Exergy Sustainability Indicators as a Tool in Industrial Ecology

RESEARCH AND ANALYSIS

Exergy SustainabilityIndicators as a Tool inIndustrial EcologyApplication to Two Gas-FiredCombined-Cycle Power Plants

Anita Zvolinschi, Signe Kjelstrup, Olav Bolland,and Hedzer J. van der Kooi

Summary

Life-cycle assessment is an established tool for industrial ecol-ogy. An analysis of the energy use in the chemical and otherenergy-intensive industries is still under discussion in this field.We argue that the concept of exergy can play a role in in-dustrial ecology, using a recent Norwegian power produc-tion policy question as illustration. The question is whether tobuild a standard natural gas- or a hydrogen-fired gas-turbinecombined-cycle power plant to meet increased needs for elec-tricity in Norway. Several indicators are relevant for this dis-cussion, and we calculate three based on exergy calculations,as proposed in the literature. The indicators are exergy renew-ability, exergy efficiency, and environmental compatibility. Weshow how these indicators can be used to evaluate paths forsustainable power production in two gas-fired combined-cyclepower plants. We found that the two plants in question wereequivalent, as judged by their exergy renewability and theirenvironmental compatibility, but not by their exergy efficiency.This indicator favored the standard power plant, possibly incombination with carbon dioxide (CO2) sequestration in a de-pleted gas reservoir. The analysis suggested that the presentsituation for power production in gas-fired combined-cyclepower plants is such that one may have to choose in gen-eral between power production with a high exergy efficiency,but low renewability indicator, or the opposite, low exergy effi-ciency and high renewability indicator. The general importanceof exergy analysis was demonstrated by this example. It en-ables communication between different professional groups.The technological details, understood by the engineers, canbe transposed to meaningful aggregated indicators for deci-sion makers.

Keywords

abatementdecision makingenvironmental compatibilitygas-fired power plantrenewabilitythermodynamics

e-supplement available on the JIEWeb site

Address correspondence to:Signe KjelstrupDepartment of ChemistryNorwegian University of Scienceand Technology—NTNUNO-7491 Trondheim, Norway<[email protected]><www.indecol.ntnu.no>

© 2007 by the Massachusetts Institute ofTechnology and Yale University

Volume 11, Number 4

www.mitpressjournals.org/jie Journal of Industrial Ecology 85

R E S E A R C H A N D A N A LYS I S

Introduction

Industrial ecology is meant to be “an agentof change” for industry by providing princi-ples and tools that can foster change towarda more sustainable society. Principles such as“industrial metabolism” (Ayres and Simonis1994), “closed industrial ecosystem” (Frosch andGallopoulos 1989), “technological food webs”(Graedel and Allenby 1995), and “industrial sym-biosis” (Ehrenfeld and Gertler 1997) are nowused as general principles in industrial ecology.Tools, such as life-cycle assessment and net en-ergy analysis, generate comparative studies of ma-terial and energy conversion processes. However,these tools are accounting for energy needs interms of lower heating values (LHVs).1 The sec-ond law of thermodynamics is then not consid-ered. As a result, the various qualities of energyare not calculated, and can thus not be comparedor studied in energy converting processes.

Several articles have presented the impor-tance of understanding the quality of energyneeds in industrial systems, and furthermore, theneed to discuss where and how the energy qualityor exergy degrades in these systems; see, for ex-ample, the references by Wall (1977), Connellyand Koshland (1997), Cornelissen (1997),Lowenthal and Kastenberg (1998), de SwaanArons and colleagues (2004), and Sciubba(2004). The benefits of exergy-based metrics overthe metrics based on material or energy account-ing were illustrated by Yi and colleagues (2004).This article aims to continue this discussion.

We shall see that the results of exergy anal-ysis can be aggregated into three exergy-basedindicators in industrial ecology. We argue thatthese indicators will quantify different aspects ofsustainable power production, namely its exergyrenewability, exergy efficiency, and the exergyrequirements for abatement of carbon dioxide(CO2) emissions. We demonstrate the useful-ness of these indicators by examining a recentNorwegian policy problem dealing with addi-tional electricity generation in Norway. Norwayhas agreed to comply with the Kyoto Protocoltarget. This means that the CO2 emissions mustbe reduced by about 3 millions tonnes2 of CO2

equivalents compared with the 1996 level, andby about 12 million tonnes of CO2 equivalents

compared with the projected figure for 2010 (seeNorwegian Ministry of the Environment [1998]).The proposal to build two natural gas-fired powerplants at Karstø and Kollsnes shall be evaluated inthis context. Thus, construction of power plantswith some kind of CO2 abatement option is nec-essary. Therefore, Bolland and colleagues (2001)and later Ertesvag and colleagues (2005) haveproposed a power plant that burns hydrogen in-stead of natural gas, using hydrogen that is ob-tained by a natural gas reforming process (Bollandet al. 2001, Ertesvag et al. 2005). This plant istherefore taken as one of the options. As the otherplant we take the standard IEA power plant (IEAGreenhouse Gas R&D Programme 2000).

