Environ. Sci. Technol. 2006, 40, 6844-6851
An Improved Calculation of the Exergy of Natural Resources for Exergetic Life Cycle Assessment (ELCA) B R A M D E M E E S T E R , † J O D E W U L F , * ,† ARNOLD JANSSENS,‡ AND HERMAN VAN LANGENHOVE† Research Group EnVOC, Faculty of Bioscience Engineering, Ghent University, Coupure Links 653, B-9000 Ghent, Belgium, and Buildings and Climatic Control, Department of Architecture and Urban Planning, Faculty of Engineering, Ghent University, J. Plateaustraat 22, B-9000 Ghent, Belgium
The focus in environmental research is shifting from emission abatement to critical process analysis, including assessment of resource consumption. The exergy theory offers a thermodynamic methodology to account for the consumption of natural resources. However, exergy data on mineral resources available in the literature are inadequate to apply to exergetic life cycle analysis, due to incompleteness, inconsistencies, and a dated thermochemical basis. An uncertainty assessment of the data has to be performed as well. In this work, three recent thermochemical databases were applied to evaluate the chemical exergy of 85 elements and 73 minerals, 21 of which had not yet been quantified in the literature. The process required the choice of a new reference species for aluminum. Muscovite was selected, giving rise to a chemical exergy of 809.4 kJ/mol for aluminum. The theory proved to be robust for the exergy of chemical elements, as exergy values differing by 1.2% on average from most recent literature were found. On the contrary, the exergy values for minerals differed by factors up to 14 from literature values, due to the application of recent thermochemical values and consistently selected reference species. The consistent dataset of this work will enable straightforward resource intake evaluation through an exergetic life cycle assessment.
Introduction With respect to the environmentally friendly design of technology, emphasis is gradually shifting from emission control to critical analysis of resource consumption. The endof-pipe approach and abatement techniques are increasingly complemented with upstream process adjustments and input choices. This is illustrated by the increased use of renewables: next to reduced greenhouse gas emissions and local pollution, decreased resource depletion and an improved security of supply are major advantages (1). It is in this context that a suitable methodology to account for resource consumption in life cycle assessment (LCA)shistorically focused on emissionssis still a subject of debate (2). * Corresponding author phone: +32 9 264 59 49; fax: +32 9 264 62 43; e-mail:
[email protected]. † Research Group EnVOC, Faculty of Bioscience Engineering. ‡ Buildings and Climatic Control, Department of Architecture and Urban Planning, Faculty of Engineering. 6844
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 21, 2006
Industrial processes aresgiven that emissions are under controlsgoverned by two boundary conditions: economics and thermodynamics (3). The latter is able to assess both process efficiency and resource input (4). Consequently, it is suggested by several authors to apply thermodynamicss more specifically exergy accountingsto resource consumption in life cycle approaches (5-8). Exergy accounting found its origin in production analysis, in terms of efficiency (9, 10) and resource accounting (11), yielding applications on process level (12, 13), sector level (14, 15), and national level (16-18). The concept raised interest for sustainability assessment purposes (19-21); it can be applied to industrial ecology principles (22, 23) and to life cycle assessmentssleading to exergetic life cycle assessment (ELCA) (24-26). Exergy combines the first and second law of thermodynamics to evaluate the quality of energy. The concept can be expanded to materials, accounting for the intrinsic net energy of a substance. The theoretical framework of this method was developed in the 1960s-1980s (27-30). The work of Szargut et al. (11) offers an elaborate theoretical background on the exergy evaluation of resources; current developments still stem from that basis. The theory defines exergy as the minimal work necessary to produce a material with its specified state in a reversible way from common materials in the environment. It evaluates to what extent a resource stands out from its environment from a physical chemical point of view. Local environmental conditions can be defined for processes of regional impact. For industrial processes envisaged in LCA, global conditions of 1 standard atmosphere (101325 Pa) and 298.15 K, together with average geophysical chemical characteristics, are assumed. Exergy can be divided into four major components: potential, kinetic, physical, and chemical exergy. Potential and kinetic energy can ideally be completely converted into work; thus the potential, respectively kinetic exergy equals, e.g., potential energy of water in a hydropower storage basin or wind kinetic energy. Physical exergysobtained by reversible physical processesscan be calculated straightforwardly (9); e.g. the exergy evaluation of geothermal energy sources and solar radiation. The chemical exergy reflects the resource’s deviation in chemical composition from the reference environment. For the majority of natural resources, this chemical exergy is the most important contribution to its exergetic value. For the calculation of the chemical exergy of a resource (Figure 1), one considers a reference compound in the natural environment for each chemical element in the resource material, e.g., O2 for O, Cl-(aq) for Cl, SiO2 for Si. These ground states are the most probable products of the interaction of the elements with other common compounds in the natural environment and show typically high chemical stability. A reference species can be selected from the atmosphere (gaseous compounds), seawater (dissolved ionic compounds), and the earth’s upper crust (solid compounds). The pure reference compound’s exergy valuesas it is valueless in its environmental concentrationsis governed by geochemical data: its relative occurrence in the natural environment. From the reference species’ exergies, the chemical exergy of any resource substance can be calculated through thermochemistry, requiring free energies of reaction to be known. If the calculation proves the exergy values determined from the initially selected reference compound to be negative, this ground state is to be replaced for the element under consideration. Calculations then have to be repeated based on a newly selected ground state. 10.1021/es060167d CCC: $33.50
