F EAT U R E C G G176 Green Chemistry December 1999 This journal is © The Royal Society of Chemistry 1999 diagram. A lost work analysis reveals internal losses due to process imperfections, whereas an energy analysis has to ascribe losses to waste material and heat streams leaving the process (i.e., physical streams). The input-side of such a process is represented by natural resources (primary resources).No distinction has been made between resources used as feedstock and resources applied as fuel; they all are quantified by their theoretical workpotential, the universal measure. The output- side is represented by the theoretical work-potential of the desired product(s) and of recovered useful heat (in the form of a steam credit). By comparing the total amount of work-potential entering and leaving the process, the loss of work-potential is revealed, which is either due to processinefficiencies or material/heat release to the environment, so internal and external losses are lumped.All data have been taken from published literature (Table 1). ntroduction The process industry is a large consumer of raw materials, which are utilised both as feedstock for its numerous products and as energy source to drive its numerous processes. In the scope of sustainable development, consensus on the limited availability of our natural resources, and on the need for closed cycles in our ecosphere has grown.Hence, the current approach of our process industry is in question. One might raise the question ‘how sustainable are we?’ In answering this question, first, we have to fall back to the efficiency question.There is a need for quantitative figures on the efficiency with which natural resources are consumed. Such quantitative figures can be provided by the known thermodynamic concept of lost work, indicating the discrepancy between the ideal thermodynamic situation and our current process technology. This paper is partly based on the work presented by Hinderink et al.(1996a). Lost work in the process industry Lost work analysis has so far drawn the most attention in the energy-systems area, where heat is converted to power or electricity. It has less penetrated into the chemical process industry, perhaps because of its greater complexity. Work-potential of materials (chemical exergy) The theoretical work-potential of an energy-source depends on the extent to which it deviates from equilibrium with its natural surrounding.When heat and material streams are ideally brought in equilibrium with their surroundings, a maximum amount of work is available, which is often called ‘exergy’. If such an equilibration is not carried out ideally— which is always the case for real processes—a less than maximum amount of work is available, the remainder being lost, i.e.lost work. The precise calculation of the work-potential of material streams has been described by Szargut et al. (1988), and Hinderink et al. (1996b). Lost-work analysis (exergy analysis) In real material conversion processes, primary materials (natural resources) are converted to consumer materials and heat.These processes do not proceed ideally, so part of the work-potential of the primary materials will be lost. To obtain a ‘feeling’ for lost work in the process industry, the production processes of several ‘large-quantity’ products have been analysed. This paper only focuses on input and output streams of the conversion process; production of the natural resources, transportation, and storage, are excluded from the system boundary, because most often the largest part of lost work is occurred in the conversion step.Elaborate lost work analyses are given, for example, by Hinderink et al. (1996c) and Wall (1988). Figure 1 illustrates the general result of a lost-work analysis of a material conversion process in a so-called Grassmann On the efficiency and sustainability of the process industry I P.Hinderink of the Process Design Center in Breda, and H. J. van der Kooi and J. de Swaan Arons, both from the Laboratory of Applied Thermodynamics and Phase Equilibria at the Delft University of Technology (all in The Netherlands) explore the use of the thermodynamic concept of lost work for discussing the sustainability of the process industry.Figure 1 Generic Grassmann diagram for a material and energy conversion process.Green Chemistry December 1999 G177 This journal is © The Royal Society of Chemistry 1999 The processes Table 1 gives a ‘thermodynamic blueprint’ of some large-scale production processes. The numerical values presented refer to the technology level of the 1970s or 1980s and are based on primary (natural) resources only, e.g.natural gas and air. Because only primary resources are allowed to enter the processes, several sub-processes can be present inside the system, e.g. for the generation of