- Email: [email protected]
Dynamic System for Life Cycle Inventory and Impact Assessment of Manufacturing Processes
Available online at www.sciencedirect.com
ScienceDirect Procedia CIRP 15 (2014) 531 – 536
21st CIRP Conference on Life Cycle Engineering
Dynamic System for Life Cycle Inventory and Impact Assessment of Manufacturing Processes Remo A. P. Filletia*, Diogo A. L. Silvab, Eraldo J Silvaa, Aldo R. Omettob Laboratory for Advanced Processes and Sustainability (LAPRAS), University of São Paulo, Av. Trabalhador São Carlense, 400, São Carlos - 13566Ǧ690, Brazil b Laboratory of Operations Management, University of São Paulo, Av. Trabalhador São Carlense, 400, São Carlos - 13566Ǧ690, Brazil
a
* Corresponding author. Tel.: +55-16-3373-8645 ; fax: +55-16-3373-9425. E-mail address: [email protected]
Abstract Life cycle impact assessment is a crucial phase of the life cycle assessment (LCA) technique, once it identifies and quantifies the environment impacts of products/processes life cycle. This phase, however, is performed in a static way, becoming a bottleneck on a dynamic LCA of manufacturing processes. This paper presents a method to perform a web based dynamic life cycle inventory and impact assessment for manufacturing processes, in order to improve the data quality and accuracy of environmental aspects and potential impacts. The method implementation is based on combining Labview and GaBi softwares, as well as MTConnect® standard.
© 2014 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
© 2014 The Authors. Published by Elsevier B.V. (http://creativecommons.org/licenses/by-nc-nd/3.0/). Selection and peer-review under responsibility of the International Scientific Committee of the 21st CIRP Conference on Life Cycle Selection and peer-review under responsibility of the International Scientific Committee of the 21st CIRP Conference on Life Cycle Engineering in the person of the Conference Chair Prof. Terje K. Lien.
Engineering in the person of the Conference Chair Prof. Terje K. Lien
Keywords: Sustainable Manufacturing; Environmental Performance Assessment; Life Cycle Impact Assessment; Manufacturing Process.
1. Introduction Green manufacturing is part of the sustainable manufacturing and can be defined as a form of pollution prevention that integrates environmental considerations in the production of goods, employing environmentally-friendlier manufacture processes, conserving energy and natural resources, and reducing negative environmental impacts (Dornfeld 2013). Advances towards green manufacturing have been achieved through better management practices based on the use of environmental management tools such as the Life Cycle Assessment – LCA (Garetti and Taisch 2012). LCA is very important on the identification of hotspots during the manufacture production phase and on the development of cleaner production strategies. An LCA study is composed of four phases: definition of goal and scope, life cycle inventory analysis (LCI), life cycle impact assessment (LCIA) and interpretation (ISO 2006a,b).
Besides its importance and benefits, the use of LCA to subsidize the green manufacturing has limitations. Finnveden et al. (2009) and Thorn et al. (2011) pointed out the problems associated with the uncertainties about LCI data, which are collected on an aggregated way (i.e. black-box processes), and also by a static way (i.e. they are not updated during production real-time). For Thorn et al. (2011) the companies that have interest on LCA shall perform the LCI on a dynamic way in order to decrease the uncertainties. In other words, it is necessary to collect the inventory data directly from the productive processes, and in real-time. Santos et al. (2011) gave an example that the electric power consumption on manufacturing processes is not constant, what, by itself, provides a motivation to manage this consumption at real-time basis. Kellens et al. (2011) proposed a global methodology called CO2PE! Institute UPLCI (Unit Process Life Cycle Inventory), which provides a systematic orientation to perform an LCI for manufacturing unit processes. The methodology was proposed in order to standardize and to reduce the uncertainties of LCI
2212-8271 © 2014 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/). Selection and peer-review under responsibility of the International Scientific Committee of the 21st CIRP Conference on Life Cycle Engineering in the person of the Conference Chair Prof. Terje K. Lien doi:10.1016/j.procir.2014.06.024
532
Remo A.P. Filleti et al. / Procedia CIRP 15 (2014) 531 – 536
data. However, this methodology does not include the LCIA and Interpretation phases of a complete LCA study. Besides, the life cycle perspective is limited by the use phase, do not accounting the others life cycle phases. In this sense, it becomes increasingly important a higher independence level of LCI data from the LCI databases. These databases are composed by mean and aggregated processes data and in many times does not include specific characteristics of the real unit process, as the used technology and consumables, and also the consumption and emission variations during the process working steps. Accordingly, some studies in literature propose solutions in order to overcome this gap. According to Geyer et al. (2003), at manufacture level, data on all aspects of the value chain needs to be integrated and processed in real-time to support both day-to-day business processes and long-term decision-making. For green manufacturing approaches, many studies have been performed about real-time LCI for manufacturing processes, especially for measurements of electric power consumption of machine tools and components. Kara et al. (2011) described an approach to the monitoring of electrical power within manufacturing processes, classifying electricity consuming in three levels: factory, department, and unit process. Behrendt et al (2012) and Vikhorev et al. (2013) proposed an advanced energy management procedure for monitoring energy consumption data in real-time, identifying the energy used by each unit process. Sun and Li (2013) studied a method to estimate the opportunity for real-time energy control of manufacturing systems by adjusting machines to lower energy consumption levels without sacrificing system throughput. However, energy savings/reductions in manufacturing are not enough (Ahn et al. 2013), making necessary to extend research on the topic by the inclusion of the following subjects to the realtime monitoring systems: x The consumption study of other energy resources (e.g. for fossil sources) and materials, as well the local manufacturing emissions (Kellens et al. 2011; Pusavec et al. 2010); x The inclusion of the LCI dataset of the consumed and monitored resources during manufacturing processes (Narita et al. 2006; Silva et al. 2012); x The inclusion of all life cycle stages (e.g. raw material supply, rehabilitation and end-of-life strategies), in addition to the use phase (Garetti and Taisch 2012; Santos et al. 2011); x Besides the LCI, the inclusion of the LCIA and Interpretation LCA phases for the monitoring systems (Garetti and Taisch 2012; Joung et al. 2012). Therefore, the aim of this paper is to present a method to perform a web-based dynamic life cycle inventory (LCI) and impact assessment (LCIA) of manufacturing processes. The method includes the aforementioned subjects to eliminate the uncertainness related to the use of general databases when evaluating manufacture unit processes.
