CIRP Annals - Manufacturing Technology 61 (2012) 587–609

Contents lists available at SciVerse ScienceDirect

CIRP Annals - Manufacturing Technology

jou rnal homep age : ht t p: // ees .e lse vi er . com /ci r p/ def a ult . asp

Towards energy and resource efficient manufacturing:

A processes and systems approach
Joost R. Duflou (2)a,*, John W. Sutherland (1)b, David Dornfeld (1)c, Christoph Herrmann d,
Jack Jeswiet (1)e, Sami Kara (1)f, Michael Hauschild (1)g, Karel Kellens a
a
Centre for Industrial Management, Department of Mechanical Engineering, KU Leuven, Belgium
b
Division of Environmental and Ecological Engineering, Purdue University, West Lafayette, United States
c
Laboratory for Manufacturing and Sustainability, University of California, Berkeley, United States
d
Institute of Machine Tools and Production Technology (IWF), Technische Universität Braunschweig, Germany
e
Department of Mechanical and Materials Engineering, Queen’s University, Kingston, Canada
f
School of Mechanical & Manufacturing Engineering, University of New South Wales, Australia
g
Department of Manufacturing Engineering and Management, Technical University of Denmark, Lyngby, Denmark

A R T I C L E I N F O A B S T R A C T

Keywords: This paper aims to provide a systematic overview of the state of the art in energy and resource efficiency
Energy efficiency
increasing methods and techniques in the domain of discrete part manufacturing, with attention for the
Manufacturing
effectiveness of the available options. For this purpose a structured approach, distinguishing different
Sustainable development
system scale levels, is applied: starting from a unit process focus, respectively the multi-machine, factory,
multi-facility and supply chain levels are covered. Determined by the research contributions reported in
literature, the de facto focus of the paper is mainly on energy related aspects of manufacturing. Significant
opportunities for systematic efficiency improving measures are identified and summarized in this area.
ß 2012 CIRP.

1. Introduction (e.g. metals, fossil oil-derived materials, and water) as well as

significant amounts of energy, resulting in substantial stress on
Manufacturing plays an indispensible role within the global the environment. Manufacturing also releases solid, liquid, and
economy. Not only does manufacturing provide the goods needed gaseous waste streams that can result in damage to the
by consumers and industries worldwide, it also accounts for a environment.
significant portion of the employment, community presence, and
economic strength. The industry sectors (including mining, 1.1. Scope
manufacturing, and construction) account for nearly one-quarter
of all jobs globally, with over 650 million individuals employed in While all the resources consumed and wastes produced by
the sectors [201]. Job growth within the industry sectors has been manufacturing affect the environment to a greater or lesser extent,
vibrant over the last decade, especially in the developing world, the focus of the present work is on the efficient and effective
with 130 million industry jobs created between 1999 and 2009 as utilization of resources, and in particular energy resources.
supply chains have become increasingly globalized [96,201]. While To illustrate the importance of industry with respect to the
manufacturing employment within industrialized countries has energy demand, consider, for example, the amount of energy
remained at best flat over the last few decades, the productivity of associated with the four principal end-uses: residential, commer-
the sector underwent extraordinary growth; for example, between cial, industrial, and transportation. Residential, commercial, and
1980 and 2005, the annual global production of aluminum more transportation energy consumption represent 14%, 7%, and 27% of
than doubled from 15.4 Tg to 37.3 Tg [207]. Given its importance in the global total, with industrial usage constituting the balance
terms of employment and wealth creation (in 2007, industry (51%) [204]. While these percentages change from country to
constituted approximately 20% of the GDP of the OECD countries country, and notably from developing to industrialized nations (in
[147]), manufacturing will continue to play a vital role in the global 2007 the energy consumed by non-OECD countries exceeded that
economy. Unfortunately, manufacturing also must accept respon- of the OECD countries for the first time), the sheer size of industrial
sibility for placing increasing pressure on the environment. energy consumption (250 quadrillion BTU or 264 EJ in 2007)
Industrial activity, and in particular the manufacturing sector, suggests that this is an issue of importance [204]. Manufacturing
has a large environmental burden associated with it. Manufac- activities dominate industrial energy consumption; Schipper [174]
turing consumes both renewable and non-renewable materials reports that manufacturing is responsible for 84% of energy-related
industry CO2 emissions and 90% of industry energy consumption.
Sizeable energy consumers, and consequently large CO2
* Corresponding author. emitters, within the manufacturing sector include petroleum
E-mail address: Joost.Dufl[email protected] (J.R. Duflou). refining, primary metals processing, non-metallic minerals

0007-8506/$ – see front matter ß 2012 CIRP.

http://dx.doi.org/10.1016/j.cirp.2012.05.002
588 J.R. Duflou et al. / CIRP Annals - Manufacturing Technology 61 (2012) 587–609

processing, chemical production, and paper products manufac- 1.3. Resource efficient and effective manufacturing
turing. These sectors are characterized by semi-continuous
processes, for which material and energy demands form major This contribution addresses both the efficiency and effective-
cost factors. The efficiency of these processes has received ness with which we use energy and material resources, and some
significant attention as an optimization objective in both brief explanation of the distinction between efficiency and
academic and industrial research environments for many years. effectiveness is warranted. Efficiency refers to the amount of
This is somewhat different in the sectors focused on discrete part resources required to produce a given level of output, where it is to
and product manufacturing, where attention became visible be understood that it is desirable to minimize the amount of
mainly over the last decade. This paper is therefore predominantly resources to achieve a given output level. Effectiveness is focused on
focused on the industrial sectors active in discrete part/product making wise choices with respect to how resources are used. To
manufacturing. Until recently, companies in these sectors illustrate the difference between efficiency and effectiveness,
investing in new machine tools primarily took functional consider the energy consumed in a grinding operation. Efficiency
performance and investment cost into account as selection improvements might include changing the wheel type or employ-
criteria. ing a different cutting fluid in order to reduce the energy required
Perhaps driven by concerns related to environmental sustain- for a given operation. An effectiveness approach might change the
ability, a trend towards more environmentally benign manufac- process plan entirely to use a dramatically smaller amount of
turing can be observed today. In addition to more stringent energy, with the resulting process plan not needing the grinding
regulations (e.g. emission standards, worker exposure standards, operation. In short, efficiency is doing things right, and effective-
and banned materials), additional motivating factors to switch to ness is doing the right things.
more environmentally benign manufacturing solutions include
competitive economic advantages (e.g. conservation of energy, 1.4. Manufacturing system organization
water, and materials, and reduced waste treatment and disposal
costs) and proactive green behavior (e.g. corporate image, ISO From the perspective of the organization of the system,
14001 certification [97], and eco-labeling) [68]. manufacturing activities can be considered as being composed
Discrete part manufacturing industries emit CO2 indirectly of multiple levels, from the level of the individual devices where
through the consumption of electricity and directly through plant- unit processes take place, through to that of the enterprise,
based use of fossil fuels. It is becoming increasingly apparent that incorporating all the activities in the manufacturing system,
manufacturers play a critical, multi-faceted role in dictating the including supply chain externalities [165]. In the context of this
material and energy resources in modern society. Not only do the paper five levels are distinguished:
processes designed and employed directly by manufacturers have
a sizable environmental impact, but the product design decisions  Device/unit process: Individual device or machine tool in the
made by manufacturers also control the energy and resource manufacturing system, which is performing a unit process.
intensive production of materials and chemicals and the energy Support equipment of the unit process is included here, such as
and resources consumed by products across their life cycle. gage systems and device level oil-circulating systems.
Moreover, as supply chains become more globalized, energy  Line/cell/multi-machine system: Logical organization of devices in
intensive operations are increasingly being outsourced to devel- the system that are acting in series or parallel to execute a
oping nations that often utilize more CO2 intensive energy specific activity (such as manufacturing a part or assembly).
resources. Support equipment for the collection of devices is included here,
such as chip conveyers and tool cribs.
1.2. Objective  Facility: Distinct physical entity housing multiple devices, which
may or may not be logically organized into lines, cells, etc.
The environmental challenges that we face in the near future Support equipment required at the facility level are also included
are significant, and every part of society must proactively respond, here, such as power generators, water purifiers, and HVAC
including manufacturing. Opportunities exist across the manu- systems.
facturing enterprise for more efficient usage of energy and material  Multi-factory system: Different facilities whose proximity to one
resources, and in particular improved material utilization. The another allows them to make use of possible synergies in terms
scope of opportunities ranges from the process to the factory level of reuse of waste and lost energy streams.
to the entire enterprise, and this paper will present research efforts  Enterprise/global supply chain: The entire manufacturing system,
that cover this scope, although attention will be limited to discrete consisting of all the individual facilities, the infrastructure
part manufacturing processes and systems. required to support the facilities, as well as the transportation
To support the design and realization of products and their and supply chain externalities.
subsequent distribution, utilization, and disposition, some envir-
onment-oriented software tools have been developed, e.g. life
cycle assessment software. Unfortunately, manufacturing data in 1.5. Paper overview
support of such software tools are often not available, not
representative of the situations faced by manufacturers, or based Depending on the scale of the considered system, different
on unrealistic assumptions. With this in mind, this paper will also strategies can be applied to optimize efficiency and effectiveness of
begin to address this deficiency. manufacturing systems. In order to offer a systematic review of
This paper has the ambition to provide a state of the art perceived opportunities and tested strategies, the paper is
assessment of the methodologies and technologies that can be structured according to an increasing system scale, largely
called upon at different levels within the production system to corresponding to the levels distinguished above.
contribute to a significant impact reduction. Perhaps owing to In Section 2, the individual process level is studied. The
climate change, energy independence, and energy cost concerns, material–process interaction, as achieved by stand-alone machine
recent research work by many researchers has focused on energy- tools performing individual operations or a series of operations on
related issues. Thus, the paper has a strong focus on energy-related a workpiece, is the focus. The unit processes studied in this section
efforts. However, other resource streams are also covered. These, of form the smallest building blocks for the composition of
course, also represent indirect energy consumption through manufacturing systems. Choice of process principle and machine
embodied energy, but also have other associated wide ranging tool selection are typical measures that can be considered at this
environmental effects in their production and end-of-life (EOL) level. For machine tool designers, component selection and
treatment stages. optimized control of subsystems offer opportunities.
 J.R. Duflou et al. / CIRP Annals - Manufacturing Technology 61 (2012) 587–609 589

Possible interactions and synergies between different machine the total impact for cutting machine tools and press brakes
tools are considered in Section 3. Multi-machine ecosystems can respectively, the use stage however systematically proves to be the
allow reuse of energy and material flows through proper planning dominant contributor to the total life cycle environmental impact
and control. of machine tools. Therefore, Section 2 of this paper focuses on the
Section 4 focuses on spatial and temporal considerations at energy and resource consumption during the machine tool use
factory level. Energy management aspects linked to production stage. While the impact assessment of the use stage is covered in
planning and the influence of factory layout, support facilities and Sections 2.1–2.3, the potential for environmental improvement is
building technology on plant performance are reviewed here. treated in Section 2.4.
Wherever considerable energy and material waste flows cannot
be resolved at an in-house scale, industrial symbiosis among 2.1. State of affairs in unit process data availability
different manufacturing facilities can provide opportunities, as
studied in Section 5. 2.1.1. Coverage of unit process impact by existing LCI databases
In a globalized economy, supply chain decisions can have The Ecoinvent database, as supplied by the Ecoinvent Centre
significant consequences in terms of required transport efforts and [61], is one of the most widely consulted sources of consistently
the nature of local supplies of energy and other resources. The and transparently documented life cycle inventory (LCI) data. In
sensitivity of the environmental impact of manufacturing for these this database, in contrast to materials production, manufacturing
factors is analyzed in Section 6. processes, as used for discrete part manufacturing, are less well
In order to provide the reader with a realistic view of the documented in terms of the overall environmental impact. On the
magnitude of the energy and resource efficiency affecting one hand, the coverage of the wide range of available manufactur-
measures at the respective scale levels, examples and quantified ing processes is limited to more conventional processes, such as
effects are reported in the different sections. While these can only turning, milling and casting, while commonly used processes, such
be considered as illustrations of the system sensitivity for the as, for example, electrical discharge machining and additive
tested strategies at different scales, it is the hope of the authors that manufacturing processes, are lacking in the database. On the
such examples contribute to a more balanced view on prioritiza- other hand, most of the available data on manufacturing processes
tion among the wide range of identified efficiency increasing are incomplete: the focus is often limited to theoretical energy
measures. consumption, and data on emissions are rarely found [185]. For
other LCI databases, such as the U.S. Life-Cycle Inventory Database
2. Unit process level [146], similar observations can be made.
It is to be noted that various industrial sectors provide LCI-data
Discrete part manufacturing processes are defined as produc- about their products. Organizations like the International Iron and
tion processes in which the output can be identified and is Steel Institute [64], the European Aluminium Association [56] and
measurable in distinct units rather than by weight or volume as in Plastics Europe [54] provide extensive LCI data for respectively
process industry. In literature, different taxonomies are available steel, aluminum and plastic half products, while the BUWAL
to order the wide variety of discrete part manufacturing processes database covers a wide range of packaging materials [22]. Despite
[46,198]. Fig. 1 shows the typical system boundaries of a unit the very useful work of the respective sector organizations, the
process. scope of these LCI databases is typically limited to primary material
In the content of this paper, the boundaries of a unit process production (e.g. sheets, foils, and profiles) and recycling processes.
typically coincide with individual machine tools as the smallest A comprehensive overview of the state of the art in environmental
unit of which production systems are composed. In this approach, impact assessment of unit processes is provided by Duflou et al.
hybrid workstations, combining multiple processes on a single [53].
machine structure, can be considered as the sum of multiple
individual unit processes that can be analyzed separately. 2.1.2. Sectorial interest in the environmental performance of machine
While some preliminary environmental studies for machine tools
tools in material removal processes (e.g. turning and milling) Research contributions by different machine tool builders
indicate that more than 90% of the environmental impact is due to demonstrate the increasing interest in the environmental (mainly
the consumption of electrical energy during the use stage [23], energetic) performance of the equipment they develop, [40,
other authors reporting detailed analysis observe significantly 136,213]. CECIMO, the European Machine Tool Builder Association,
larger contributions of the machine tool construction [9,42, launched a self-regulatory initiative for supporting the identifica-
172,229]. With a minimum of approximately 83% and 60% of tion of measures to improve the energy and resource efficiency of
the machine tools of their members [23]. The ongoing ISO
standardization efforts ‘Environmental evaluation of machine
tools’ (ISO 14955) and ‘Automation systems and integration –
Environmental and energy efficiency evaluation method for
manufacturing systems’ (ISO 20140) as well as the Life Cycle
Initiative of UNEP–SETAC [194] are likely to reinforce the trend
towards more environmentally conscious manufacturing.

