the scope of this paper, a dispersion model to measure the actual ozone concentration level for a given location can supplement the integrated model. This study uses an input±output model integrated with air emission coef®cients to estimate the environmental impact associated with an industry's output. The I±O model is extended to represent the interaction between economic activities and the environment. While industrial activity is measured in monetary values of output, air emission coef®cients are estimated from the total chemical emission as a direct by-product of the economic processes, hence measurement is closely related to production technology and volume of output. This study adopted net emissions for computing emission coef®cients. Net emissions are pollutants that are not captured for further processing, recycling and treatment but are discharged to the environment. The coef®cients are achieved by dividing net emissions by total output on an industry-by-industry basis (see Cumberland and Stram, 1972; James, 1985). The US Environmental Protection Agency (USEPA) regulates air pollutant emissions, and sets standards (USEPA, 1993). USEPA sets national ambient air quality standards to specify the level at which air pollutants can be safely tolerated. Based on the health safety emission standards, some compounds and chemicals are listed as hazardous. Many of the hazardous air pollutants are also volatile organic compounds that are regulated to control ambient ozone concentration. The national emission standard for hazardous air pollutants was established to provide an ample margin of safety to protect public health. Health risk is associated with duration of exposure. Acetaldehyde, benzene,
87
chlorine, naphthalene, phenol, toluene, and xylene are identi®ed as hazardous air pollutants for both odor and health risk as concentration increases (Bradstreet, 1995). Volatile organic compounds (VOCs) and nitrogen oxides undergo chemical reaction after emission creates ambient ozone concentration, and result in the formation of urban smog (Sher, 1998). These pollutants affect human health, vegetation, and building structures (Powe and Willis, 1998). Environmental epidemiological studies, though not extensive, indicate health problems due to exposure to environmental contaminants in the form of VOCs released from manufacturing industries (Krupnick et al., 1990). Inhalation of VOCs and non-volatile organic compounds, and ingestion of foods that are exposed to contaminants can result in debilitating health conditions. Benzene can cause leukemia, neurotoxic symptoms, anemia; toluene may cause disfunction of the nervous system, and eye irritation. Heavy metals such as cadmium, chromium, lead, manganese, and zinc may also cause anemia, neurological defects, and kidney dysfunction (Canter, 2000; Liu, 2000). Carel (1998) discussed the environmental effects of pollutants on the human body, especially the effect of long-term exposure.
Model and estimation procedure A two-component integrated economic-environmental model is adopted to assess the economicenvironmental impact of industrial activities in Jefferson County: the ®rst component is an economic activity model, regional I±O matrix, and the second component is a physical air emission model for selected industries. James et al. (1978); Forsund (1985); Leontief (1986); James (1994); and Lave et al. (1995) used different versions of the I±O model.
I. Regional I±O model A standard static input±output Leontief model is used to describe regional output, employment, and income: X I--Aÿ1 Y
5
where: Xvector of industry total output, (I±A)ÿ1 Leontif matrix, Iidentity matrix, Amatrix of inter-industry coef®cients, Yvector of ®nal demand.
88
E. Kebede et al.
II. Physical emission model The physical emission model is the link between the industry's output and the volume of chemical emission. The basic I±O model is extended to cover the generation of residuals. All the assumptions of input±output are extended to the physical model by assuming that each industry generates residuals in ®xed proportion to the industry's output. The generation of residues is a function of total output, intermediate inputs and ®nal demand: R f X
6
R is a matrix of pollutant emission by industry, where the elements are rkj, k type of pollutant by industry j. The generation of pollutants is, assumed identical to net emission of pollutants, that is, remedial activities are internalized in the industry technology. A matrix of pollution emission coef®cients for a dollar of output from each industry can be computed from the pollution discharge, matrix R, and total industrial output. This matrix is incorporated into the I±O model indicating the different types of chemical pollutant emission accompanying the ®nal demand of an industry. The equation is as follows: P EX
7
P E I--Aÿ1 Y
8
where P is a vector of direct and indirect chemical pollution emission with elements pk representing chemical pollutant type k associated with a unit increase in ®nal demand, E is a matrix of pollution emission coef®cients, with elements ekj, representing a measure of chemical pollutant type k, generated per unit of output of industry j (i.e. ekjrkj/Xj) where rkj is the total pollutant k and Xj is total output of industry, and E (I±A)ÿ1 is direct and indirect pollutant emission coef®cients. The solution of Equation (8) provides the estimate of the total pollution emission-cumulative emission-generated per dollar of ®nal demand (Miller and Blair, 1985; James et al., 1978). Any change in ®nal demand by a given industry will have a multiplier effect on output level of other industries, and hence chemical pollutant emission. The model assumes a linear relationship with output and ®xed coef®cients of production, applied to the year of production. Fixed coef®cients assume that the technology for the given period, i.e. input mix of a given proportion. The coef®cients will change with the change in technology at any given time. The
assumption of the ®xed coef®cient of production is extended to the physical emission model: chemical pollutants are generated in a ®xed proportion for a given technology at a given time. The application of emission coef®cients could be complicated because industries with low emission coef®cients can generate chemical emission more hazardous than those with higher emission coef®cients.
Study area Jefferson County is part of the Birmingham Metropolitan Statistical Area (MSA). The data from the 1990s show that about 20% of establishments, and 15% of employment in manufacturing for the state are in the Birmingham MSA. Jefferson County accounts for about 77% of establishments, and 82% of employment for the manufacturing sector in Birmingham. Distribution of the manufacturing industries in the 1990s showed that total manufacturing employment was mainly in three industries in a descending order: primary metals (SIC 33), fabricated metal products (SIC 34), and industrial machinery and equipment (SIC 35) (Alabama Development Of®ce, 1990 and 1996). The selected plants are located in 11 cities in Jefferson County. Forty-six percent of these plants are located in and around Birmingham. The Jefferson County EPA ®eld of®ce has seven stations at seven different locations in the county. The stations are equipped to monitor the one-hour ozone National Ambient Air Quality Standard (NAAQS compliance of 0124 ppm). Data from these ®eld of®ces showed an increase in the number of smog warnings from 1991 to 1998 (reliable data are available only for these years). Two monitoring stations exceeded the national limit more than three times during the period 1996±1998 (Alabama Department of Environmental Management, Jefferson County Field Of®ce, 1999).
Data source The selection of industries for the analysis was based on the importance of the industry in the county's economy, and the availability of complete plant level TRI data. Forty-seven plants of the selected industries, about 67% of the plants in Jefferson County, were included in the development of pollution emission coef®cients. Industry average was used to develop the I±O economic and the
Economic-environmental modeling of point source pollution
pollution coef®cient matrix. The I±O economic model was compiled from the Minnesota IMPLAN Group (MIG) data for Jefferson County, and the State of Alabama for the year 1996. IMPLAN is an input±output database available in microcomputer software developed by USDA Forest Service and currently licensed under agreement with the Minnesota IMPLAN Group (MIG). The data for the emission model were collected from the 1996 USEPA Toxic Release Inventory System (TRIS), an online database, facility report for manufacturing industries in the State of Alabama and Jefferson County (USEPA, 1996a). Industries, classi®ed under standard industrial classi®cation (SIC) 20± 39, employing an equivalent of 10 full-time workers, and either manufacturing or processing more than 11 340 kilograms of the chemicals indexed under the toxic chemical categories by the USEPA are required to report TRI data at the end of each year. The reported chemical emissions were used to estimate air emission of toxic chemicals and metal compounds. The emission coef®cients were computed from net discharge of chemical pollutant emissions associated with individual plants as reported in the TRI database for 1996. Net discharge is total emission minus source reduction, recycling, treatment, and disposal. Industries report source reduction, on-site treatment, off-site transfer, but data were not suf®ciently detailed for each industry. Four-digit (SIC) level TRI data were collected for all the manufacturing industries in Jefferson County and in the State of Alabama.
Empirical results Table 1 shows the distribution of the reporting plants by IMPLAN code. Except for the 10 plants categorized as industrial organic chemicals (190) and paints and allied products (200), the predominant number of plants are categorized as primary and fabricated metals. This is consistent with the industrial distribution observed in the state's industrial directories for 1990 and 1996, and thus the sample of industries is representative of the industrial activities in the county. The last column shows the share of the industries as a percentage of industries in Jefferson County. These industries produce 41% of the total manufacturing output in the county. Blast furnaces and steel mills (254), iron and steel foundries (259), and aircraft (389) account for 31% of manufacturing output with the remaining industries accounting for 10%.
