Incorporating Environmental Impact Assessment into Conceptual Process Design: A Case Study Example Jeffrey R. Seaya and Mario R. Edenb a Department of Chemical and Materials Engineering, University of Kentucky, Paducah, KY 42002;
[email protected] (for correspondence) b Department of Chemical Engineering, Auburn University, Auburn, AL 36849 Published online 10 December 2008 in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/ep.10328
The purpose of this contribution is to present a case study illustrating how environmental impact assessment using the U.S. EPA Waste Reduction Algorithm has been incorporated into conceptual process design. This will be achieved via the introduction of a simple methodology for integrating environmental impact assessment into the standard heuristics of conceptual process design to ensure that environmental and sustainability goals are met while maintaining economic viability. The case study presented to illustrate the methodology is based on a process to manufacture industrially important C3 compounds from the dehydration of bio-based glycerol. The selected case study is important from the perspective of sustainability because glycerol is produced as a byproduct of the manufacture of biodiesel from vegetable oils. Ó 2008 American Institute of Chemical Engineers Environ Prog, 28: 30–37, 2009
Keywords: environmental impact assessment, WAR algorithm, process design, glycerol dehydration
INTRODUCTION
The impacts of chemical processes on the environment are becoming increasingly important to industry and to the general public. This research will illustrate how environmental impact assessment has been integrated into the standard heuristics of conceptual Ó 2008 American Institute of Chemical Engineers
30 April 2009
process development by use of a simple methodology. The developed methodology is illustrated using a case study example based on the manufacture of industrially important C3 products from bio-based glycerol. By integrating environmental impact assessment into the standard design heuristics used to develop and screen potential conceptual processes, the designer can ensure that the resulting process is not only optimized in terms of overall performance, but also is based on minimizing the environmental impact of the process.
PREVIOUS WORK IN PROCESS DESIGN AND INTEGRATION
This research will focus on the development of a structured approach to the integration of sustainable design principles, environmental impact assessment, and laboratory experimentation into conceptual process design. Early work on the development of a structured approach to process design based on hierarchical techniques and standard design heuristics has been described by Douglas [1] in his book, Conceptual Design of Chemical Processes. This concept has been further advanced in widely used process design textbooks by Seider [2] and Biegler et al., [3]. In general, these techniques rely on a knowledge-based approach. Although these techniques can be applied quite successfully to systems based on well-established chemistry, they do not address the gathering of
Environmental Progress & Sustainable Energy (Vol.28, No.1) DOI 10.1002/ep
the experimental data needed to design a process based on newly developed chemistry. The need for the inclusion of laboratory experimentation has begun to be addressed in other ongoing work in this area by Kiss et al. [4], regarding the linking of experiments to modeling in biodiesel production. Kiss et al. points out that the disconnect between the chemists who typically carry out the kinetic experiments and the design engineers who must use the data to develop conceptual process models can lead to design failures because the conditions considered in the lab may not lead to feasible conceptual processes. Another key component of process design is process integration. Process integration is a holistic approach to process design that emphasizes the unity of the process [5]. El-Halwagi defines three key components of process integration: process synthesis, process analysis, and process optimization. Energy integration, an important subset of process integration, has been introduced in recent years. Energy integration has been defined by El-Halwagi [4] as: A systematic methodology that provides a fundamental understanding of energy utilization within the process and uses this understanding in identifying targets and optimizing heat-recovery and energy-utility systems. This systematic methodology is used to determine the optimum utilization of heating and cooling energy within a process. Thermal pinch analysis is the principle tool for determining this optimum. Recently, the idea of system-based environmental management has been proposed [6]. This idea is based on the fusion of chemical engineering principles with the tools of other disciplines, such as environmental sciences, toxicology, and economics. Each of these tools is used in different ways to assess the economic performance of a process. The research presented will draw on this previous work and incorporate it into a cohesive methodology that integrates safety, environmental impacts, sustainability, and economic considerations into the standard heuristics of conceptual process design. The application of this proposed methodology will be illustrated using a process development case study.