2.2. Methodologies for the determination of unit process energy and

resource consumption

The basic procedure for a life cycle inventory (LCI) effort of a

manufacturing unit process is similar for a wide range of
manufacturing processes and can in principle cover a full process
taxonomy like DIN8580 [46,63], including auxiliary support
processes such as compressed air supply, centralized cooling
and transport systems. The consumption of energy and resources
as well as the generated waste and process emissions should be
determined and allocated to the individual production process. In
this section, energy demand is interpreted as net extraction from
Fig. 1. System boundaries of a unit process [111]. the power supply system. Conservation of energy and possible
590 J.R. Duflou et al. / CIRP Annals - Manufacturing Technology 61 (2012) 587–609

reuse in other unit processes is not taken into consideration here,

but will be considered as part of the energy and exergy analysis
covered in Section 3. Collection and documentation of LCI data can
be obtained in different ways, starting from theoretic calculations
until detailed process measurements and analysis.

2.2.1. Equations to calculate the (minimum) theoretic process energy

and resource consumption
Abele et al. [1] describe a method for life cycle inventory
analysis of production processes using theoretic equations to
calculate the energy and resource consumption as well as waste
and process emissions. Equation 1 shows the general formula for
the total energy demand during production.

Etotal ¼ Eth þ Eadditional þ Eperiphery (1)

where, Eth is the active energy theoretically needed to obtain the

physical process effect, and represents the minimum energy
demand of the production process. Eadditional and Eperiphery stand for
the additional energy demands of the machine tool (e.g. energy to Fig. 2. CO2PE! – UPLCI – Methodology [111].
cover efficiency losses, or energy for machine functions such as
central control) and peripherals (e.g. cutting fluid pump)
respectively. Abele et al. [1] provide specific equations for a wide into account based on the available average data per unit of
variety of production processes. consumable as documented in existing LCI-databases such as
Overcash et al. [150] developed a generic methodology to EcoInvent for a range of auxiliary unit processes.
gather unit process life cycle inventories (UPLCI) using rules of Besides a process dependent functional unit, a generally
engineering and industrial practice to calculate energy and mass applicable reference flow of 1 s of processing time for a specified
losses. As described below, this approach is also used as screening load level of a unit manufacturing process for a given material,
approach within the CO2PE! – Initiative. based on a working scheme of 2000 h/year including some
specified use modes is used.
2.2.2. Detailed process measurements and analysis
Within the framework of the CO2PE! – collaborative research 2.2.2.2. Process inventorisation. The in-depth approach of the
programme [30], a life cycle assessment (LCA) oriented methodol- process inventorisation is based on industrial measurements
ogy for systematic inventory analysis of manufacturing unit and includes a time, power, consumables as well as emission
processes (referred to as unit process life cycle inventories, UPLCI) study. First, a statistical time study is performed in order to
has been developed which comprises two approaches with identify the different use modes of a machine tool and their
different levels of detail [111]. respective share in the covered time span. The energy consump-
The screening approach relies on representative, publicly tion of the machine tool is obtained by measuring the total
available data and theoretical calculations for energy use, material machine tool power consumption over a specified time period. By
loss, and identification of variables for improvement, while the in- measuring individual power consumption patterns for all relevant
depth approach is subdivided into four modules, including a time active energy consuming subunits in each production mode, the
study, a power consumption study, a consumables study, and an energy and corresponding environmental impact optimization
emissions study. In this approach all relevant process inputs and potential can be identified as output towards machine tool
outputs are measured and analyzed in detail. The screening designers.
approach provides the first insight in the unit process and results in Of course, the usefulness of such data collection efforts
a set of approximate LCI-data [150]. These data serve to guide the largely depends on the completeness and quality of the data
more detailed and complete in-depth approach leading to more capturing and measurements. While Herrmann et al. [84]
accurate LCI-data as well as the identification of the potential for provide an overview of energy consumption metering proce-
energy and resource efficiency improvements of the involved dures and tools, Kara et al. [105] present an overview of the
manufacturing unit process. As far as the energy consumption evolution and the latest developments in power measurement
quantification effort is concerned, the methodology under devel- and monitoring systems. An exemplary list of metering devices
opment as part of the ISO TC39WG12 efforts [98] corresponds well is given to demonstrate how features of the different measure-
with the CO2PE! in-depth approach. Fig. 2 shows the framework of ment instruments can be matched for certain measurement
this CO2PE! – UPLCI – methodology as explained more in detail in tasks in different application levels. For automated energy
the next paragraphs. monitoring of machine tools, Vijayaraghavan and Dornfeld [212]
introduce a framework based on event stream processing to
2.2.2.1. Goal and scope definition. First, the goal and scope of the automate the monitoring and analysis of energy consumption in
study should be clearly defined and must be consistent with the manufacturing processes.
intended unit process. The most important parts of the scope Parallel to the time and power measurements, the resource
definition to be considered are the system boundaries and the consumption (e.g. compressed air, lubricants, process gasses) are
functional unit of the intended process. Furthermore, the machine measured for each process material in each production mode.
tool architecture and process parameters are investigated and all While the raw material flow does not represent manufacturing
subsystems that are considered relevant for design optimization induced impact, the amount of waste created is process dependent
are identified and located within the machine tool. and is therefore included as consumable. Finally, also an emission
As illustrated in Fig. 1, the system boundaries are set to include study takes place when relevant (e.g. mass balance abnormalities,
only the operating stage of a single, isolated manufacturing unit use of hazardous materials). This study includes gaseous, liquid
process, disregarding materials processing, production, transport, and solid emissions as well as heat losses.
maintenance and disposal of the machine tool itself. For processes Herrmann et al. [88] proposed a framework to visualize the
using consumables (e.g. compressed air, lubricants, and process environmental impact of manufacturing processes using virtual
gases) delivered by central systems, the related impacts are taken reality.
 J.R. Duflou et al. / CIRP Annals - Manufacturing Technology 61 (2012) 587–609 591

2.3. Data collection efforts: case studies

Various energy and resource consumption data collection

efforts are described in literature using one of the above mentioned
approaches. An overview of available case studies using the CO2PE!
– Screening approach can be found on the UPLCI-website [202].
This section provides a selection of case studies organized
according to the DIN 8580 taxonomy [46,63], which distinguishes
six main categories of discrete part manufacturing processes:
Primary shaping, Forming, Separating, Joining, Coating/Finishing
and Processes which change the material properties. For each of
the first three categories some case study results, based on
industrial measurements, are presented in the next paragraphs.

2.3.1. Primary shaping processes

Mognol et al. [133] investigated the energy consumption of
three rapid prototyping systems: Thermojet (3DS), FDM3000 Fig. 4. Power levels during the productive modes of an EOSINT P700 Selective Laser
(Stratasys) and EOSINT M250 Xtended (EOS). Since the electrical Sintering machine tool [110].
power consumption of these systems was found to be approxi-
mately constant over different operating modes, the authors
concluded that the manufacturing time is the determining factor setting or catalytic reactions) and clay (green sand casting), with
for the total energy consumption. green sand molds used in about 90% of part production [205]. To
More detailed studies, covering all aspects of impact generating form internal voids in the parts, chemically-bound cores are used,
energy and material flows, allow to provide a more refined which are stronger than green sand. Unlike permanent mold
analysis. A case study about the environmental impact of Selective casting, sand molds are destroyed after each use, with the majority
Laser Melting (SLM) and Selective Laser Sintering (SLS) processes of the sand usually recycled into new molds and cores. The
has been conducted at the KU Leuven [110]. In-depth time studies environmental concerns of sand casting include virgin material
were performed on three different EOSINT P760 SLS machine tools. consumption, solid wastes, and hazardous air emissions, which
The non-productive modes comprise machine tool cleaning, have been addressed over the past several decades. Sand casting
preheating and cooling down. The productive modes are respon- utilizes large amounts of sand; many foundries in the U.S. extract
sible for approximately 87% of the total production time, and can sand from the sand dunes of the Great Lakes [220]. While sand can
be subdivided further into three main modes: the laser exposure be reused, it breaks down over time. In the U.S. about ten million
mode, the recoating mode and observed other activities like filling tons of waste foundry sand are generated annually [206]. In
the feed containers. Based on 63 batches (5801 products), Fig. 3 addition, the various types of green sand and chemical binding
presents the time distribution for products made of PA2200 with a systems form a variety of air emissions, which can be harmful to
layer thickness of 120 mm. workers [12,33,138]. Dalquist and Gutowski [37,38] investigated
Fig. 4 shows the different power levels during the productive sand casting and die casting using life cycle assessment. Haapala
mode of an EOSINT P760 machine tool. Besides the laser cooler, the et al. [71,72] have established a model to compare the environ-
process heating units are the largest energy consumers. mental performance of sand casting process design alternatives.
Besides energy, SLS processes also need a process gas (typically In these case studies the contribution manufacturing opera-
N2) to create an inert working atmosphere. Since this atmosphere tions can make to impact avoidance has not been taken into
is created by an internal, compressed air driven N2-generator, the account when compared to more conventional manufacturing
compressed air consumption of 20 m3/h should be taken into processes.
account. Furthermore, based on industrial observations, only half For additive processes an example of such an effect is illustrated
of the residual powder is recyclable, which represents a waste by Morrow et al. [137]. The authors present three case studies to
material rate of approximately 45% of the input powder. reveal the extent to which Direct Metal Deposition (DMD)
For a sample batch with a total product volume of 3.3 kg and manufacturing of molds and dies can achieve reduced environ-
production time of 15 h, the major contributor to the total mental emissions and energy consumption relative to conven-
environmental impact was found to be the waste of raw material tional manufacturing procedures. It is shown that DMD’s greatest
(see Table 1). Major opportunities to reduce the environmental opportunity to reduce the environmental impact of tool and die
impact of the SLS process for polymers seem to be to identify manufacturing comes from its ability to enable laser-based
measures that result in a higher material utilization efficiency and/ remanufacturing.
or methods to regenerate the residual powders for reuse. For injection molding processes, Thiriez and Gutowski [197]
The majority of discrete part casting operations occurs through obtained specific energy consumption values of 3.39, 1.67 and
sand casting (60%) [203]. The sand casting process involves 1.46 MJ/kg for hydraulic, hybrid and all-electric machine tool
pouring of molten metal into a mold composed primarily of sand variants, respectively. For hydraulic and hybrid machine tools, the
and binders. The binders used are primarily chemical (thermal specific energy demand decreases with increasing throughput.

2.3.2. Forming processes

Several authors investigated the energy consumption of air
bending [39,52,108,172] and observed that the standby energy is
substantial. On the one hand, due to diverse operator activities, the

Table 1
Environmental parameters of sample batch for SLS processing [110].

Impact (mPts) %

Energy 120 kWh 3120 31.9

Compressed air 340 m3 1598 16.3
Fig. 3. Time distribution for SLS production modes (PA2200, layer thickness of
Waste material 10.3 kg 5068 51.8
120 mm) [110].
592 J.R. Duflou et al. / CIRP Annals - Manufacturing Technology 61 (2012) 587–609

Fig. 8. Distribution of the environmental impact during 1 h of EDM roughing [109].

Fig. 5. Time distribution for the production modes of the air bending process [52].
An empirical model to characterize the relationship between
energy consumption and process variables for chip forming
material removal processes was presented by Kara and Li [106]
and Diaz et al. [43]. Eq. (2) shows the developed model predicting
the energy consumption of turning and milling processes with
deviations of less than 10% on the measured values, which
indicates that one generic model could be adopted to describe the
specific energy consumption under various cutting conditions.

C0 þ C1
SEC ¼ (2)
MRR
where, SEC is the specific energy consumption [kJ/cm3], MRR is the
material removal rate [cm3/s] and C0, C1 are machine tool specific
coefficients.
The environmental impact of CO2 laser cutting was analyzed by
Duflou et al. [51] and Oliveira et al. [149]. In addition to energy, the
assist gases consumed in the process (e.g. nitrogen and oxygen)
Fig. 6. Power profile of a turning process [121]. and waste generated were also found to significantly contribute to
the total environmental impact of the use stage of a laser cutting
time share of the standby mode (mode 1) is very high as indicated machine tool. The laser source and chiller are responsible for more
in Fig. 5. On the other hand, the power levels during the standby than 80% of the total consumed energy.
mode are significant, ranging from 1.4 kW to 5 kW for conven- Dhanik et al. [40] as well as Kellens et al. [109] investigated the
tional hydraulic press brakes with a maximum capacity between environmental impact of the use stage of Electrical Discharge
80 and 170 ton. Similar conclusions were drawn by Shi et al. [179] Machining (EDM) processes. Of all subunits, the different dielectric
for stamping processes. pumps proved to be the dominant energy consumers.
Ingarao et al. [95] outlined the state of the art from a resource Fig. 8 summarizes the distribution of a range of different impact
efficiency point of view for sheet metal forming technologies, creating factors for 1 h of EDM roughing with a copper electrode,
covering the total life cycle starting from raw materials production on a hard metal workpiece. The environmental impact is mainly
up to recycling technologies. caused by the consumed electrical energy (40.5%) and the
dielectric fluid (43.4%).
2.3.3. Separating processes
Dahmus and Gutowski [36] showed that the cutting energy 2.4. Energy demand and resource consumption reducing strategies
consumed by a modern automatic machine tool during machining
is less than 15% of the total energy demand. Fig. 6 illustrates a Different strategies can be considered while aiming for the
typical power profile of a turning process [121], which divides the reduction of environmental impact at a unit process level. Machine
total machine tool power in four power levels: tool manufacturers can work on the process efficiency by
optimizing the machine tool design, while process planners can
 Fixed power: power demand of all activated machine components work on process parameters optimization, on machine tool
ensuring the operational readiness of the machine. selection or can consider process substitution.
 Operational power: power demand to distinctively operate
components enabling the cutting as performed in air-cuts. 2.4.1. Effects of optimized machine tool design
 Tool tip power: power demand at tool tip to remove the
workpiece material. 2.4.1.1. More efficient machine tool components. An obvious way to
 Unproductive power: power converted to heat mainly due to improve the environmental performance of machine tools is the
friction during the material removal. development and adapting of more energy efficient components.
Among many others, the Self-regulatory Initiative (SRI) of CECIMO
As shown in Fig. 7, an average breakdown of the fixed energy [23] as well as the ISO standardization effort on the environmental
consumption of machining processes is given by Li et al. [121]. evaluation of machine tools [98], list potential measures towards
more energy and resource efficient machine tool components.
Examples can be found in more efficient drives, pumps, and
spindles. Another example can be found in switching from active to
passive workpiece clamping, offering significant reductions of the
time span in which active power is required for a wide range of
processes [98].
Abele et al. [2] explored the potential of an energy optimized
spindle unit with an adapted electric drive train. An axiomatic
Fig. 7. Average fixed energy breakdown adapted from [121]. approach to identify structural improvement potential with
 J.R. Duflou et al. / CIRP Annals - Manufacturing Technology 61 (2012) 587–609 593

guidance for the optimal implementation sequence of measures

was proposed by Zein et al. [226].