89
Table 1. Number of plants and output of selected industries as a share of total manufacturing output in Jefferson County, 1996 Industry Industrial organic chemicals Paints and allied products Blast furnaces and steel mills Iron and steel foundries Secondary nonferrous primary metal products Metal cans Miscellaneous fabricated structure Miscellaneous metal works Plating and metal coating Aircraft Total a b
IMPLAN Number Output codea of plantsb sharea 190 200 254 259 263
5 5 6 6 5
001 002 019 007 0.01
273 282
4 5
002 001
288 296 389
4 4 3 47
0.02 001 005 041
IMPLAN code and output data. USEPA, TRI data base, 1996a.
Column 4 of Table 1 shows the share of the selected industry as a percentage of the industry's output at the state level. Except for secondary nonferrous primary metals (263), and miscellaneous fabricated structure (282), the rest of the industries produce a signi®cant portion of the total state output. Most of the plants are located in Birmingham and surrounding cities. These industries account for about 58 and 22 percent of the total state output and employment, respectively (Alabama Development Of®ce, 1990 and 1996). Table 2 shows total chemical and metal compound air emission by the selected industries, and all industries in Jefferson County. The chemical compounds are composed of volatile organic compounds (VOCs)Ðalcohol, benzene, chlorine, ethers/ ketones, and toluene/xylene, other VOCs. The category of other VOCs consists of VOCs that account for less than 5% of the total emission individually. Metal compounds are composed of barium, lead, manganese, zinc, and other metals. VOCs accounted for about 85% of the total air emission, while metal compounds accounted for 15%. The percentage distribution shows that other VOCs accounted for 36%, toluene/xylene accounted for 30%, zinc accounted for 13%, ethers/ketones accounted for 10%, and alcohol and benzene accounted for about 4%. These compounds are identi®ed as hazardous and high risk by (USEPA, 1996b; Lui, 2000; Canter, 2000). At the lower level of the atmosphere, VOCs are the major components of smog and are serious environmental and health hazards. The USEPA listed carbon disul®de, methanol, copper compounds, zinc compounds, ammonia, xylene,
90
E. Kebede et al.
Table 2. Chemical and metal compound air emission by selected industries as a share of emission by all industries in Jefferson County and of the selected industries (kgs) Chemical compounds Alcohol Benzene Chlorine Ethers/ketones Toluene/xylen Other VOCs1 Barium compounds Lead compounds Manganese compounds Zinc compounds Other metal compounds2 Total
Emission for all industries in the County
Emission for selected industries
Emission as share of County
Share of emission for selected industries
72 660 153 204 1 372 251 317 378 726 294 203 9 978 11 756 15 606 125 268 7611 1 317 167
14 320 30 545 1 254 77 763 236 277 285 414 9 862 8 611 10 810 104 432 3788 783 078
020 020 091 032 062 097 099 073 070 083 050 060
002 004 000 010 030 036 001 001 001 013 000 100
Source: TRI database (USEPA, 1996b). 1 Other VOCs among others include: ammonia, anthracene, creoste, diethanolamine, ethylene, formaldehyde, naphthalene, and tricholoethylene. 2 Other metal compounds include: antimony, cadmium, chromium, copper, cobalt, and nickel.
Table 3. Total chemical and metal compounds air emission of selected industries by type of chemical, Jefferson County 1996 (kgs) Ind.
Alcohol
Benzene
Ethers/ ketones
Toluene/ xylene
Other VOCs1
Barium2
Lead2
Mang2
Zinc2
Other Metals3
Total
190 200 254 259 263 273 282 288 296 389 Total
4 854 3 051 1 402 0 0 0 0 5 014 0 0 14 320
7 324 4 035 13 746 0 0 406 5 035 0 0 0 30 546
2 068 22 408 0 12 053 0 936 0 17 401 0 22 897 77 764
10 891 28 693 30 662 96 444 0 953 2 722 0 48 717 17 201 236 281
53 382 633 38 452 11 052 125 82 611 0 0 0 98 835 285 091
0 0 65 9798 0 0 0 0 0 0 9863
0 0 8153 0 454 0 0 0 5 0 8612
0 0 9 561 1 134 0 0 2 113 0 0 10 810
0 80 98 064 113 3334 0 0 0 2 842 0 104 432
0 514 1 879 366 916 0 0 0 113 0 3 788
78 518 59 415 201 983 130 960 4 829 84 907 7 759 22 528 51 677 138 934 781 509
Source: TRI database (USEPA, 1996b). 1 Other VOCs include: ammonia, anthracene, creoste, diethanolamine, ethylene, formaldehyde, naphthalene, and tricholoethylene. 2 Metal compounds. 3 Other metal compounds include: antimony, cadmium, chromium, copper, cobalt and nickel.
hydrochloric acid (acid aerosol form), toluene, manganese compounds, and methyl ethyl ketone, as the ten highest hazardous releases. Table 2 also shows emissions for the selected industries as a share of total emissions by all industries in Jefferson County. The emission of these industries is approximately 60% of the total pollution emission by industrial plants in the county. They account for a signi®cant amount of other VOCs and chlorine compounds generated in the county, 97% and 91%, respectively. A large proportion of the metal compounds is also generated by these industries: 99% for barium compounds, 83% for zinc compounds, 73% for lead compounds, 70% for manganese compounds.
Table 3 provides the chemical emission for all the plants in selected industries. Blast furnaces and steel mill industries (254), aircraft industries (389), and the iron and steel foundries (259) generate the highest volumes of pollution and accounted for about 60% of the total pollution generated by selected industries. Industrial organic chemicals (190), paints and allied products (200), metal cans (273), and plating and metal coating (296) produced the next highest volumes, while secondary nonferrous primary metals (263) and miscellaneous fabricated structure (282) produced the least. Emission by type of pollutants varied substantially across industries. Almost all industries generate some level of toluene/xylene with the highest
Economic-environmental modeling of point source pollution
91
Table 4. Estimated direct pollution emission coef®cients for selected industries for the State of Alabama and Jefferson County, 1996 Industry
190 200 254 259 263 273 282 288 296 389
Alabama
Jefferson County
Chemical1 emission (kgs)
Output2 $10 000
Direct emission coef®cient (kgs/$10 000)
Chemical1 emission (kgs)
Output2 $10 000
Direct emission coef®cient (kgs/$10 000)
1 000 287 177 654 312 669 287 151 16 942 123 994 56 112 34 850 271 210 235 860
96 8884 26 6267 219 2146 107 6309 26 5698 12 1089 49 1696 16 7388 74681 58 2081
1032 667 143 267 064 1024 114 208 3632 405
78 518 59 415 201 983 130 960 4829 84 907 7759 22 528 51 677 138 934
77508 87452 140 0891 51 3632 40624 82993 67441 13 9199 60681 34 2904
1013 679 144 255 119 1023 115 162 852 405
Figures are rounded to two decimal places. Source: 1 TRI database (USEPA, 1996b). 2 Industrial output (1996 IMPLAN data).
emission from iron and steel foundries (259) and plating and metal coating (296): 96 444 and 48 716 kgs respectively. The next highest were blast furnaces and steel mills (254), 30 662 kgs; paints and allied products (200), 28 693 kgs; aircraft industries (389), 17 201 kgs; and industrial organic chemicals (190), 10 891 kgs. Blast furnaces and steel mills (254) produce the highest volume of metals pollutants. They produce 98 064 kgs of zinc that accounts for about 76% of the total emission by the selected industries.