intermediate products, or for the generation of steam or electricity. Lost work involved with the latter type of subprocesses is handled by using commonly applied second law-efficiencies (e.g. 50% for power production via cogeneration). An example of a process with an intermediate product is the urea process. The second step, starting from ammonia, is over 90% efficient, whereas the total process— thus having ammonia just as an intermediate product—shows an efficiency of only half of this value. For the nitric acid process, the second step is the least efficient as indicated by the simplified Grassmann diagram in Figure 2.C G F E AT U R E The left-hand box overleaf gives the calculation procedure for the theoretic workpotential of natural gas. The right-hand box overleaf presents the lost-work analysis for the production of hydrogen from natural gas via the steam reforming route. Figure 2 Simplified Grassmann diagram for the nitric acid process Results of lost-work analyses strongly depend on the system boundary considered and the credit that is given to co-products and by-products.Therefore, the analysis results can vary per author. Thermodynamic efficiency Although efficiency-values can be misleading because they can be defined in numerous ways (Wall, 1977; Gong and Wall, 1997), they are easy to handle.The definition of thermodynamic efficiency applied here is the ratio between workpotential of the desired products (excluding useful heat) and the primary resources applied. Usually, thermodynamic efficiencies based on lost-work analysis do not exceed 70% when starting from primary resources. At the higher side of the efficiency- range, the production of organic products can be found, while inorganic and metallurgical processes are at the lower side of the efficiency range.A combination of low efficiency and high input of work-potential indicates the need for process improvement. An overview of absolute lost work figures for the process industry, e.g. as shown in Figure 3, is more distinct and can be of use to determine which products and/or which processes need to be reconsidered in view of sustainable development.Table 1 Results of global lost work analyses of several important production processes Thoretical work-potential [kJ/mol final product] Final Molecular Raw Data taken Technology Raw Final Steam Lost thermodynamic product weight materials from level materials product credit work efficiency * [%] hydrogen 2 natural gas/air Giacobbe et al. 1990 409 236 28 145 58 ammonia 17 natural gas/air Cremer 1980 763 338 85 340 44 aluminum 27 bauxite Szargut; Habersatter 1990 4703 888 n.a. 3815 19 methanol 32 natural gas/air Supp 1985 1136 717 80 339 63 oxygen 32 air Ullmann 1980 64 4 n.a. 60 6 urea 60 natural gas/air Cremer; Pagani 1980 1590 686 150 754 43 via ammonia nitric acid 63 natural gas/air Cremer; 1975 995 43 151 801 4 via ammonia Lowenheim et.al.copper 63.5 copper ore Szargut; Boustead 1980 1537 130 n.a. 1407 9 methane 16 830 *) excluding steam-credit Figure 3 Graphical representation of the lost work analyses.Table 3 Minimum amount of input data required for the lost-work analysis of hydrogen production input compression power 0.02 kWh = 0.072 MJ natural gas* consumption 4141 kCal = 17.33 MJ water 1.78 kg output export steam 1.2 kg hydrogen 1 Nm3 *) natural gas is used both as feedstock and as energy source process, this rule is broken; hardly any fraction of the natural gas adopted as high quality raw material ends up in the nitric acid.Also, the direct use of high quality natural gas for low quality heating purposes does not coincide with the idea of sustainability.Towards sustainability In the discussion about how we should set up our chemical process industry in the (near) future, the sustainability issue is of prime importance. Sustainability in the ecological sense means that we do not place an intolerable load on the ecosphere and that we maintain the natural basis for F E AT U R E C G G178 Green Chemistry December 1999 This journal is © The Royal Society of Chemistry 1999 Efficient use of high quality resources Figure 4 shows lost work figures for the utilisation of natural gas for various purposes. From this picture it can be concluded that it is the best to use natural gas for those chemical processes in which it eventually ends up in the desired products. The basic rule behind this conclusion is that the degradation of work-potential has to be delayed as long as possible.This rule facilitates the choice of chemical route and raw materials. For the nitric acid The theoretical workpotential of natural gas A typical composition of Dutch natural gas is given in Table 2, which also presents its pure chemical work-potential. The average delivery pressure of Dutch natural gas is about 60 bara, representing a (physical) work-potential