We aim to compare these plants using ex-ergy analysis (i.e., mapping of energy quality useand degradation in all process steps), and threeexergy-based indicators that reflect important as-pects of the sustainability of a system (i.e., re-newability, efficiency, and recycling of materials).The aim is to bring out in a systematic mannerarguments that favor one type of power plantabove the other, in a specific way, and in a waythat will be also useful for other power produc-tion systems. Before the systems are calculatedin detail, we present the exergy-based indicatorsand the methodology employed. Subsequently,we describe, analyze and characterize the two op-tions for electricity generation. Finally, we discussissues that could help us to understand the use-fulness and the meaning of the present study.

Three Exergy-Based Indicators

Exergy measures the energy quality of any ma-terial flow or stock. Unlike mass and energy, ex-ergy is not conserved; in every successive processstep exergy degrades. Identifying exergy degrada-tion during materials processing allows us to in-terpret its causes. Improvement options can thenbe evaluated and compared.

The three indicators intended to quantifythe sustainability of power production paths are(a) exergy renewability (α), which considers theexergy input of renewable resources (E in,renewable)compared to the total of all exergy inputs (Ein);(b) exergy efficiency (ε), which shows howwell energy quality is preserved in the system;and (c) environmental compatibility (ζ ), which

86 Journal of Industrial Ecology


considers the exergy needed to abate preferablyall emissions and all used products. These indi-cators were first introduced by Dewulf and col-leagues (2000), who also provided illustrations oftheir use in the evaluation of the sustainabilityof electricity generation in a cogeneration powerplant and a photovoltaic power plant (Dewulfet al. 2000).

The exergy of solar radiation and solar exergystored in biomass are considered as renewable re-sources for exergy, while the exergy of fossil fuelsis nonrenewable. After this classification is made,the exergy renewability indicator is calculated as

α = Ein,renewable

Ein(1)

In agreement with the second law of thermo-dynamics, during any real process there is an en-ergy quality loss due to the necessary forces thatdrive the process in the desired direction. Thisloss is called exergy loss. It follows directly thatan important strategy for improving the sustain-ability of systems is to reduce the rate of exergyloss or entropy production or, in other words, toincrease the exergy efficiency. The exergy effi-ciency (ε) is defined as the ratio of the utilizedexergy output (Eout) and the exergy needed toobtain that, (E in)

ε = Eout

Ein(2)

In the present case, the used exergy output isthe net electric power made available for societyby the power plant, and the exergy input is theexergy content of the natural gas as it is in thereservoir, plus any possible solar exergy input orother inputs.

Given that we do not want to increase emis-sion levels by industrial production, all used prod-ucts and wastes must be abated at approximatelythe same rate as they are produced. Buffering ofproducts and wastes in the environment may benecessary, however. When natural gas is used asa feedstock for a power plant, the product is elec-tric power, and as waste we have, among othergases, carbon dioxide. Closure of the carbon cy-cle requires that we return carbon to the state ithad in the reservoir, in other words, to naturalgas. Only in this manner can we speak of powerproduction without CO2 emissions.

The environmental compatibility was definedby Dewulf and colleagues (2000) as follows:

ζ = Ein

Ein + Ein,abatement(3)

where E in is the exergy input to the productionprocesses, the same as given in equations (1) and(2), and E in,abatement is the total exergy inputrequired to abate all emissions and wastes andeventually all used products. An environmentalcompatibility indicator of unity means that noabatement is required by the outputs of the pro-duction processes; the production processes areentirely compatible with the environment. Avalue lower than unity means that the outputshave to be or are abated; the production pro-cesses are partially compatible with the environ-ment. Because different levels of abatement canbe considered, the indicator should be used for as-sessing alternative processes with the same levelof abatement. In the present study no abatementis needed for the electricity, as its end product isheat in the natural environment. Exergy inputsare needed to abate all emissions, but we shallconsider only the emission of CO2. A completereturn of CO2 to natural gas is unfortunately nottechnically feasible, so we are limited to dealingwith two approximations of a closed cycle.

In the ideal situation, one may say that man-made systems should have perfectly closed mate-rial cycles and should perform these cycles in amost efficient way (i.e., reversibly) according tothe second law of thermodynamics. The only in-put to the system should be renewable exergyfrom the sun, and the only output should beheat (entropy). In such a situation, all exergy-based indicators defined here are unity. This sit-uation cannot be achieved in reality, however. Itis still possible to use the indicators to examinethe question posed in the Introduction: Can onepower plant be favored over the other when CO2

abatement processes are included in the analysiscompletely or partially? As we shall see, the indi-cators are able to point to a path toward more sus-tainable power production in gas-fired combined-cycle power plants.

Zvolinschi et al., Exergy Sustainability Indicators as a Tool in Industrial Ecology 87


Figure 1 Partial closure of the carbon used in Plants A and B combined with Option C (a) or combinedwith Option D (b).

Power Plants with CO2Abatement Options to BeCompared

Both power plants use natural gas from a NorthSea gas reservoir. The plants are designed to havethe same exhaust compositions to allow for a faircomparison between them. This is achieved byamine absorption processes, which are integratedwith the power plants. The power production ca-pacity of both plants is the same, 400 megawatts(MW), so the technologies of the plants fit withthis production capacity.

(a) Plant A:This plant is a standard power plant(IEA Greenhouse Gas R&D Programme2000), which is a gas-turbine combined-cycle plant that burns natural gas. Amineabsorption processes are integrated withthe gas-turbine combined cycle to captureCO2 from the exhaust gases.