2. Overview of environmental impact analyzers There are few studies on the literature that developed integrated systems for collection and evaluation of environmental performance of manufacturing processes based on LCA. Ramani et al. (2010) explained that the ability to characterize environmental impacts of manufacturing processes via LCA is limited, due the deficiencies on collecting accurately LCI data at real time basis from the shop-floor. This limitation has been a strong motivator for many researchers to propose environmental impact analyzers of manufacturing processes. Narita et al. (2006) pointed out the need of integration between a dynamic LCI data collecting of manufacturing processes and LCA software and databases, coming up with an evaluation system of environmental burden for machining operation. The system is composed of an estimator, a database and an analysis block. The LCI dataset evaluated included electric consumption of machine tool components, lubricant oil, cutting fluid and other factors. Environmental impacts to global warming potential were automatically calculated and in real-time by system developed. However, the authors do not evaluated the whole process life cycle, and focused only on Other impact the use phase of machining operation. categories besides the global warming potential were not studied. Jiang et al. (2012) developed an environmental performance analyzer of manufacturing processes based on LCA. The method was encoded in a software (Manufacturing Environmental Performance Evaluation – MEPE) to provide insights into how manufacturing decisions affect the environment. The MEPE consists of several modules: a measurement normalizer (by a functional unit), a weight calculation module based on analytic hierarchy process (AHP) technique), and an impact matrix forming (inventory data is ranked into categories i.e. raw material consumption, airborne emissions, liquid waste, so on). The following supporting databases are included: process inventory (a total of 24 processes, such as casting, forming and shaping, machining, etc.), standard evaluation criterion and knowledge database. Although LCIA phase is not included in the software, the LCA interpretation phase is presented, outputting quantitative scores for each criterion and operation in the process plan. Results can be detailed to determine the worst environmental criterion and the poorest performing manufacture process. 3. Description of the proposed dynamic life cycle inventory and impact assessment method In order to perform a web-based dynamic LCI and LCIA of manufacturing processes, a dynamic data collection step of the resource consumption and local emissions per unit process is mandatory. In addition, the chosen platform to categorize the acquired data must be vendor-neutral and has to be simple, general and usable over the Internet. The MTConnect ® standard (Sobel 2012) (AMT 2011) categorizes the shop-floor data in an Extensible Markup Language (XML), encoding the data in a way that both machines and humans can read. Born to be for the shop-floor the same breakthrough as USB
Remo A.P. Filleti et al. / Procedia CIRP 15 (2014) 531 – 536
standard is for computer peripherals in terms of compatibility, MTConnect improves the easiness of data gathering from machine tools among different proprietary CNC systems. Major benefits are the reduction of automation efforts and system scalability. In addition, gathering data in a XML format increases the potential for integration with other web servers, as well as LCI and LCIA software, such as Gabi. Figure 1 introduces the overview of the proposed method with two subsystems: manufacturing unit process and client application.
533
3.3. Client Application and Inventory and Impact Databases The Client Application is the key software on the proposed method. It performs the interconnection between the categorized and collected data from each unit process with the background LCA databases, providing the LCIA results of the emissions and consumed resources. Dynamic information about the potential environmental impacts of the unit process is shown according to the chosen LCIA method (e.g. CML, EDIP, ReCiPe, USETox, etc.) and impact categories under interest.
3.1. Manufacturing Unit Process 3.4. User Interface The manufacturing unit process represents the process to me monitored. Data related to in loco resources consumption (e.g. electricity, compressed air, cutting fluid) and to emissions (e.g. mist and noise) has to be measured and stored. If measuring devices are not MTConnect complaints, an adapter has to be developed to translate the output signals into the standard format. In the proposed system, the Adapter is installed in a local computer, integrated to the machine tool being monitored. The local computer is connected with the Ethernet and reachable through TCP/IP protocol. Local adapters can be eliminated if measuring devices installed are already compatible with the MTConnect standard. 3.2. Agent Software The Agent software is hosted at the main server and manages the information from all process devices presented in one or more unit processes. This software also collects, arranges, and stores the updated data from the adapters (Sobel 2012) (AMT 2011). Output data is structured according to their physical units (i.e. kWh for energy, m3 for air volume, liter for cutting fluid, dB for noise emission), time, source instrument/measurement device and process unit label. The Agent also performs the communication with the Client Application software, providing data as XML file according to the standard, when requested.
The User Interface has access to the dynamic LCI and LCIA results generated by the Client Application through the Ethernet/Internet connection. Different levels of data access and report generation can be provided, by defining user rules according to the user hierarchy. 4. System implementation for a machining process evaluation To demonstrate the potential of the dynamic method, a laboratory setup of a cylindrical plunge grinding for a shaft external diameter was selected as a case study. . The shafts were made by mild steel (SAE 4140) and were machining by turning in a previous operation. The required steps to implement the method are described as follow below. 4.1. Grinding Unit Process Measuring devices were coupled in a grinding machine ZEMA G-800HS (Fanuc 180i CNC controller) to acquire the overall machine electric energy consumption (electric multivariable meter), compressed air consumption (air flow meter), cutting fluid consumption (fluid flow meter) and noise emission (decibel meter).