2.4.1.2. Technological changes. Instead of the gradual improvement

of current machine tool technologies and devices, a shift to
alternative, innovative technologies can yield significant environ-
mental gains. For example, adopting new laser source technolo-
gies, such as new generations of fiber and diode lasers, to replace
conventional CO2-lasers, holds the promise of an increase from
around 12% for CO2-lasers to approximately 30% in laser source
efficiency, while simultaneously resulting in lower output power
requirement levels for most materials due to more favorable
absorption in the near infrared wavelength range for most
commonly processed materials [52].

2.4.1.3. Waste recovery within a machine tool. Another way to Fig. 10. Specific energy demand for milling processes as function of the material
removal rate (MRR) [43].
improve the energy and resource efficiency of manufacturing
processes is the recovery of waste streams and heat losses within a
machine tool. A recent patent [213] describes a system for modes (e.g. switching to a less energy consuming stand-by level).
recovering heat losses of a laser cooler system using a Sterling Switching from conventional, continuously active hydraulic
engine that drives a compressed air generator. A filter system systems for air bending towards servomotor driven pumps with
separates nitrogen and/or oxygen from the compressed air that can direct control of hydraulic pistons, allows the required standby
be used as assist gas within the machine tool itself. energy to be reduced significantly (up to 65%) as the servomotors
Another example related to material resource efficiency can be only need to be activated during the actual bending operation
found in the commonly applied recycling and remelting of runners [39,108,172].
by casting processes.
A practical method to separate and collect un-sintered powder 2.4.2.2. Reducing idle production times. Besides the required power
materials for polymer laser sintering processes has been developed level of machine tools, the second parameter in energy demand is
by Dotchev and Yussof [49]. In order to control the input material time. Therefore reducing the unproductive idle time during a
quality and use the fresh powder more efficiently, the authors manufacturing operation contributes to an increase in overall
suggest using different grades according to the melt flow rate of the energy efficiency. Using functional states for common suppliers,
recycled powder. Schmitt et al. [176] proposed a self-optimization model to control
In terms of energy recovery the kinetic energy recovery system the energy consumption depending on the usage profile of a
(KERS) is a possible method of improving energy efficiency. machine tool, automatically switching to the most optimal
Investigations by Diaz et al. [41] show reductions of the average production mode.
energy consumption of high-speed cutting processes up to 25%
depending on workpiece geometry and machining time. 2.4.2.3. Optimized process parameters. The selection of process
parameters can have a significant influence on the consumed
2.4.1.4. Integrated or central peripherals. Consumables such as energy and resources. Diaz et al. [41,43] as well as Mori et al. [136]
compressed air and process gases can be produced within the illustrated that the energy consumption for drilling and face/end
machine tool itself or delivered from central support systems, milling can be reduced by setting the cutting conditions (cutting
causing different environmental impacts depending on the chosen speed, feed rate and cutting depth) high, thereby shortening the
solution. The same consideration applies for local or centralized machining time, yet within a value range which does not
supply of cooling and lubrication facilities. A comprehensive compromise tool life and surface finish. Fig. 10 shows the specific
literature review about compressed air energy use, savings and energy demand for milling processes as function of the material
payback period of energy efficient strategies was compiled by removal rate (MRR).
Saidur et al. [170]. An overview of different derived energy saving For deep hole machining, the power consumption can be
opportunities at the machine level is shown in Fig. 9. A reduced with an adaptive pecking cycle, which executes pecking as
comprehensive overview of compressed air systems design needed by sensing cutting load. Finally, synchronization of the
considerations is provided in Section 4.3. spindle acceleration/deceleration with the feed system during a
rapid traverse stage can reduce the energy consumption up to 10%.
2.4.2. Effects of optimized process control Similar results were presented for drilling processes by Neuge-
bauer et al. [140].
2.4.2.1. Selective actuation of non-continuously required devices. An Furthermore, Neugebauer et al. [142] defined preferential
easy and highly effective way to reduce the electrical energy working spaces with limited power consumption for robot based
consumption can be found in selectively shutting down devices of assembly systems.
which the functionality is not required in specific operational
2.4.2.4. Energy and resource efficient process modeling and plan-
ning. Various models for energy and resource efficient process
modeling, planning and scheduling have been presented in
literature. Sheng et al. [178] were among the first to also include
environmental factors in their multi-objective process planning
models in addition to traditional criteria such as production rate,
quality and costs. Over the last decade, several variants of models
have been proposed, all defining the most energy efficient or
ecological process or process chain, using the available data of
different manufacturing unit processes and auxiliary equipment
Fig. 9. Relative importance of different energy saving measures for compressed air
(e.g. robots and compressed air systems). Among many others,
systems. examples at unit process as well as system level can be found in
Adapted from [170]. [8,44,57,86,141,144,218].
594 J.R. Duflou et al. / CIRP Annals - Manufacturing Technology 61 (2012) 587–609

2.4.3. Effects of process/machine tool selection adapted metalworking fluid systems. A comparative life cycle
assessment of water and gas-based systems shows that delivery of
2.4.3.1. Process selection. Energy and resource efficiency compar- lubricants in air rather than water can reduce solid waste by 60%,
ison and deduced selection of alternative manufacturing processes water use by 90% and aquatic toxicity by 80%, while virtually
could help to reduce the total environmental product impact. For eliminating occupational health concerns.
example, Pecas et al. [152] developed a model to compare the life
cycle performance of mold manufacturing techniques suitable for 3. Multi-machine level
low production volumes. The model includes three independent
but essential and complementary performance criteria: economic, A production system is a combination of different production
technical, and environmental. Two candidate technologies were processes and typically is composed of diverse machines for
evaluated: one involving a mold made of a spray metal shell processing or transportation as well as personnel. All these
backfilled with resin and aluminum powder and another based on production factors are being planned and controlled by a production
the machining of aluminum. While the former mold offers the best management system. All involved technical equipment, with certain
solution in terms of economic and environmental aspects, the energy and media consumption profiles, result in cumulative load
latter has a better technical performance. curves for such measures as power or compressed air for the whole
Pusavec et al. [156,157] present cryogenic and high pressure jet process chain or the plant respectively.
assisted machining (HPJAM), using liquid nitrogen as a coolant, as
viable machining technologies offering a cost-effective route to 3.1. Multi-machine ‘ecosystems’
improve economic and environmental performance in comparison
to flood cooling in conventional machining. A multi-machine ‘ecosystem’ describes a network of machines
Brockhoff and Brinksmeier [20] as well as Reinhardt et al. [166] within a factory. The connection of these machines can be in a
proposed grind-hardening as a resource efficient alternative to the parallel organization, like in a job shop production, or in a sequence
traditional hardening processes using induction. organization, like a process chain. Due to the structure of the
Klocke et al. [112] concluded that the directed supply of network, the output of one process may be the input for another.
lubricoolant, with supply pressure as the dominant process The qualitative condition of the output is not relevant. Thus, by-
parameter; could lead to significant reduction of tool temperature products, scrap, energetic emissions (waste heat, waste air,
up to 30% compared to conventional flood cooling. mechanical energy, etc.) or waste material, which arise in the
production of products in one process, are inputs on the multi-
2.4.3.2. Optimal machine tool capacity. For most manufacturing machine level (Fig. 12). In this context the energetic or physical
processes the fixed power level, which corresponds to a non- flows can be classically recycled within the process chain, but also
loaded machine tool in stand-by mode, has a significant be reused in another process chain nearby. At the end of these
contribution to the total power consumption. Therefore, a proper multiple use stages, the further non-useable energetic or physical
selection of the right equipment (and related maximum capacity) output leaves the system boundary.
could reduce the energy consumption. An example can be found in In the following, methods and tools are presented to capture
Fig. 11 which shows potential energy savings up to 50% by and track energy and material flows in a multi-machine ecosystem.
optimizing the selection of a laser cutting machine tool [52]. Target Furthermore, examples are presented of implementation possibi-
here should be to use machine tools as near as possible to their lities in such ecosystems.
maximum capacity.
3.1.1. Energy flow and machine ecosystems
2.4.3.3. Optimal resource consumption. In addition to energy The multi-machine ecosystem can be described as an energy
consumption, the consumption of resources, such as lubricants, cascade system with the purpose of completely utilizing energy
compressed air, and process gasses also causes environmental flows (Fig. 13) [67,77,182]. A cascade example is the thermal
impact. In the domain of lubricants, several researchers investi- down-flow of heat. The thermal energy flows from a higher
gated strategies such as Dry Machining and Minimum Quantity temperature level to the ambient temperature. During this
Lubrication, aiming to avoid or limit resource consumption (e.g. downfall, the heat can be utilized in multiple processes [4].
[7,216]). Thereby the quality of the energy flow decreases gradually with
In parallel, environmentally benign fluids have been developed increasing duration of utilization.
for operations which still require lubricants (e.g. [5,81,85,
119,148,221]). A comprehensive overview of important develop- 3.1.1.1. Modeling methodologies of energy flows in an ecosystem. The
ments of new, environmentally benign lubricants for metal foundation for description, acquisition and analysis of all energetic
forming processes was published by Bay et al. [13]. flows into and out of a multi-machine ecosystem are the energy,
Clarens et al. [27] provided a quantitative assessment of
emissions and energy consumption as well as a semi quantitative
assessment of health impacts associated with environmentally

Fig. 11. Power consumption in cutting mode in function of the laser output level for
7 different laser cutting machines (3 configurations combined with different CO2
laser sources with indication of the maximum achievable output level) [52]. Fig. 12. Multi-machine ecosystem.
 J.R. Duflou et al. / CIRP Annals - Manufacturing Technology 61 (2012) 587–609 595

considered manufacturing processes are laser cutting as a

subtractive process, the Selective Laser Sintering as an additive
process and bending as mass-conserving process. For each process,
the input and output variables are determined and subsequently
form the basis for the exergy analysis. The conclusion of the
examination is that system analyses on the basis of exergy
efficiencies are able to appropriately present the resource
efficiency of the involved manufacturing processes.

Fig. 13. Energy cascade system [182].