2006 American Chemical Society Published on Web 09/29/2006
FIGURE 1. Calculation scheme for the chemical exergy of resources, resulting in a life cycle approach for processes. Starting from the reference environment, the literature provides sufficient data to calculate the exergy content of organic substances, e.g., fossil feed stocks (9). Although numerous mineral ores are relevant to resource accounting of industrial processes in a life cycle approach, a sound literature basis for the calculation of the exergy of inorganic compounds is missing. A list of 73 minerals which are important in life cycle context can be found in the Supporting Information (Table S1), comprising their industrial applications, mineral formulas, and literature exergy data availability. The list includes metal ore constituents, minerals used as such in industry (e.g., fertilizers), and important petrological minerals. Exergy data for these minerals are inadequate to perform well-established exergetic life cycle assessment for industrial products. In the first place, the exergy values of 21 minerals are not available in the literature. Given the appropriate thermochemical data (the Gibbs free energy of formation), the exergy of a mineral can be calculated from the exergy of the constituent elements. This was illustrated by Finnveden et al. (7), but for 10 minerals only. The listed exergy values of Finnveden, combined with the tables of Szargut et al. (11), as they are the most important sources for mineral exergy data, display important lack of data for minerals relevant to industrial processes. Second, since the used reference environment originates from the late 1980s (11), several refinements of the selection procedure of reference compounds have been proposed. Valero et al. (31) give a critical overview of different reference environments and suggest an update of the selection procedure of Szargut (32). However, for 10 of the 49 elementssAu, Ba, Ca, Co, Cr, F, Mg, P, Sr, Usoriginally comprised in Szargut et al. (11), recommended reference compounds have been altered in later updates (32, 33). The effect of changing the reference for 10 elements on the exergy of minerals has not yet been assessed. Third, the available exergy values are calculation results based on outdated thermochemical data sources; some go back to the 1950s. Very recently, Szargut et al. (33) proposed an updated set of elements exergies. It should, however, be noted that only geochemical data were updated, not thermochemical data (Gibbs free energies of formation), as they are identical to those implemented previously (32), apart
from sillimanite. New extensive thermochemistry databases have been made available meanwhile. The determination of mineral exergies should benefit from the progress in thermochemistry. The fact that no uncertainty information is included in the evaluation of the exergy of minerals up to present is a last issue in the literature’s shortcomings with respect to exergy data. This uncertainty assessment is a key tool in concluding about the significance and sensitivity of an assessment’s outcome. Undoubtedly, uncertainties are connected to the exergy values tabulated. Model uncertainties (deviations from ideal behavior), approximations (the earth’s mean composition), and errors in data (experimental error margin of thermochemical data) can affect the accuracy and precision of exergy values. The objective of this paper is to present a scientifically sound and accurate calculation of the exergy of inorganic natural resources, enabling the implementation in life cycle approaches. Calculations will be based on the most recent geochemical and thermochemical data and start from the latest set of recommended reference species. After recomputing the standard chemical exergy of 85 elements in order to enhance previous updates, the exergy of 73 mineralssof which 21 were previously not available in the literatureswill be established. If necessary, calculation will be iterated from newly selected reference species. A first assessment of uncertainties of exergy values originating from uncertainty of thermochemical data will be performed.
Materials and Methods Thermochemical Calculation Methods. The chemical exergy of any species can be calculated from the exergy values of the reference compounds, considering its reference reaction. This is a reaction that involves, besides the target compound, only reference species. Given the standard Gibbs free energy of the reference reaction ∆G0r (kJ/mol), the chemical exergy 0 of a compound i, bch,i (kJ/mol), is computed by (11) 0 bch,i ) ∆G0r +
∑ν b
0 k ch,k
(1)
k
0 where νk, bch,k are the number of moles and the standard chemical exergy (kJ/mol) of the kth reference species, respectively. 0 denotes that reference temperature T0 (298.15 K) and standard pressure p0 (1 standard atmosphere or 101325 Pa) of the reference system are assumed. E.g., calculating the exergy of the compound CO, one has to apply the exergy of reference species CO2 (for carbon) and O2 (for oxygen) and thermochemical data of the oxidation reaction of CO to CO2. This equation can be applied to any chemical reaction, provided that the exergies of the reaction compounds apart from the target compound are previously established. In order to calculate the reaction free energy ∆G0r , one extracts the enthalpy of formation H0f (kJ/mol) and absolute entropy S0 (kJ/mol K) for both reagents and reaction products from thermochemical databases. ∆G0r is then obtained through
∆G0r ) ∆H0f - T0∆S0 ) (
∑H T (∑ S
∑H - ∑S
0 f,products 0 0 products
0 f,reagents) 0 reagents) (2)
The exergy of reference species differs from zero when its concentration is different from its environmental concentration. For the reference reaction, the chemical exergy of the 0 pure reference compound bch,k , e.g. pure O2, is computed based on its abundance only, no thermochemical data is required. For gaseous and solid reference species, the 0 chemical exergy bch,k of a reference species k issassuming ideal mixing behaviorsexpressed as (11) VOL. 40, NO. 21, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6845
0 bch,k ) - RT0lnxk
(3)