in addition, which is calculated as follows assuming ideal gas behaviour: Exfysical = R T0 .ln [ P P0 ] kJ/mol Since the natural gas is a mixture, the work-potential of mixing should be subtracted from the workpotential of the gas. Assuming ideal gas behaviour, this mixture-term is given by: Exmix = R T0 .S [yi ln yi] = 0.95 kJ/mol The total theoretic work-potential thus becomes 876 kJ mol–1. Conversion to normal cubic meters then gives a work-potential of 39.1 MJ Nm–3 (1 kmol gas = 22.4 Nm3). The ratio between the theoretical work-potential (Ex) of natural gas and its lower heating value (LHV) is thus 1.0455. Lost work analysis of hydrogen production based on steam reforming of natural gas Because the processes described in this paper are not simulated (rigorously), the accuracy with which the lost work analyses have been carried out is submissive to the quality of data available in public literature. Especially for processes involving intermediate products, e.g.the production of nitric acid from natural gas via ammonia, data from various sources have been merged making the analysis more complex than the hydrogen-case presented in this box.Nevertheless, the values indicated in Table 1 in the main text are fairly indicative. Here, we present the lost work analysis of a hydrogen production process. Table 3 gives the data applied, which are taken from Giacobbe et al. (1992). Hydrogen is produced by conventional steam reforming of natural gas; purification of the hydrogen product is done by pressure swing adsorption. Table 4 presents the results of the lost work analysis.From these results, the thermodynamic efficiency of hydrogen production is 10.53/18.26 = 0.58. The work-potential equivalent of export steam Export heat is given in the public literature in many ways. To convert the export steam— which is given in mass by Giacobbe et al.—to theoretical work-potential, it is assumed that saturated water at 60 bar (Tsat = 275 °C) is evaporated and superheated to 500 °C by process heat, thereby recovering work-potential.On the basis of the thermodynamic relation Exfys = DH 2 T0·DS and using data from the steam tables, this change of state results in an uptake of 1.06 MJ of workpotential per kg of steam. Figure 4 Lost work figures for the utilisation of natural gas for various purposes Table 2 A typical composition and the chemical work-potential of Dutch natural gas Chemical Composition work-potential *) Contribution Component mole% [kJ/mole] [kJ/mole] methane 90 830 747 ethane 8 1496 120 nitrogen 1.5 0.7 0.01 carbon dioxide 0.5 20 0.1 Total 100 867 *) based on the reference environment defined by Szargut et al., (1988)C G Green Chemistry December 1999 G179 This journal is © The Royal Society of Chemistry 1999 F E AT U R E Chemical routes The work-potential of chemical components can be calculated from thermodynamics. This work-potential can be considered as the minimum work needed to synthesize the specific component from constituents of its surrounding.It has been shown that in practice the production of desired chemicals requires far more work than indicated by the work-potential of this desired product. In other words, the work-potential entering and leaving such processes do not balance, so work is lost. The challenge of our process industry is to limit the losses, while still being able to let our processes run with sufficient speed.In past decades, increasing energyefficiency was accomplished mainly by complex heat-integration within existing chemical processes requiring considerable investments. The lost-work analyses described by Hinderink et al. (1996c), however, shows that the chemical reaction step largely determines the overall thermodynamic efficiency.Chemical reactions have been found by us to be a notorious source of lost work. If chemical processes are developed from scratch by state-of-the-art methods—i.e. by structured process synthesis procedures—attention can be paid to the core of the process, i.e., the chemical reactions or the chemical reactor. Then, a significant improvement of energy efficiency and process economics can be achieved simultaneously (Harmsen et al., 1999).Losses resulting from chemical reactions can be viewed similarly to losses resulting from heat exchange. The driving force for heat transfer is the temperature gradient, which not only determines the rate of transfer, but also the degree of devaluation of work-potential. Chemical reactions do also proceed along a gradient from high to low chemical affinity.On flowing along this gradient, heat is released and work-potential is lost. The relation between the Gibbs free energy of reaction and lost work is linear. This relationship has been established by Denbigh (1956) and was discussed by Hinderink et al. (1996b). This insight is of prime importance