3.1.1.2. Example of energy flows in a multi-machine ecosystem. In
the following, examples are presented where the exergy concept is
used to analyze and describe a technical system.
exergy and entropy concepts. These three concepts are defined by Creyts et al. [34] use the exergy analysis of a metal machining
the first and second law of thermodynamics. The first law describes process to understand and to present the energy flows in a multi-
that energy cannot be destroyed or created, it is conserved, and it machine ecosystem. The used model consists of a process model to
can just change form. The energy only serves as a carrier of quality, determine the waste stream characteristics and an exergy model
and it is the quality that is consumed during the conversion of that performs the extended exergy analysis on the resulting waste
energy [214]. The second law describes that energy has quality and streams. The examined process is the machining of an aluminum
quantity. Quality and quantity of energy or ‘work potential’ can be workpiece with a water–oil emulsion. In addition to the actual
described by the properties exergy and entropy [10,15,78,214]. machining process, particularly the treatment and disposal of
The energy flows into and out of a system along paths of mass waste streams (such as metal working fluid, cutting swarf, etc.)
flow, heat transfer and work [15]. The exergy is that part of energy were exergetically examined. After the modeling and calculation of
that is convertible into all other forms of energy; it represents the the existing process, the influence of the input parameters on the
‘useful’ energy respectively the technical working capacity of the exergy was examined by a sensitivity analysis. Besides the analyses
streams that flow to the system [15,214]. In contrast to energy, of the influence of changed process parameters, Creyts et al. [34]
exergy can be destroyed. The amount of destroyed exergy is perform an exergetical comparison of the used technologies for the
proportional to the generated entropy. In order to analyze a treatment and disposal of waste streams with alternative process
machine, an energy analysis, based on an energy balance according configurations and technologies. Both examinations showed that
to the first law, can be carried out. The disadvantage is that in the exergy analysis is a suitable tool to integrate environmental
discrete manufacturing, the energy analysis often only measures criteria into the planning stages of industrial process system
one resource: electrical energy and other process connected design. Further examples are given, amongst others, by Szargut
resources are not taken into consideration. Furthermore, no et al. [191,192], Brodyansky et al. [21], Kotas [114], Meyer [131],
information on the system energy degradation is available. To Dinçer et al. [47], Gutowski et al. [69] and Saiganesh et al. [171].
overcome these disadvantages, an exergy based method can be While an energy analysis can be used to quantify the efficiency of
used to analyze a technical system. standalone machine tools (Section 2), opportunities for improve-
The exergy method allows the complete measure of in- and ment at the level of energy reuse are often requiring a systems
outgoing system streams and the derivation of information about approach involving multiple unit processes. In most cases, the
energy inflation in the system [15,78]. Thus, a minimization of all industrial waste heat is considered as source for energy flow
exergy losses leads to a minimum energy consumption of the cascades. An example of cascading energy utilization is the use of
process [162]. Furthermore, it can be derived that the exergy is the industrial waste heat of a melting furnace as an input flow for a heat
property that has an economic value [134]. The concept allows the treatment process. This is the case for example for the use of hot flue
identification of energetic discharges. Beside the reduction of these gases of one process to preheat metal bolts, which have to be
energetic discharges, the concept can identify potentials for energy processed, in a different process. Therefore, the flue gases flow in the
based machine ecosystems. counter direction of the material flow. Thus the energy contained in
The exergy analysis allows the rational comparison of different the flue gas is reused [14,154]. Another option is the heat recovery of
energy forms and different process structures to identify optimal hot off gases in steel foundries. Due to constructive measure at the
process parameters. In order to fulfill this task, the exergy analysis furnaces, a twin shell system can be built for heat recovery. In this
calculates exergy losses on the basis of entropy functions of all system, iron melts in one half of the furnace while the other half is
process streams. The height of the exergy losses depends on the being loaded. The hot off gases from the melting side are piped to the
difference of the current system state to the state of the other side to pre-heat the charge. Precondition in this context is that
environment, the so-called ‘dead state’. In the dead state a the temperature level of waste heat is higher than the required
thermodynamic equilibrium of the considered system with its process temperature of the heat utilization. Furthermore, there
environment exists. Furthermore, there is no possibility of a should be a temporal concurrence between heat supply and demand
spontaneous change within the system or the environment, nor and the heat source and sink should be located at close distance [14].
can there be an interaction between them resulting in work [134]. Beside the multi-machine utilization of industrial waste heat, there
When using the exergy analysis to improve process structures, the is a possibility to supply the corporate internal heating network via a
components of the process system have to be analyzed and the heat exchanger with thermal energy in cases where ambient
whole system has to be compared with various different process temperature requires conditioning of the work environment.
alternatives [162]. Thus, the exergy analysis allows to determine Another example of a multi-machine ecosystem is the
the location, type and magnitude of the dissipation of energy generation of electricity from industrial waste heat by using a
[3,15,21,191]. Rankine cycle (RC) [6,186]. The Rankine cycle is a thermodynamic
The advantage of the exergy analysis is that it can help to locate cycle with the greatest efficiency for the conversion of low
energetic discharges in a system, which either are not identified or temperature heat into electricity. A modified version is the organic
misevaluated by the energy analysis. Moreover, these discharges rankine cycle (ORC). In this case an organic fluid is used instead of
can be analyzed by calculating the exergy destruction rates under water as the working fluid. Especially for low temperature
changed process parameters and afterwards comparing it with applications the ORC has more benefits than the RC due to better
rates of the original process [134]. specific properties of the organic fluids. Larjola [117] demonstrated
Renaldi et al. [167] presents a comparison of different exergy that the ORC-process can use the exhaust air of a process (e.g. gas
efficiency definitions for the comparison of three different turbine, blast furnace, etc.) as input. The output of an ORC-process
manufacturing processes on the basis of an exergy analysis. The is electric power, produced from the given heat source. In order to
596 J.R. Duflou et al. / CIRP Annals - Manufacturing Technology 61 (2012) 587–609

determine the electrical output of the ORC-process, the exergy Gomes et al. [65] investigated the use of compartmentalized
analysis can be used to calculate the cycle process. metalworking fluids for treatment and cleaning of foundry sand. In
The utilization of a thermoelectric generator (TEG) can be this case, fluid emulsion (water and mineral oil) that was ready to
considered as another possibility in recovering industrial waste heat be disposed was instead divided through a separation process into
[16]. The TEG is based on solid-state thermoelectric materials which the respective components of water and mineral oil. The treated
are able to directly convert heat into electricity due to the Seebeck water was used within the company as a substitute for tap water.
effect. It is a potentially preferable approach especially if there are The treated water has, compared to the tap water, higher values in
difficulties in effectively transporting the waste heat from the source terms of hardness, alkalinity, nitrates, etc., but was however still
to a separate energy conversion system [58]. The efficiency of TEG suitable for use as cleaning fluid of foundry sand. But Gomes et al.
devices can be expressed by means of the dimensionless figure of [65] also point out that the implementation of this cascade
merit ZT. At present, the implementation of TEG is still limited in utilization is limited. On the one hand due to limited availability of
appliances where the durability and maintenance-free operation are treated water, and on the other hand due to the limited economic
among the most important criteria [16]. Applications in more and environmental potential in consequence of the high intensity
abundant products, such as vehicle exhaust systems, are foreseeable of energy and resource consumption during the treatment process.
in the near future along with an increase in TEG performance. A
study by the U.S. Department of Energy concluded that applications 3.2. Process chain design and control
in aluminum smelting, glass manufacture and cement production
can be practical at a ZT value of two [80]. As an illustration, in one of The load profiles of single machines add up to a cumulative load
the latest research outputs in TEG, researchers managed to achieve a profile for the process chain and determine the embodied energy of
ZT value of 1.8 at 850 K [155]. a product. As different studies show, besides other relevant
Another case of recovering energy is usage of industrial waste objectives (e.g. output) the specific energy and resource con-
heat for preheating necessary media in other processes. As sumption behavior of a process chain can be significantly
example, [196] assesses the effect of reusing heat that is occurring influenced by its specific technical configuration (design) and
in the context of large scale compressed air generation for a control [82]. This includes the individual selection/combination of
weaving mill. It is used to preheat steam, which is needed for the processes/machines and their interlinkage (e.g. process chain
sizing process just before the weaving, resulting in significant structure, buffers) as well as aspects like batch sizes, scheduling of
energy savings of over 20%. orders (e.g. start time, capacity allocation) or speed of production.
Methods and tools for supporting an energy aware design and
3.1.2. Material flow and machine ecosystems control of process chains are provided by [17,82,153,173,218,219].
Similar to the energy flows, the material flows in a multi- An example for influencing the process chain energy demand
machine ecosystem can be utilized to increase material efficiency. through an organizational measure (in this case shifting of an
The material based output of one process is the input for another. order) is shown in Fig. 14 [163].
A possibility to describe the material flows in a multi-machine In general three different perspectives can be considered when
environment is the use of the input–output analysis. This concept talking about energy aware process chain design and control.
provides the foundation to identify potentials of material based Firstly, the electrical work can be reduced, for example by optimal
machine ecosystems and is described in more detail in Section 6.2. utilization of equipment and avoiding energy waste in idling
With regard to the identification of input–output flows on the unit machines or the selection of appropriate machines for the specific
process level, a description and detection tool is the material/ manufacturing task. A second aspect, the avoidance of consump-
energy input–output analysis concept. In this analysis the tion peaks is an important issue. From an economic perspective,
monetary units are substituted by material and/or energy units peaks should be avoided since they may cause cost surcharges in
[11,118]. The concept is commonly used to analyze the flows an electricity bill. However, peaks of consumption are also not
between different industrial sectors of an economy. In this context favorable in context of dimensioning and control of supporting
the method records material and/or energy flows among the units processes (e.g. the compressed air system) while this may lead to
within the company or the supply chain [31,73,145,187]. Thus, this higher energy consumption at the end as well. Finally, another
concept offers a starting position for the material flow-based option – again rather from an economic point of view – is the
analysis of a multi-machine ecosystem. shifting of consumption from day to night, because of less
expensive energy price rates (e.g. base time at night) [199].
3.1.2.1. Example of resource flows in a multi-machine ecosystem. In
the following, two examples of a material flow-based multi- 3.3. Process chain simulation
machine ecosystem are presented. Further examples are given,
among other, by Munoz and Sheng [139], Konijn et al. [113], As pointed out above, the energy and resource consumption of
Sutherland and Gunter [189] and Xue et al. [223]. process chains is rather dynamic depending on the state of the
Logožar et al. [126] present a first step into the direction of a
multi-machine ecosystem in the aluminum production process. As
part of the melting process and the following processing steps to
produce aluminum tubes, rods and bars, aluminum scrap is
generated. To increase the resource and energy efficiency the scrap
was collected within the different units and reinstated as an input
flow in the melting process. The advantage of this approach is that
the remelting of recycled aluminum requires only about 5–10% of
the energy used for primary production [160].
Another example of even more direct reuse of waste flows is
presented by Tekkaya et al. [66,151,193]. The authors investigated
the reuse of aluminum AA-6060 milling and turning chips in direct
hot extrusion. This research has shown that using billets made of
AA-6060 chips can lead to similar mechanical and microstructural
properties as use of conventional cast aluminum billets. Investiga-
tion of material removal by cutting or drilling from the extruded
profiles has even demonstrated improved properties due to a
reduced chip length. Fig. 14. Influence of PPC on energy demand [163].
 J.R. Duflou et al. / CIRP Annals - Manufacturing Technology 61 (2012) 587–609 597

Fig. 15. Paradigms for simulating energy flows in manufacturing systems [86,196].

machines and their interactions. To cope with those dynamics

when designing and controlling process chains, simulation is a
promising approach [82,86]. Simulation can be used for single
process chains and, if incorporating several chains and additional
related aspects like technical building services, the factory as a
whole (Section 4). Available commercial manufacturing system
simulation tools do not consider these aspects yet. However, in
research first approaches can be found (e.g. [45,70,79,90,92,
101,102,125,128,161,181,184,213,215,217,222]). They aim at aug-
menting material flow oriented discrete event simulation (DES) Fig. 16. Production facility as a complex control system [82].

with environmentally relevant energy and resource flows. A

comprehensive overview and discussion of those approaches can The complexity and sophistication in the organization of manu-
be found in [196]. In general three different basic paradigms can be facturing systems and processes necessitate a keen understanding
distinguished, as shown in Fig. 15, which differ regarding the of the organization for accurate environmental analysis. To assist
embedding of evaluation schemes or the interfaces with other in this effort, manufacturing can be broken into ‘levels of study’
simulation tools. across two orthogonal frameworks, spanning organizational and
This inherent logic also leads to quite specific characteristics temporal levels.
with distinctive advantages or drawbacks of each approach. The different levels of manufacturing systems as proposed by
Simulation approaches pursuing paradigm A offer relatively good [165] were provided in the overview in Section 1.4. The primary
coverage of manifold energy and resource flows, comprehensive focus here is on the third level – the facility. Herrmann and Thiede
evaluation schemes with relatively low modeling/simulation and [82] present a holistic view on a production plant (facility)
good transferability. As a drawback, certain energy oriented considering three main partial systems: the production system
dynamics and interdependencies cannot be considered in detail. itself (with interlinked machines and personnel controlled through
In contrast, paradigm B allows very detailed analysis of different production management), the technical building services (TBS)
subsystems and fields of action, but leads to quite complex models and the building shell. The partial systems together with their
which require knowledge and effort and are hardly transferable. dynamic interdependencies are shown in Fig. 16.
Paradigm C as a ‘one-stop solution’ can basically overcome some An equally compelling orthogonal view of manufacturing can
drawbacks of the other paradigms. However, the user is potentially be made based on temporal activities – through the design to
restricted by possible limitations of the utilized simulation manufacturing life cycle. This starts with product or process
environment which often do not allow combining the necessary design, and continues through the design of the manufacturing
logic to integrate dynamic energy consumption and the strong process, process optimization, and finally post-process finishing
discrete event and material flow oriented perspective [86]. and abatement. These temporal levels characterize a degree of
control over the total environmental impact. They also affect
4. Factory level facility level consumption and impacts. These four levels are as
follows [165]:
Most studies and experience indicate that, to be complete, one
must consider the operation of the factory, at a factory or plant level,  Product design: The earliest in design and manufacturing. At this
as part of the efforts to reduce the impact and increase the stage there is the most opportunity to influence environmental
effectiveness of manufacturing. Advances towards the next gen- impact and decisions throughout all future stages. At this level,
eration of manufacturing require the development and promotion of critical decisions on part precision, materials, and design for
‘a holistic understanding of manufacturing’ [93,94]. This requires the assembly/recycling are made. Here there is scope to design the
capability of holistic simulation involving technical building services product as well as its manufacturing process to satisfy specific
and building climate, production machines/material flow, and requirements in all applied criteria.
production management, including production planning and  Process design: The product design is fixed; however here a
scheduling. With this simulation capability, linking energy effi- manufacturing process to suit this design is created. Flexibility to
ciency to the other important parameters in manufacturing is optimize the system is limited to known tools and processes that
critical. Synergies should be identified and encouraged, for example work with the specified design. Here there is extensive control
better buildings, well optimized support facilities, effective main- over the performance of the process in all the criteria as allowed
tenance, diagnostics, improved process quality, better safety and by the product design.
new service concepts, all the way up to new business models.  Process adjustments: The basic manufacturing process is fixed but
small changes to the process through process parameter
4.1. Factory level considerations selection and optimization are used to control the critical
features such as precision, burr formation, and energy or
4.1.1. Defining levels of production in a facility – spatial and temporal consumable demand.
A key requirement for insuring that analysis and optimization  Post-processing: Post-process finishing and abatement processes
are actually effectively employed for planning and scheduling is to are used in controlling the part-precision and the environmental
allow information to flow to where it is needed, e.g. for enterprise impact; at this level there is no control over the process as it has
level and cross-enterprise optimization of energy consumption. already been designed.
598 J.R. Duflou et al. / CIRP Annals - Manufacturing Technology 61 (2012) 587–609

Fig. 17. An integrated view of manufacturing design levels and the decisions they
contain. Arrows represent the flow of information from one decision to another.
Adapted from [165]. Fig. 18. Schematic of strategy for increasing energy efficiency in manufacturing
companies spanning process to facility level consumption and impacts [82].