with R the molar gas constant (8.3145 J mol-1 K-1) (34) and xk the molar fraction of the reference species k in the air, respectively, in the upper crust. For gaseous species, the molar fraction can be replaced by xk ) pk0/p0, in which pk0 is the species’ conventional mean partial pressure in the atmosphere (Pa). In the case of reference species dissolved in seawater, the 0 standard chemical exergy bch,el of the element is calculated assuming the formation of the reference ion in a standard electrochemical cell with hydrogen electrode (11). This scheme yields
{
0 bch,el ) j - ∆G0f +
1 2
∑ν b
0 zbch,H 2
0 l ch,l
l
}
- RT0[2.303z(pH) + ln(m0γ)]
(4)
where j is the number of reference ions formed from one molecule of the element under consideration, ∆G0f is the standard Gibbs energy of formation of one molecule of reference species from the element (kJ/mol), z is the number of positive electronic charges in the reference ion, νl is the number of moles of the lth additional element in the reference 0 ion, bch,l is the specific standard chemical exergy of the lth additional element (kJ/mol), pH is the assumed pH of seawater (8.1), m0 is the standard molarity of the reference species in seawater, and γ is the activity coefficient of the reference species in seawater on the molarity scale. Note that eq 4 is an application of the two preceding eqs 1 and 3. A more detailed description about the calculation of the standard chemical exergy can be found in refs 11 and 32. This work makes use of up to date quantitative (geo-) chemical data (x0i, m0 , γ) of Szargut et al. (33), which is based on the work of Valero et al. (31) and geochemical research of the past decade. Thermochemical data (∆G0r ) used are described in the next section. The initially selected reference compounds are taken from the most recent set of Szargut et al. (33), which is identical to the preceding set of Szargut (32). Thermochemical Databases. Two internally consistent thermodynamic datasets have been implemented: Chatterjee et al. (35) and Holland et al. (36, 37). Both databases present enthalpy of formation (H0f ) and absolute entropy (S0) values of a limited number of compounds: Chatterjee et al. included 148 species containing 16 chemical elements, Holland et al. included 189 species and 14 elements. These two datasets have been made internally consistent, incorporating phase equilibrium relation constraints, reversal reaction constraints, and limits set by experimental ranges. Both methods also yield confidence intervals on the obtained values, as well as correlations between the values, enabling a reliable calculation of uncertainties on mineral reactions. The datasets were established through different methodologies: Chatterjee et al. used a Bayesian technique to combine data and constraints. Holland et al. applied a leastsquares regression method on the enthalpies only and determined the entropies by estimation techniques, reasoning that this is more reliable. The corpus of thermochemical data is to a large extent the same, the reaction constraints used often differ. These differences in approach account for the differences in results. The uncertainties have been computed differently also. Holland et al. only cite uncertainties for the enthalpies, not for the entropiessas the latter are estimated. Chatterjee et al. quote uncertainties for both sets of characteristics, but assume zero uncertainty for the enthalpy of certain “reference” compounds (one per element 6846
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 21, 2006
included). This undoubtedly yields lower uncertainties for the other minerals. One may conclude that mineral reaction uncertainties on Gibbs free energy calculated from each of both datasets are rather minima than maxima. Next to the internally consistent databases, HSC Chemistry (38) is applied in calculation. This database combines a vast amount of sources from the literature to establish a large set of uncorrelated thermodynamic properties, including enthalpy and entropy values, of a wide variety of compounds (more than 17 000 species comprising all chemical elements). The data are continuously updated and extended; in this work version 5.2 has been used. This kind of database enables the pragmatic and fairly accurate application of the described theory in order to determine the exergy value of any thermodynamically characterized compound. Uncertainty information is not included; compound data have been subdivided in reliability classes on a scale from 1 (good quality) to 4 (poor). To enhance the quality of the data in calculation, HSC values have been replaced by the internally consistent data if available. Instead of starting from the reference reaction of a compound, the chemical exergy of the elements was established first involving internally consistent data, subsequently enthalpy of formation and entropy of the target compound were taken from HSC. E.g., first the exergy of calcium Ca was calculated from its reference species calcite (CaCO3) using Holland et al. (37). Afterward the enthalpy of formation H0f and entropy S0 values of gypsum (CaSO4‚2H2O) were collected from HSC for the determination of gypsum’s exergy. Some thermochemical properties could be found neither in the internally consistent databases, nor in HSC Chemistry. This was the case for some rare elements’ reference species. For these elements (Er, Ho, Lu, Nb, Sm, Tl, Tm, and Yb) the Gibbs energy values of Szargut et al. (33) have been adopted without updating. Also for 5 minerals from Table 1 (illite, Na-montmorillonite, vermiculite, ulexite, and pentlandite) no data were found in the abovementioned sources. They were individually collected from the literature. The free energy of formation of ulexite was found in Chen et al. (39) and that of pentlandite was found in Waldner et al. (40). For illite, montmorillonite, and vermiculite, the Gibbs energies of formation are taken from Vieillard (41). It should be noted that, although the reference environment has conventionally the pressure of one standard atmosphere, all data sources quote thermochemical properties at a pressure of 1 bar (105 Pa). As the influence of pressure on the chemical exergy of liquids and solids is small, exergy values obtained were considered to be valid at 1 standard atmosphere. For gases, a correction is applied in calculations to account for the difference in pressure. All results shown below however, are those at the conventional reference pressure of 1 standard atmosphere.