(b) Plant B:This plant is a hydrogen-fired power plant,which is an integrated system between anatural gas reforming plant for hydrogenproduction and a gas-turbine combined-cycle plant where hydrogen is burned(Bolland et al. 2001). Before hydrogen isburned, the gas mixture of hydrogen andCO2 is separated in an amine chemical ab-sorption plant.

The outlets of Plants A and B are the same by thischoice. We have chosen to analyze two optionsfor CO2 emission abatement. They are as follows:

(c) Option C, consisting of three steps thatconvert the captured CO2 from Plant Aor B into synthetic gas, a pipeline-qualitymethane-rich gas. The steps are (1) photo-synthetic conversion of CO2 into biomass(Legrand 1993), (2) anaerobic digestionof biomass (Legrand 1993) and (3) biogasprocessing (Kapoor and Yang 1989).

(d) Option D, with only one step, namely se-questration of CO2 from Plant A or B intoa depleted natural gas reservoir. This op-tion is motivated by the feasibility of CO2

sequestration (White 2003) and the lim-ited exergy efficiency of CO2 conversionto biomass (Bisio and Bisio 1998).

The different plants and processes are illustratedin figure 1. Full details on the natural gas pro-cessing in the power plants, and the abatementsoptions are given in appendices A (power plants)and B (abatement options), which are availableas an e-supplement on the Journal’s Web site.

Methodology

The gas-fired, gas-turbine, combined-cyclepower plants were modeled using GTPRO soft-ware from Thermoflow (Elmasri 1996), and thehydrogen production plant and the amine absorp-tion plant were modeled using PRO/II software



from PRO/II (SMISCI 2001). Detailed processdescriptions, processes stream data and bound-ary conditions are given in appendices A and B.The software provided material- and energy-flowsheets for the processes of interest. Exergy in-puts and outputs of all individual streams werecalculated from the data in these flow sheets us-ing standard methods (Szargut et al. 1988, Kotas1995). The environmental reference temperaturewas 15◦C and the pressure was 1.01 bar.3 Thechemical exergy of each stream was computed inan in-house program, using the standard exergyvalues of components, proposed by Szargut et al.(1988). Options C and D were likewise calcu-lated using Microsoft Excel. Technical details ofthese options can be found in appendix B of thee-supplement. Making sure that mass and energyconservation principles were obeyed, the exergyloss in each process was determined by subtract-ing the exergy output from the input. Data forthe modeling of the CO2 amine absorption pro-cess were taken from the article by Sander (1991).These data provided the input needed for the cal-culation of exergy-based indicators.

Exergy analysis localizes and quantifies the en-ergy quality degradation or exergy loss due to massand energy conversions. It is normal in exergyanalysis to present results in Grassmann diagrams(Grassmann 1984), which are exergy utilizationmaps, and this practice was followed here for thepower plants.

Neither the exergy analysis nor the exergy-based indicators considered the construction anddismantling phases of any subsystem (see Discus-sion). No economic assessment was performed.

Results

Power Plants

A summary of the exergy calculations of thepower plants is presented in the first two columnsof table 1 and in Grassmann diagrams in figures2 and 3. The detailed data behind the table andthe diagrams are given in appendices A and B.

Table 1 gives normalized results using thestandard IEA power plant as a reference. Thepower available for the society from this plantis 400 MW. The Grassmann diagrams give anoverview of the actual numbers for the two plants.

The thickness of the arrows in these diagrams (inMW) is proportional to the exergy content of astream, and the gray space is the exergy loss in acertain process (in MW). The uncertainty in theexergy data is estimated to be ±3%. The error isdue to assumptions made in the calculations, suchas the assumptions of ideal gas and of equilibriumreactors. The error concerns the precision of asingle value. The accuracy in the ratios is similar.

Table 1 shows then that the ratio of naturalgas exergy input to net power output is 2.0 inPlant A, and it is 2.1 in Plant B. We can alsosee that the main reason for this difference is theexergy losses in the hydrogen plant. As can beseen from the Grassmann diagrams (figures 2 and3), the two technological options appear ratherdifferent, with Plant B being much more complexthan Plant A. The diagrams show the distributionof the natural gas exergy input, and of the exergylosses, in the power plants.

The major sources of lost exergy in Plant A,according to figure 1, are the gas turbine(4.3 kWh/kg natural gas), the amine absoptionprocess (0.8 kWh/kg natural gas), and the steamturbine (0.3 kWh/kg natural gas). In Plant B, thelosses are more spread out. Again the gas turbinehas most of the losses (2.9 kWh/kg natural gas),with the reformer as the number two loss source(1.0 kWh/kg natural gas).

We see that amine absorption needs a substan-tial exergy input. In Plant A this exergy comesfrom thermal and electric energy in the com-bined cycle. In Plant B, the best option is towithdraw the thermal energy from the medium-pressure steam produced in the hydrogen plant.In doing so, the electric power production in thecombined cycle of Plant B is less affected. For fur-ther technical details on plants and options, seeappendix A in the e-supplement. We found thatCO2 absorption by amine causes an exergy loss of5.3% of the natural gas chemical exergy input toPlant A. In Plant B this loss was only 1.1%.