Fig. 17 illustrates the interaction between the temporal and propose a general multi-objective mixed integer linear program-
spatial levels described by Reich-Weiser et al. [165]. Moving up ming formulation for optimizing a shop schedule. The method
and to the right in the figure means a loss of decision-making considers both productivity (e.g. make span) and energy (e.g. peak
flexibility and, hence, an inability to control outcomes based on load and carbon footprint) related criteria. This new methodology
planning. considers the speed of operation as an independent variable, which
From these hierarchies – which span temporal and organiza- can be varied to affect the peak load and energy consumption. This
tional levels – one can get a sense of the complexity involved in is most useful in situations in which the production machine tools
information capture and transfer in manufacturing systems. For often consume significant amounts of energy even in idle mode
effective decision-making, one needs to understand both what (and perhaps little additional energy when processing). So, idle
quality and quantity of information needs to pass between the time must be considered in the product flow determination.
levels and how decisions early on will percolate through the spatial Energy and resource utilization along the process chain for an
and temporal levels. automotive gear train manufacturing facility were studied by
The view shown in Fig. 17 points out two critical needs: firstly, Schlosser et al. [175]. This included both the core processes (such
factory wide planning and scheduling methodologies with the as machining, heat treating, grinding, deburring, and washing) as
ability to accommodate complex interactions, and secondly and well as the various resources provided to the chain by the facility
complementary to this, monitoring and data communication (such as plant ventilation, water, electricity, and compressed air)
strategies and methodologies to be able to track the facility and facility wide provision of consumables for such a high volume
performance over these spatial and temporal axes. production operation (e.g. process water, central oil supply,
nitrogen and other gases, and propane). The components of this
4.1.2. Production planning and scheduling at the factory level production facility addressed were at the first three levels of
The literature on factory production planning and scheduling manufacturing, corresponding to Sections 2–4 in this paper. The
methodologies is rich. More recently, researchers have started broader enterprise aspects were not included.
addressing the inclusion of energy and other environmental
metrics in the methodologies [57,82]. Herrmann and Thiede [82] 4.2. Energy management: load control and peak load minimization
point out that in addition to the energy consumed directly by
production machinery, there are a number of building services that The previous section described strategies for balancing energy,
also account for energy consumption in support of production and consumables and production constraints at various levels within
the demand for higher productivity (cost, quality, and time) must the factory. To achieve this, the primary concerns are with energy
be balanced with facility management energy optimization since management and, as a corollary, monitoring and control of energy
the cost of energy can be high enough to alter the plan for best consumption. The ability to actually manage these resources is
facility operation. This can be done by use of simulation tools dependent upon knowing the present state of consumption,
according to a stepwise procedure. historical consumption and the responsiveness of the system to
Fig. 18, illustrates the procedure starting from the process chain changes (planned or otherwise).
level of detail, including cycle times and availability. Next, analysis One of the goals of facility energy management is minimization
of production machines with respect to all relevant energy and of peak load surcharges (cost-management) and the appropriate
consumables input to or output from the machine (e.g. cooling sizing of infrastructure and the power distribution as well as
water, compressed air, heat) are accounted for. The historic load meeting constraints of infrastructure and technical building services
profile for the facility and reality of pricing for peak and off peak (e.g. heating, ventilation and air conditioning). This requires data for
energy consumption (electricity, natural gas, oil, etc.) are to be process and system characterization as well as continuous
documented. Finally, taking into consideration the interdepen- monitoring. There are, not surprisingly, different approaches to this.
dencies of the machines, building and facility, the best operating As mentioned in Section 2.2, Kara et al. [105] provide an up to
environment and production plan can be determined based on date overview and background in the developments and chal-
simulation. The authors mention that some of the measures for lenges in electricity monitoring and metering systems and
energy efficiency may conflict with manufacturing target criteria associated standards. Focusing on reducing the energy consump-
like throughput or availability. Further work [195] describes the tion of machine tools or other production machinery as a means for
energy oriented manufacturing system simulation approach in significantly improving the environmental performance of man-
detail. More examples of this methodology and a review of other ufacturing systems, Vijayaraghavan and Dornfeld [212] proposed a
related work were presented in [86]. methodology, based on a communication interface standard
Challenged by the potential conflict between process efficiency MTConnect for manufacturing, for monitoring energy consump-
(time based) and factory efficiency (energy based), Fang et al. [57] tion patterns in larger systems. It is vital to correlate energy usage
 J.R. Duflou et al. / CIRP Annals - Manufacturing Technology 61 (2012) 587–609 599

traffic and material flow as well as location and insulation can have
a large impact on the energy consumption for the structure
containing the manufacturing operations.
Commercial software is available for comprehensive factory
layout and process and material flow simulation (e.g. Autodesk
Factory Design Suite, Siemens FactoryCAD and FactoryFLOW) help
create more efficient factory layouts based on digital models of the
factory. These tools allow engineers to analyze existing and
proposed plant and system layouts for more efficient material
flows and scheduling and routing in production systems. Energy,
materials and other resource flows can also be tracked. Unlike the
optimization tools discussed earlier in this section, these factory
design tools are meant to compliment architectural design
functions for a complete facility design.
The concept of life cycle management of production facilities
was introduced [74] to address the design, installation, operation,
adaptation and disposal of production facilities. Topics addressed
in this research include maintenance, re-use or reconfiguring of
production facilities or removal and reconstruction at another
Fig. 19. Schematic of various energy control loops at the different levels of the location. As with product life cycles, the life cycle of a production
facility [211].
facility deserves attention.

with the operations being performed in the manufacturing system. 4.3.1. Energy efficient building technology
However, since this can be challenging due to complexity of The term ‘green factory’ often refers to the factory buildings
manufacturing systems and the vast number of data sources, event designed and built to minimize energy use and to recover resources
stream processing techniques were applied to automate the [48]. Industrial buildings often consume and waste a huge amount of
monitoring and analysis of energy consumption in manufacturing energy in terms of electricity and heat. In particular, when a factory
systems. The technique is extensible to all levels described above. or a building is designed as a new structure, energy efficient building
Adding a contextual aspect to the power demands in industry as technologies can easily be implemented. However, in existing
part of industrial ‘smart metering’ was also proposed [87]. Energy buildings, small and inexpensive changes can lead to large energy
consumption for various industrial processes varies dramatically savings and improve the energy efficiency of the whole building. A
and thus the time dynamic nature of process energy, and the clear example of a factory designed to be energy and resource
energy of subordinate processes and machinery, must be under- efficient, with respect to the building use, is the Ford Motor Company
stood. This can come from metering. Rouge Plant [59].
The event stream processing identifies trends in consumption The following impacts are recognized as significantly con-
of critical resources. Using trend analysis the consumption of tributing to energy conservation.
electric energy, gas, compressed air, technical heat and water can The facades of a building have a number of functions including
be monitored, predicted and used to balance demand if that thermal and sound insulation as well as protection from humidity.
capability is present. For example, if the estimated workload is In particular, thermal insulation for production facilities is not only
rising above a set maximum value within a given time period (e.g. important from a sustainability point of view, but also to provide
15 min), consumers with low priority (e.g. redundant machines or thermal stability for the processing equipment.
heating treatment furnaces with a high thermal capacity) can be Technically, the fenestration includes every opening in a
switched off temporarily to meet the predefined limits. building’s outer shell, e.g. windows, doors and gates. The energy
One must either determine in advance the strategy for energy loss in heating or cooling through these openings can add up to
efficient operation of production systems through a priori optimiza- 30%. In the last decade or so, windows have become much more
tion, or, if the architecture is capable thereof, process controls can sophisticated. Single pane windows have been replaced by double,
decide if it is possible to reduce energy via local control and set triple or even quadruple panes. The layers are separated by
components to energy-optimal states. Verl et al. [211] describe insulating materials, as inert gases, to reduce their thermal
information flows that generate energy control loops at different conductivity. A low emissivity coating on the glass can reduce
levels based on model information, and propose a matching heat loss in the winter and prevent unwanted heat from entering in
communication and control infrastructure. These mechanisms and summer.
control allow the use of models automatically and coherently on all For production facilities, gates or material ports usually tend to
required levels of detail and abstraction. The energy consumption cause the highest energy loss inside the fenestration, since they are
model information is communicated only between the controls of a oftentimes left open to permit a frequent and rapid material flow.
resource and sub-resources. Each control must understand the Rapid action doors open and close rapidly for each crossing and
interdependencies with and within its direct neighbors. The system therefore prevent energy losses. This effect can be enhanced if two
ensures that communication and computational loads are limited rapid action doors are arranged consecutively as a lock.
and balanced even for large systems. Fig. 19, from Verl et al. [211], The inside of a building should be illuminated as much as
schematically shows the various energy control loops that can be possible with daylight, since this is the most natural and energy
realized according to the proposed decentralized approach. efficient type of light. The minimum intensity of illumination
cannot always be achieved only through daylight. Therefore,
4.3. Factory design and operation influences artificial lighting is often necessary. In order to reduce the energy
consumption, a sensor, which measures the intensity of illumina-
Diaz et al. [42] analyzed the energy consumption and CO2 tion, can be installed. A controller in connection with a dimmer
emissions for each life cycle stage of two milling machine tools in regulates the artificial light such that the overall intensity of
different environments. Besides the machine tool energy, HVAC illumination is permanently above the minimum threshold.
and lighting of the working hall were found to be significant,
consuming 40–65% of the total use stage energy. 4.3.2. Example of ‘green factory’ technology – compressed air analysis
Beyond the manufacturing floor and the production systems, In a factory, compressed air is often regarded as the fourth
the design of the factory, its physical layout, construction and utility, after electricity, natural gas and water, in facilitating
600 J.R. Duflou et al. / CIRP Annals - Manufacturing Technology 61 (2012) 587–609

production activities. In manufacturing plants, compressed air is focuses on the ‘smoothness’ of production and constructing a
widely used for actuating, cleaning, cooling, drying parts, and process with the capability to produce the required results by
removing metal chips. However, the cost of compressed air designing out process inconsistency (or ‘muri’). This is to be done
production is one of the most expensive and least understood while trying to maintain as much flexibility as possible since
processes in a manufacturing facility [168]. The cost of electric excessive constraints or rigidity often induce waste (as in excessive
power used to operate an air compressor continuously for a year set up/change over time or high minimum lot sizes, requiring extra
(about 8200 h) is usually greater than the initial price of the inventory or inducing poor response to customer needs, i.e. poor
equipment [107]. Per unit of energy delivered, compressed air is response to ‘pull’.)
often more expensive than the other three utilities. As an example of the second type of lean manufacturing, Comau
Besides the cost issues, compressed air production consumes a [29] recently introduced a ‘smart assembly’ cell focused on the
large amount of energy. In absolute terms it is estimated that about production of high precision complex assemblies as in valve trains
3–9% of total energy consumed in the U.S. in 1997 is for air for auto engines. Citing the large number of individual machines
compression in manufacturing [35]. Yuan et al. [225] detail an and process steps used in traditional valve train assembly (some 72
example of compressed air use in automotive manufacturing parts in one case at a cycle time from 25 to 30 s per machine and
facilities. Compressed air is used relatively indiscriminately in several minutes per assembly in total along with a large capital
automotive manufacturing due to its ease of setup. There is no investment), Comau’s smart machine replaces the entire line by
need for additional maintenance or special machines; the task can four operations and a total cycle time of 54 s. The cell, designed for
be accomplished by adding piping. In addition, as a form of energy, 325,000 cylinder head annually, requires only 223 sq. meters of
compressed air represents no fire or explosion hazard; it is clean floor space compared to 753 sq. meters. Mori Seiki announced a
and safe and often regarded as totally ‘green’ [32]. In a large similar system as well for a smart machining cell [135].
production facility, plant air is the more typical supply and suffers Although the Comau cell design was likely not motivated by
from a number of limitations, including complexity of system, low green manufacturing concerns, the cell will have an impact on
efficiency (typically less than 60% of the total compressed air energy consumption by nature of the reduced number of stand-
consumed contributes directly to the goods and services for which alone processes and, importantly, the tremendous reduction in
production was intended [60]; leaks are a major problem in plant floor space. And, since idle time is a big energy consumption
air supply), energy storage limitations, unstable system pressure, concern in standalone machine tools, the higher utilization will
and associated high cost. This low efficiency translates into excess reduce idle time.
energy consumption. The associated carbon footprint for energy One of the main tools for the first type of ‘lean’ is the Value
use in providing compressed air was analyzed in [100]. Stream Map (VSM) – charting exactly the material and information
Common supply patterns for compressed air include the flow in the system. Rother and Shook [169] define a value stream as
following, plant air: the whole plant is supplied with compressed ‘all the actions (both value added and non-value added) currently
air from the air house, with pipes spread out in the plant to provide required to bring a product through the main flows essential to
compressed air to all facilities, point-of-use: each machine is every product: (1) the production flow from raw material into the
exclusively supplied by an independently installed air compressor, arms of the customer, and (2) the design flow from concept to
and local generation: a certain number of machines are grouped launch.’ This can be a basis for introducing green manufacturing
together and supplied by an air compressor. concepts to the enterprise. It starts with a very careful (and often
An investigation of the compressed air supply methodologies to tedious) assessment of the present state of the production system.
determine which of the three supply options mentioned above This means outlining the processes and flow with the key
could be the most cost effective considering energy and related interconnections and relationships, and collecting process data
operating costs, including equipment purchase and operation, was for each process. Examples of the process data include: cycle time,
performed for an automotive component production plant [225]. changeover time, uptime (on demand machine availability),
Compressed air in the facility studied was used to supply operating production batch sizes, number of operators, number of product
air to a number of CNC machine tools in the facility. An variations, pack size, working time (minus breaks), and scrap rate.
environmental analysis and a cost of ownership study were made Many of these characteristics have green implications, meaning
for the alternative system layouts. The analysis found that a ‘local they are predictors of energy or resource consumption – like cycle
generation’ option would be preferable when compared with point time, which can help define process energy, or scrap rate, which is
of use and plant air options considering both cost and energy an indication of efficiency of conversion of resources into products.
efficiency. Employment of local generation instead of plant air Although the procedures for VSM are well established, the use
could potentially save $2000–$3200 dollars and 95,000 kWh each of VSM in green manufacturing analysis is not so well defined. The
year on the CNC milling machines. Meanwhile, local generation best approach is to use the concepts of VSM and lean to
offers numerous advantages over plant air with regard to compliment the development and operation of efficient manu-
reliability, simplicity, leakage prevention, and flexibility. Local facturing operations with the requirements of reduced energy and
generation is supplied by relatively short pipelines, which may resource utilization – leading towards green and sustainable
lead to a significant reduction of losses due to leaks. Extra local manufacturing. If one reviews Ohno’s ‘seven wastes to be avoided’
compressors may be connected to the CNC machines in parallel, [123] and, realizing that this was developed before the current
which automatically builds a great deal of redundancy into the concern about the environment and green manufacturing was so
system. Furthermore, the scale of local generation compressors commonly of interest, one can observe that many map directly
enables greater flexibility as machines and processes change. onto green manufacturing practice. For example, the ‘wastes’ of
producing more than is needed, or storing/inventorying more than
4.4. Lean and green production influences at the facility level needed, unnecessary transportation, unnecessary work steps or
processes – all can be converted into wasted resources, energy, and
Green manufacturing concepts are often linked to other lean other consumables or the indirect efforts of these wastes (such as
practices in industry. Lean manufacturing is defined by Holweg floor space and HVAC costs, additional tooling and the manu-
[91] as a production practice that ‘considers the expenditure of facture and operation of it, unneeded raw materials and the
resources for any goal other than the creation of value for the end associated imbedded energy, transport, storage and recycling).
customer to be wasteful, and thus a target for elimination’. There
are a number of approaches to ‘lean’. The first approach is 5. Multi-factory level – industrial symbiosis
nominally the elimination of waste and uses tools that assist in
uncovering waste in the process and system. A second approach is Moving a further step up the manufacturing hierarchy from the
more aligned with the Toyota Production System (TPS), which factory level, the interaction between industrial facilities comes
 J.R. Duflou et al. / CIRP Annals - Manufacturing Technology 61 (2012) 587–609 601