Results and Discussion Selection of Reference Species. The calculation of the chemical exergy of minerals based on the initially selected reference species yields negative exergy values for aluminum compounds, even though these reference species have been recommended (32, 33). Negative exergy means that work would have to be spent to bring a substance to what is regarded as the least valuable stateswith zero exergy. This is inconsistent and has to be rectified by selecting a new reference species. Minerals previously showing negative exergy should then be considered as the new reference. Negative exergy was only observed for aluminum compounds, for which the application of the reference sillimanite (Al2SiO5) gave unsatisfactory results. Figure 2 shows calculation results of the chemical exergy of aluminum containing minerals referred to sillimanite (and muscovite, see below), based on the abovementioned thermochemical data sources.
TABLE 1. Standard Chemical Exergy of Minerals (kJ/mol) and Uncertainty Information, Calculated from Refs 35-38, Compared to Literature Data from Refs 7 and 11, and Recommended Values for Applications (kJ/mol and MJ/kg)a calculated exergy of minerals (kJ/mol) mineral name
based on Roine, 2002
literature data
based on based on Chatterjee et al., 1998 Holland and Powell, 1998 20.0 -
{1} {1} {1}
9.0 129.8 401.8
{1} {1} {1}
(0.6
{2} {1} {1} {2} {1} {1} {1}
93.8 -
{1}
135.4
(0.5
{1} {1} {1}
-
31.2 20.6 24.6 2494.0 166.4 134.2 94.8 3084.5 655.1 1535.7 reference species 10.5 126.0 742.4 reference species 95.8 670.5 1677.3 25.8 689.2 667.8 932.2 1618.9 1905.3 706.8 reference species 131.7 reference species 251.8 174.0 747.8
fluorapatite hydroxyapatite sylvite kieserite
reference species 63.7 {2} 20.0 {1} 37.6 {2}
chrysotile colemanite tincal ulexite albite anortite orthoclase illite barite wollastonite montmorillonite spodumene nepheline vermiculite anhydrite gypsum
25.7 55.3 89.7 25.2 71.7 18.7 reference species 41.8 39.1 27.6 21.5 25.6
corundum halite diopside enstatite ferrosillite clinochlore muscovite dolomite kaolinite opal fayalite
42.4 14.5 40.3 19.2 118.2 0.1 reference species 21.4 16.8 316.5 237.3
ref
kJ/mol
MJ/kg
Metal ore minerals
{1}a {1} {1} {1} {1} {1} {1} {1} {1} {1}
boehmite diaspore gibbsite stibnite chromite black cobalt oxide gray cobalt oxide bornite chalcocite chalcopyrite hematite limonite magnetite galena braunite pyrolusite rhodochrosite cinnabar molybdenite pentlandite theophrastite vysotskite cooperite sperrylite rhenium(IV)sulfide rhodium(III)sulfide acanthite cassiterite ilmenite rutile uraninite (U3O8) uraninite (UO2) sphalerite
kJ/mol
recommended exergy of minerals
21.0 -
(0.8
(0.1 (0.4
8.9 128.9 -
(2.3
(1.9
(0.2 (0.3
195.3 NA 209.5 NA 129.1 38.2 52.8 NA 791.8 2054
(11) (7) (11) (11) (11) (11) (7)
9.9 121.6 743.7 NA
(7) (11) (11)
81.8 674.8 1723.1 NA 25.5 NA NA NA NA NA 709.5
(11) (11) (11)
(11)
131.4
(11)
-
218.3 162.9 747.6
(11) (11) (11)
-
63.3 19.6 NA
(7) (11)
61.3 NA NA NA 105.5 218.3 99.9 NA
(11)
135.1
(0.3
(11)
31.2 0.260 21.0 0.175 24.6 0.158 2494 7.34 166.4 0.744 134.2 0.557 94.8 1.27 3085 6.15 655.1 4.12 1536 8.37 17.7 0.111 8.88 0.100 128.9 0.557 742.4 3.10 401.8 0.664 35.2 0.405 93.8 0.816 670.5 2.88 1677 10.48 6619* 8.57 25.8 0.278 689.2 4.98 667.8 2.94 932.2 2.70 1619 6.47 1905 6.31 706.8 2.85 43.1 0.286 135.1 0.890 21.1 0.264 251.8 0.299 174.0 0.644 747.8 7.67
Fertilizers
{2} {2} {2} {1} {1} {1} {1}
-
Minerals in building products 26.4 (0.3 29.5 10.8 (0.4 11.3 64.9 (0.8 65.9 -
{1} {1}
41.0 39.7 26.4 -
{1} {1} {1} {1} {2} {2}
41.5 44.5 22.9 120.5 0.7
{1} {1} {2} {1}
20.1 15.5 239.3
{1} {1}
(0.1 (0.1 (1.7
42.5 27.2 -
Miscellaneous (0.7 42.9 (0.1 43.8 (0.1 23.0 (0.2 120.3 (4.2 13.5 (0.2 (0.8 (0.4
23.2 14.6 238.4
(0.5
(2.0 (1.9
(0.2 (2.1
(1.9 (0.3 (0.03 (0.2 (2.0 (0.4 (1.9 (0.4
23.6 NA NA NA NA 8.2 8.6
(11) (11) (11) (11)
(11) (11)
200.4 14.3 8.0 22.0 161.7 NA
(11) (11) (7) (11) (11)
15.1 197.8 NA 236.2
(11) (11) (11)
13.2 63.7 20.0 37.6
0.026 0.127 0.268 0.272
29.5 55.3 89.7 87.8* 11.3 65.9 18.7 40.8* 29.9 42.5 40.1* 39.7 27.2 43.6* 21.5 25.6
0.106 0.269 0.235 0.217 0.043 0.237 0.0670 0.106 0.128 0.366 0.109 0.213 0.191 0.110 0.158 0.149
42.9 14.5 43.8 23.0 120.3 13.5 12.3 23.2 14.6 316.5 238.4
0.420 0.248 0.202 0.229 0.912 0.0243 0.0308 0.126 0.057 4.05 1.17
VOL. 40, NO. 21, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6847
TABLE 1. (Continued) calculated exergy of minerals (kJ/mol) mineral name forsterite quartz fluorspar magnesite talc pyrite sillimanite thenardite calcite
based on Roine, 2002 46.0 reference species 34.4 25.0 reference species 1433.3 41.4 18.1 reference species
based on Chatterjee et al., 1998
based on Holland and Powell, 1998
Miscellaneous, continued (0.2 32.2
{1}
34.1
{1} {1}
8.5
(0.3
{1} {1} {1}
41.9 -
(0.4
literature data
recommended exergy of minerals
kJ/mol
ref
kJ/mol
MJ/kg
32.2 1.4 34.4 9.7 21.4 1433 43.7 18.1 18.6
0.229 0.0228 0.440 0.115 0.0565 11.95 0.270 0.127 0.186
(0.1
74.9
(11)
9.7
(0.2
11.4 37.9
(11) (11)
43.7 -
(1.9
1428.7 15.4 21.4
(11) (11) (11)
a Uncertainty information is expressed as reliability class (between {}) for results based on ref 38 and 95% confidence intervals for results based on refs 35-37. NA: Not available in literature. - : Not available in the respective database. * : Thermochemical data from other sources (see text).