Emission coef®cients The industrial output at producer prices, and the industry TRI data for 1996 were used to compute the emission coef®cients. Emission coef®cients are the link between economic activity and the environment, and indicate the level of pollution emission per $10 000 of industry output. Emission coef®cients for the State of Alabama and Jefferson County are presented in Table 4 for the selected industries. The emission coef®cient of the State of Alabama was computed to compare and test the validity of the coef®cients at the county level. The emission coef®cients for the State of Alabama and Jefferson County are comparable, except for secondary and nonferrous primary metals, N.E.C (263), miscellaneous metal works (288), and plating and metal coating (296) industries. For these three industries, there were data differences because of aggregation and conversion from SIC code to
IMPLAN code. The IMPLAN code included several SIC numbers in the output data that created some discrepancy between the state and the county coef®cients. The industries with the highest emission coef®cients are metal cans (273), organic chemicals (190), metal plating and coating (296), and paints and allied products (200) at 1023, 1013, 852, and 679 kgs of pollutants per $10 000, respectively. The industries with the lowest pollution per unit of output are respectively, secondary nonferrous primary metals (263), miscellaneous metal works (288), iron and steel foundries (259), aircraft (389), 147, 161, 255, and 405 kgs per $10 000. As opposed to total emission (Table 3), blast furnaces and steel mills (254) have the lowest emission coef®cients, 144 kgs per $10 000 output. Some of the industries with the highest total pollution have the lowest coef®cients, indicating that they had higher total output than industries with the lower volume of pollution but higher emission coef®cients. Direct emission coef®cients by type of chemical were computed to examine the speci®c emission and the hazard level of each industry per unit of output. Table 5 shows that industrial organic chemicals (190), metal cans (273), miscellaneous fabricated structures (282), and aircraft (389) produce only VOCs. These chemicals are easily assimilated in the atmosphere and are the main sources of smog, odor, and fog. Paints and allied products (200), blast furnaces and steel mills (254), iron and steel foundries (259), secondary nonferrous primary metals (263), miscellaneous metal works (288), and plating and metal coating (296) generate both VOCs and
92
E. Kebede et al.
Table 5. Industry
Estimated direct pollution emission coef®cients for selected industries by type of chemical (kgs/$10 000) Alcohol/ benzene
Ethers/ ketones
Toluene/ xylene
Other VOCs1
Barium2
Lead2
Manganese2
Zinc2
Other metals3
Total
157 081 011 0 0 005 075 036 0 0
027 256 0 024 0 011 0 125 0 067
141 328 022 186 0 011 040 0 803 050
689 007 028 022 003 995 0 0 0 288
0 0 0 020 0 0 0 0 0 0
0 0 006 0 011 0 0 0 0 0
0 0 007 003 0 0 0 001 0 0
0 001 070 0 082 0 0 0 047 0
0 006 001 001 023 0 0 0 002 0
1013 679 144 255 119 1023 115 162 851 405
190 200 254 259 263 273 282 288 296 389
Figures are rounded to two decimal places. 1 Other VOCs include: ammonia, anthracene, creoste, diethanolamine, ethylene, formaldehyde, naphthalene, and tricholoethylene. 2 Metal compounds. 3 Other metal compounds include: antimony, cadmium, chromium, copper, and nickel.
Table 6. Direct and indirect pollution emission coef®cients of selected industries, Jefferson County, 1996 Industry
Direct emission coef®cienta
Direct and indirect emission
Emission multiplier
10130 6794 1442 2550 1189 10231 1150 1618 8516 4052
10601 7308 1916 3013 1513 10690 1425 1948 8994 4187
1046 1075 1328 1181 1273 1045 1239 1204 1056 1033
190 200 254 259 263 273 282 288 296 389 a
See Table 4.
metal compounds including barium, lead, manganese, and zinc compounds. Metal compounds take the form of particulates, and tend to settle in the lower levels of the atmosphere, and are major sources of acid rain that affect vegetation and soil, surface and ground water contamination, and deposits on structures. The inter-industry linkages in the I±O model re¯ect the direct and indirect effect of each industry on the economy. The estimated coef®cients in Table 4 indicate only the direct pollution emission generated by each industry in 1996. Table 6 provides the direct and indirect pollution emission coef®cients for each industry in 1996. The I±O model is used to compute the direct and indirect emission coef®cient. Comparison of the direct pollution emission and the direct and indirect emission coef®cients (cumulative emission) for 1996 show that aircraft industry (389) had the lowest percentage increase of 33%, blast furnaces and steel mills (254) have the highest (328%). The percentage
increase re¯ects the indirect linkages; the multiplier effect, of the industry on the total economy. While blast furnaces and steel mills (254) have the highest indirect increase, the next highest increase is secondary nonferrous primary metal industry (263) at 273%. The indirect volume of pollution increased by 239% for miscellaneous fabricated structures (282) and 204% for miscellaneous metal works (288). Iron and steel foundries (295) increased indirect pollution emission by 181%. The remaining industries increased indirect pollution by 75% or less.
Summary and conclusion Jefferson County, especially Birmingham and the surrounding cities, is the center of manufacturing industries, and also has the highest level of smog warnings in the State of Alabama. The majority of the industries are classi®ed as primary and fabricated metal products. This study selected 10 industries that consisted of 47 plants, about 67% of the plants in the county. These industries include: industrial organic chemicals (190); paints and allied products (200); blast furnaces and steel mills (254); iron and steel foundries (259); secondary nonferrous primary metals (263), metal cans (273); miscellaneous fabricated structure (282) and miscellaneous metal works (282); plating and metal coating (296); and the aircraft industry (389). The majority of the plants are categorized as primary and fabricated metals. These selected industries account for 41% of the county's total output. They accounted for 97% of the volatile organic compounds (VOCs), that are the major source of smog and odor,
Economic-environmental modeling of point source pollution
acid rain, contamination of surface and ground water, and deposits on structures. The distribution shows that industries (254), (259), and (389) account for 31% of output, and 60% of the total volume of pollution by the selected ten industries. Emission coef®cients indicate the relationship between the industry's contribution to output of the economy and the associated impact on the environment. Direct emission coef®cients, emission per $10 000 output, show that the industries with highest total volume of chemical emission (254), (259), and (389) had a lowest emission coef®cient, i.e. higher at output relative to volume of pollution emission. Industries (190), (200), (263), and (296) had highest emission coef®cients, i.e. lowest output relative to volume of pollution emission. All of the industries in the sample generate VOCs, however, the direct emission coef®cients by type of chemical show that (254) and (259) with the lowest emission coef®cients generate the highest VOCs, especially toluene/xylene, and the highest level of metal compounds, especially zinc. These chemicals are referred as hazardous by USEPA. Furthermore, cumulative emission from inter-industry linkages using 1996 data show that there is an increase of 33% to 328% in the total emission by each industry group. The aircraft industry (389) had the lowest increase 33%. These percentage increases illustrate that an isolated treatment of the plant will underestimate the level of total pollutants generated by an industry. An increase in the ®nal demand (i.e. exports) for the products of each industry will have a direct and indirect effect on the commodity output, and emission of VOCs and metal compounds (hazardous chemicals). In conclusion it is safe to say that these industries individually and collectively impact the ozone ambient concentration, increase the odor, and smog level in Jefferson County in and around Birmingham. Based on the economic-environmental impact industries with high levels of output and low chemical emission coef®cients may be favored as part of an overall development strategy; however, they may also generate hazardous waste. The impact of these industries might be low at low levels of production, but with expansion and increase in volume of output, there will be an increase in emissions. The results of this study are limited by the assumptions made and the accuracy of the data used. Estimates from the study must be interpreted cautiously. The output data used are industrial averages. TRI data include only a portion of emission because discharge data from small plants, and for some industries, are unavailable. Despite the
93
shortcomings of the study the TRI data set is the only source of industrial chemical discharge. Thus while the analysis is only partial, environmental implications are general. However, the economicenvironmental analysis developed in this study provides a useful analytical tool for direct and cumulative emission estimation, and generates insights on the complexity in the choice of industries, based on economic and environmental impacts.
Acknowledgements The authors would like to thank an anonymous reviewer and Dr Curtis M. Jolly for constructive suggestion and comments. The Study is funded by a USDA/CSREES capacity building grant.