into scope. This interaction may occur between suppliers and

customers in the supply chain, as described in Section 6 of this
keynote, but it may also occur between economically independent
companies exchanging and utilizing flows of materials or energy to
their mutual benefit, increasing the overall output from the given
input of materials and resources compared to the situation where
each facility would treat its waste streams in the conventional way,
i.e. often just landfilling them. The mutualism in the interaction
has led to the introduction of the term ‘industrial symbiosis’ from
the examples of mutualistic symbiotic relationships between
organisms of different species in natural ecosystems, in which
these exchange flows of materials or protection for work in a way
that is mutually beneficial.

5.1. Definition of industrial symbiosis

A widely used definition of industrial symbiosis has been

formulated by Chertow [24]: ‘Industrial symbiosis engages tradi-
tionally separate industries in a collective approach to competitive
advantage involving physical exchange of materials, energy, water,
and/or by-products. The keys to industrial symbiosis are collabora-
tion and the synergistic possibilities offered by geographic
proximity.’ Fig. 20. Industrial symbiosis at Kalundborg.
Adapted from [190].
The definition focuses on the fact that the industries are
separate, i.e. cleaner production activities involving different
departments or subsidiaries of one company are not considered waste flows totals around 2.9 mio tons [26]. The roots of the
industrial symbiosis (in the context of this paper these are covered industrial symbiosis collaboration at Kalundborg go back to the
in Section 3 at the level of multi machine ecosystems). The nature 1970s when some of the partners started collaborating on the
of the engagement is described in the definition as ‘exchange of reuse of water, but it was not until 1989 that the extent of the
physical flows of materials, energy or water’. In their discussion of symbiotic relationships was realized, and the term industrial
Australian experience with industrial symbiosis, van Beers et al. symbiosis coined.
[209] expand this point of the definition to include the joint use of Since then, other examples of industrial symbiosis have been
utility infrastructure for energy or water production or wastewater identified or developed. A more comprehensive and geographically
treatment. The geographic proximity of the participants in the extended example is a network of collaborations that was
industrial symbiosis is emphasized in the definition, and indeed, discovered in the Austrian region of Styria in the 1990s, comprising
most known examples of industrial symbiosis involve companies a steel plant, power plants, paper producing industries, textile
that are closely co-located. There are, however, examples that industry, cement industry and stone and ceramic industry as the
challenge this part of the definition in discussing the Australian main partners [177]. Examples from the mining and minerals
minerals industry; Schwarz and Steininger [177] report on a large industry in Australia focus on reuse of water, chemicals and
Austrian network of industrial waste recycling in the province of mineral wastes [209]. Other examples are centered around forest
Styria, where some of the participants are at distances that and pulp and paper industries where residual products are used for
necessitate transport of the commodities through pipelines or road energy production, and chemicals are recycled (e.g. [129,183]), or
transport. Chertow’s definition also emphasizes that industrial agro-industry systems like a Chinese example bordering between
symbiosis is a collective approach, meaning that several industries industrial ecology and supply chain integration. Here, an original
are normally involved. Again, the purpose is to distinguish the sugar refinery activity has been expanded with industries using the
efforts beyond simple recycling activities, as carried out by scrap two main by-product streams from the sugar production: spent
dealers for example. As a pragmatic way of deciding on this point, molasses which are used in an alcohol plant, the residuals from
Chertow later writes: ‘To distinguish industrial symbiosis from which are used in fertilizing industry, and bagasse, the fibrous
other types of exchanges, my colleagues and I have adopted a ‘3–2 residues of the sugar canes, which are used in paper production
heuristic’ as a minimum criterion. Thus, at least three different [227,228].
entities must be involved in exchanging at least two different In East Asian countries like China, Korea and Singapore, where
resources to be counted as a basic type of industrial symbiosis. By formal planning is more institutionalized, and in the United States,
involving three entities, none of which is primarily engaged in a industrial symbiosis typically occurs in the form of eco-industrial
recycling oriented business, the 3–2 heuristic begins to recognize parks [26,124,228]. An eco-industrial park was defined for the U.S.
complex relationships rather than linear one-way exchanges.’ [26]. EPA by the Field Book for the Development of Eco-Industrial Parks
[127] as: ‘a community of manufacturing and service businesses
5.2. Practical examples of industrial symbiosis seeking enhanced environmental and economic performance
through collaboration in managing environmental and resource
The most well-known example remains the industrial ecosys- issues including energy, water, and materials. By working together,
tem constructed around a co-generation power plant in Kalund- the community of businesses seeks a collective benefit that is
borg, Denmark (Fig. 20) [190]. Multiple flows are involved, but the greater than the sum of the individual benefits each company
main ones are centered around the exchange of energy flows and would realize if it optimized its individual performance only’. In
different qualities of water back and forth between power plant the United States the US President’s Council on Sustainable
and the other main partners: an oil refinery, two biotech Development, USPCSD in 1996 recommended that ‘Federal and
companies (pharmaceuticals and enzymes), and the municipality state agencies should assist communities that want to create eco-
of Kalundborg (district heating for 5000 homes). In addition there industrial parks that cluster businesses in the same area to create
are flows of flue gas treatment products between the power plant new models of industrial efficiency, cooperation, and environ-
and a plasterboard producer (gypsum) and a local cement industry mental responsibility’ [208]. At that time, 15 eco-industrial parks
(fly ash), and between the biotech companies and local farmers were planned in the U.S., but until now few, if any, have shown
(yeast slurry and sludge). On an annual basis, the exchange of successful and lasting examples of industrial symbiosis [26].
602 J.R. Duflou et al. / CIRP Annals - Manufacturing Technology 61 (2012) 587–609

5.3. Benefits realized through industrial symbiosis by-products, where it is unlikely that disposal at high costs would
continue in the case where no symbiosis were established.
It is clear that an industry will not voluntarily enter and remain Typically, alternative solutions would be developed, eliminating
in an industrial symbiotic relationship with one or more other or reducing the flow or finding other routes of utilizing the flows at
industries if it does not entail clear economic benefits for the lower costs or even commercializing them. Industrial development
company. The economic benefits can be direct through: is full of examples on utilization of what was once a waste stream.
Indeed, this kind of activity has been the nucleus around which
 Reduced costs of raw materials on the input side, replacing costly several of the known examples of industrial symbiosis have grown.
virgin resources or materials by waste that is free or very cheap Finally, the net environmental impact of the symbiosis is
to acquire from the symbiosis network. determined as the sum of the impacts associated to the three
 Reduced costs for the treatment of waste streams (on the output processes, conversion, substitution and avoidance, including both
side) as these are taken over by partners in the industrial direct impacts and indirect impacts from other activities induced
symbiosis network. elsewhere in society by these processes. In order to determine the
overall benefit in terms of environmental impacts and resource
In addition to these financial benefits, there are more indirect use, there is a need to apply a life cycle perspective on the induced
benefits through: changes and identify any problem shifting to other parts of the
involved product chains and between categories of environmental
 Better access to and supply security for resources that are impact [18,75,183].
restricted. This is particularly the case for water, but in the future
it is also potentially the case for waste flows containing other 5.4. Hurdles and enhancers, constraints and concerns
strategic resources like metals.
 Brand value, satisfying important stakeholders among custo- Examining the existing examples of industrial symbiosis, it
mers, environmental regulators, local community members and becomes clear that they are often constructed around one or a few
employees [209]. central actor(s), with several more peripheral partners exchanging
few and minor streams. In areas where industrial water is a scarce
For most companies, the economic and strategic benefits will be resource, the exchange and reuse of this resource can be a central
the drivers, but the question can be posed whether there are also element in the symbiosis. When a central partner is a power plant,
environmental benefits, and if so, how large these are. some of the partners will typically utilize waste heat and steam
Industrial symbiosis typically allows an industry to substitute a from the power plant. Depending on the nature of the partners,
high quality virgin resource (material, energy carrier or water) by a other flows that are exchanged in the symbiosis will involve waste
lower quality substitute (waste) that per definition is available in products, that can be reused and replace other resources for the
excess of demand in the region. This substitution involves three users, often using (down cycling) organic materials as fuels in heat
processes: conversion, substitution and avoidance [210]. production. In the systems described above, examples of
exchanged waste streams include gypsum, waste paper, textile
 Conversion involves collection of the waste materials, energy or waste, fertilizing sludge, used solvents, wood residuals and
water from the symbiosis partner, processing into a suitable hydrogen gas.
feedstock, and transport to the site of application. The question can be posed which factors determine the
 Substitution involves the use of the generated alternative emergence of industrial symbiotic networks. Boons et al. [18]
feedstock, including possible adjustments of the process to proposed the following list of societal mechanisms for enhancing
compensate for lower quality inputs. the development of industrial symbiosis:
 Avoidance relates to the reduced or eliminated disposal of the
waste that has been recovered.  Coercion, forcing industries to adopt industrial symbiosis
through legislation.
The conversion processes entail overall negative environmental  Imitation, when industries adopt industrial symbiosis routines
impacts as they require facilities, energy and possibly other and concepts that they see in other industries for reasons of
resources and ancillary substances. Substitution, on the other status or to deal with uncertain situations (e.g. access to
hand, has indirect environmental benefits from helping to avoid resources).
resource extraction and processing, but it may also cause negative  Private interest government, where a group of industries choose to
direct environmental impacts, since the use of lower grade inputs collectively adopt industrial symbiosis voluntarily, because of
often requires more processing energy and produces more waste. the threat of legislation if they remain inactive.
Avoidance normally has only positive environmental impacts due  Demonstration projects, where policy actors initiate projects
to the reduced disposal of the avoided flows of waste or heat [210]. experimenting with industrial symbiosis, and actively spread the
To quantify the environmental impacts of the industrial results as ‘best practice’ to accelerate its diffusion.
symbiosis, typically a comparison is performed to a reference  Training and professionalization of individuals in the concepts
system without industrial symbiosis. In this way, annual savings and routines of industrial ecology who then subsequently start to
for some of the central flows of the industrial symbiosis at apply these in their work environment.
Kalundborg were quantified to 0.5 mio m3 surface water,  Altering boundary conditions, to stimulate actors within regional
23 mio m3 of seawater and 1.3 mio GJ thermal energy (steam industrial systems to self-organize into industrial symbiosis
and hot water). In a similar approach, Chertow and Lombardi [25] networks.
demonstrated savings of both costs and effluent discharges for an
industrial symbiosis involving substitution of oil-fired on-site While Boons et al. [18] put some faith in regulatory
steam generation in Guayama, Puerto Rico (USA), at a petro- involvement in the development of industrial symbiosis, Chertow
chemical complex by steam import from a coal-fired power [26] is less optimistic in this regard. Comparing the US
station. Emissions of SO2, NOx, and PM10 were reduced, but experience, trying to plan eco-industrial parks or industrial
emissions of CO and CO2 increased. symbioses, with the Kalundborg symbiosis case, the author draws
The approach holds some methodological problems [210]: the two conclusions on why the latter has produced a successful and
benefits may be considered confidential and hence difficult to lasting industrial symbiosis and the former not: ‘First, we see that
determine because they are commercially sensitive. The use of the rather than resulting from planning or a multi stakeholder
stand-alone operation without symbiosis as reference scenario is process such as the ones pursued through the U.S. President’s
questionable, particularly in the case of large streams of waste or Council on Sustainable Development, the Kalundborg symbiosis
 J.R. Duflou et al. / CIRP Annals - Manufacturing Technology 61 (2012) 587–609 603

existing processes is improved by industrial symbiosis (otherwise

this would not occur in the first place) and hence this important
element in the overall incentive to introduce new and cleaner
technology is weakened. In this way, industrial symbiosis has the
potential to act both as a driver and a barrier for the development
towards a sustainable industrial society, depending on the context
and the nature of the involved industries.