Carbon Mass Balance

The only chemical element cycle studied hereis the carbon cycle. The carbon mass flow rates (inkilograms of carbon per second [kg-C/s]) for eachpower plant with and without their combinationswith abatement Options C and D are given in



Table 1 Exergy inputs (renewable and nonrenewable), losses (internal and external), and outputs (netelectric power, CO2 , synthetic gas) for Plant A and B with and without Option C or D

A B A B A B

Plant/Stream or process Without Option C or D With Option C With Option D

Exergy inputsNonrenewNatural gas 800.8 859.2 960.4 992.8 1,035.2 1,108.0Monoethanolamine (MEA) 32.4 57.6 38.8 63.2 42.0 75.0Nutrients (NH3, P2O5, K2O) — — 1.6 1.6 — —Carbon molecular sieve — — 0.8 0.8 — —

RenewSolar radiation — — 761.6 808.4 — —Seawater 0.0 0.0 0.0 0.0 0.0 0.0Make-up water 4.1 8.8 9.0 9.6 5.5 11.6Algae, biofermenters — — 0.4 0.4 — —

Total exergy input 837.3 925.6 1,772.6 1,876.8 1,082.6 1,194.6

Exergy lossesInternHydrogen plant — 126.8 — 146.8 — 172.4Power plant 335.5 291.6 402.4 338.0 464.8 377.6CO2 amine absorption 42.6 9.3 66.0 12.0 71.2 16.0CO2 to biomass conversion — — 351.3 367.3 — —Biomass digestion — — 6.5 7.2 — —Biogas processing — — 21.0 27.4 — —CO2 sequestration — — — — 70.9 107.8ExternCleaned exhaust gases 4.3 8.3 2.0 2.0 5.5 10.8Waste heat in seawater 4.0 7.6 0.0 0.0 5.2 10.0Total exergy loss 386.4 443.6 849.2 900.7 617.6 694.6Gross power output (a) 421.2 391.2 487.3 463.3 509.7 520.4Auxiliary power for powerplant (b)

8.8 4.9 8.8 5.0 8.8 5.2

Power from fuel expansion (c) 12.9 17.3 15.5 21.1 16.7 22.3Exergy needed for CO2

abatement:In CO2 amine absorption 25.3 3.6 28.0 4.2 30.8 4.7In Option C or D — — 66.0 75.2 86.8 132.8Exergy OutputsNet electricity (a−b+c−d) 400.0 400.0 400.0 400.0 400.0 400.0CO2 18.5 24.4 4.4 5.2 23.0 25.0Monoethanolamine (MEA) 32.4 57.6 38.8 63.2 42.0 75.0Carbon molecular sieve — — 0.8 0.8 — —Fertilizers — — 22.8 30.5 —Synthetic gas (SG) — 456.6 476.4 —Total exergy output 450.9 482.0 923.4 976.1 465.0 500.0

Note: All numbers are in megawatts (MW). The uncertainty in the numbers is estimated to be ±3%. Renew = renewable;Nonrenew = nonrenewable; Extern = external; Intern = internal; NH3 = ammonia; P2O5 = phosphorus pentoxide;K2O = potassium oxide; CO2 = carbon dioxide.



Figure 2 Grassmann diagram (in MW) for mapping the exergy losses in Plant A.

Figure 3 Grassmann diagram (in MW) for mapping the exergy losses in Plant B.

table 2. Here, the amount of carbon taken fromthe natural gas reservoir is given in the first col-umn for Plant A (upper part of the table) andPlant B (lower part of the table). Inflow contri-butions are given in the three first columns, andoutflow values are given in the last three.

The table shows, for instance, that only77% of a total of 13.3 kg-C/s was convertedinto either synthetic gas (7 kg-C/s) or fertilizers(3.3 kg-C/s) in Plant A with Option C. The re-maining part was found in the cleaned exhaustgases released to the atmosphere (1 kg-C/s) and



Table 2 Input (in) and output (out) of the carbon mass flow rate (MC , in kg-C/s) for Plant A (A), Plant B(B), Plant A and B with Option C (A+C and B+C, respectively), and Plant A and B with Option D (A+Dand B+D, respectively)

MC,in(kg-C/s) MC,out (kg-C/s)

Plant + Option A A+ C A+ D A A+ C A+ D

Inputs: Natural gas 10.79 13.30 13.95Air 0.06 0.07 0.09Outputs: Exhaust gas 1.00 1.01 1.49CO2 9.85 2.06 12.55Synthetic gas — 7.0 —Fertilizers — 3.3 —Total 10.85 13.37 14.04 10.85 13.37 14.04

Plant + Option B B + C B + D B B + C B + D

Inputs: Natural gas 11.83 14.60 15.25Air 0.08 0.09 0.10Outputs: Exhaust gas 1.11 1.03 1.62CO2 10.80 3.26 13.73Synthetic gas — 7.16 —Fertilizers — 3.24 —Total 11.91 14.69 15.35 11.91 14.69 15.35

in the CO2-rich gas obtained from biogas pro-cessing (2 kg-C/s). In Plant A with Option D,almost 90% of a total of 14 kg-C/s was recoveredas sequestered gas (12.5 kg-C/s). The remainingpart was found in the cleaned exhaust gases (1.5kg-C/s).

The mass flow rate of carbon that was ex-tracted from the natural gas reservoir and used inPlant B (11.8 kg-C/s) was 1.0 kg-C/s higher thanin Plant A (10.8 kg-C/s). The same differencein the input flow rates was found when the CO2

captured in Plants A and B was abated by OptionC or D.

We see that the recycling of carbon is notcomplete and that Option D stores by far themost of the carbon extracted from the natural gasreservoir.