EVALUATING POTENTIAL ENVIRONMENTAL IMPACTS
For the case study presented, the tool used to calculate the potential environmental impact, PEI, of the proposed conceptual processes is the Waste Reduction (WAR) algorithm, developed by the U.S. Environmental Protection Agency [7, 8]. The PEI of a given quantity of material or energy can be defined as the effect this material or energy would have if it were emitted directly to the environment [7, 8]. For the purposes of this study, only the PEI of the streams leaving the process is considered. It is assumed that each of the conceptual process design alternatives would subject to the same regulatory environmental constraints, so the PEI after waste treatment would not provide a clear picture of the true impact of the production process itself.
There are some important limitations to using the WAR algorithm for an analysis of this type. Although the WAR algorithm provides an effective means of comparing process options with a similar basis, it cannot evaluate the impacts of switching from a crude oil to a renewable, biomass derived feed stock. Therefore one should use caution when directly comparing the PEI calculated for cases based on utilizing sustainable feed stocks with that calculated for processes based on utilizing crude oil derived feed stocks. The use of biomass derived feed stocks have significant advantages regarding carbon dioxide generation, therefore their use as a feed stock has a potentially significant advantage over crude oil derived feed stocks in terms of environmental impact. Because this analysis does not include a complete lifecycle assessment, any benefit gained by this switch in feed stocks will not be captured.
SUSTAINABILITY AND PROCESS DESIGN
Often the criteria used in developing conceptual options for a proposed process is based solely on its economic viability in terms of return on investment. Although the importance of economics should not be minimized, the environmental impacts of a new process should not be ignored. There are a multitude of benefits for minimizing the environmental impacts of a process: Potential new environmental regulations. Emissions trading opportunities. Potential increased costs of ‘‘after the fact’’ means of emissions abatement. By taking an integrated approach, an optimized design that considers viability in terms of both economic performance and minimum environmental impacts can be proposed. A flow chart has been generated describing the general steps included in conceptual process design activities. This flow chart is illustrated in Figure 1. In this generalized flowchart, the basic steps from the selection of the initial chemistry to final optimized conceptual design are included. The inclusion of environmental impact assessment and determination of economic viability are included in Steps 2 and 4, application of simulation and optimization methodology. This additional methodology is described in a second flowchart, which is illustrated in Figure 2. In this flowchart, the simulation and modeling activities of process design are grouped with environmental impact assessment. This is done to ensure that design choices that lead to unacceptable results with regard to potential environmental impacts are eliminated at an early stage. Additionally, by applying this methodology, process risk is also addressed. This risk is not simply limited to process safety, but also included the risk of potential releases to the environment because of upset conditions. By incorporating this methodology into the process design hierarchy as illustrated in Figure 1, the final conceptual will be not only economically optimized, but also require fewer external safeguards to
Environmental Progress & Sustainable Energy (Vol.28, No.1) DOI 10.1002/ep
April 2009 31
Figure 1. Flowchart of proposed process integration
methodology.
minimize the environmental impacts, because of both normal operation and upset conditions. This proposed methodology covers the entire range of activities included in conceptual process design. This methodology has been applied to a process development project based on the sustainable manufacture of specialty chemical products from bio-based glycerol. This process development project will be presented as a case study on the application of the proposed methodology to conceptual process design. In particular, the portion of the methodology focused on environmental impact assessments, Step F in Figure 2, will be described in detail. PROCESS DEVELOPMENT CASE STUDY
A Growing Market for Biodiesel In recent years, much attention has been focused on the impact of human activities on the global environment. The effects of using of crude oil derived fuels and feed stocks on the global climate, and the importance of pursuing the use of sustainable raw material feed stocks has been well documented [9, 10]. In terms of energy production, biodiesel is the only alternative with an overall positive lifecycle energy balance [11]. Therefore use of biodiesel and its byproducts may have a positive impact on global climate change. In addition, according to the U.S. Department of Energy’s 2003 World Energy Report, at 32 April 2009
Figure 2. Simulation and optimization methodology
flowchart.
current rates of consumption, crude oil reserves may be depleted in 80 to 120 years [12]. This provides an incentive for replacing crude oil derived fuels sources with sustainable sources, such as biodiesel. Recent estimates predict that the demand for biodiesel will grow from 6 to 9 million metric tons per year in the United States and from 5 to 14 million metric tons per year in the European Union in the next few years [13]. However, for every 9 kg of biodiesel produced, 1 kg of crude glycerol is formed as a byproduct [14]. Beacause of its high viscosity, glycerol must be removed from the biodiesel product, thus reducing the carbon utilization. Therefore, the identification of novel industrial uses for this glycerol is important to the economic viability of biodiesel [15]. Case Study Background The process investigated in this research is the catalytic dehydration of glycerol using an acid catalyst. The reason why this process has been chosen is that glycerol dehydration can be used as a renewable feedstock for industrial C3 chemical products that are
Environmental Progress & Sustainable Energy (Vol.28, No.1) DOI 10.1002/ep
Figure 3. Glycerol dehydration process block flow diagram.