References Alabama Development Of®ce. Alabama Industrial Directory 1990±1996. Alabama Development Of®ce, Alabama Center for Commerce, Montgomery, Alabama. Alabama Department of Environmental Management (1999). Exceedances the Birmingham Ozone Non Attainment Area 1991±1998. Unpublished material, Jefferson County Of®ce. Bergh, Jeroen C. J. M. van den (1996). Ecological Economics and Sustainable Development, Theory, Methods, and Application. Cheltenham, UK; Brook®eld, USA: Edward Elgar. Bradstreet, Jeffery W. (1995). Hazardous Air Pollutants, Assessment, Liabilities, and Regulatory Compliance. Park Ridge, New Jersey, USA: Noyes Publication. Canter, L. W. (2000). VOCs and HAPs emission from chemical plants. In Air Emission (D. H. F. Lui and L. W. Canter, eds), pp. 20±22. Boca Raton, Florida: Lewis Publishers. Carel, Rafael S. (1998). Health Aspects of Air Pollution. In Handbook of Air Pollution from Internal Combustion Engines (Fran Sher, ed.), pp. 42±64. Boston, London: Academic Press. Cumberland, John H. and Stram, Bruce N. (1972). Effects of Economic Development Upon Water Resources. Technical Report No.18, College Park, Maryland: University of Maryland, Bureau of Business and Economic Research. Dixon, John A. and Hufschmidt, Maynard M. (eds) (1986). Economic Evaluation Techniques for the Environment, A Case Study Workbook. Baltimore, USA: Johns Hopkins Press Ltd. Forsund, Finn R. (1985). Input±output models, national economic models, and the environment. In Handbook of Natural Resource and Energy Economics, Vol. I. (A. V. Kneese and J. L. Sweeney, eds). pp. 325±341. Amsterdam: Elsevier Science Publishers, B.V. Isard, Walter (1972). Ecological-Economic Analysis for Regional Development. New York: The Free Press.
94
E. Kebede et al.
James, David (1994). The Application of Economic Techniques in Environmental Impact Assessment. The Netherlands: Kluwer Academic Publishers. James, David (1985). Environmental economics, Industrial process models, and regional residual management. In Handbook of Natural Resource and Energy Economics, Vol. I A. V. Kneese and J. L. Sweeney (eds), pp. 271±324. Amsterdam: Elsevier Science Publishers, B.V. James, D., Jansen, H. and Opschoor, J. B. (1978). Economic Approaches to Environmental Problems. Amsterdam: Elsevier Science Publishers, B.V. Krupnick, A. J., Harrington, W. and Ostro, B. (1990). Ambient ozone and acute health effects: evidence from daily data. Journal of Environmental Economics and Management 18, 1±18. Lave, Lester B., Cobas-Flores, Elisa, Hendrickson, Christ T. and McMichael, F. C. (1995). Using input±output analysis to estimate economy-wide discharges. Environmental Science and Technology 29, 420±426. Leontief, Wassily (1986). Input±Output Economics 2nd Edn. New York: Oxford University Press. Lipnowski, I. F. (1976). An input±output analysis of environmental preservation. Journal of Environmental Economics and Management 3, 205±214. Liu, D. H. F. (2000). HAPs and synthetic organic chemical manufacturing industries. In Air Pollution (D. H. F. Lui and L. W. Canter, eds), pp. 24±29. Boca Raton, Florida: Lewis Publishers. Miller, Ronald E. and Blair, Peter, D. (1985). Input± output analysis foundations and extensions, pp. 236± 265. Englewood Cliffs, New Jersey: Prentice-Hall, Inc. Minnesota Implan Group (MIG, Inc) (1999). 1996 IMPLAN Data, MIG Inc, 1725 Tower Dr W Suite 140, Stillwater, MN 55082.
Ngandu, Mudiayi and Kebede, Ellene (1996). Foreign Investment, Jobs, Equity, Rural Development, and the Emerging Automobile Industry in the Southeast. Research Working Paper #96-1, Tuskegee, Alabama: George Washington Carver Agricultural Experiment Station and The School of Agriculture and Home Economics, Tuskegee University. Perring, Charles (1987). Economy and Environment, pp. 15±80. Cambridge, NY: Cambridge University Press. Powe, N. A. and Willis, K. G. (1998). Industrial location and residential disamenity: a case study of the chemical industry in Castleford, England. Journal of Environmental Management 53, 17±29, doi:10.1006/ jema.1998.0193. Rainey, Daniel V. and McNamara, Kevin T. (1999). Taxes and location decision of manufacturing establishment. Review of Agricultural Economics 21, 86±98. Schmenner, Roger W. (1982). Making Business Location Decisions. Englewood Cliffs, New Jersey: Prentice-Hall. Sher, Fran (1998). Environment aspect of air pollution. In Handbook of Air Pollution from Internal Combustion Engines Fran Sher (ed.). pp. 27±41. Boston, London: Academic Press. Victor, P. A. (1972). Pollution, Economy, and Environment. London: George Allen and Unwin. United States Environmental Protection Agency (USEPA). (1996a). 1996 Toxic Release Inventory (TRI) Database (http:// www.rtk.net). United States Environmental Protection Agency (USEPA). (1996b). State Toxic Release Inventory Report (http://www.epa.gov/tri/tri96/data/index.htm). United States Environmental Protection Agency (USEPA). (1993). The 1990 Clean Air Act, USEPA, Of®ce of Air Quality Planning and Standard, EPA-4±K-93-001.
1
Economic-environmental modeling of point source pollution in Jefferson County, Alabama, USA Ellene Kebede*À, Dean F. SchreinerÁ and Gobena HulukaÀ À
Department of Agricultural and Environmental Sciences, Tuskegee University, Tuskegee, Alabama, USA Á Department of Agricultural Economics, Oklahoma State University, Stillwater, Oklahoma, USA Received 9 June 2000; accepted 6 November 2001
This paper uses an integrated economic-environmental model to assess the point source pollution from major industries in Jefferson County, Northern Alabama. Industrial expansion generates employment, income, and tax revenue for the public sector; however, it is also often associated with the discharge of chemical pollutants. Jefferson County is one of the largest industrial counties in Alabama that experienced smog warnings and ambient ozone concentration, 1996±1999. Past studies of chemical discharge from industries have used models to assess the pollution impact of individual plants. This study, however, uses an extended Input±Output (I±O) economic model with pollution emission coef®cients to assess direct and indirect pollutant emission for several major industries in Jefferson County. The major ®ndings of the study are: (a) the principal emission by the selected industries are volatile organic compounds (VOC) and these contribute to the ambient ozone concentration; (b) the direct and indirect emissions are signi®cantly higher than the direct emission by some industries, indicating that an isolated analysis will underestimate the emission by an industry; (c) while low emission coef®cient industries may suggest industry choice they may also emit the most hazardous chemicals. This study is limited by the assumptions made, and the data availability, however it provides a useful analytical tool for direct and cumulative emission estimation and generates insights on the complexity in choice of industries. & 2002 Elsevier Science Ltd. All rights reserved.
Keywords: economic-environment, chemical emission, ozone, volatile organic compounds, cumulative emission.
Introduction Industry expansion is a frequent strategy designed to exploit regional resource potential and enhance economic growth, employment, income, and local government tax revenue. Cities are centers of ®nancial, technical and other services, pools of skilled labor, and location of supplier industries and hence attract more manufacturing activities. Studies (Schmenner, 1982; Rainey and McNamara, 1999) have identi®ed these as factors that in¯uence industrial location decisions. Literature * Corresponding author. Email: [email protected] 0301±4797/02/$ ± see front matter
has shown that once the automobile industry started locating assembly plants in the Southeast during the 1980s and the 1990s it subsequently attracted parts suppliers (Ngandu and Kebede, 1996). Economic growth based on industrial expansion, however, is often accompanied by discharge of chemical pollutants to the air, water, and land that can affect human health and the environment. Pollution emission is a by-product of industrial activities that is not usually considered in the valuation of commodity output and economic growth. The reason is a lack of common method for the inclusion of pollution emission is such analyses. Integrated modeling can be used to assess # 2002 Elsevier Science Ltd. All rights reserved.