Exergy Inputs for CO2 Abatement inOption C

Option C consisted of three steps. The detailsof each step are given in appendix B, but a sum-mary is given in figure 4a (Plant A) and figure 4b(Plant B). The exergy needed for the three stepsin Option C was taken from the power plants

themselves. The same type of exergy input inOption D is shown in table 1; see the two lastcolumns. The exergy losses in the processes arealso shown in this table.

The exergy needed to convert CO2 into syn-thetic gas by Option C was 66 MW for Plant Aand 75 MW for Plant B. The marine areas thatare required for converting the CO2 coming fromPlants A and B into biomass, are 6,100 and 6,500hectares (ha), respectively.

When the CO2 emission is abated by seques-tration (Option D), the total exergy needed forabatement was 87 MW for Plant A and 133 MWfor Plant B.

Exergy-Based Indicators

The exergy-based indicators were calculatedfrom table 1, using equations (1)–(3). All terms inthese equations are given in MW in the first threecolumns of table 3, and the indicators for PlantsA and B with Options C and D alternatively areshown in the last three columns of table 3.

The exergy renewability indicators of bothplants are near zero, α = 0.005 (Plant A) andα = 0.009 (Plant B). The numbers refer to the



Figure 4 Exergy flows (in MW) in Plant A with Option C (a) and in Plant B with Option C (b). Exergylosses are also shown in boxes.

Table 3 Total exergy input (Ein), renewable exergy input (Ein,renewable), useful exergy output (Eout), exergyinput for CO2 abatement (Ein,abatement), and exergy-based indicators

Ein Eout Ei n,r enewabl e Ei n,abatement

System (MW) (MW) (MW) (MW) α ε ζ

Plant A 837 400 4 25 0.005 0.48 0.97Plant A with Option C 1,772 400 771 94 0.435 0.22 0.95Plant A with Option D 1,082 400 6 118 0.005 0.37 0.90Plant B 926 400 9 4 0.009 0.43 0.98Plant B with Option C 1,877 400 818 80 0.435 0.21 0.96Plant B with Option D 1,194 400 12 138 0.009 0.33 0.89

Note: α, exergy renewability; ε, exergy efficiency; and ζ , environmental compatibility. The uncertainty in the indicatorsis estimated to be ±3%.

operational phase of the plants. The low valuescan be explained by almost no input of renewableresources with exergy content different from zero(the specific chemical exergy of seawater that isneeded in a large quantity is almost zero). Theexergy renewability increases when we attemptto close the carbon cycle by Option C. The valuebecomes 0.44 for both power plants with OptionC. The main reason for this gain in α comparedwith that for Plant A and B alone, is the ex-ergy input from the sun. When Plants A or Bare combined with Option D the value of α doesnot change, as almost all exergy inputs are fromnonrenewable sources.

The exergy efficiency, ε, is always higher forPlant A than for Plant B; see the sixth column intable 3. The difference in ε between Plant A andB is 0.05 units. This difference remains if Option

D is included; see column six, rows 3 and 6, intable 3. It becomes smaller (0.01) if Option C isused.

The environmental compatibility (ζ ) was cal-culated for the various degrees of CO2 capture.Already Plants A and B contain an abatementprocess, the CO2 capture by means of an amineabsorption process. The values of ζ were pairwisethe same within the accuracy given in the num-bers: For Plants A and B alone they were 0.97and 0.99, respectively. For the plants combinedwith Option C they were 0.96 and 0.94, and withOption D they were 0.90 and 0.89.

Discussion

The exergy indicators shall first be discussedper se, before we use them in an overall evaluation



of the particular problem posed, namely that ofwhich power plant to build, Plant A or B aloneor with abatement Options C or D. We continueto give some general perspectives related to theuse of these indicators in industrial ecology.

The Meaning of the Exergy-BasedIndicators

We have seen from the results in table 3 thatall options discussed here are far from being re-newable. The exergy renewability indicators forPlants A and B alone or combined with OptionD are for all practical purposes zero. The only wayto raise this number is to use Option C in combi-nation with the plants. Only these combinationspoint to a more sustainable power production. Inthe ideal situation α = 1, we found values near0.4. The high value of the exergy renewabilityfor the power plants in combination with Op-tion C may be an argument for developing thisalternative.

The efforts to approximately close the carboncycle, and thus to attain a higher exergy renewa-bility, do not come without drawbacks. The risein α to 0.4 is accompanied by a drop in the exergyefficiency, from 0.48 to 0.22 in Plant A and from0.43 to 0.21 for Plant B. Also, it should be remem-bered that only a fraction of the total amount ofcarbon (77%) is taken care of in this option (cf.table 2). Clearly the exergy efficiency is signifi-cantly reduced when the renewability increases.