Literature values of Szargut et al. (11), although sillimanite was used as ground state as well, deviate substantially from calculation results in this study; Szargut’s values are about 90 kJ/mol per aluminum atom in the chemical formula higher than the sillimanite based values in this study. This is due to a difference of free energy of formation for sillimanite. This has been obviated for the calculation of the exergy of the element aluminum (33, 42). The effect of the updated Gibbs energy of formation on mineral exergiessnegative calculation resultssrequires the selection of a new aluminum reference species. Kaolinite (Al2Si2O5(OH)4), muscovite (KAl3Si3O10(OH)2), and clinochlore (Mg5Al2Si3O10(OH)8)sshowing the lowest chemical exergy referred to sillimaniteswere examined as candidate reference compounds for aluminum. To determine the molar fraction xk of these minerals in the upper crust, needed in eq 3, Szargut (32) suggested the equation
xk )
niciM0 li
(5)
in which ni is the mean molecular fraction (mol/g) of the ith element in the earth’s upper crust, ci is the molar fraction of the ith element appearing in the reference k, M0 is the mean molecular mass of the upper crust (g/mol), and li is the number of atoms of the element i in one molecule of the reference species k. The most recent values for M0 and nAl (33) are 143.4 g/mol and 2.98 10-3 mol/g, respectively. On the basis of the abundance of rock types in the earth’s upper crust (43) and average mineral content (44) reported by Wedepohl, cAl values of kaolinite, muscovite, and clinochlore were estimated to be 0.01, 0.05, and 0.005 respectively. The application of each of the candidate reference species yields a chemical exergy of aluminum of about 810 kJ/mol: 810.1, 809.4, and 811.5 kJ/mol based on kaolinite, muscovite, and clinochlore, respectively. Although the relative abundances in the upper crust of the candidate reference species differ by one order, the chemical stabilitys∆G0r in eq 1sappears to be of overriding importance in the calculation of the exergy of aluminum. Respective figures vary within the uncertainty margins relative to thermochemical errors; these errors are discussed more in detail below. Therefore muscovite was selected as the new aluminum reference species, based on its higher abundance. It should indeed be the most probable product of interaction of aluminum with the environment, due to its relatively high abundance and chemical stability. The resulting value of the chemical exergy of aluminum is 809.4 kJ/mol (based on the thermochemical values of Holland and Powell, see selection principles below). This adaptation thus has a rather limited 6848
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 21, 2006
FIGURE 2. Aluminum compounds: chemical exergy calculated from reference species sillimanite (closed symbols) and muscovite (open symbols) and literature values (11). influence on the exergy value of pure aluminum (+1.5% relative to Szargut et al. (33)). The selection of muscovite as aluminum reference instead of sillimanite results in an exergy increase of about 15 kJ/mol per aluminum atom comprised in the chemical formula of a compound (Figure 2). The resulting exergies of the aluminum compounds referred to muscovite are included in Table 1 (see below). Chemical Exergy of Elements and Their Reference Species in the Environment. Detailed figures on the standard chemical exergy of 85 chemical elements and their reference species can be found in the Supporting Information (Table S2). Calculations are based on HSC Chemistry (38) data unless internally consistent data from Holland et al. (36, 37) were available, which was the case for only eight elements (Al, C, Ca, Fe, H, Mg, Si, and Ti). The majority of standard chemical exergy values do not differ by more than 3% compared to Szargut et al. (33). Even the exergy of aluminumsdespite the shift in reference speciessstays within this margin. Principally only rare
elements show a relative exergy difference that exceeds this threshold: Dy, Eu, Gd, La, Nd, Rh, and Tb. Within the group of more common elements, only the value for bismuth shows a considerable difference: 251.1 instead of 274.9 kJ/mol. These deviations are due to differences in enthalpy of formation H0f and entropy S0 values. However, it proves that the application of recent thermochemical data does not result in major changes of chemical exergy values of the elements: on average only 1.2%. Our findings at this point are in accordance with the conclusion of Rivero et al. (42). Chemical exergies calculated from solid reference species tend to have higher relative differencess1.7% on average compared to Szargut et al. (33) and even 2.6% on average regarding to Szargut (32)sthan those calculated from reference species dissolved in seawatersan average of 0.7% regarding refs 32 and 33. Gaseous reference species based values show no differences as used data were principally unchanged. There are three reasons for higher differences for solid reference based elements. The first reason is that geochemical data (M0,n0,i) have been updated since the 1980s, explaining also the difference between the early and recent values of Szargut. Second, recent thermochemical data show larger differences with respect to values available in the 1980s for solid compounds than for aqueous species, the