86
E. Kebede et al.
economic and cumulative environmental impacts of industrial expansion on a regional scale. Cumulative environmental effects arise from technical and economic linkages between industrial activities as each industry requires other supportive industries to meet input requirements. One of the services of the environment is the ability to assimilate wastes from human and economic activities. However, accumulation of pollutants in a given area can exceed the assimilative capacity of the environment. Smog is a warning sign of the exhaustion of the assimilative capacity of the atmosphere. Jefferson County, in the State of Alabama, USA, especially the city of Birmingham and its surrounding area, has major industrial activities and the highest smog index in the state of Alabama (Alabama Department of Environmental Management, 1999). These industries individually and collectively affect the environment, especially through the creation of smog. In spite of the noted effects of industrial expansion on the environment, local governments make tremendous sacri®ces to attract these industries. Industry choice can be a complicated decision-making process. Industry choices usually are made after consideration of economic and environmental impact. The economic and environmental impact can be measured by total value of output and total emission per dollar value of output, total emission as a ratio of total output. What is often omitted in measurement consideration is the hazardous nature of the emission. The objective of this study is to use a general analytical framework to assess the environmental and economic effects of major manufacturing industries in Jefferson County. The major industrial activities and inter-industry linkages are identi®ed, and we attempt to quantify total air pollutant emissions, and their contribution to the smog level, and also to examine industries, based on the hazardous nature of the emission.
Conceptual framework Economic activities in general take raw materials from the environment, transform them into useful products, and discharge waste back into the environment. Both producers and consumers use environmental services and generate wastes. Wastes generated by consumption and production processes are treated and recycled; however, it is not generally possible to recover all the residuals, and hence there is a ¯ow of residual discharges back into the environment. Economic waste is a function of
the technology used in the production process. The mass of inputs to the intermediate product sector equals product supplied to ®nal demand plus emission to the environment (Perring, 1987; James, 1985). If we represent industrial output as: X
N X
xi ;
i1 . . . N
1
i1
With a given production function: xi f ai , ki , ei
2
Pollution generation function: ri f ai , ki , ei A
N X i1
ai ;
K
N X i1
ki ;
R
N X i1
3
ri ;
T
N X
ti ,
i1
4 where xi is production function of industry i, ri is pollution generation function of industry i, ai is input requirement of industry i, ki is technology used by industry i, ei is an indicator of environmental condition, air, soil, weather, and water quality, and ti is the level of pollution reduction activity of industry i. Pollution emission is assumed, therefore, to be a function of the regular inputs, level of pollution reduction activity, and the environmental condition. Pollution emission increases with output, but may be reduced dependent on the state of the technology and attempts made at pollution abatement. Environmental condition is an important indicator for assessing the technology used and capacity utilization. It can be measured in terms of concentration of residuals at receptor points, air, water, and land. This paper addresses only industrial air emissions. Integrated economic-environmental models incorporate physical effects, production processes, and interactions of economic activities. The approach describes the economic relationship between inputs and outputs of different activities, and relates physical interaction with the environment in terms of resource use and waste emission levels. The models used to explore and jointly determine economic-environmental systems are developed to avoid the free disposal assumption (Lipnowski, 1976; Victor, 1972; Bergh, 1996). The economic-environmental model, based on the extension of input±output models has been suggested and applied in several regional studies (James et al., 1978; Leontief, 1986; Dixon and
Economic-environmental modeling of point source pollution
Hufschmidt, 1986). Isard (1972) employed input± output to depict the interrelationships among ecological variables, the consumption of resources and ¯ow of matter back to the environment in an economic-ecological framework. Input±output provides an opportunity to observe the direct and the indirect effects associated with an increase or a decrease of the ®nal demand of an industry. Inter-industry linkages show that an increase in production increases the demand for inputs, and the related pollution emission by input suppliers. Isolated treatment of industries may have a minor effect on the environment, but the cumulative result from incremental activities through the interactions of industries may be signi®cant. An integrated model allows us to obtain a proxy of the cumulative environmental effects arising from the technical and economic linkages between industries (James et al., 1978). Although beyond the scope of this paper, a dispersion model to measure the actual ozone concentration level for a given location can supplement the integrated model. This study uses an input±output model integrated with air emission coef®cients to estimate the environmental impact associated with an industry's output. The I±O model is extended to represent the interaction between economic activities and the environment. While industrial activity is measured in monetary values of output, air emission coef®cients are estimated from the total chemical emission as a direct by-product of the economic processes, hence measurement is closely related to production technology and volume of output. This study adopted net emissions for computing emission coef®cients. Net emissions are pollutants that are not captured for further processing, recycling and treatment but are discharged to the environment. The coef®cients are achieved by dividing net emissions by total output on an industry-by-industry basis (see Cumberland and Stram, 1972; James, 1985). The US Environmental Protection Agency (USEPA) regulates air pollutant emissions, and sets standards (USEPA, 1993). USEPA sets national ambient air quality standards to specify the level at which air pollutants can be safely tolerated. Based on the health safety emission standards, some compounds and chemicals are listed as hazardous. Many of the hazardous air pollutants are also volatile organic compounds that are regulated to control ambient ozone concentration. The national emission standard for hazardous air pollutants was established to provide an ample margin of safety to protect public health. Health risk is associated with duration of exposure. Acetaldehyde, benzene,
87
chlorine, naphthalene, phenol, toluene, and xylene are identi®ed as hazardous air pollutants for both odor and health risk as concentration increases (Bradstreet, 1995). Volatile organic compounds (VOCs) and nitrogen oxides undergo chemical reaction after emission creates ambient ozone concentration, and result in the formation of urban smog (Sher, 1998). These pollutants affect human health, vegetation, and building structures (Powe and Willis, 1998). Environmental epidemiological studies, though not extensive, indicate health problems due to exposure to environmental contaminants in the form of VOCs released from manufacturing industries (Krupnick et al., 1990). Inhalation of VOCs and non-volatile organic compounds, and ingestion of foods that are exposed to contaminants can result in debilitating health conditions. Benzene can cause leukemia, neurotoxic symptoms, anemia; toluene may cause disfunction of the nervous system, and eye irritation. Heavy metals such as cadmium, chromium, lead, manganese, and zinc may also cause anemia, neurological defects, and kidney dysfunction (Canter, 2000; Liu, 2000). Carel (1998) discussed the environmental effects of pollutants on the human body, especially the effect of long-term exposure.
Model and estimation procedure A two-component integrated economic-environmental model is adopted to assess the economicenvironmental impact of industrial activities in Jefferson County: the ®rst component is an economic activity model, regional I±O matrix, and the second component is a physical air emission model for selected industries. James et al. (1978); Forsund (1985); Leontief (1986); James (1994); and Lave et al. (1995) used different versions of the I±O model.
I. Regional I±O model A standard static input±output Leontief model is used to describe regional output, employment, and income: X I--Aÿ1 Y
5
where: Xvector of industry total output, (I±A)ÿ1 Leontif matrix, Iidentity matrix, Amatrix of inter-industry coef®cients, Yvector of ®nal demand.
88
E. Kebede et al.
II. Physical emission model The physical emission model is the link between the industry's output and the volume of chemical emission. The basic I±O model is extended to cover the generation of residuals. All the assumptions of input±output are extended to the physical model by assuming that each industry generates residuals in ®xed proportion to the industry's output. The generation of residues is a function of total output, intermediate inputs and ®nal demand: R f X
6
R is a matrix of pollutant emission by industry, where the elements are rkj, k type of pollutant by industry j. The generation of pollutants is, assumed identical to net emission of pollutants, that is, remedial activities are internalized in the industry technology. A matrix of pollution emission coef®cients for a dollar of output from each industry can be computed from the pollution discharge, matrix R, and total industrial output. This matrix is incorporated into the I±O model indicating the different types of chemical pollutant emission accompanying the ®nal demand of an industry. The equation is as follows: P EX
7
P E I--Aÿ1 Y
8
where P is a vector of direct and indirect chemical pollution emission with elements pk representing chemical pollutant type k associated with a unit increase in ®nal demand, E is a matrix of pollution emission coef®cients, with elements ekj, representing a measure of chemical pollutant type k, generated per unit of output of industry j (i.e. ekjrkj/Xj) where rkj is the total pollutant k and Xj is total output of industry, and E (I±A)ÿ1 is direct and indirect pollutant emission coef®cients. The solution of Equation (8) provides the estimate of the total pollution emission-cumulative emission-generated per dollar of ®nal demand (Miller and Blair, 1985; James et al., 1978). Any change in ®nal demand by a given industry will have a multiplier effect on output level of other industries, and hence chemical pollutant emission. The model assumes a linear relationship with output and ®xed coef®cients of production, applied to the year of production. Fixed coef®cients assume that the technology for the given period, i.e. input mix of a given proportion. The coef®cients will change with the change in technology at any given time. The
assumption of the ®xed coef®cient of production is extended to the physical emission model: chemical pollutants are generated in a ®xed proportion for a given technology at a given time. The application of emission coef®cients could be complicated because industries with low emission coef®cients can generate chemical emission more hazardous than those with higher emission coef®cients.