for the development of future chemical routes. For more sustainable chemical routes, chemical gradients should be reduced, or should be counterbalanced by chemical reactions proceeding against their gradient.In this view, there is an analogy between heat-pinch and reaction-pinch. Table 4 Lost work analysis of hydrogen production by steam reforming of natural gas* input natural gas 17.33 3 1.0455 = 18.12 MJex / Nm3 of hydrogen water negl.air negl. power 0.072 / hex (=50%) = 0.144 MJex / Nm3 of hydrogen Total = 18.26 output hydrogen 236 kJ/mole = 10.53 MJex / Nm3 of hydrogen steam-credit 1.2 3 1.06 = 1.27 MJex / Nm3 of hydrogen Total = 11.80 lost work 6.46 MJex / Nm3 of hydrogen 145 kJ/mole 72.4 GJ/ton *) for electricity/power production, an exergetic efficiency of 50% has been assumed (cogeneration process) life.The complexity of the chemical industry with its numerous products made that we have lost sight of the associated ecological impact of these products’ lifecycles: when you produce something, you also produce long-term effects. Improvement of thermodynamic efficiency is frequently but erroneously considered as the contribution to sustainability. An increase of thermodynamic efficiency, however, has to do with a lowering of the rate with which our non-renewable natural resources are consumed, whereas sustainability implies utilisation of renewable resources such as biomass or solar energy.Although the figures presented in this paper are quite indicative, they do not reflect the degree of sustainability of the processes but rather their efficiency.In other words, the degree to which they use renewable materials is not given explicitly. Actually, a chemical conversion process can be 100% efficient when all non-renewable work-potential (exergy) ends up in the desired product(s). In real processes, exergy is lost. Hence, more exergy enters the process than leaves it. This excess of exergy entering the process to make it proceed has to originate from renewable sources, such as solar exergy, in order to contribute to sustainability.Such a ‘balanced’ process industry can still lead to exhaustion of our natural resources if the exergy-content of the nonrenewable end-products is not utilized at the end of the product-life. If we go one step further, our starting materials should be renewable, e.g. carbon dioxide and water.Real sustainable systems/chains need to be circular with respect to matter (outputs become inputs). The driving force for such a sustainable system must stem from solar energy, which is essentially available in large quantity (Figure 5). Wall (1977) and Gong and Wall (1997) give elaborate discussions on exergy and sustainability. Looking into this matter more carefully, we have come to the insight that the extent to which a process contributes to sustainability can be characterised by three parameters.The first parameter is the one we have discussed before, the thermodynamic efficiency of the process. The second parameter needs to reflect the extent to which use has been made of renewable resources. Finally, a third parameter is needed to indicate the extent to which circles have been closed.For more details we refer to Van Den Berg et al. (1999). Figure 5 Flows of energy, exergy, and matter on earth (free from Wall, 1977)F E AT U R E C G G180 Green Chemistry December 1999 This journal is © The Royal Society of Chemistry 1999 Examples of reactions proceeding with a large gradient are the production of nitric acid by the partial oxidation of ammonia and the conversion of H2S into elemental sulphur and steam.The development of new chemical processes should thus focus on exergy-neutral reactions. In addition, one should also look for sophisticated utilisation of the Gibbs energy of reaction (Harmsen and Hinderink, 1999), e.g. by using it directly to drive a separation, i.e. reactive distillation.Of course, all future chemistry will still be submitted to the laws of nature. Renewable resources Although Figure 3 is fairly indicative, the graph lacks information on which part of the primary work-potential is renewable. If, for example, the energy needed for the metallurgical processes comes from hydro-energy, our conclusions with respect to the improvement-potential of these processes will have to be adapted.Conclusions Some general observations from lost work analyses are: l The extent of lost-work differs from product to product and from process to process and depends largely on how skilled we are. l Processes showing a large steam credit, although useful, should be distrusted, because this implicitly means that more primary work-potential is applied than actually needed.l All second law improvements basically have to come from postponing as long as possible the devaluation of workpotential; keep the quality of energy and matter