The exergy efficiencies give a clear messagewhen it comes to comparing Plants A and Balone. There is a reason to prefer A to B, froman exergy efficiency point of view, because Ahas a higher value of ε. So, it is not irrelevantwhether we abate CO2 by amine absorptionbefore or after the power production, as judgedfrom this indicator alone. A comment on ourchoice of alternatives to be compared is now inplace. We have chosen as base cases plants thatalready have a certain degree of CO2 abatementin the form of an amine absorption process. Theexergy efficiencies of Plants A and B without theamine absorption process were 0.55 and 0.47,respectively (not shown in table 1 or 3). Withoutthe amine absorption processes, the end states ofPlants A and B are different, however, meaningthat these configurations cannot be compared in

a fair way. When we compare the plants that havethe amine absorption process, which we can do ina fair way, we find that Plant A is better from anexergy efficiency point of view. This conclusionwill not change if we combine the plants withOption D; the exergy efficiency will sink to 0.37(Plant A) or 0.33 (Plant B). When the plants arecombined with Option C, their exergy efficiencyindicators become the same (ε = 0.22), a very lowvalue. A shift of the system boundaries to includemonoethanolamine and fertilizer production willnot improve the indicators for Plant B.

The environmental compatibility indicator(ζ ) does not give any argument to favor A over B,whether the plants stand alone or are combinedwith Option D or C.

The above analysis was carried out withoutconsidering the exergy needs and losses to buildand dismantle the power plants. Lombardi (2003)showed that these assumptions were good (within1%) for combined-cycle power plants of the samesize based on coal gasification. For the hydrogen-fired power plant, with its extra equipment, theassumption may be less good but of the same orderof magnitude.

The Political Debate: Which PowerPlant to Build

The indicators used here are obtained withrelatively high accuracy. This is an advantage fora decision-making process. The environmentalcompatibility did not give a preference for anyof the power plants, so a decision to build oneof them should use arguments connected to theother indicators. The exergy efficiency indica-tors gave the message that Plant A without extraabatement processes beyond the amine process ispreferable to Plant B with amine absorption ofCO2. The choice for Option C comes with thedrawback of a very large reduction in the exergyefficiency, a rather high area needed, and an in-complete abatement of the emissions. The ques-tion that emerges from this discussion is whetherit then pays to abate by sequestration (Option D)or by the amine absorption process alone. The dif-ference in environmental compatibility betweenPlant A and Plant A with Option D is less than10%. The reason for this difference is the extrapower needed to store CO2 in the depleted gas



reservoir. The storage problem of CO2 capturedfrom the amine solution is on the other hand notsolved in the case given by what we have calledPlant A.

From the above considerations we appear tobe in a trade-off situation, a typical situation fordecision makers. The analysis tells that we are ina situation in which we have to choose betweenpower production with high exergy efficiency butlow exergy renewability (Plant A with Option D),or the opposite, low exergy efficiency and a highexergy renewability indicator (Plant A with Op-tion C). Plant B does not perform better than ei-ther of these options and can be ruled out. Such atrade-off situation may be common for situationswhen fossil fuels are the resources. We showedabove that exergy and exergy indicators can offera quantitative basis for the analysis of gains andlosses. This input may be useful for decision mak-ers in making their weighted decisions but alsofor the governmental sector in determining theirresearch funding priorities.

General Perspectives

The Kyoto protocol can be seen as a first ef-fort of the international community to commititself to reduce CO2 emissions worldwide. Thepresent example shows that it is possible to an-alyze in a systematic manner alternative routesto power production in gas-fired combined-cyclepower plants with various degrees of CO2 abate-ment. By taking into account the second law ofthermodynamics, we can establish a scale for en-ergy quality, namely the exergy scale. The qual-ity of energy in its various forms is calculatedas exergy. This then allows a comparison of anyenergy-converting process, exemplified here bytwo power production systems. In the first place,the detailed exergy analysis given in Grassmanndiagrams and appendices A and B, is useful. Inthe second place, the exergy-based indicators areuseful.

For the professional engineer, exergy analysiscan be used to localize points where research ef-forts should be concentrated in order to increaseexergy efficiency. The analysis allows a compari-son and assessment of different material and en-ergy conversion systems with the same functionon the same scale. It helps industrial ecology to

adhere to the first law as well as to the secondlaw of thermodynamics. Engineering profession-als can assess the potential for improvements ofa system and transmit the potential to the indus-trial ecologist, who will then be informed aboutthe limits of the indicators.

The technological details are given in appen-dices A and B. It is clear from these appendicesthat there are numerous technological details be-hind the indicators. Such a wealth of technicaldetail is difficult to comprehend for nonengi-neering professionals. It is in this context thatexergy-based indicators first proposed by Dewulfand colleagues (2000) can be useful, as we haveillustrated with our example. They can includeall technical details on a meaningful aggregatedlevel, and they are therefore much more accessi-ble to decision makers.

We saw in the definition of the exergy indi-cators that they all have an ideal limit of unity.These are unrealistic limits of operation due tothe second law of thermodynamics. Any natu-ral or man-made system has an exergy efficiencybelow unity (or 100%). The conversion of solarenergy in plants is never without exergy losses.It is therefore beyond expectation that any man-made system can reach the ideal limit for ε. Asa consequence the environmental compatibilityis always affected by exergy losses and the exergyinputs needed to abate emissions, waste, and usedproducts. We must face the constraint of the sec-ond law of thermodynamics and recognize thatconstructing multilevel material cycles with anexergy renewability of unity is essentially impos-sible. Yet we still need a practical way towardsustainability and a methodology for assessingthe way. This is why we would like to name theexergy indicators sustainability indicators, becausethey give such a direction. They do not becomeunity, but they should be as high as possible. Arealizable target for the exergy efficiency in single-process units was discussed by Bejan and Tondeur(1998), Nummedal and colleagues (2003), andJohannessen and Kjelstrup (2005). If produc-tion is carried out with such a practical limit onthe exergy loss (namely with minimum entropyproduction), the exergy efficiency of many pro-cesses can be raised beyond present-day levels. Itfollows that the need for exergy for abatementprocesses in attempts to close material cycles



becomes smaller. If at the same time, fossil fu-els are replaced by renewable energy sources, wewill be on a path that will increase all indicators.This is a path toward sustainable power production.