latter being apparently more precisely thermochemically characterized at the time. Third, solid reference species can include elements which themselves have solid reference species, redoubling thus abovementioned differences. Indeed, the adjustment to the value of fluorapatite predominantly depends on the change in exergy of calcium. Szargut (32) already warned that the uncertainty of eq 3 for solid reference species is relatively high. His recommendation is to avoid including elements with solid reference species into the (solid) reference of other elements; for Si though this is assumed to be acceptable. This has been brought into practice by Szargut (32), shifting the reference of magnesium from dolomite to talc, thus avoiding the exergy of magnesium to be dependent on the exergy of calcium. However, Szargut’s recommendation implicitly prefers precision over accuracy. As the reference environment is a reflection of the natural environment, the most probable compounds (not resulting in negative exergy values) should be selected. Over time, precision also will improve as new thermochemical and geochemical data are made available. The evaluation shows that, due to the elevated exergy value of pure elements, absolute differences of 10 kJ/mol and more have only limited relative effect on the elements’ standard chemical exergies. The theory proves to be robust and the implementation of recent source data leads to refinements, rather than to big changes. Accurate geochemical and thermodynamic data should not only be applied to the exergy of pure elements, but can have a more pronounced effect on mineral exergies. In this respect, it is important to study the effect of recent source data on the exergy of a wide range of minerals, to come to an adjusted reference environment. Chemical Exergy of Mineral Resources. Table 1 shows calculation results for 73 minerals which are relevant as industrial resources. Comparison of the results obtained in this work, calculated from the three thermochemical databases, show the influence of different H0f and S0 values. First, it can be seen that the results from the two internally consistent databases are very similar. Differences are in the order of 1-3 kJ/mol, except for clinochlore. For the latter the difference is about 13 kJ/mol. One should notice that this is mainly due to a relative difference of 0.15% between the enthalpy values of clinochlore from the databases. This case illustrates the importance of accurate thermochemical values. Second, exergy values obtained from HSC Chemistry diverge from those obtained from the internally consistent databases,
a few by 10-15 kJ/mol. In most cases though, the results are in good accordance. If several exergy values are available for a particular mineral, a recommendation is to be made. HSC Chemistry (38), the most complete dataset, covers most of the minerals’ thermochemical properties. However, as the thermochemical data are collected from hundreds of sources, it is virtually impossible for such a large database to be internally consistent. The figures used indeed have been obtained through different experimental methods and include the inherent errors. To create an internally consistent database, different sources have to be evaluated critically, combining the values with phase relations and reaction reversal data. This is not feasible for databases of that magnitude. The selection of the most accurate data is therefore based on rather simple principles, preferring the more recent sources and excluding discrepancies through simple evaluations. The variety of data sources and experimental methods also makes a uniform uncertainty assessment impossible. Therefore, internally consistent data are used in this work as primary source whenever possible, supplying both refined thermochemical data and uncertainty values. Holland et al. (37) is the primary source, as it is more recent than Chatterjee et al. (35). The latter dataset is the secondary source, supplying values for minerals missing in the former. The properties for the remaining minerals have been taken from HSC Chemistry (38), as it supplies values for minerals composed of elements not included in the internally consistent datasets. The uncertainty information in Table 1 obviously demonstrates that the exergy values are often not within the 95% confidence range of the other set(s), although the only differences in calculation are the thermochemical sources. This confirms the assumption that the uncertainties produced by the datasets are rather small. Note that the two internally consistent databases yield rather different uncertainties. These findings can be explained by differences in experimental thermochemical data used and varying algorithms applied to make the databases consistent. The scatter in results from the different databases proves to give a better indication on uncertainties that may arise from implementing various data sources. Although the differences are limited in most cases to 5 kJ/mol, values up to 16.5 kJ/mol can be found, e.g., for magnesite, albite, forsterite, and clinochlore: 16.5, 14.4, 13.8, and 13.4 kJ/mol, respectively. In the case of clinochlore, this difference gives rise to a factor of ten between the exergy values based on different databases. It can be estimated that errors originating