Fig. 1. Overview of the dynamic LCI and LCIA method
534
Remo A.P. Filleti et al. / Procedia CIRP 15 (2014) 531 – 536
The energy meter was installed at the main power supply of the machine tool, acquiring its electric information, such as power, power consumption, power factor, voltages and currents. The device has a RS-485 serial output, which can be converted to a RS-232 or USB by an adapter device. The air flow meter was installed to the main compressed air supply and its measures can be read by an electric current output (4-20 mA). The fluid flow meter was installed to the fluid tank output and its measures are shown through an indicator device at the machine. Its measures also can be acquired by a 0-10V electric voltage output from the indicator device. The noise meter was installed close to the machine tool, inside the worker perimeter during operation. Its measures can be acquired by a 0-10V electric voltage output. The hardware device for the data collection step was composed by a data acquisition kit (NI cDAQ – National Instruments), and a serial RS-485/RS-232 converter. The Adapter software was developed using the Labview software, once the measuring devices were not MTConnect compliant.
in kg, and ADP, in kg Sb-eq). ADP and GWP results were obtained based on the CML (2001) method. Table 1. Electric power consumption Electric Power Consumption
Consumption (MJ/worked piece)
Real Grinding Unit Process
0,221
Standard Grinding Unit Process
1,112
4.2. Agent Software, Databases, Client Application and User Interface Agent software was the most recent release of MTConnect® Institute, MTConnect C++ Agent Version 1.2.0.18. The LCI and LCIA results were accessed through the system model generated from the GaBi software for the case studied. Different LCI databases as well LCIA methodologies can be integrated according to the goal and scope definition. Besides, Gabi has import and export data functions by XML files (e.g. ecoSpold format). However, GaBi presents a static and manual structure of LCI and LCIA data generation, that could not be used, by itself, as the Client Application. One solution for this issue would be the development of auxiliary software, which would automate GaBi data income, generation and outcome. 4.3. Dynamic and Static Data Acquisition Differences In order to demonstrate the differences between a dynamic and a static LCA of unit processes, two different scenarios were evaluated. The first scenario was to conduct the dynamic LCA of the grinding machining using the proposed method as shown in Figure 1. The second scenario was the static LCA of a general grinding machine process available at the GaBi software professional database version 6.0. These scenarios were compared for the electric power consumption aspect, from the Brazilian electricity mix, and assuming a grinding removal amount of 5,82 g of metal SAE 4140 per processed piece. Table 1 presents the inventory results of electric power consumption for the two scenarios analyzed and Figure 2 shows the results of water consumption and LCIA comparing the two cases in terms of global warming potential (GWP, in kg CO2-eq.), and abiotic depletion potential (Water ,
Fig. 2. Environmental performance assessment of the two scenarios analyzed.
Both LCIA and LCI results of the two systems showed considerable differences, which can be explained by the aspects and considerations adopted on each system when a dynamic and a static LCA is performed. The data acquired from the real time system consisted on a measurement of all the operation modes of a cylindrical plunge grinding cycle, including the starting, warm-up, standby, grinding and dressing modes. It also included the fluid refrigeration control (extra cooling).
Remo A.P. Filleti et al. / Procedia CIRP 15 (2014) 531 – 536
The time study of these operation modes were also acquired, for one grinding cycle and then extrapolated for a working day of 16h, considering that there is no time for maintenance and that the machine have to be shut down, initialized and warmed up once every day. From this perspective, it was calculated, as described by the UPLCI Methodology (Kellens et. al 2011), the normalized power consumption for one second of the grinding machine operation (Table 2). Table 2. Power consumption of 1s of machine operation Operation Mode
Time of Operation (s/day)
Electric Energy (MJ/day)
Starting Mode
1800
1,044
Warm-Up Mode
3000
19,622
Stand-By Mode
24000
124,330
Dressing Mode
4800
67,0134
Grinding Mode
24000
331,227
Extra Cooling
-
93,968
Total Normalized
16 h
0,011 MJ/s
535
consumption as well as environmental impacts potential were totally different for both systems due to the different limitations and assumptions discussed when conducting a dynamic and a static LCA of unit processes. Besides, the proposed acquisition system (i.e. the devices and software) can easily be installed on almost all types of CNC machining tools, once its installation process has a low level of complexity and does not require modifications at the machine CNC or PLC (Programmable Logic Controller). This embracing aspect associated with the use of the MTConnect protocol allows the standardization and collection of shopfloor machines data, as well their availability on the internet. This can be an important way, for instance, to make possible an online green manufacturing management of systems and to show relevant manufacture environmental indicators. This web based dynamic method under development can enable the generation of LCA results at different levels (unit process, production cells, product production, whole plant, whole product life cycle) and perspectives (unit process optimization, industrial symbiosis, formulate best practices and Ecodesign guidelines for machine tools, etc. ). Acknowledgements