6. Supply chain level

A supply chain can be defined as ‘a set of three or more entities

(organizations or individuals) directly involved in the upstream
and the downstream flows of products, services, finances, and/or
Fig. 21. The progression of industrial symbiosis relationships.
Adapted from [24]. information from a source to a customer’ [83,130]. In contrast to
the organizational/spatial hierarchy covered in the previous
sections, the focus here is on the provided good or service and
emerged from self-organization initiated in the private sector to its flow through the system. Transformation steps taking place in
achieve certain goals., such as cost reduction, revenue enhance- single companies/factories are therefore interlinked by transpor-
ment, business expansion, and securing long-term access to tation processes entailing environmental impacts and use of non-
water and energy. This implies that the symbiosis was not ‘seen’ renewable resources. Based on this, sustainable supply chain
by outsiders because the exchanges emerged from the invisible management has been defined as ‘management of raw materials
hand of the market rather than direct government policy or and services from suppliers to manufacturer/service provider to
involvement. Second, once a revelation was made, a coordinative customer and back with improvement of the social and environ-
function was found to be helpful in organizing more exchanges mental impacts explicitly considered’ [143,158].
and moving them forward’. Also in the Austrian and Australian
examples of successful industrial symbiosis networks, there has 6.1. Country specific variables
been little active regulatory involvement in the development
[177,209], and a conclusion seems to be that the industrial Based on the definition of the manufacturing system and
symbiosis is a difficult thing to plan and stage by policy measures. efficiency indicators, three major country specific factors with an
Chertow [26] concludes that ‘. . .many of the successful industrial influence on energy efficiency can be identified: climate, distances,
ecosystems . . . did not arise in the ways pursued by the PCSD. One and energy sources and associated price structure. Considered as
feature that several of these have in common is the experience of a national averages, other factors, such as worker qualification,
quiet period where firms engaged in exchanges among them- degree of automation, age of machines, also significantly differ
selves, unconscious of a bigger picture, followed by an act of between countries. However, these factors are not dependent on
discovery that revealed the pattern of existing symbiotic geographic location. In contrast to the first set of factors, they are
exchanges and the resulting environmental benefits’. In the transferable and their actual state depends on the specific case of
Danish case however, the role of public policy cannot be each manufacturing company.
completely dismissed since a rather ambitious environmental
policy must have played a motivating role in the development of 6.1.1. Energy source and energy cost
the Kalundborg symbiosis, with green taxation leading to high The different means of electricity production and provision vary
prices of both waste treatment and water extraction and later also strongly in terms of technology, use of fuels and impacts on the
energy. Building on the observation that a period of self- environment. Conventional thermal energy generation by incin-
organization before being discovered as industrial symbiosis eration of non-renewable resources, such as coal or gas are among
and marketed and expanded as such, was instrumental for the the most important contributors to emissions of greenhouse gases.
formation of the successful cases, Chertow [26] presents the Nuclear power generation contributes to the dispersion of radio-
model in Fig. 21 for the progression of industrial symbiosis nuclides and increased background radiation in the environment
relationships. She proposes a role for government and policy in: and has unresolved problems with the storage of radioactive
waste. Renewable sources, such as wind, water or solar power all
 Help to identify and publicize already existing kernels of tap the energy inflow from the sun. Wind and sun are not
cooperative activity in the form or symbiotic relationships, associated with any emissions in the use stage, but have issues like
 Support such emerging kernels and help them expand by land occupation and visual impacts that influence their propaga-
involving other industries in the region, and tion. Hydroelectricity has impacts on riverine ecosystems and the
 Provide incentives to catalyze the formation of new kernels by creation of huge dams at hydropower stations causes submersion
identifying promising precursors to symbiosis, but abstain from of existing vegetation that leads to substantial emissions of the
‘supporting projects, through public or private investment, that greenhouse gas methane in countries like Brazil. Fig. 22 shows the
have much wishful thinking but no tangible kernels to roast’ [26]. energy mix composition for the electricity net generation in
different countries worldwide [204].
With its name derived from a biological analogy and its central Significant differences can be observed between countries
role in the whole concept of industrial ecology [24], industrial largely depending on conventional thermal energy generation
symbiosis with its exchange of waste streams between indepen- with certain high GHG emissions, such as Australia or Saudi Arabia,
dent industries would seem to be an important element in the way and countries that already mainly switched to renewable energy
towards a more sustainable industrial society. But a relevant sources, like Brazil or Norway. Thus, energy consumption in
concern is whether the focus on exchanging waste streams moves specific countries is associated with a specific environmental
the focus from a more traditional pollution prevention principle impact depending on the sources.
where waste should be prevented rather than treated and wasteful The energy price is another important variable that needs to be
production processes should be substituted. And whether the considered in this category. As depicted in Fig. 23, since 2000
contractually based exchanges of waste flows in this way serve to electricity prices show a steady trend upwards which is mainly
perpetuate a wasteful production rather than motivate develop- caused by rising price of fossil fuels (e.g. oil and gas) [83].
ment of cleaner processes that avoid the waste production Studies show that this general trend is very likely to continue in
altogether. At least it is clear that the economic performance of the coming years despite temporal fluctuations due to changing
604 J.R. Duflou et al. / CIRP Annals - Manufacturing Technology 61 (2012) 587–609

6.1.3. Local climate

Referring to the definition of the Intergovernmental Panel of
Climate Change (IPCC), climate is ‘the average weather or more
rigorously the statistical description in terms of the mean and
variability of relevant quantities over a period of time. These relevant
quantities are most often surface variables such as temperature,
precipitation and wind’ [83]. Depending on the geographical location
on the globe, the climate naturally differs between regions. In the
case of air conditioned production, the local climate naturally has an
impact on energy consumption of technical building services (TBS).
For instance, while striving to keep certain production conditions
constant, higher outside temperatures are likely to cause higher
energy consumption for cooling. However, general statements about
the actual effect are difficult to formulate as they strongly depend on
the very specific case (e.g. dimension and isolation of building shell,
Fig. 22. Electricity generation in 2008 by type and country (top 20 countries) [204]. configuration of TBS, actual progression of temperature and
radiation, internal heat loads, as described in Section 4).

6.2. Global supply chains

In the context of global supply chains, the focus is not only on

the energy efficiency of manufacturing systems, but also on the
energy embodied into producing a product and the associated
environmental impact. At this stage, a product life cycle view is
critical since each product life cycle stage can be carried out in
different geographical locations. Therefore, the aforementioned
country specific variables may vary significantly along the product
life cycle. As a result, ‘embodied energy’ is used more widely as an
objective measure since the types of primary energy used may vary
Fig. 23. Evolution of electricity prices in European countries [62,83].
from electricity to coal and petrol [103,104].
In this context, there are two principal approaches to address
the problem of including environmental impact considerations in
supply chain design decisions. The first one is the extension of
economic conditions. However, Fig. 23 also underlines that energy traditional supply chain design with environment as an additional
prices strongly differ between countries due to the specific objective. In this context, traditional business indicators, such as cost,
structure of energy supply (e.g. availability of energy sources), time and quality are evaluated concurrently with the environmental
market conditions (supply/demand and competition situation) and objectives. Then multi-criteria optimization techniques are used to
political/legal background (e.g. taxes and subsidies). Therewith optimize the supply chain network design. Implementation in a
local energy prices are naturally a major location factor, which company of the ISO 14001 standard [97] on environmental
directly affects the cost structure of manufacturing companies. management supports the availability of data on environmental
parameters and also gives requirements on the management of the
6.1.2. Transportation supply chain [99]. Use of one of the existing LCA tools (e.g. SimaPro,
On the supply chain level, especially the distances to be covered Umberto or GaBi) may support estimation of the embodied energy
during transport are of major influence for the energy consump- and the associated environmental impact [75,116].
tion. This is realized using different modes of transportation, e.g. As a second approach, simplified input–output tools combined
truck, rail, air, or ship. The input of these processes is mainly with detailed process LCA analysis can be used to evaluate the
energy, i.e. the fuel for the different transportation modes. The embodied energy of different manufacturing supply chains [104].
output is on the one hand the transportation performance Later, other objectives can be added, such as time and cost, in order
described by the mass of the freight transported over a specific to do trade-off analysis in between various conflicting objectives.
distance. On the other hand, especially the use of non-renewable One of the challenges of process LCAs is the so-called whirlpool
fuels is associated with emissions to the environment, such as CO2 effect caused by (nearly) infinite feedback-loops in the system. As
and other greenhouse gas emissions, and NOx and particulate an example, production of manufacturing equipment requires
matter with negative impacts on human health. Besides the freight energy (e.g. coal), but to produce this energy manufacturing
task in terms of lot sizes and volume, especially the distances industry is needed to produce tools that can be used to extract the
between the single actors in a supply chain and the mode of fossil fuels and so forth. Modeling of these loops typically involves
transportation are of particular importance for the energy a cut-off after a few iterations, but at the macro-economic level,
consumption in supply chains. Actors within a supply chain statistical input–output analysis (IOA) is a powerful tool to include
may either be located within the same country as others, or in whirlpool effects between the different sectors of the economy in
another country since single manufacturing steps are nowadays the analysis of the impacts from industrial production.
allocated globally based on availability and costs of resources, Input–output analysis in the current form was founded by
production factors, human resources and expertise as well as based Leontief [120] on the basis of Quesnay’s concept ‘Tableau
on the location of markets and customers [28]. The specific fuel Economique’ [159]. The method describes and analyzes the
type and fuel efficiency of the chosen transport mode is a interactions among different sectors of a network [188,200]. In
determining factor for the energy consumption in supply chains as this network, a general equilibrium model of all components of a
well as the associated environmental impact. Furthermore, the given economic system is formulated. The interdependence among
location also determines available modes of transportation. An the different parts of a given system is described by a set of linear
island location, such as Australia or Japan, allows only air or ship equations [115]. Leontief’s original method considered only the
freight transport for international trade. On the contrary, inter- monetary flow from a given amount of output backward to the
national trade within Europe can mainly be conducted using input and it is used for the examination of economic activities
domestic shipping, rail, and road transportation [83]. [11,50]. In recent years, through various scientific contributions,
 J.R. Duflou et al. / CIRP Annals - Manufacturing Technology 61 (2012) 587–609 605

the economic input–output tables have been combined with predicting energy and resource flows for larger manufacturing
information on resource use and environmental emissions linked systems.
to the economic flows to support analysis of the environmental  Technical building services can consume considerable amounts
impacts of different sectors of the economy [19,55,76,118, of energy: it is obvious that factory layout and facility
132,224]. The environmentally enhanced input–output analysis optimization need careful attention in a design stage.
(EIO) has the strength of solving the whirlpool effect but the  Just like residential buildings, production facilities need to be
quality of the environmental information is deficient for many constructed according to state of the art building physics
environmental impacts. The emission of CO2 is one of the best principles, thus minimizing energy inputs for HVAC conditioning
covered aspects and this has been used to investigate the of the work environment, while taking into account local climate
consequences of moving production from industrialized nations conditions.
to developing countries such as China [89,122,180].  Similar to the multi-machine strategy, production planning can
Another possible approach is to use a hybrid of product LCA and be optimized at a facility wide level in order to limit the total
EIO analysis [164]. In this case, a matrix approach can be used to energy consumption.
determine what needs to be measured (either time or cost and either
green or sustainability), what was the geographic scope (local, At multi-facility level the following conclusions can be drawn
regional, or global scope), and what was the manufacturing scope with respect to industrial symbiosis opportunities:
(product, machine/device, facility/line/cell, supply chain, or life
cycle scope). Then the iterative financial hybrid life cycle assessment  The geographic co-location of production plants with possible
(LCA) can be used in order to set up key areas as contributors to the synergies in terms of waste streams facilitates the exchange of
GHG emissions and to estimate the electricity emissions in a the physical flows that are involved. As such industrial symbiosis
particular location while the transportation emissions of goods may help to reduce industry’s need for treatment of solid and
between locations can be estimated by multiple sources. liquid waste and increase the overall efficiency both in terms of
energy and material resources.
7. Conclusions  Reported industrial symbiosis examples indicate that the
principle has mainly been tested in process industry and is
A wide variety of considerations with relation to environmental hardly recognized in discrete manufacturing sectors.
impact reducing measures in general and energy and resource  The focus on exchanging waste streams may increase the risk for
efficiency in specific has been discussed and identified methods reduced attention for traditional pollution prevention activities
and techniques for analysis and system optimization reviewed. At aiming to avoid the generation of waste in the first place, e.g.
each system level a number of generic conclusions can be drawn. through internal recycling.
 The most effective way of strengthening industrial symbiosis is
For the unit process level it can be concluded that:
to increase the economic motivation. Regulatory means thus
include taxation of the use of virgin resources and higher fees for
 Redesign of machine tools and selective control can significantly
the treatment of waste.
increase the energy efficiency without affecting productivity.
Case studies illustrate improvements with a factor 1.2–3. The
Considering the supply-chain level the choice for specific
obvious lack of priority for energy efficiency among machine tool
production locations clearly influences the energy embedded in
builders leaves on average substantial space for improvement
a product and the related environmental impact:
requiring only well-known methods and techniques.
 When allocating machine tools during process and production
 While beyond the control of individual manufacturing compa-
planning, an appropriate choice at near nominal capacity level is
nies, regional electricity generation practices have a major
highly recommendable. Case studies demonstrate an influence
influence on the environmental impact associated with manu-
on the energy requirements of up to a factor 2.
facturing operations (a factor 6 between countries with
 Optimization of process parameter settings and well optimized
minimum and maximum impact of average electricity genera-
control allow to reduce the energy consumption. A factor 1.1 is a
tion impact can be observed).
typical order of magnitude of the achievable improvements.
 The production location also determines the need for transport of
For this level systematic impact assessment methods have been raw materials and finished products as well as the transport
developed that allow to expect systematic data collection and LCI mode, resulting in corresponding energy demands.
database contributions over the next decade. This forms a first step  Depending on the type of operations performed, local climate
towards transparency in comparison of machine tool environ- conditions can significantly influence the energy need.
mental efficiency and ultimately a machine tool eco-labeling
system. External visibility of achieved efficiency levels through It is obvious that the different opportunities for efficiency
such label should help to motivate machine tool builders to work improvement identified at the respective distinguished levels can
towards more energy and resource efficient solutions. be combined in an integrated effort for impact reduction and
economy improvement. However, the possible measures are not
Related to multi-machine systems the following major conclu- mutually independent: energy or material resource saving
sions can be formulated: measures at unit process level will, for example, typically result
in a decrease in possible energy reuse at multi-machine level. An
 Energy and material resource reuse opportunities can often be overall reduction in electrical energy consumption will reduce the
identified through exergy cascading strategies. Examples with a impact of choice of location for a production facility in absolute
factor 1.2 efficiency improvement have been reported. However, terms.
this strategy has not yet been extensively explored. Most of the demonstrated available technological solutions for
 Peak power and total energy consumption can be minimized by efficiency improvement only require well-known engineering
means of planning and optimization methods supported by methods for implementation. The authors conservatively estimate
simulation techniques. that combining such techniques with the reviewed intelligent
planning and control optimization methods, offers a potential for
At the factory level important conclusions are: global energy consumption reduction in manufacturing with at
least 50%. A thorough optimization effort at all considered levels
 When exceeding the level of multi machine chains, simulation could result in more substantial improvements though. It is
techniques become predominant to master the complexity of evident however that the replacement of machine tools and the
606 J.R. Duflou et al. / CIRP Annals - Manufacturing Technology 61 (2012) 587–609