from uncertainty with respect to the composition of the earth’s crust are in the same order. Indeed, the use of updated geochemical data by Szargut et al. (33) yields changes that in most cases are limited to 10 kJ/mol with respect to Szargut’s values (32), although adjustments of 11.58 (MnO2) and 23.79 (SrCO3) kJ/mol are found as well. It is clear that uncertainty assessment may be improved in the future, provided that more detailed information on earth’s mineral composition, including uncertainties, will come available. These findings should avoid a false impression of precision in the use of exergy values. Furthermore, an assessment of all uncertainties is essential for a scientific evaluation of the outcome of exergy analyses. Table 1 includes the results of exergy calculations of 21 minerals previously not available in the literature, although they are very relevant to exergetic life cycle approaches. Among them, e.g., clay minerals used in building materials and processes and main ore minerals for nickel, platinum group metals, and boron. If literature values are availables from Szargut et al. (11) and Finnveden et al. (7), a comparison can be made. As the reference species of 11 elements have been modified since the publication of the literature values, shifts in the chemical exergy of minerals can be expected. This is mostly expressed as an increase in mineral exergy VOL. 40, NO. 21, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6849
(e.g., barite (Ba), gypsum (Ca), the cobalt oxides (Co), fluorspar (F), uraninite (U), etc.) as the new references were recommended to avoid negative exergy values. Noted exceptions are magnesium compoundsse.g., chrysotile (-32 kJ/mol)sand aluminum compoundsse.g., gibbsite (-185 kJ/mol), which show a decrease in exergy relative to literature. For the magnesium compounds this is due to the application of talc as reference compound instead of dolomite for reasons cited above. The differences between this study and the literature for the aluminum compounds are explained above. Note that the differences give rise to exergy values which are up to a factor of 14 (kaolinite) smaller than those cited in the literature. Other large differences are governed by the implementation of recent thermochemical data (e.g., chalcocite, chalcopyrite, molybdenite, etc.) as no substantial adjustments were made to the reference species of the constituent elements. These findings prove that chemical exergies of mineralssunlike the chemical exergy of elementss are substantially affected by shifts in reference species and differences in thermochemical data. The above emphasizes the need for the establishment of an update for the exergy values of minerals, as the cited literature is being used as a reference up to present.
Acknowledgments This work was supported by a scholarship from the Special Research Fund of Ghent University (BOF 011D17104). This financial support is highly appreciated. We thank Marc Verhaege (Laboratory of Non-Ferrous Metallurgy and Electrometallurgy, Ghent University) for providing the thermochemical data from HSC Chemistry 5.2.
Supporting Information Available Detailed list of the minerals used in calculation, comprising their industrial applications, mineral formulas, and literature exergy data availability (Table S1). Table S2 lists the results of standard chemical exergy calculations of 85 elements and their reference species. This material is available free of charge via the Internet at http://pubs.acs.org.
Literature Cited (1) Uyterlinde, M. CO2 emission reduction from high renewable penetration; Presented at Renewable Energy for Europe Research in Action; European Commission: Brussels, Belgium, November 2005. (2) Finnveden, G. The resource debate needs to continue. Int. J. Life Cycle Assess. 2005, 10, 372. (3) Whitesides, G. M. What is the scope of green chemistry? Picking the problems; Presented at Joint meeting of the 2nd International Conference on Green and Sustainable Chemistry and the 9th Annual Green Chemistry and Engineering Conference; Washington DC, 2005. (4) Dewulf, J.; Van Langenhove, H. Integrating industrial ecology principles into a set of environmental sustainability indicators for technology assessment. Resour. Conserv. Recycl. 2005, 43, 419-432. (5) Ayres, R. U.; Ayres, L. W.; Martinas, K. Exergy, waste accounting, and life-cycle analysis. Energy 1998, 23, 355-363. (6) Cornelissen, R. L.; Hirs, G. G. The value of the exergetic life cycle assessment besides the LCA. Energy Convers. Manage. 2002, 43, 1417-1424. (7) Finnveden, G.; Ostlund, P. Exergies of natural resources in lifecycle assessment and other applications. Energy 1997, 22, 923931. (8) Hau, J. L.; Bakshi, B. R. Promise and problems of emergy analysis. Ecol. Model. 2004, 178, 215-225. (9) Kotas, T. J. The Exergy Method of Thermal Plant Analysis; Butterworths: London, 1985. (10) Moran, M. J. Availability Analysis, A Guide to Efficient Energy Use, Corrected Edition; The American Society of Mechanical Engineers: New York, 1989. (11) Szargut, J.; Morris, D. R.; Steward, F. R. Exergy Analysis of Thermal, Chemical and Metallurgical Processes; Hemisphere Publishing: New York/Springer-Verlag: Berlin, 1988. 6850
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 21, 2006