currently manufactured via the catalytic partial oxidation of propylene. Some of these products include, 1,2-propanediol, 1,3-propanediol, acrolein, hydroxyacetone, and acrylic acid [14, 16–18]. The process for each of these products is similar, differing only with regard to the feed composition, catalyst, and reactor operating conditions. The process described in this contribution is generally applicable to any of these products. Previously published work on this research project has presented how an optimized conceptual process was developed based on the glycerol dehydration reaction [19–21]. In the process considered for this research, the primary feedstock is considered to be crude glycerol from a biodiesel production process. This crude glycerol will contain some fraction of water, which must be considered in the process design. Additionally, the previous research has indicated that dilute solutions of glycerol lead to higher reactor conversions than do more concentrated feed
solutions. In this process, inert nitrogen will be used to dilute the glycerol. A block flow diagram of the proposed conceptual process for glycerol dehydration is illustrated in Figure 3. This glycerol dehydration process will be used to illustrate how environmental impact assessment has been incorporated into conceptual process design. However, to have a point of comparison, a conceptual process design for the propylene based process must also be developed. A block flow diagram for a typical propylene based process is illustrated in Figure 4 [22, 23]. Reactor Optimization Results Based on the glycerol dehydration process previously described, laboratory experiments were carried out to determine the optimum operating conditions. The integration of laboratory experiments with process simulation is an important part of the process
Environmental Progress & Sustainable Energy (Vol.28, No.1) DOI 10.1002/ep
April 2009 33
Figure 4. Propylene partial oxidation process block flow diagram.
integration methodology illustrated in Figure 1. A laboratory scale model of the glycerol dehydration reactor was constructed to conduct the experiments. A schematic of this reactor is illustrated in Figure 5. Statistical Design of Experiments (DOE) techniques were applied to determine the laboratory conditions at which to operate the glycerol dehydration reactor. Three variables were identified that had a significant influence on the reaction yield: Glycerol to water ratio in the crude glycerol feed, space velocity in the reactor, and operating temperature. Using a BoxBehnken technique, 13 experimental points were chosen. Based on the results of the analysis of the experimental results, six additional experimental points were added. These results were then analyzed to determine the conditions leading to the maximum reaction yield. In general, it was found that lower values for operating temperature, space velocity, and glycerol to water ratio led to higher yield. 34 April 2009
Overall Process Performance Results To select the final operating conditions for the glycerol dehydration process, the performance of the entire process was evaluated for each of the experimental cases. Considerations such as operating risk, as evaluated by a process hazard analysis, energy utilization, capital costs, operating costs, and environmental impacts are all part of the assessment metrics of a proposed conceptual process and are included in determining the process performance index. Each of these considerations is evaluated as part of the simulation and optimization methodology illustrated in Figure 2. Using the process simulation to model each of the experimental cases, the overall performance index was determined. To accomplish this, each case was put on an equal basis in terms of raw material feed rate. However, putting all cases on an equal basis for raw material feed leads to a different product rate for
Environmental Progress & Sustainable Energy (Vol.28, No.1) DOI 10.1002/ep
Figure 5. Schematic of laboratory equipment used in case study example.
Figure 6. PEI plot of GLY : WTR Ratio versus space-
Figure 7. PEI plot of space-velocity versus tempera-
velocity.
ture.