This normalized consumption was multiplied by the time spend to perform a grinding operation at the worked piece (20 seconds). It is important to highlight that the laboratory setup acquired the information of only one kind of grinding operation (cylindrical plunge grinding). The standard grinding data acquired from the inventory GaBi software database, however, hold different aspects and considerations. They represent a mean data from different kinds of machines and grinding processes, as well different operational modes and periods of measurement. It also has different inputs compared with the real time system. Besides the removed material in both systems was the same, the electricity consumption (Table 1) as well the results of environmental performance (Figure 2) were totally different for each system because of the different approaches previously discussed when conducting a dynamic and a static LCA of unit processes. . This fact emphasizes that in some cases the use of a mean data from standard LCA databases may not represent in a satisfactory way the reality of manufacturing processes monitored in real-time. 5. Conclusions The data quality and the accuracy of an LCA study is of paramount importance for the reliability of results and can interfere significantly the decision-making process for green manufacturing purposes. . The proposed dynamic method, in development process provided a good idea on how to conduct an LCA of unit processes in real-time. The developed dynamic method was applied to a real case study of grinding machining and the results were compared with a conventional grinding process available on the GaBi software. The results showed that the electricity and water
The authors would like to thank the CAPES (“Coordenação de Aperfeiçoamento de Pessoal de Nível Superior”, 9331/13-1), CNPq (“Conselho Nacional de Desenvolvimento Científico e Tecnológico”) and the FAPESP (“Fundação de Amparo à Pesquisa do Estado de São Paulo”, 2013/06736-9) for supporting this research. References [1] Dornfeld AD. Green manufacturing: fundamentals and applications (green energy and technology). 1st ed. Berkeley: Springer; 2013. [2] Garetti M, Taisch M. Sustainable manufacturing: trends and research challenges. Prod Plan Control: Manag Oper 2012; 23:83-104. [3] ISO (International Organization of Standardization). 14040: environmental management – life cycle assessment - principles and framework, International Organization of Standardization, Geneva; 2006. [4] ISO (International Organization of Standardization). 14044: environmental management – life cycle assessment – requirements and guidelines, International Organization of Standardization, Geneva; 2006a. [5] Finnveden G, Hauschild MZ, Ekvall T, Guinée J, Heijungs R, Hellweg S, Koehler A, Pennington D, Suh S. Recent developments in Life Cycle Assessment. Journal of environmental management 2009; 91:1-21. [6] Thorn MJ, Kraus JL, Parker DR. Life-cycle assessment as a sustainability management tool: strengths, weaknesses, and other considerations. Environ Qual Manag 2011; 20:1-10. [7] Santos JP, Oliveira M, Almeida FG, Pereira JP, Reis A. Improving the environmental performance of machine-tools: influence of technology and throughput on the electrical energy consumption of a press-brake. J Clean Prod 2011; 19:356–364. [8] Kellens K, Dewulf W, Overcash M, Hauschild MZ, Duflou JR. Methodology for systematic analysis and improvement of manufacturing unit process life-cycle inventory (UPLCI)—CO2PE! initiative (cooperative effort on process emissions in manufacturing). Part 1: Methodology description. Int J Life Cycle Assess 2011; 17:69-78. [9] Geyer A, Scapolo F, Boden M, Döry T, Ducatel K. The future of manufacturing in Europe 2015-2020 - the challenge for sustainability 2003; Thecnical report series, EUR 20705. [10] Kara S, Bogdanski G, Li W. Electricity metering and monitoring in manufacturing systems. In: Proceedings of the 18th CIRP International
536
Remo A.P. Filleti et al. / Procedia CIRP 15 (2014) 531 – 536
Conference on Life Cycle Engineering. Technische Universität Braunschweig: Springer-Verlag; 2011. p.1-10. [11] Behrendt T, Zein A, Min S. Development of an energy consumption monitoring procedure for machine tools. CIRP Ann Manuf Technol 2012; 61:43–46. [12] VikhorevK, Greenough R, Brown N. An advanced energy management framework to promote energy awareness. J Clean Prod 2013; 43:103– 112. [13] Sun Z, Li L. Opportunity estimation for real-time energy control of sustainable manufacturing systems. IEEE Trans Automat Sci Eng 2013; 10:38–44. [14] Ahn S-H, Chun D-M, Chu W-S. Perspective to green manufacturing and applications. Int J Prec Eng Manufact 2013;14:873–874. [15] Pusavec F, Krajnik P, Kopac J. Transitioning to sustainable production – Part I: application on machining technologies. J Clean Prod 2010; 18:174–184. [16] Narita H, Kawamura H, Norihisa T. Development of Prediction System for Environmental Burden for Machine Tool Operation( כ1st Report , Proposal of Calculation Method for Environmental Burden). JSME Int J 2006; 49:1188–1195. [17] Silva EJ, Ometto AR, Rozenfeld H, Silva DAL, Pigosso DCA, Reis VRA. Prototypal implementation of a remanufacturing oriented griding machine. In: Proceedings of the 19th CIRP International Conference on Life Cycle Engineering. University of California at Berkeley: SpringerVerlag; 2012. p.257-262. [18] Joung CB, Carrell J, Sarkar P, Feng SC. Categorization of indicators for sustainable manufacturing; Ecol Indic 2013; 24:148–157. [19] Ramani K, Ramanujan D, Bernstein WZ, Zhao F, Sutherland JW, Handwerker C, Choi JK, Kim H, Thurston D. Integrated sustainable life cycle design: a review. J Mech Des 2010; 132:1-15. [20] Jiang Z, Zhang H, Sutherland JW. Development of an environmental performance assessment method for manufacturing process plans. The Int J Advanced Manufact Technol 2012; 58:783–790. [21] AMT (The Assossiation For Manufacturing Technology). ® , 2011. Available at: Getting Started with MTConnect. Access: 11/01/2013. [22] Sobel W. MTConnect® Standard Part 1 - Overview and Protocol. 1.2.0 - Final, 2012. Available at: . Access: 11/01/2013. [23] Sobel W. MTConnect® Standard Part 1 - Overview and Protocol. 1.2.0 - Final, 2012. Available at: . Access: 11/01/2013. [24] Machado AR, Abrão AM, Coelho RT, Silva MB. Theory of the Machining of the Materials: 1ed. Blucher Ltda Publishing House, São Paulo; 2009.