(12) Berthiaume, R.; Bouchard, C. Exergy analysis of the environmental impact of paving material manufacture. Trans. Can. Soc. Mech. Eng. 1999, 23, 187-196. (13) Costa, M. M.; Schaeffer, R.; Worrell, E. Exergy accounting of energy and materials flows in steel production systems. Energy 2001, 26, 363-384. (14) Masini, A.; Ayres, L. W.; Ayres, R. U. An Application of Exergy Accounting to Five Basic Metal Industries; CMER: Fontainebleau, France, 2001. (15) Rivero, R. Application of the exergy concept in the petroleum refining and petrochemical industry. Energy Convers. Manage. 2002, 43, 1199-1220. (16) Ayres, R. U.; Ayres, L. W.; Warr, B. Exergy, power and work in the US economy, 1900-1998. Energy 2003, 28, 219-273. (17) Ertesvag, I. S. Energy, exergy, and extended-exergy analysis of the Norwegian society 2000. Energy 2005, 30, 649-675. (18) Wall, G. Exergy conversion in the Swedish society. Resour. Energy 1987, 9, 55-73. (19) Cornelissen, R. L. Thermodynamics and sustainable development. Ph.D. Thesis, Universiteit Twente, The Netherlands, 1997. (20) Dewulf, J.; Van Langenhove, H.; Mulder, J.; van den Berg, M. M. D.; van der Kooi, H. J.; de Swaan Aronsqq, J. D. Illustrations towards quantifying the sustainability of technology. Green Chem. 2000, 2, 108-114. (21) Gong, M.; Wall, G. On exergy and sustainable development Part 2: Indicators and methods. Exergy Int. J. 2001, 1, 217-233. (22) Connelly, L.; Koshland, C. P. Exergy and industrial ecology Part 1: An exergy-based definition of consumption and a thermodynamic interpretation of ecosystem evolution. Exergy Int. J. 2001, 1, 146-165. (23) Dewulf, J. P.; Van Langenhove, H. R. Quantitative assessment of solid waste treatment systems in the industrial ecology perspective by exergy analysis. Environ. Sci. Technol. 2002, 36, 1130-1135. (24) Dewulf, J.; Van Langenhove, H. Assessment of the sustainability of technology by means of a thermodynamically based life cycle analysis. Environ. Sci. Pollut. Res. 2002, 9, 267-273. (25) Michaelis, P.; Jackson, T.; Clift, R. Exergy analysis of the life cycle of steel. Energy 1998, 23, 213-220. (26) Wall, G.; Gong, M. On exergy and sustainable development Part 1: Conditions and concepts. Exergy Int. J. 2001, 1, 128145. (27) Fratzscher, W.; Brodjanskij, V. M.; Michalek, K. Exergie: Theorie und Anwendung; VEB Deutscher Verlag fu ¨ r Grondstoffenindustrie: Leipzig, 1986. (28) Kameyama, H.; Yoshida, K.; Yamauchi, S.; Fueki, K. Evaluation of reference exergies for the elements. Appl. Energy 1982, 11, 69-83. (29) Szargut, J. Analysis of cumulative exergy consumption. Int. J. Energy Res. 1987, 11, 541-547. (30) Wall, G. Exergy - a useful concept within resource accounting; Report no. 77-42; Institute of Theoretical Physics: Go¨teborg, Sweden, 1977. (31) Valero, A.; Ranz, L.; Botero, E. Exergetic Evaluation of Natural Mineral Capital (1) Reference Environment and Methodology. In Proceedings of 15th International Conference on Efficiency, Cost, Optimization, Simulation and Environmental Impact of Energy Systems (ECOS 2002); Berlin, Germany, July 2002. (32) Szargut, J. Chemical exergies of the elements. Appl. Energy 1989, 32, 269-286. (33) Szargut, J.; Valero, A.; Stanek, W.; Valero, A. Towards an international reference environment of chemical exergy. In Proceedings of 18th International Conference on Efficiency, Cost, Optimization, Simulation and Environmental Impact of Energy Systems (ECOS 2005); Trondheim, Norway; Vol. 1, 409-417, June 2005. (34) Mohr, P. J.; Taylor, B. N. CODATA. recommended values of the fundamental physical constants: 2002. Rev. Mod. Phys. 2005, 77, 1-107. (35) Chatterjee, N. D.; Kruger, R.; Haller, G.; Olbricht, W. The Bayesian approach to an internally consistent thermodynamic database: theory, database, and generation of phase diagrams. Contrib. Mineral. Petrol. 1998, 133, 149-168. (36) Holland, T. J. B.; Powell, R. An internally consistent thermodynamic data set for phases of petrological interest. J. Metamorph. Geol. 1998, 16, 309-343. (37) Holland, T. J. B.; Powell, R., An internally consistent thermodynamic data set for phases of petrological interest: recent update, dataset 5.5 (22 November 2003); http://www.earthsci.unimelb.edu.au/tpg/thermocalc/ (accessed May 2006). (38) Roine, A. HSC Chemistry, 5.2; Outokumpu Research Oy: Pori, Finland, 2002.
(39) Chen, R. Y.; Li, J.; Xia, S. P.; Gao, S. Y. Thermochemistry of ulexite. Thermochim. Acta 1997, 306, 1-5. (40) Waldner, P.; Pelton, A. D. Critical thermodynamic assessment and modeling of the Fe-Ni-S system. Metall. Mater. Trans. B 2004, 35, 897-907. (41) Vieillard, P. A new method for the prediction of Gibbs free energies of formation of hydrated clay minerals based on the electronegativity scale. Clays Clay Miner. 2000, 48, 459-473. (42) Rivero, R.; Garfias, M. Standard chemical exergy updated. Part II. In Proceedings of 17th International Conference on Efficiency, Cost, Optimization, Simulation and Environmental Impact of
Energy Systems (ECOS 2004); Guanajuato, Mexico; Vol. 2, 773785, July 2004. (43) Wedepohl, K. H. The composition of the continental crust. Geochim. Cosmochim. Acta 1995, 59, 1217-1232. (44) Wedepohl, K. H., Ed. Handbook of Geochemistry; Springer: Berlin, 1969; Vol. 1.
Received for review January 25, 2006. Revised manuscript received August 25, 2006. Accepted August 28, 2006. ES060167D
VOL. 40, NO. 21, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6851