each case. To address the discrepancy in the mass and energy balances caused by different product rates, the results for the each of the cases were calculated per mass of product. An analysis on each of the process options including heat integration based on thermal pinch analysis, the implementation of inherently safe design practices and an assessment of the operating risk, as described in the simulation and optimization methodology illustrated in Figure 2 was conducted. From the results of the analysis of the overall process, including separation systems and waste treatment systems, the trends of which conditions lead to a minimization of the performance index function were evaluated. The trends indicate that, as with the results of the
product yield, the optimal performance occurs at the lowest values of temperature and flow rate, however, unlike the yield results, the optimal performance occurs at the highest value of glycerol : water ratio. At this point, an analysis of the potential environmental impacts can be conducted. As illustrated in the simulation and optimization methodology, this analysis is completed after the optimization of the overall process performance. This is done to ensure that the chosen optimum process is not based on operating conditions that lead to poor environmental performance, relative to the performance of the other process options. If the optimized process does vary significantly from the other process options, modification can be made during conceptual process design
Environmental Progress & Sustainable Energy (Vol.28, No.1) DOI 10.1002/ep
April 2009 35
These results illustrate that the region chosen as the performance optimum, that is, the region of low temperature, low space velocity and high glycerol : water ratio is in the region of low environmental impacts. Although the point corresponding to the performance optimum is not the minimum PEI point, it is close to the minimum. Because of the level of accuracy of the PEI calculation, this is deemed to be an acceptable result. Only if the performance optimum were far from the minimum PEI would further study be warranted. CONCLUSIONS
Figure 8. PEI plot of GLY : WTR ratio versus tempera-
ture.
to reduce its environmental impacts, without degrading its overall process performance. Potential Environmental Impact Results As described in the simulation and optimization methodology illustrated in Figure 2, the potential environmental impacts for each proposed conceptual process must be considered. The WAR algorithm is a good tool to use for this task because it is simple to apply though a spreadsheet with an interface to a process simulator. Although the WAR algorithm is not the only available measure of environmental performance, it was the best option for use during the conceptual phase of chemical process development for two important reasons. First, it provides a broad ‘‘snapshot’’ of environmental performance over a wide range of indicators. The WAR algorithm is a measure of 8 environmental indicators: Human toxicity: ingestion; human toxicity: dermal/inhalation; aquatic toxicity; terrestrial toxicity; global warming; ozone depletion; photochemical oxidation; and acidification [7, 8]. This broad spectrum makes the WAR algorithm applicable to a wide range of chemical processes. Secondly, the WAR algorithm is easy to use with regard to evaluating process options at the conceptual design stage. Once conceptual design has begun, radical changes to processes chemistry are usually not possible. Therefore, the final PEI value calculated using the WAR algorithm is less important than the relative difference among several potential process options. This makes it an appropriate tool to use in conceptual process design. Because a process simulation was prepared for the proposed conceptual process, determining the input values for the application of the WAR algorithm was simple. The simulation package used to model the process for this conceptual design case study is Aspen Plus [24]. By using the simulation data import functionality within the WAR algorithm graphical user interface available for download from the US EPA, the PEI of each of the reactor operating cases could quickly be evaluated. These results are illustrated in Figures 6–8. 36 April 2009
In conclusion, this research has illustrated how the previous work done in the field of process design and integration can be expanded to include sustainability and an assessment of potential environmental impacts. A case study example has been used to illustrate how environmental impact assessment can be incorporated into the standard heuristics of conceptual process design. By applying the methodology presented in this work, it has been shown that environmental impact assessment can be made an integral part of conceptual process design. The ability to include environmental impacts into the overall evaluation of conceptual process options will only become more important as energy costs increase and additional scrutiny from the public is applied to chemical processes. Therefore, by including the potential environmental impacts into the optimization process, it is possible to choose designs that are optimized for all aspects of process performance, not just for economic viability. ACKNOWLEDGMENTS
Funding and facilities for this research were provided by the Evonik Degussa GmbH Health and Nutrition Business Unit. Additional equipment was provided by the University of South Alabama. The authors would like to also thank Prof. Tom Thomas of the University of South Alabama and Evonik Degussa intern students Maria Schley, Astrid Roesner, Mareike Schaum, Stephan Adelmann, and Holger Werhan for their contributions to this research. LITERATURE CITED
1. Douglas, J.M. (1988). Conceptual design of chemical processes, New York: McGraw-Hill. 2. Seider, W.D. (1999). Process design principles, New York: Wiley. 3. Biegler, T., Grossmann, I.E., & Westerberg, A.W. (1997). Systematic methods of chemical process design, Prentice-Hall: Upper Saddle River, NJ. 4. Kiss, A.A., Alexandre, C.D., & Rothenberg, G. (2006). Linking experiments to modeling in biodiesel production. In W. Marquardt & C. Pantelides (Eds.), Computer-aided chemical engineering, Amsterdam, The Netherlands: Elsevier, (Volume 21A, pp. 731– 736). 5. El-Halwagi, M.M. (1997). Pollution prevention through process integration: systematic design tools, Academic Press: San Diego, CA. 6. Clift, R. (2005). Sustainable development and its implications for chemical engineering, Chemical Engineering Science (Volume 61, pp. 4179–4187).