ScienceDirect Procedia CIRP 15 (2014) 531 – 536
21st CIRP Conference on Life Cycle Engineering
Dynamic System for Life Cycle Inventory and Impact Assessment of Manufacturing Processes Remo A. P. Filletia*, Diogo A. L. Silvab, Eraldo J Silvaa, Aldo R. Omettob Laboratory for Advanced Processes and Sustainability (LAPRAS), University of São Paulo, Av. Trabalhador São Carlense, 400, São Carlos - 13566Ǧ690, Brazil b Laboratory of Operations Management, University of São Paulo, Av. Trabalhador São Carlense, 400, São Carlos - 13566Ǧ690, Brazil
a
* Corresponding author. Tel.: +55-16-3373-8645 ; fax: +55-16-3373-9425. E-mail address: [email protected]
Abstract Life cycle impact assessment is a crucial phase of the life cycle assessment (LCA) technique, once it identifies and quantifies the environment impacts of products/processes life cycle. This phase, however, is performed in a static way, becoming a bottleneck on a dynamic LCA of manufacturing processes. This paper presents a method to perform a web based dynamic life cycle inventory and impact assessment for manufacturing processes, in order to improve the data quality and accuracy of environmental aspects and potential impacts. The method implementation is based on combining Labview and GaBi softwares, as well as MTConnect® standard.
© 2014 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
© 2014 The Authors. Published by Elsevier B.V. (http://creativecommons.org/licenses/by-nc-nd/3.0/). Selection and peer-review under responsibility of the International Scientific Committee of the 21st CIRP Conference on Life Cycle Selection and peer-review under responsibility of the International Scientific Committee of the 21st CIRP Conference on Life Cycle Engineering in the person of the Conference Chair Prof. Terje K. Lien.
Engineering in the person of the Conference Chair Prof. Terje K. Lien
Keywords: Sustainable Manufacturing; Environmental Performance Assessment; Life Cycle Impact Assessment; Manufacturing Process.
1. Introduction Green manufacturing is part of the sustainable manufacturing and can be defined as a form of pollution prevention that integrates environmental considerations in the production of goods, employing environmentally-friendlier manufacture processes, conserving energy and natural resources, and reducing negative environmental impacts (Dornfeld 2013). Advances towards green manufacturing have been achieved through better management practices based on the use of environmental management tools such as the Life Cycle Assessment – LCA (Garetti and Taisch 2012). LCA is very important on the identification of hotspots during the manufacture production phase and on the development of cleaner production strategies. An LCA study is composed of four phases: definition of goal and scope, life cycle inventory analysis (LCI), life cycle impact assessment (LCIA) and interpretation (ISO 2006a,b).
Besides its importance and benefits, the use of LCA to subsidize the green manufacturing has limitations. Finnveden et al. (2009) and Thorn et al. (2011) pointed out the problems associated with the uncertainties about LCI data, which are collected on an aggregated way (i.e. black-box processes), and also by a static way (i.e. they are not updated during production real-time). For Thorn et al. (2011) the companies that have interest on LCA shall perform the LCI on a dynamic way in order to decrease the uncertainties. In other words, it is necessary to collect the inventory data directly from the productive processes, and in real-time. Santos et al. (2011) gave an example that the electric power consumption on manufacturing processes is not constant, what, by itself, provides a motivation to manage this consumption at real-time basis. Kellens et al. (2011) proposed a global methodology called CO2PE! Institute UPLCI (Unit Process Life Cycle Inventory), which provides a systematic orientation to perform an LCI for manufacturing unit processes. The methodology was proposed in order to standardize and to reduce the uncertainties of LCI
2212-8271 © 2014 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/). Selection and peer-review under responsibility of the International Scientific Committee of the 21st CIRP Conference on Life Cycle Engineering in the person of the Conference Chair Prof. Terje K. Lien doi:10.1016/j.procir.2014.06.024
532
Remo A.P. Filleti et al. / Procedia CIRP 15 (2014) 531 – 536
data. However, this methodology does not include the LCIA and Interpretation phases of a complete LCA study. Besides, the life cycle perspective is limited by the use phase, do not accounting the others life cycle phases. In this sense, it becomes increasingly important a higher independence level of LCI data from the LCI databases. These databases are composed by mean and aggregated processes data and in many times does not include specific characteristics of the real unit process, as the used technology and consumables, and also the consumption and emission variations during the process working steps. Accordingly, some studies in literature propose solutions in order to overcome this gap. According to Geyer et al. (2003), at manufacture level, data on all aspects of the value chain needs to be integrated and processed in real-time to support both day-to-day business processes and long-term decision-making. For green manufacturing approaches, many studies have been performed about real-time LCI for manufacturing processes, especially for measurements of electric power consumption of machine tools and components. Kara et al. (2011) described an approach to the monitoring of electrical power within manufacturing processes, classifying electricity consuming in three levels: factory, department, and unit process. Behrendt et al (2012) and Vikhorev et al. (2013) proposed an advanced energy management procedure for monitoring energy consumption data in real-time, identifying the energy used by each unit process. Sun and Li (2013) studied a method to estimate the opportunity for real-time energy control of manufacturing systems by adjusting machines to lower energy consumption levels without sacrificing system throughput. However, energy savings/reductions in manufacturing are not enough (Ahn et al. 2013), making necessary to extend research on the topic by the inclusion of the following subjects to the realtime monitoring systems: x The consumption study of other energy resources (e.g. for fossil sources) and materials, as well the local manufacturing emissions (Kellens et al. 2011; Pusavec et al. 2010); x The inclusion of the LCI dataset of the consumed and monitored resources during manufacturing processes (Narita et al. 2006; Silva et al. 2012); x The inclusion of all life cycle stages (e.g. raw material supply, rehabilitation and end-of-life strategies), in addition to the use phase (Garetti and Taisch 2012; Santos et al. 2011); x Besides the LCI, the inclusion of the LCIA and Interpretation LCA phases for the monitoring systems (Garetti and Taisch 2012; Joung et al. 2012). Therefore, the aim of this paper is to present a method to perform a web-based dynamic life cycle inventory (LCI) and impact assessment (LCIA) of manufacturing processes. The method includes the aforementioned subjects to eliminate the uncertainness related to the use of general databases when evaluating manufacture unit processes.