We have seen above how the laws of thermo-dynamics, as embraced by exergy analysis, can beused to give information that can help decisionmakers in their choice between future paths forpower production. The three exergy-based sus-tainability indicators, recently proposed in theliterature, were central in this context. Giventhis background, we argue that industrial ecologyshould adopt these indicators as another tool tobe used to analyze choices in terms of their envi-ronmental consequences and to promote choicesthat are more sustainable. The indicators can bewidely used. In particular, they can be used to es-tablish a scale to measure power production sys-tems. Nerverthless, other indicators are neededas sustainability is more than renewability, effi-ciency, and evironmental compatibility (Mayeret al. 2004).

Conclusions

We have demonstrated that some recentlyproposed exergy sustainability indicators can beused to assess the sustainability of options forpower production, and we have proposed thatthe indicators become tools within the industrialecology framework. We have shown this using theexample of a natural gas- and a hydrogen-firedcombined-cycle power plant, an actual powerproduction policy problem in Norway. Aggre-gated information for decision makers was ob-tained from three exergy-based indicators.

Our analysis reached the following conclu-sions. The two plants with abatement options in-cluded were equivalent, as judged by their exergyrenewability indicator and by their environmen-tal compatibility alone, but the exergy efficien-cies of all systems studied gave a clear message.The option to deal with CO2 abatement beyondamine absorption, using solar energy, halved theexergy efficiency of both power plants. Therefore,sequestration of CO2 in a depleted gas reservoirseemed the most viable path. But also this system(Plant A with Option D) lowered the exergy ef-ficiency significantly. The analysis suggested thatthe present situation for power production from

fossil fuels is such that we may have to choosebetween power production with high exergy ef-ficiency, but a low renewability indicator, or theopposite, low exergy efficiency and a high exergyrenewability indicator.

The importance of exergy analysis is that itenables communication between different pro-fessions. The technological details, understoodby the engineers, can be translated to a meaning-ful aggregated level, a level that can be compre-hended by decision makers.

We propose naming the exergy indicators sus-tainability indicators, because they show which di-rection to take toward more and more sustainablepower production. Initial efforts have been madeto set realizable targets for the exergy efficiency indifferent industrial processes (Wall 1977 and deSwaan Arons et al. 2004). More work is needed tocontinue this effort and to promote the necessarychanges in technology.

Acknowledgment

Anita Zvolinschi acknowledges financial sup-port from the Research Council of Norway.

Notes

1. The lower heating value (also known as net calorificvalue or LHV) of a fuel is defined by the AmericanPetroleum Institute (API) as the amount of heatreleased by combusting a specified quantity of fueland returning the temperature of the combustionproducts to 150◦C. The Gas Producers SuppliersAssociation (GPSA) defines the LHV as the en-thalpy of all combustible products, minus the en-thalpy of the fuel at the reference temperature (APIuses 25◦C; GPSA uses 60◦F or 15.56◦C), minus theenthalpy of the stoichiometric oxygen at the refer-ence temperature, minus the heat of vaporizationof the water vapor content of the combustion prod-ucts. Both definitions show that the LHV is calcu-lated using thermodynamic concepts such as heatand enthalpy, but not entropy. This means that theLHV does not consider the second law of thermo-dynamics.

2. The term tonne refers to metric tonne. It is ameasurement of mass equal to 1,000 kilograms(kg, SI). The similar Imperial units and UnitedStates customary units are spelled ton in English;1 tonne ≈ 0.98 Imperial or long tons ≈ 1.10 shorttons.



3. One bar ≈ 0.987 atmospheres (atm) ≈ 14.50pounds per square inch (psi).

References

Ayres, R. U. and U. Simonis. 1994. Industrialmetabolism: Restructuring for sustainable develop-ment. Tokyo: United Nations University Press.

Bejan, A. and D. Tondeur. 1998. Equipartition, opti-mal allocation, and the constructal approach topredicting organization in nature. Revue Generalede Thermique 37(3): 165–180.

Bisio, G. and A. Bisio. 1998. Some thermodynamics re-marks on photosynthetic energy conversion. En-ergy Conversion and Management 39(8): 741–748.

Bolland, O., I. S. Ertesvag, and D. Speich. 2001. Ex-ergy analysis of gas turbine combined cycle withCO2 capture using auto-thermal reforming of nat-ural gas. In Proceedings of Power Generation andSustainable Development Conference 2001. Liege,Belgium: University of Liege, Department ofPower Generation.

Connelly, L. and C. P. Koshland. 1997. Two aspectsof consumption using an exergy-based measure ofdegradation to advance the theory and implemen-tation of industrial ecology. Resource Conservationand Recycling 19(3): 199–217.

Cornelissen, R. L. 1997. Thermodynamics and sus-tainable development: The use of exergy analysisand the reduction of irreversibility. Ph.D. Thesis,University of Twente, Enschede, Netherlands.

de Swaan Arons, J., H. J. van der Kooi, and K. Sankara-narayanan. 2004. Efficiency and sustainability in theenergy and chemical industries. New York, Basel:Marcel Dekker.