Environmental Progress & Sustainable Energy (Vol.28, No.1) DOI 10.1002/ep
7. Young, D.M., & Cabezas, H. (1999). Designing sustainable processes with simulation: The waste reduction (WAR) algorithm. Computers and Chemical Engineering, 23, 1477–1491. 8. Young, D., Scharp, R., & Cabezas, H. (2000). The waste reduction (WAR) algorithm: Environmental impacts, energy consumption, and engineering economics, Waste Management, 20, 605–615. 9. Hoel, M., & Kverndokk, S. (1996). Depletion of fossil fuels and the impacts of global warming, Resource and Energy Economics, 18, 115–136. 10. Vicente, G., Martinez, M., & Aracil, J. (2004). Integrated biodiesel production: A comparison of different homogeneous catalysts systems, Bioresource Technology, 92, 297–305. 11. Buczek, B., & Czepiriski, L. (2004). Applicability of used rapeseed oil for production of biodiesel, Inform, 15, 3. 12. U.S. Department of Energy (DOE). (2003). World energy report, Washington, DC. 13. Blume, A.M., & Hearn, A.K. (2007). The evolution of biodiesel, Biofuels, 20–23. 14. Chiu, C.-W., Dasari, M.A., & Suppes, G.J. (2006). Dehydration of glycerol to acetol via catalytic reactive distillation, AIChE Journal, 52. 15. U.S. Department of Agriculture (USDA). (2003). EU: Biodiesel industry expanding use of oilseeds, Washington, DC. 16. Antal, M.J., Jr., Mok, W.S.L., Roy, J.C., Raissi A.T., & Anderson, D.G.M. (1985). Pyrolytic sources of hydrocarbons from biomass, Journal of Analytical and Applied Pyrolysis. 17. Neher, A., Haas, T., Arntz, D., Klenk, H., & Girke, W. (1995). Process for the production of acrolein, United States Patent, 5,387,720.
18. Ramayya, S., Brittain, A., DeAlmeida, C., Mok, W., & Antal, M.J., Jr. (1987). Acid-catalysed dehydration of alcohols in supercritical water, Fuel, 66, 1364– 1371. 19. Seay, J., Eden, M., D’Alessandro, R., & Weckbecker, C. (2006). Sustainable production of industrial chemical products from bioresources. In W. Marquardt, & C. Pantelides (Eds.) Computer-aided chemical engineering, Amsterdam, the Netherlands: Elsevier, (Volume 21A, pp. 961– 966). 20. Seay, J., Eden, M., D’Alessandro, R., Thomas, T., Redlingshoefer, H., Weckbecker, C., & Huthmacher, K. (2007). Integration of process modeling with laboratory experiments in conceptual design: Bio-based glycerol dehydration case study. In V. Plesu & P.S. Agachi (Eds.), Computer-aided chemical engineering, Amsterdam, the Netherlands: Elsevier, (pp. 485–491). 21. Seay, J., Werhan, H., Eden, M., D’Alessandro, R., Thomas, T., Redlingshoefer, H., Weckbecker, C., & Huthmacher, K. (2008). Integrating laboratory experiments with process simulation for reactor optimization, Computer Aided Chemical Engineering (Paper 168). 22. Weigert, W.M., & Haschke, H. (1976). Acrolein and derivatives. In J. McKetta, (Ed.), Encyclopedia of chemical processing and design, New York: Marcel Dekker, Inc. 23. Etzkorn, W.C., Pedersen, S.E., & Snead, T.E. (2001). Acrolein and derivatives. Kirk-Othmer encyclopedia of chemical technology, New York: Wiley, (pp 232–251). 24. Aspen Technology. (2006). Aspen Engineering Suite Users Manual, Cambridge, MA.
Environmental Progress & Sustainable Energy (Vol.28, No.1) DOI 10.1002/ep
April 2009 37