2. Overview of environmental impact analyzers There are few studies on the literature that developed integrated systems for collection and evaluation of environmental performance of manufacturing processes based on LCA. Ramani et al. (2010) explained that the ability to characterize environmental impacts of manufacturing processes via LCA is limited, due the deficiencies on collecting accurately LCI data at real time basis from the shop-floor. This limitation has been a strong motivator for many researchers to propose environmental impact analyzers of manufacturing processes. Narita et al. (2006) pointed out the need of integration between a dynamic LCI data collecting of manufacturing processes and LCA software and databases, coming up with an evaluation system of environmental burden for machining operation. The system is composed of an estimator, a database and an analysis block. The LCI dataset evaluated included electric consumption of machine tool components, lubricant oil, cutting fluid and other factors. Environmental impacts to global warming potential were automatically calculated and in real-time by system developed. However, the authors do not evaluated the whole process life cycle, and focused only on Other impact the use phase of machining operation. categories besides the global warming potential were not studied. Jiang et al. (2012) developed an environmental performance analyzer of manufacturing processes based on LCA. The method was encoded in a software (Manufacturing Environmental Performance Evaluation – MEPE) to provide insights into how manufacturing decisions affect the environment. The MEPE consists of several modules: a measurement normalizer (by a functional unit), a weight calculation module based on analytic hierarchy process (AHP) technique), and an impact matrix forming (inventory data is ranked into categories i.e. raw material consumption, airborne emissions, liquid waste, so on). The following supporting databases are included: process inventory (a total of 24 processes, such as casting, forming and shaping, machining, etc.), standard evaluation criterion and knowledge database. Although LCIA phase is not included in the software, the LCA interpretation phase is presented, outputting quantitative scores for each criterion and operation in the process plan. Results can be detailed to determine the worst environmental criterion and the poorest performing manufacture process. 3. Description of the proposed dynamic life cycle inventory and impact assessment method In order to perform a web-based dynamic LCI and LCIA of manufacturing processes, a dynamic data collection step of the resource consumption and local emissions per unit process is mandatory. In addition, the chosen platform to categorize the acquired data must be vendor-neutral and has to be simple, general and usable over the Internet. The MTConnect ® standard (Sobel 2012) (AMT 2011) categorizes the shop-floor data in an Extensible Markup Language (XML), encoding the data in a way that both machines and humans can read. Born to be for the shop-floor the same breakthrough as USB
Remo A.P. Filleti et al. / Procedia CIRP 15 (2014) 531 – 536
standard is for computer peripherals in terms of compatibility, MTConnect improves the easiness of data gathering from machine tools among different proprietary CNC systems. Major benefits are the reduction of automation efforts and system scalability. In addition, gathering data in a XML format increases the potential for integration with other web servers, as well as LCI and LCIA software, such as Gabi. Figure 1 introduces the overview of the proposed method with two subsystems: manufacturing unit process and client application.
533
3.3. Client Application and Inventory and Impact Databases The Client Application is the key software on the proposed method. It performs the interconnection between the categorized and collected data from each unit process with the background LCA databases, providing the LCIA results of the emissions and consumed resources. Dynamic information about the potential environmental impacts of the unit process is shown according to the chosen LCIA method (e.g. CML, EDIP, ReCiPe, USETox, etc.) and impact categories under interest.
3.1. Manufacturing Unit Process 3.4. User Interface The manufacturing unit process represents the process to me monitored. Data related to in loco resources consumption (e.g. electricity, compressed air, cutting fluid) and to emissions (e.g. mist and noise) has to be measured and stored. If measuring devices are not MTConnect complaints, an adapter has to be developed to translate the output signals into the standard format. In the proposed system, the Adapter is installed in a local computer, integrated to the machine tool being monitored. The local computer is connected with the Ethernet and reachable through TCP/IP protocol. Local adapters can be eliminated if measuring devices installed are already compatible with the MTConnect standard. 3.2. Agent Software The Agent software is hosted at the main server and manages the information from all process devices presented in one or more unit processes. This software also collects, arranges, and stores the updated data from the adapters (Sobel 2012) (AMT 2011). Output data is structured according to their physical units (i.e. kWh for energy, m3 for air volume, liter for cutting fluid, dB for noise emission), time, source instrument/measurement device and process unit label. The Agent also performs the communication with the Client Application software, providing data as XML file according to the standard, when requested.