Dewulf, J., H. van Langenhove, J. Mulder, M. M. D.van den Berg, H. J. van der Kooi, and J. deSwaan Arons. 2000. Illustration towards quantify-ing the sustainability of technology. Green Chem-istry 2(3): 108–114.

Ehrenfeld, J. R. and N. Gertler. 1997. Industrial ecologyin practice: The evolution of interdependence atKalundborg. Journal of Industrial Ecology 1(1): 67–79.

Elmasri, M. 1996. GTPRO Users Manual. Sudbury,MA: Thermoflow, Inc.

Ertesvag, I., H. Kvamsdal, and O. Bolland. 2005. Exergyanalysis of a gas-turbine combined-cycle powerplant with precombustion CO2 capture. Energy30(1): 5–39.

Frosch, R. and N. Gallopoulos. 1989. Strategies formanufacturing. Scientific American 261(3): 142–155.

Graedel, T. and B. R. Allenby. 1995. Industrial ecology.Englewood Cliffs, NJ: Prentice Hall.

Grassmann, P. 1984. Thermodynamik des Lebensaus dem Blickwinkel der technischen Ther-modynamik und die Exergie. [Engineering-thermodynamics and thermodynamics of life, theexergy.] Naturwissenschaften 71(7): 335–341.

Johannessen, E. and S. Kjelstrup. 2005. Numerical ev-idence for a highway in state space for reactorswith minimum entropy production. Chemical En-gineering Science 60(5): 1491–1495.

IEA Greenhouse Gas R&D Programme. 2000. Leadingoptions for the capture of CO2 emissions at powerstations. Report no. PH3/14, IEA GreenhouseGas R&D Programme. Amsterdam, Netherlands:Storck Engineering Consultancy.

Kapoor, A. and R. T. Yang. 1989. Kinetic separation ofmethane/carbon dioxide by adsorption on molec-ular sieve carbon. Chemical Engineering Science44(8): 1723–1733.

Kotas, T. J. 1995. The exergy method of thermal plantanalysis. Melbourne, FL: Krieger Publishing.

Legrand, R. 1993. Methane from biomass systems anal-ysis and CO2 abatement potential.Biomass andBioenergy 5(3–4): 301–316.

Lombardi, L. 2003. Life cycle assessment comparisonof technical solutions for CO2 emissions reduc-tion in power generation. Energy Conversion andManagement 44(1): 93–108.

Lowenthal, M. D. and W. E. Kastenberg. 1998. In-dustrial ecology and energy systems: A first step.Resources, Conservation and Recycling 24(1): 51–63.

Mayer, A. L., H. W. Thurston, and C. W. Powlowski.2004. The multidisciplinary influence of commonsustainability indices. Frontiers in Ecology and theEnvironment 2(8): 419–426.

Norwegian Ministry of the Environment. 1998. Norwe-gian Climate Change Policy—Overview. Whitepaper. Report to the Storting (Norwegian parlia-ment), Report No. 29. Oslo, Norway: Ministry ofthe Environment.

Nummedal, L., S. Kjelstrup, and M. Costea. 2003. Min-imizing the entropy production rate of an exother-mic reactor with a constant heat-transfer coeffi-cient: The ammonia reaction. Industrial and Engi-neering Chemistry Research 42(5): 1044–1056.

Sander, M. T. 1991. Econamine FG feasibility study.Contract No. T–122.013. Prepared for Statoil byFluor Daniel, Inc.

Sciubba, E. 2004. From engineering economics to ex-tended exergy accounting: A possible path frommonetary to resource-based costing. Journal of In-dustrial Ecology 8(4):19–40.



SMISCI (Simulation Sciences). 2001. PRO-IIUsers Guide. Brea, CA: Simulation Sciences(SMISCI)–An Invensys Company.

Szargut, J., D. R. Morris, and F. R. Steward. 1988. Ex-ergy analysis of thermal, chemical, and metallurgicalprocesses. Berlin: Springer-Verlag.

Wall, G. 1977. Exergy–A useful concept within resourceaccounting. Report no. 77-42. Goteborg, Sweden:Institute of Theoretical Physics.

White, C. M. 2003. Separation and capture of CO2

from large stationary sources and sequestration ingeological formations—coal beds and deep salineaquifers. Journal of the Air & Waste ManagementAssociation 53(6): 645–715.

Yi, H.-S., L. J. Hau, N. U. Ukidwe, and B. R. Bakshi.2004. Hierarchical thermodynamic metrics forevaluating the environmental sustainability of in-

dustrial processes. Environmental Progress 23(4):302–314.

About the Authors

Anita Zvolinschi was a doctoral student in theIndustrial Ecology Programme at the NorwegianUniversity of Science and Technology (NTNU) inTrondheim, Norway, at the time this article was writ-ten. She is now working as a senior process engineer atStatoil ASA, Norway. Signe Kjelstrup is a professor inthe Chemistry Department, also at NTNU, where OlavBolland is a professor in the Department of Energy andProcess Engineering. Hedzer van der Kooi is an assis-tant professor in the Laboratory for Applied Thermody-namics of the Delft University of Technology in Delft,the Netherlands.


Documents

Exergy Sustainability Indicators as a Tool in Industrial Ecology