Study area Jefferson County is part of the Birmingham Metropolitan Statistical Area (MSA). The data from the 1990s show that about 20% of establishments, and 15% of employment in manufacturing for the state are in the Birmingham MSA. Jefferson County accounts for about 77% of establishments, and 82% of employment for the manufacturing sector in Birmingham. Distribution of the manufacturing industries in the 1990s showed that total manufacturing employment was mainly in three industries in a descending order: primary metals (SIC 33), fabricated metal products (SIC 34), and industrial machinery and equipment (SIC 35) (Alabama Development Of®ce, 1990 and 1996). The selected plants are located in 11 cities in Jefferson County. Forty-six percent of these plants are located in and around Birmingham. The Jefferson County EPA ®eld of®ce has seven stations at seven different locations in the county. The stations are equipped to monitor the one-hour ozone National Ambient Air Quality Standard (NAAQS compliance of 0124 ppm). Data from these ®eld of®ces showed an increase in the number of smog warnings from 1991 to 1998 (reliable data are available only for these years). Two monitoring stations exceeded the national limit more than three times during the period 1996±1998 (Alabama Department of Environmental Management, Jefferson County Field Of®ce, 1999).
Data source The selection of industries for the analysis was based on the importance of the industry in the county's economy, and the availability of complete plant level TRI data. Forty-seven plants of the selected industries, about 67% of the plants in Jefferson County, were included in the development of pollution emission coef®cients. Industry average was used to develop the I±O economic and the
Economic-environmental modeling of point source pollution
pollution coef®cient matrix. The I±O economic model was compiled from the Minnesota IMPLAN Group (MIG) data for Jefferson County, and the State of Alabama for the year 1996. IMPLAN is an input±output database available in microcomputer software developed by USDA Forest Service and currently licensed under agreement with the Minnesota IMPLAN Group (MIG). The data for the emission model were collected from the 1996 USEPA Toxic Release Inventory System (TRIS), an online database, facility report for manufacturing industries in the State of Alabama and Jefferson County (USEPA, 1996a). Industries, classi®ed under standard industrial classi®cation (SIC) 20± 39, employing an equivalent of 10 full-time workers, and either manufacturing or processing more than 11 340 kilograms of the chemicals indexed under the toxic chemical categories by the USEPA are required to report TRI data at the end of each year. The reported chemical emissions were used to estimate air emission of toxic chemicals and metal compounds. The emission coef®cients were computed from net discharge of chemical pollutant emissions associated with individual plants as reported in the TRI database for 1996. Net discharge is total emission minus source reduction, recycling, treatment, and disposal. Industries report source reduction, on-site treatment, off-site transfer, but data were not suf®ciently detailed for each industry. Four-digit (SIC) level TRI data were collected for all the manufacturing industries in Jefferson County and in the State of Alabama.
Empirical results Table 1 shows the distribution of the reporting plants by IMPLAN code. Except for the 10 plants categorized as industrial organic chemicals (190) and paints and allied products (200), the predominant number of plants are categorized as primary and fabricated metals. This is consistent with the industrial distribution observed in the state's industrial directories for 1990 and 1996, and thus the sample of industries is representative of the industrial activities in the county. The last column shows the share of the industries as a percentage of industries in Jefferson County. These industries produce 41% of the total manufacturing output in the county. Blast furnaces and steel mills (254), iron and steel foundries (259), and aircraft (389) account for 31% of manufacturing output with the remaining industries accounting for 10%.
89
Table 1. Number of plants and output of selected industries as a share of total manufacturing output in Jefferson County, 1996 Industry Industrial organic chemicals Paints and allied products Blast furnaces and steel mills Iron and steel foundries Secondary nonferrous primary metal products Metal cans Miscellaneous fabricated structure Miscellaneous metal works Plating and metal coating Aircraft Total a b
IMPLAN Number Output codea of plantsb sharea 190 200 254 259 263
5 5 6 6 5
001 002 019 007 0.01
273 282
4 5
002 001
288 296 389
4 4 3 47
0.02 001 005 041
IMPLAN code and output data. USEPA, TRI data base, 1996a.
Column 4 of Table 1 shows the share of the selected industry as a percentage of the industry's output at the state level. Except for secondary nonferrous primary metals (263), and miscellaneous fabricated structure (282), the rest of the industries produce a signi®cant portion of the total state output. Most of the plants are located in Birmingham and surrounding cities. These industries account for about 58 and 22 percent of the total state output and employment, respectively (Alabama Development Of®ce, 1990 and 1996). Table 2 shows total chemical and metal compound air emission by the selected industries, and all industries in Jefferson County. The chemical compounds are composed of volatile organic compounds (VOCs)Ðalcohol, benzene, chlorine, ethers/ ketones, and toluene/xylene, other VOCs. The category of other VOCs consists of VOCs that account for less than 5% of the total emission individually. Metal compounds are composed of barium, lead, manganese, zinc, and other metals. VOCs accounted for about 85% of the total air emission, while metal compounds accounted for 15%. The percentage distribution shows that other VOCs accounted for 36%, toluene/xylene accounted for 30%, zinc accounted for 13%, ethers/ketones accounted for 10%, and alcohol and benzene accounted for about 4%. These compounds are identi®ed as hazardous and high risk by (USEPA, 1996b; Lui, 2000; Canter, 2000). At the lower level of the atmosphere, VOCs are the major components of smog and are serious environmental and health hazards. The USEPA listed carbon disul®de, methanol, copper compounds, zinc compounds, ammonia, xylene,
90
E. Kebede et al.
Table 2. Chemical and metal compound air emission by selected industries as a share of emission by all industries in Jefferson County and of the selected industries (kgs) Chemical compounds Alcohol Benzene Chlorine Ethers/ketones Toluene/xylen Other VOCs1 Barium compounds Lead compounds Manganese compounds Zinc compounds Other metal compounds2 Total
Emission for all industries in the County
Emission for selected industries
Emission as share of County
Share of emission for selected industries
72 660 153 204 1 372 251 317 378 726 294 203 9 978 11 756 15 606 125 268 7611 1 317 167
14 320 30 545 1 254 77 763 236 277 285 414 9 862 8 611 10 810 104 432 3788 783 078
020 020 091 032 062 097 099 073 070 083 050 060
002 004 000 010 030 036 001 001 001 013 000 100
Source: TRI database (USEPA, 1996b). 1 Other VOCs among others include: ammonia, anthracene, creoste, diethanolamine, ethylene, formaldehyde, naphthalene, and tricholoethylene. 2 Other metal compounds include: antimony, cadmium, chromium, copper, cobalt, and nickel.
Table 3. Total chemical and metal compounds air emission of selected industries by type of chemical, Jefferson County 1996 (kgs) Ind.
Alcohol
Benzene
Ethers/ ketones
Toluene/ xylene
Other VOCs1
Barium2
Lead2
Mang2
Zinc2
Other Metals3
Total
190 200 254 259 263 273 282 288 296 389 Total
4 854 3 051 1 402 0 0 0 0 5 014 0 0 14 320
7 324 4 035 13 746 0 0 406 5 035 0 0 0 30 546
2 068 22 408 0 12 053 0 936 0 17 401 0 22 897 77 764
10 891 28 693 30 662 96 444 0 953 2 722 0 48 717 17 201 236 281
53 382 633 38 452 11 052 125 82 611 0 0 0 98 835 285 091
0 0 65 9798 0 0 0 0 0 0 9863
0 0 8153 0 454 0 0 0 5 0 8612
0 0 9 561 1 134 0 0 2 113 0 0 10 810
0 80 98 064 113 3334 0 0 0 2 842 0 104 432
0 514 1 879 366 916 0 0 0 113 0 3 788
78 518 59 415 201 983 130 960 4 829 84 907 7 759 22 528 51 677 138 934 781 509
Source: TRI database (USEPA, 1996b). 1 Other VOCs include: ammonia, anthracene, creoste, diethanolamine, ethylene, formaldehyde, naphthalene, and tricholoethylene. 2 Metal compounds. 3 Other metal compounds include: antimony, cadmium, chromium, copper, cobalt and nickel.
hydrochloric acid (acid aerosol form), toluene, manganese compounds, and methyl ethyl ketone, as the ten highest hazardous releases. Table 2 also shows emissions for the selected industries as a share of total emissions by all industries in Jefferson County. The emission of these industries is approximately 60% of the total pollution emission by industrial plants in the county. They account for a signi®cant amount of other VOCs and chlorine compounds generated in the county, 97% and 91%, respectively. A large proportion of the metal compounds is also generated by these industries: 99% for barium compounds, 83% for zinc compounds, 73% for lead compounds, 70% for manganese compounds.