The User Interface has access to the dynamic LCI and LCIA results generated by the Client Application through the Ethernet/Internet connection. Different levels of data access and report generation can be provided, by defining user rules according to the user hierarchy. 4. System implementation for a machining process evaluation To demonstrate the potential of the dynamic method, a laboratory setup of a cylindrical plunge grinding for a shaft external diameter was selected as a case study. . The shafts were made by mild steel (SAE 4140) and were machining by turning in a previous operation. The required steps to implement the method are described as follow below. 4.1. Grinding Unit Process Measuring devices were coupled in a grinding machine ZEMA G-800HS (Fanuc 180i CNC controller) to acquire the overall machine electric energy consumption (electric multivariable meter), compressed air consumption (air flow meter), cutting fluid consumption (fluid flow meter) and noise emission (decibel meter).
Fig. 1. Overview of the dynamic LCI and LCIA method
534
Remo A.P. Filleti et al. / Procedia CIRP 15 (2014) 531 – 536
The energy meter was installed at the main power supply of the machine tool, acquiring its electric information, such as power, power consumption, power factor, voltages and currents. The device has a RS-485 serial output, which can be converted to a RS-232 or USB by an adapter device. The air flow meter was installed to the main compressed air supply and its measures can be read by an electric current output (4-20 mA). The fluid flow meter was installed to the fluid tank output and its measures are shown through an indicator device at the machine. Its measures also can be acquired by a 0-10V electric voltage output from the indicator device. The noise meter was installed close to the machine tool, inside the worker perimeter during operation. Its measures can be acquired by a 0-10V electric voltage output. The hardware device for the data collection step was composed by a data acquisition kit (NI cDAQ – National Instruments), and a serial RS-485/RS-232 converter. The Adapter software was developed using the Labview software, once the measuring devices were not MTConnect compliant.
in kg, and ADP, in kg Sb-eq). ADP and GWP results were obtained based on the CML (2001) method. Table 1. Electric power consumption Electric Power Consumption
Consumption (MJ/worked piece)
Real Grinding Unit Process
0,221
Standard Grinding Unit Process
1,112
4.2. Agent Software, Databases, Client Application and User Interface Agent software was the most recent release of MTConnect® Institute, MTConnect C++ Agent Version 1.2.0.18. The LCI and LCIA results were accessed through the system model generated from the GaBi software for the case studied. Different LCI databases as well LCIA methodologies can be integrated according to the goal and scope definition. Besides, Gabi has import and export data functions by XML files (e.g. ecoSpold format). However, GaBi presents a static and manual structure of LCI and LCIA data generation, that could not be used, by itself, as the Client Application. One solution for this issue would be the development of auxiliary software, which would automate GaBi data income, generation and outcome. 4.3. Dynamic and Static Data Acquisition Differences In order to demonstrate the differences between a dynamic and a static LCA of unit processes, two different scenarios were evaluated. The first scenario was to conduct the dynamic LCA of the grinding machining using the proposed method as shown in Figure 1. The second scenario was the static LCA of a general grinding machine process available at the GaBi software professional database version 6.0. These scenarios were compared for the electric power consumption aspect, from the Brazilian electricity mix, and assuming a grinding removal amount of 5,82 g of metal SAE 4140 per processed piece. Table 1 presents the inventory results of electric power consumption for the two scenarios analyzed and Figure 2 shows the results of water consumption and LCIA comparing the two cases in terms of global warming potential (GWP, in kg CO2-eq.), and abiotic depletion potential (Water ,
Fig. 2. Environmental performance assessment of the two scenarios analyzed.
Both LCIA and LCI results of the two systems showed considerable differences, which can be explained by the aspects and considerations adopted on each system when a dynamic and a static LCA is performed. The data acquired from the real time system consisted on a measurement of all the operation modes of a cylindrical plunge grinding cycle, including the starting, warm-up, standby, grinding and dressing modes. It also included the fluid refrigeration control (extra cooling).
Remo A.P. Filleti et al. / Procedia CIRP 15 (2014) 531 – 536
The time study of these operation modes were also acquired, for one grinding cycle and then extrapolated for a working day of 16h, considering that there is no time for maintenance and that the machine have to be shut down, initialized and warmed up once every day. From this perspective, it was calculated, as described by the UPLCI Methodology (Kellens et. al 2011), the normalized power consumption for one second of the grinding machine operation (Table 2). Table 2. Power consumption of 1s of machine operation Operation Mode
Time of Operation (s/day)
Electric Energy (MJ/day)
Starting Mode
1800
1,044
Warm-Up Mode
3000
19,622
Stand-By Mode
24000
124,330
Dressing Mode
4800
67,0134
Grinding Mode
24000
331,227
Extra Cooling
-
93,968
Total Normalized
16 h
0,011 MJ/s
535
consumption as well as environmental impacts potential were totally different for both systems due to the different limitations and assumptions discussed when conducting a dynamic and a static LCA of unit processes. Besides, the proposed acquisition system (i.e. the devices and software) can easily be installed on almost all types of CNC machining tools, once its installation process has a low level of complexity and does not require modifications at the machine CNC or PLC (Programmable Logic Controller). This embracing aspect associated with the use of the MTConnect protocol allows the standardization and collection of shopfloor machines data, as well their availability on the internet. This can be an important way, for instance, to make possible an online green manufacturing management of systems and to show relevant manufacture environmental indicators. This web based dynamic method under development can enable the generation of LCA results at different levels (unit process, production cells, product production, whole plant, whole product life cycle) and perspectives (unit process optimization, industrial symbiosis, formulate best practices and Ecodesign guidelines for machine tools, etc. ). Acknowledgements