high. In view of sustainability: l The best source for work-potential (exergy) is a renewable resource l A more complete thermodynamic analysis of processes does not only deal with the efficiency but also with the extent to which renewable resources have been used and to which extent ecological cycles have been closed.l An intuitive approach towards efficiency and sustainability is in our view unacceptable given the importance and urgency of these issues. Therefore, it is advisable to benefit from the quantitative power of the extended thermodynamic analysis proposed.References Berg, van den, M. M. D.; Kooi, van den, H. J.; Swaan Arons, de, J. (1999) A thermodynamic basis for sustainability, In ‘Proceedings of Efficiency, cost, optimization, simulation, and environmental aspects of energy systems, ECOS ‘99, June 8–10, Tokyo, Japan (Ed. Ishida et al.), pp. 270–275. Boustead, F.; Hancock, G. F. (1979) Handbook of industrial energy-analysis, John Wiley & Sons (New York).Cremer, H. (1980) Thermodynamic balance and analysis of syngas and ammonia plant, ACS Symposium Series, 122, ASME (Washington). Denbigh, K. G. (1956) The second law efficiency of chemical processes, Chem. Eng. Sci., 6(1), 1–9. Giacobbe, F. G.; Iaquaniello, G.; Loiacono, O. (1992) Increase hydrogen production, Hydrocarbon Process., 3, 69–72. Gong, M.; Wall, G.(1997) On exergetics, economics and optimization of technical processes to meet environmental conditions, in Proceedings of Int. Conf. on Thermodynamic Analysis and Improvement of Energy Systems (Ed. C. Ruixian), TAIES ’97 Beijing (China), pp. 453-460. Habersatter, K. (1991) Ökobilanz von Packstoffen stand 1990, Schriftereihe Umwelt, 132, BUWAL (Bern). Harmsen, G.J.; Hinderink, A. P.; Sijben, J.; Gottschalk, A.; Schembecker, G. (1999) Industrially applied process synthesis method creates synergy between economy and sustainability, 5th International Conference on Foundations of Computer-Aided Process Design (FOCAPD’99), July 18–23, 1999, Colorado, USA. Harmsen, G. J.; Hinderink, A. P. (1999) We want less: Process intensification by process synthesis methods, 3rd Int.Conf. on Process Intensification for the chemical industry, Ed. by A. Green, Publication No. 38, Antwerp, Belgium, October. 25–27., 1999, pp. 23-28. Hinderink, A. P.; Van der Kooi, H. J.; De Swaan Arons, J. (1996a) De exergie rekening: Veel chemische processen staan in het rood, Chemisch Magazine, 5, 191–193 (in Dutch). Hinderink, A. P.; Kerkhof, F. P.J. M.; Lie, A. B. K.; De Swaan Arons, J.; Van Der Kooi, H. J., (1996b) Exergy analysis with a flowsheeting simulator, Part 1: Theory; calculating exergies of material streams, Chem. Eng. Sci. 51(20), 4693–4700. Hinderink, A. P.; Kerkhof, F. P. J. M.; Lie, A. B. K.; De Swaan Arons, J.; Van Der Kooi, H. J. (1996c) Exergy analysis with a flowsheeting simulator: Part 2, Application: Synthesis gas production from natural gas, Chem.Eng. Sci., 51(20), 4701–4715. Lowenheim, F. A.; Moran, M. K. (1975) Industrial Chemicals, Wiley & Sons (New York). Pagani, G. (1982) New process gives urea with less energy, Hydrocarbon Process., 11, 87–91. Supp, E. (1985) Improved methanol production and conversion technologies, Energy Progr. 5(3), 127–130. Szargut, J.; Morris, D.R.; Steward, F. R. (1988) Exergy analysis of thermal, chemical, and metallurgical processes, Hemisphere Publishing Corp. (New York). Ullmann’s Encyklopaedie der technischen Chemie, 1981, band 20, Verlag Chemie (Weinheim). Wall, G. (1977) Exergy-a useful concept within resource accounting, Report No. 77-42, Institute of Theoretical Physics, Chalmers Univ. of Technology, Göteborg, Sweden. Wall, G. (1988) Energy flows in industrial processes, Energy 13(2), 197–208. Mr. A.P. Hinderink is currently a process engineer/consultant with Process Design Center B.V. where he is involved in many industrial process synthesis projects. He holds a Bachelor of Chemical Technology degree and a Master of Technological Design degree from the Delft University of Technology. The foundation of this paper was laid in 1994, when he worked with the group of professor De Swaan Arons. After a career of nearly 20 years in the industry, Jakob de Swaan Arons became a full professor at the Delft University of Technology. His specific interests are phase equilibria at high pressures and in the critical region, and the efficiency and sustainability of industrial processes. Hedzer van der Kooi has worked for more than 30 years in the Applied Thermodynamics and Phase Equilibria group within the Department of Chemistry at the Delft University of Technology. After several studies on phase equilibria, for example for natural gas processing, urea production and reversed micelles, he worked for about 10 years on process improvement and recently also on the definition and realisation of sustainability.