Table 3 provides the chemical emission for all the plants in selected industries. Blast furnaces and steel mill industries (254), aircraft industries (389), and the iron and steel foundries (259) generate the highest volumes of pollution and accounted for about 60% of the total pollution generated by selected industries. Industrial organic chemicals (190), paints and allied products (200), metal cans (273), and plating and metal coating (296) produced the next highest volumes, while secondary nonferrous primary metals (263) and miscellaneous fabricated structure (282) produced the least. Emission by type of pollutants varied substantially across industries. Almost all industries generate some level of toluene/xylene with the highest
Economic-environmental modeling of point source pollution
91
Table 4. Estimated direct pollution emission coef®cients for selected industries for the State of Alabama and Jefferson County, 1996 Industry
190 200 254 259 263 273 282 288 296 389
Alabama
Jefferson County
Chemical1 emission (kgs)
Output2 $10 000
Direct emission coef®cient (kgs/$10 000)
Chemical1 emission (kgs)
Output2 $10 000
Direct emission coef®cient (kgs/$10 000)
1 000 287 177 654 312 669 287 151 16 942 123 994 56 112 34 850 271 210 235 860
96 8884 26 6267 219 2146 107 6309 26 5698 12 1089 49 1696 16 7388 74681 58 2081
1032 667 143 267 064 1024 114 208 3632 405
78 518 59 415 201 983 130 960 4829 84 907 7759 22 528 51 677 138 934
77508 87452 140 0891 51 3632 40624 82993 67441 13 9199 60681 34 2904
1013 679 144 255 119 1023 115 162 852 405
Figures are rounded to two decimal places. Source: 1 TRI database (USEPA, 1996b). 2 Industrial output (1996 IMPLAN data).
emission from iron and steel foundries (259) and plating and metal coating (296): 96 444 and 48 716 kgs respectively. The next highest were blast furnaces and steel mills (254), 30 662 kgs; paints and allied products (200), 28 693 kgs; aircraft industries (389), 17 201 kgs; and industrial organic chemicals (190), 10 891 kgs. Blast furnaces and steel mills (254) produce the highest volume of metals pollutants. They produce 98 064 kgs of zinc that accounts for about 76% of the total emission by the selected industries.
Emission coef®cients The industrial output at producer prices, and the industry TRI data for 1996 were used to compute the emission coef®cients. Emission coef®cients are the link between economic activity and the environment, and indicate the level of pollution emission per $10 000 of industry output. Emission coef®cients for the State of Alabama and Jefferson County are presented in Table 4 for the selected industries. The emission coef®cient of the State of Alabama was computed to compare and test the validity of the coef®cients at the county level. The emission coef®cients for the State of Alabama and Jefferson County are comparable, except for secondary and nonferrous primary metals, N.E.C (263), miscellaneous metal works (288), and plating and metal coating (296) industries. For these three industries, there were data differences because of aggregation and conversion from SIC code to
IMPLAN code. The IMPLAN code included several SIC numbers in the output data that created some discrepancy between the state and the county coef®cients. The industries with the highest emission coef®cients are metal cans (273), organic chemicals (190), metal plating and coating (296), and paints and allied products (200) at 1023, 1013, 852, and 679 kgs of pollutants per $10 000, respectively. The industries with the lowest pollution per unit of output are respectively, secondary nonferrous primary metals (263), miscellaneous metal works (288), iron and steel foundries (259), aircraft (389), 147, 161, 255, and 405 kgs per $10 000. As opposed to total emission (Table 3), blast furnaces and steel mills (254) have the lowest emission coef®cients, 144 kgs per $10 000 output. Some of the industries with the highest total pollution have the lowest coef®cients, indicating that they had higher total output than industries with the lower volume of pollution but higher emission coef®cients. Direct emission coef®cients by type of chemical were computed to examine the speci®c emission and the hazard level of each industry per unit of output. Table 5 shows that industrial organic chemicals (190), metal cans (273), miscellaneous fabricated structures (282), and aircraft (389) produce only VOCs. These chemicals are easily assimilated in the atmosphere and are the main sources of smog, odor, and fog. Paints and allied products (200), blast furnaces and steel mills (254), iron and steel foundries (259), secondary nonferrous primary metals (263), miscellaneous metal works (288), and plating and metal coating (296) generate both VOCs and
92
E. Kebede et al.
Table 5. Industry
Estimated direct pollution emission coef®cients for selected industries by type of chemical (kgs/$10 000) Alcohol/ benzene
Ethers/ ketones
Toluene/ xylene
Other VOCs1
Barium2
Lead2
Manganese2
Zinc2
Other metals3
Total
157 081 011 0 0 005 075 036 0 0
027 256 0 024 0 011 0 125 0 067
141 328 022 186 0 011 040 0 803 050
689 007 028 022 003 995 0 0 0 288
0 0 0 020 0 0 0 0 0 0
0 0 006 0 011 0 0 0 0 0
0 0 007 003 0 0 0 001 0 0
0 001 070 0 082 0 0 0 047 0
0 006 001 001 023 0 0 0 002 0
1013 679 144 255 119 1023 115 162 851 405
190 200 254 259 263 273 282 288 296 389
Figures are rounded to two decimal places. 1 Other VOCs include: ammonia, anthracene, creoste, diethanolamine, ethylene, formaldehyde, naphthalene, and tricholoethylene. 2 Metal compounds. 3 Other metal compounds include: antimony, cadmium, chromium, copper, and nickel.
Table 6. Direct and indirect pollution emission coef®cients of selected industries, Jefferson County, 1996 Industry
Direct emission coef®cienta
Direct and indirect emission
Emission multiplier
10130 6794 1442 2550 1189 10231 1150 1618 8516 4052
10601 7308 1916 3013 1513 10690 1425 1948 8994 4187
1046 1075 1328 1181 1273 1045 1239 1204 1056 1033
190 200 254 259 263 273 282 288 296 389 a
See Table 4.
metal compounds including barium, lead, manganese, and zinc compounds. Metal compounds take the form of particulates, and tend to settle in the lower levels of the atmosphere, and are major sources of acid rain that affect vegetation and soil, surface and ground water contamination, and deposits on structures. The inter-industry linkages in the I±O model re¯ect the direct and indirect effect of each industry on the economy. The estimated coef®cients in Table 4 indicate only the direct pollution emission generated by each industry in 1996. Table 6 provides the direct and indirect pollution emission coef®cients for each industry in 1996. The I±O model is used to compute the direct and indirect emission coef®cient. Comparison of the direct pollution emission and the direct and indirect emission coef®cients (cumulative emission) for 1996 show that aircraft industry (389) had the lowest percentage increase of 33%, blast furnaces and steel mills (254) have the highest (328%). The percentage
increase re¯ects the indirect linkages; the multiplier effect, of the industry on the total economy. While blast furnaces and steel mills (254) have the highest indirect increase, the next highest increase is secondary nonferrous primary metal industry (263) at 273%. The indirect volume of pollution increased by 239% for miscellaneous fabricated structures (282) and 204% for miscellaneous metal works (288). Iron and steel foundries (295) increased indirect pollution emission by 181%. The remaining industries increased indirect pollution by 75% or less.