This normalized consumption was multiplied by the time spend to perform a grinding operation at the worked piece (20 seconds). It is important to highlight that the laboratory setup acquired the information of only one kind of grinding operation (cylindrical plunge grinding). The standard grinding data acquired from the inventory GaBi software database, however, hold different aspects and considerations. They represent a mean data from different kinds of machines and grinding processes, as well different operational modes and periods of measurement. It also has different inputs compared with the real time system. Besides the removed material in both systems was the same, the electricity consumption (Table 1) as well the results of environmental performance (Figure 2) were totally different for each system because of the different approaches previously discussed when conducting a dynamic and a static LCA of unit processes. . This fact emphasizes that in some cases the use of a mean data from standard LCA databases may not represent in a satisfactory way the reality of manufacturing processes monitored in real-time. 5. Conclusions The data quality and the accuracy of an LCA study is of paramount importance for the reliability of results and can interfere significantly the decision-making process for green manufacturing purposes. . The proposed dynamic method, in development process provided a good idea on how to conduct an LCA of unit processes in real-time. The developed dynamic method was applied to a real case study of grinding machining and the results were compared with a conventional grinding process available on the GaBi software. The results showed that the electricity and water
The authors would like to thank the CAPES (“Coordenação de Aperfeiçoamento de Pessoal de Nível Superior”, 9331/13-1), CNPq (“Conselho Nacional de Desenvolvimento Científico e Tecnológico”) and the FAPESP (“Fundação de Amparo à Pesquisa do Estado de São Paulo”, 2013/06736-9) for supporting this research. References [1] Dornfeld AD. Green manufacturing: fundamentals and applications (green energy and technology). 1st ed. Berkeley: Springer; 2013. [2] Garetti M, Taisch M. Sustainable manufacturing: trends and research challenges. Prod Plan Control: Manag Oper 2012; 23:83-104. [3] ISO (International Organization of Standardization). 14040: environmental management – life cycle assessment - principles and framework, International Organization of Standardization, Geneva; 2006. [4] ISO (International Organization of Standardization). 14044: environmental management – life cycle assessment – requirements and guidelines, International Organization of Standardization, Geneva; 2006a. [5] Finnveden G, Hauschild MZ, Ekvall T, Guinée J, Heijungs R, Hellweg S, Koehler A, Pennington D, Suh S. Recent developments in Life Cycle Assessment. Journal of environmental management 2009; 91:1-21. [6] Thorn MJ, Kraus JL, Parker DR. Life-cycle assessment as a sustainability management tool: strengths, weaknesses, and other considerations. Environ Qual Manag 2011; 20:1-10. [7] Santos JP, Oliveira M, Almeida FG, Pereira JP, Reis A. Improving the environmental performance of machine-tools: influence of technology and throughput on the electrical energy consumption of a press-brake. J Clean Prod 2011; 19:356–364. [8] Kellens K, Dewulf W, Overcash M, Hauschild MZ, Duflou JR. Methodology for systematic analysis and improvement of manufacturing unit process life-cycle inventory (UPLCI)—CO2PE! initiative (cooperative effort on process emissions in manufacturing). Part 1: Methodology description. Int J Life Cycle Assess 2011; 17:69-78. [9] Geyer A, Scapolo F, Boden M, Döry T, Ducatel K. The future of manufacturing in Europe 2015-2020 - the challenge for sustainability 2003; Thecnical report series, EUR 20705. [10] Kara S, Bogdanski G, Li W. Electricity metering and monitoring in manufacturing systems. In: Proceedings of the 18th CIRP International
536
Remo A.P. Filleti et al. / Procedia CIRP 15 (2014) 531 – 536
Conference on Life Cycle Engineering. Technische Universität Braunschweig: Springer-Verlag; 2011. p.1-10. [11] Behrendt T, Zein A, Min S. Development of an energy consumption monitoring procedure for machine tools. CIRP Ann Manuf Technol 2012; 61:43–46. [12] VikhorevK, Greenough R, Brown N. An advanced energy management framework to promote energy awareness. J Clean Prod 2013; 43:103– 112. [13] Sun Z, Li L. Opportunity estimation for real-time energy control of sustainable manufacturing systems. IEEE Trans Automat Sci Eng 2013; 10:38–44. [14] Ahn S-H, Chun D-M, Chu W-S. Perspective to green manufacturing and applications. Int J Prec Eng Manufact 2013;14:873–874. [15] Pusavec F, Krajnik P, Kopac J. Transitioning to sustainable production – Part I: application on machining technologies. J Clean Prod 2010; 18:174–184. [16] Narita H, Kawamura H, Norihisa T. Development of Prediction System for Environmental Burden for Machine Tool Operation( כ1st Report , Proposal of Calculation Method for Environmental Burden). JSME Int J 2006; 49:1188–1195. [17] Silva EJ, Ometto AR, Rozenfeld H, Silva DAL, Pigosso DCA, Reis VRA. Prototypal implementation of a remanufacturing oriented griding machine. In: Proceedings of the 19th CIRP International Conference on Life Cycle Engineering. University of California at Berkeley: SpringerVerlag; 2012. p.257-262. [18] Joung CB, Carrell J, Sarkar P, Feng SC. Categorization of indicators for sustainable manufacturing; Ecol Indic 2013; 24:148–157. [19] Ramani K, Ramanujan D, Bernstein WZ, Zhao F, Sutherland JW, Handwerker C, Choi JK, Kim H, Thurston D. Integrated sustainable life cycle design: a review. J Mech Des 2010; 132:1-15. [20] Jiang Z, Zhang H, Sutherland JW. Development of an environmental performance assessment method for manufacturing process plans. The Int J Advanced Manufact Technol 2012; 58:783–790. [21] AMT (The Assossiation For Manufacturing Technology). ® , 2011. Available at: Getting Started with MTConnect