Summary and conclusion Jefferson County, especially Birmingham and the surrounding cities, is the center of manufacturing industries, and also has the highest level of smog warnings in the State of Alabama. The majority of the industries are classi®ed as primary and fabricated metal products. This study selected 10 industries that consisted of 47 plants, about 67% of the plants in the county. These industries include: industrial organic chemicals (190); paints and allied products (200); blast furnaces and steel mills (254); iron and steel foundries (259); secondary nonferrous primary metals (263), metal cans (273); miscellaneous fabricated structure (282) and miscellaneous metal works (282); plating and metal coating (296); and the aircraft industry (389). The majority of the plants are categorized as primary and fabricated metals. These selected industries account for 41% of the county's total output. They accounted for 97% of the volatile organic compounds (VOCs), that are the major source of smog and odor,
Economic-environmental modeling of point source pollution
acid rain, contamination of surface and ground water, and deposits on structures. The distribution shows that industries (254), (259), and (389) account for 31% of output, and 60% of the total volume of pollution by the selected ten industries. Emission coef®cients indicate the relationship between the industry's contribution to output of the economy and the associated impact on the environment. Direct emission coef®cients, emission per $10 000 output, show that the industries with highest total volume of chemical emission (254), (259), and (389) had a lowest emission coef®cient, i.e. higher at output relative to volume of pollution emission. Industries (190), (200), (263), and (296) had highest emission coef®cients, i.e. lowest output relative to volume of pollution emission. All of the industries in the sample generate VOCs, however, the direct emission coef®cients by type of chemical show that (254) and (259) with the lowest emission coef®cients generate the highest VOCs, especially toluene/xylene, and the highest level of metal compounds, especially zinc. These chemicals are referred as hazardous by USEPA. Furthermore, cumulative emission from inter-industry linkages using 1996 data show that there is an increase of 33% to 328% in the total emission by each industry group. The aircraft industry (389) had the lowest increase 33%. These percentage increases illustrate that an isolated treatment of the plant will underestimate the level of total pollutants generated by an industry. An increase in the ®nal demand (i.e. exports) for the products of each industry will have a direct and indirect effect on the commodity output, and emission of VOCs and metal compounds (hazardous chemicals). In conclusion it is safe to say that these industries individually and collectively impact the ozone ambient concentration, increase the odor, and smog level in Jefferson County in and around Birmingham. Based on the economic-environmental impact industries with high levels of output and low chemical emission coef®cients may be favored as part of an overall development strategy; however, they may also generate hazardous waste. The impact of these industries might be low at low levels of production, but with expansion and increase in volume of output, there will be an increase in emissions. The results of this study are limited by the assumptions made and the accuracy of the data used. Estimates from the study must be interpreted cautiously. The output data used are industrial averages. TRI data include only a portion of emission because discharge data from small plants, and for some industries, are unavailable. Despite the
93
shortcomings of the study the TRI data set is the only source of industrial chemical discharge. Thus while the analysis is only partial, environmental implications are general. However, the economicenvironmental analysis developed in this study provides a useful analytical tool for direct and cumulative emission estimation, and generates insights on the complexity in the choice of industries, based on economic and environmental impacts.
Acknowledgements The authors would like to thank an anonymous reviewer and Dr Curtis M. Jolly for constructive suggestion and comments. The Study is funded by a USDA/CSREES capacity building grant.
References Alabama Development Of®ce. Alabama Industrial Directory 1990±1996. Alabama Development Of®ce, Alabama Center for Commerce, Montgomery, Alabama. Alabama Department of Environmental Management (1999). Exceedances the Birmingham Ozone Non Attainment Area 1991±1998. Unpublished material, Jefferson County Of®ce. Bergh, Jeroen C. J. M. van den (1996). Ecological Economics and Sustainable Development, Theory, Methods, and Application. Cheltenham, UK; Brook®eld, USA: Edward Elgar. Bradstreet, Jeffery W. (1995). Hazardous Air Pollutants, Assessment, Liabilities, and Regulatory Compliance. Park Ridge, New Jersey, USA: Noyes Publication. Canter, L. W. (2000). VOCs and HAPs emission from chemical plants. In Air Emission (D. H. F. Lui and L. W. Canter, eds), pp. 20±22. Boca Raton, Florida: Lewis Publishers. Carel, Rafael S. (1998). Health Aspects of Air Pollution. In Handbook of Air Pollution from Internal Combustion Engines (Fran Sher, ed.), pp. 42±64. Boston, London: Academic Press. Cumberland, John H. and Stram, Bruce N. (1972). Effects of Economic Development Upon Water Resources. Technical Report No.18, College Park, Maryland: University of Maryland, Bureau of Business and Economic Research. Dixon, John A. and Hufschmidt, Maynard M. (eds) (1986). Economic Evaluation Techniques for the Environment, A Case Study Workbook. Baltimore, USA: Johns Hopkins Press Ltd. Forsund, Finn R. (1985). Input±output models, national economic models, and the environment. In Handbook of Natural Resource and Energy Economics, Vol. I. (A. V. Kneese and J. L. Sweeney, eds). pp. 325±341. Amsterdam: Elsevier Science Publishers, B.V. Isard, Walter (1972). Ecological-Economic Analysis for Regional Development. New York: The Free Press.
94
E. Kebede et al.
James, David (1994). The Application of Economic Techniques in Environmental Impact Assessment. The Netherlands: Kluwer Academic Publishers. James, David (1985). Environmental economics, Industrial process models, and regional residual management. In Handbook of Natural Resource and Energy Economics, Vol. I A. V. Kneese and J. L. Sweeney (eds), pp. 271±324. Amsterdam: Elsevier Science Publishers, B.V. James, D., Jansen, H. and Opschoor, J. B. (1978). Economic Approaches to Environmental Problems. Amsterdam: Elsevier Science Publishers, B.V. Krupnick, A. J., Harrington, W. and Ostro, B. (1990). Ambient ozone and acute health effects: evidence from daily data. Journal of Environmental Economics and Management 18, 1±18. Lave, Lester B., Cobas-Flores, Elisa, Hendrickson, Christ T. and McMichael, F. C. (1995). Using input±output analysis to estimate economy-wide discharges. Environmental Science and Technology 29, 420±426. Leontief, Wassily (1986). Input±Output Economics 2nd Edn. New York: Oxford University Press. Lipnowski, I. F. (1976). An input±output analysis of environmental preservation. Journal of Environmental Economics and Management 3, 205±214. Liu, D. H. F. (2000). HAPs and synthetic organic chemical manufacturing industries. In Air Pollution (D. H. F. Lui and L. W. Canter, eds), pp. 24±29. Boca Raton, Florida: Lewis Publishers. Miller, Ronald E. and Blair, Peter, D. (1985). Input± output analysis foundations and extensions, pp. 236± 265. Englewood Cliffs, New Jersey: Prentice-Hall, Inc. Minnesota Implan Group (MIG, Inc) (1999). 1996 IMPLAN Data, MIG Inc, 1725 Tower Dr W Suite 140, Stillwater, MN 55082.
Ngandu, Mudiayi and Kebede, Ellene (1996). Foreign Investment, Jobs, Equity, Rural Development, and the Emerging Automobile Industry in the Southeast. Research Working Paper #96-1, Tuskegee, Alabama: George Washington Carver Agricultural Experiment Station and The School of Agriculture and Home Economics, Tuskegee University. Perring, Charles (1987). Economy and Environment, pp. 15±80. Cambridge, NY: Cambridge University Press. Powe, N. A. and Willis, K. G. (1998). Industrial location and residential disamenity: a case study of the chemical industry in Castleford, England. Journal of Environmental Management 53, 17±29, doi:10.1006/ jema.1998.0193. Rainey, Daniel V. and McNamara, Kevin T. (1999). Taxes and location decision of manufacturing establishment. Review of Agricultural Economics 21, 86±98. Schmenner, Roger W. (1982). Making Business Location Decisions. Englewood Cliffs, New Jersey: Prentice-Hall. Sher, Fran (1998). Environment aspect of air pollution. In Handbook of Air Pollution from Internal Combustion Engines Fran Sher (ed.). pp. 27±41. Boston, London: Academic Press. Victor, P. A. (1972). Pollution, Economy, and Environment. London: George Allen and Unwin. United States Environmental Protection Agency (USEPA). (1996a). 1996 Toxic Release Inventory (TRI) Database (http:// www.rtk.net). United States Environmental Protection Agency (USEPA). (1996b). State Toxic Release Inventory Report (http://www.epa.gov/tri/tri96/data/index.htm). United States Environmental Protection Agency (USEPA). (1993). The 1990 Clean Air Act, USEPA, Of®ce of Air Quality Planning and Standard, EPA-4±K-93-001.