redesign and upgrading of facilities this will require cannot be [24] Chertow MR (2000) Industrial Symbiosis, Literature and Taxonomy. Annual
Review of Energy and the Environment 25:313–337.
expected to happen overnight. Furthermore, the impact of the [25] Chertow MR, Lombardi DR (2005) Quantifying Economic and Environmental
manufacture of investment goods itself should be taken into Benefits of Co-Located Firms. Environmental Science and Technology
account when considering early replacement of existing infra- 39(17):6535–6541.
[26] Chertow MR (2007) Uncovering Industrial Symbiosis. Journal of Industrial
structure. Ecology 11(1):11–30.
As a general recommendation for further research, the authors [27] Clarens A, Zimmerman J, Keoleian G, Hayes K, Skerlos S (2008) Comparison of
observed a significantly less intensive research activity with Life Cycle Emissions and Energy Consumption for Environmentally Adapted
Metalworking Fluid Systems. Environmental Science and Technology
respect to resource efficiency utilization in comparison with the 42(22):8534–8540.
attention spent to energy efficiency. Taking the non-continuous [28] Cohen MA, Mallik S (1997) Global Supply Chains – Research and Applications.
nature of discrete part manufacturing and the time independency Production and Operations Management 6(3):193–210.
[29] Comau, 2009, online http://WardsAuto.com/ar/comau_smart_assem-
of material reuse options into account, material efficiency
bly_091109/index.html (accessed 09.11.09).
improvement still offers substantial research potential. [30] CO2PE! – Initiative. (2011) Cooperative Effort on Process Emissions in Manu-
facturing. online http://www.mech.kuleuven.be/co2pe! (accessed 21.12.11).
[31] Costanza R (1980) Embodied Energy and Economic Valuation. Science
Acknowledgements 210:1219–1224.
[32] Cox R (1996) Compressed Air – Clean Energy in a Green World. Glass
The authors want to explicitly acknowledge the input received International 19(2):50–52.
[33] Crandell, G.R., Schifo, J.F., Mosher, G., 2006, CERP Organic HAP Emission
from numerous CIRP colleagues and from researchers in the
Measurements for Iron Foundries and Their Use in Development of an AFS
CO2PE! community. The authors want to acknowledge the support HAP Guidance Document, AFS Transactions, Paper 06-031(10).
of IWT-Vlaanderen through its PhD grant no. 091232. [34] Creyts JC, Carey VP (1999) Use of Extended Energy Analysis to Evaluate the
Environmental Performance of Machining Processes. Institution of Mechan-
ical Engineers Part E Journal of Process Mechanical Engineering 213(4):
247–264.
References [35] Curtner KL, O’Neill PJ, Winter D, Bursch P (1997) Simulation-based Features
of the Compressed Air System Description Tool. XCEEDTM, Proceedings of
[1] Abele A, Anderl R, Birkhofer H (2005) Environmentally-Friendly Product Devel- International Building Performance Simulation Association Conference, Prague.
opment – Methods and Tools. ISBN: 1-85233-903-9 (Chapter 3). [36] Dahmus J, Gutowski T (2004) An Environmental Analysis of Machining.
[2] Abele E, Sielaff T, Schiffer A, Rothenbücher S (2011) Analyzing Energy Con- Proceedings of ASME International Mechanical Engineering Congress and R&D
sumption of Machine Tool Spindle Units and Identification of Potential for Exposition, 13–19.
Improvements of Efficiency. 18th CIRP LCE Conference, Braunschweig, 280–285. [37] Dalquist S, Gutowski T (2004) Life Cycle Analysis of Conventional Manufac-
[3] Ahern JE (1980) The Exergy Method of Energy Systems Analysis, Wiley, New turing Techniques: Sand Casting. Proceedings of ASME 2004 International
York. ISBN-10: 0471054941. Mechanical Engineering Congress and Exposition (IMECE2004), Anaheim, CA,
[4] Akisawa A, Kang YT, Shimazaki Y, Kashiwagi T (1999) Environmentally 631–641.
Friendly Energy System Models Using Material Circulation and Energy Cas- [38] Dalquist, S., Gutowski, T., 2004, Life Cycle Analysis of Conventional Manu-
cade – The Optimization Work. Energy 24(7):561–578. facturing Techniques: Die Casting, LMP Working Paper LMP-MIT-TGG-03-12-
[5] Alberts M, Kalaitzidou K, Melkote S (2009) An Investigation of Graphite 19-2004.
Nanoplatelets as Lubricant in Grinding. International Journal of Machine Tools [39] Devoldere T, Dewulf W, Deprez W, Willems B, Duflou JR (2007) Improvement
and Manufacture 49(12–13):966–970. Potential for Energy Consumption in Discrete Part Production Machines. 14th
[6] Al-Rabghi O, Beirutty M, Akyurt M, Najjar Y, Alp T (1993) Recovery and CIRP LCE Conference, Tokyo, 311–316.
Utilization of Waste Heat. Heat Recovery Systems & CHP 13(5):463–470. [40] Dhanik S, Xirouchakis P, Perez R (2011) A System for Resource Efficient
[7] Aurich JC, Herzenstiel P, Sudermann H, Magg T (2008) High-Performance Dry Process Planning for Wire EDM. 18th CIRP LCE Conference, Braunschweig,
Grinding Using a Grinding Wheel with a Defined Grain Pattern. CIRP Annals – 219–224.
Manufacturing Technology 57(1):357–362. [41] Diaz N, Helu M, Jarvis A, Tönissen S, Dornfeld D, Schlosser R (2009) Strategies
[8] Avram OI, Xirouchakis P (2011) Evaluating the Use Phase Energy Require- for Minimum Energy Operation for Precision Machining. Proceedings of
ments of a Machine Tool System. Journal of Cleaner Production 19(6–7): MTTRF 2009 Annual Meeting.
699–711. [42] Diaz N, Helu M, Jayanathan S, Chen Y, Horvath A, Dornfeld D (2010) Envir-
[9] Azevedo M, Oliveira M, Peraira JP, Reis A (2011) Comparison of two LCA onmental Analysis of Milling Machine Tool Use in Various Manufacturing
Methodologies in the Machine Tool Environmental Performance Improve- Environments. IEEE International Symposium on Sustainable Systems and
ment Process. 18th CIRP LCE Conference, Braunschweig, 575–580. TechnologyArlington.
[10] Baehr HD, Kabelac S (2009) Thermodynamik: Grundlagen und technische [43] Diaz N, Redelsheimer E, Dornfeld D (2011) Energy Consumption Charater-
Anwendungen, 14. aktualisierte Aufl., Springer Verlag, Berlin. ization and Reduction Strategies for Milling Machine Tool Use. 18th CIRP LCE
[11] Bailey R, Allen JK, Bras B (2004) Applying Ecological Input–Output Flow Conference, Braunschweig, 263–267.
Analysis to Material Flows in Industrial Systems: Part I: Tracing Flows. Journal [44] Dietmair A, Verl A (2009) A generic Energy Consumption Model for Decision
of Industrial Ecology 8(1–2):45–68. Making and Energy Efficiency Optimization in Manufacturing. International
[12] Bates CE, Scott WD (1977) Decomposition of Resin Binders and the Relation- Journal of Sustainable Engineering 2(2):123–133.
ship Between Gases Formed and the Casting Surface Quality – Part 3. AFS [45] Dietmair A, Verl A (2010) Energy Consumption Assessment and Optimisation
Transactions 85:209–226. in the Design and Use Phase of Machine Tools. 17th CIRP LCE Conference, Hefei,
[13] Bay N, Azushima A, Groche P, Ishibashi I, Merklein M, Morishita M, Nakamura 116–121.
T, Schmid S, Yoshida M (2010) Environmentally Benign Tribo-Systems for [46] DIN8580. (2003) Manufacturing Processes – Term and Definitions, Deutches
Metal Forming. CIRP Annals – Manufacturing Technology 59(2):760–780. Institut Fur Normung E.V..
[14] Bayerisches Landesamt für Umweltschutz. (2005) Effiziente Energieverwen- [47] Dinçer I, Rosen MA (2007) Exergy: Energy, Environment, and Sustainable
dung in der Industrie – Teilprojekt ‘‘Metallschmelzberiebte’’ Effiziente Energie- Development, Elsevier. ISBN: 978-0-08-044529-8.
nutzung in Nicht-Eisen-Metall-Schmelzbetrieben, Augsburg. (in German). [48] Dornfeld D (2012) Green Manufacturing: Fundamentals and Applications,
[15] Bejan A (2001) Fundamentals of Exergy Analysis, Entropy Generation Mini- Springer. ISBN: 978-1-4419-6015-3.
mization, and the Generation of Flow Architecture. International Journal of [49] Dotchev K, Yussof W (2009) Recycling of Polyamide 12 Based Powders in the
Energy Research 26(7):545–565. Laser Sintering Process. Rapid Prototyping Journal 15(3):192–203.
[16] Bell LE (2008) Cooling, Heating, Generating Power, and Recovering Waste [50] Duchin F (2004) Input–Output Economics and Material Flows, Rensselear
Heat with Thermoelectric Systems. Science 321:1457–1461. Polytechnic Institute, Troy, NY.
[17] Bonneschky, A., 2002, Energiekennzahlen in PPS-Systemen, Diss. u.d.T.: [51] Duflou JR, Kellens K, Devoldere T, Deprez W, Dewulf W (2010) Energy Related
Bonneschky, A.: Integration energiewirtschaftlicher Aspekte in Systeme Environmental Impact Reduction Opportunities in Machine Design: Case
der Produktionsplanung und -steuerung, Dissertation, Technische Universi- Study of a Laser Cutting Machine. International Journal of Sustainable Man-
tät Cottbus. ufacturing 2(1):80–98.
[18] Boons F, Spekkink W, Mouzakitis Y (2011) The Dynamics of Industrial [52] Duflou JR, Kellens K, Renaldi Dewulf W (2011) Environmental Performance of
Symbiosis: A Proposal for a Conceptual Framework Based Upon a Compre- Sheet Metal Working Processes. Key Engineering Materials 473:21–26.
hensive Literature Review. Journal of Cleaner Production 19(9–10):905–911. [53] Duflou JR, Kellens K, Dewulf W (2011) Unit Process Impact Assessment for
[19] Breuil JM (1992) Input–Output Analysis and Pollutant Emissions in France. Discrete Part Manufacturing: A State of the Art. CIRP Journal of Manufacturing
The Energy Journal 13(3):173–184. Science and Technology 4(2):129–135.
[20] Brockhoff T, Brinksmeier E (1999) Grind-Hardening: A Comprehensive View. [54] Eco Profiles. (2012) Eco profiles: life-cycle analysis, PlasticsEurope, Association
CIRP Annals – Manufacturing Technology 48(1):255–260. of plastics Manufacturers . online http://www.plasticseurope.org/plastics-
[21] Brodyansky VM, Sorin MV, Le Goff P (1994) The Efficiency of Ind. Processes: sustainability/eco-profiles.aspx (accessed 10.01.12).
Exergy Analysis and Optimization, Elsevier, Amsterdam. ISBN-10: [55] Environmental Technology Laboratory. (2006) LCA Database based on 2000
0444899960. Input–Output Tables for Japan, Corporate Research & Development Center,
[22] BUWAL database, Bundesamt für Umwelt, Wald und Landschaft, 2006. Toshiba Corporation.
[23] CECIMO. (2009) Concept Description for CECIMO’s Self-regulatory Initiative [56] European Aluminium Association, 2008, Environmental Profile Report for the
(SRI) for the Sector Specific Implementation of the Directive 2005/32/ European Aluminium Industry, Life Cycle Inventory Data for Aluminium
EC. Online: http://www.ecodesign-info.eu/document-tations/Machi- ne_ Production and Transformation Processes in Europe.
tools_VA_20Oct09.pdf.
 J.R. Duflou et al. / CIRP Annals - Manufacturing Technology 61 (2012) 587–609 607

[57] Fang K, Uhan N, Zhao F, Sutherland J (2011) A New Shop Scheduling Approach [89] Herrmann IT, Hauschild M (2009) Effect of Globalisation on Carbon Footprint
in Support of Sustainable Manufacturing. 18th CIRP LCE Conference, Braunsch- of Products. CIRP Annals – Manufacturing Technology 58(1):13–16.
weig, 305–310. [90] Hesselbach J, Martin L, Herrmann C, Thiede S, Lüdemann B, Detzer R (2008)
[58] Fleurial JP (2009) Thermoelectric Power Generation Materials: Technology Energieeffizienz durch optimierte Abstimmung zwischen Produktion und
and Application Opportunities. Journal of the Minerals Metals and Materials technischer Gebäudeausrüstung. 15th CIRP LCE Conference, Sydney, 624–629.
Society 61(4):79–85. [91] Holweg M (2007) The Genealogy of Lean Production. Journal of Operations
[59] Ford, 2011, online http://media.ford.com/article_display.cfm?article_id= Management 25(2):420–437.
28471 (accessed 01.07.11). [92] Hornberger M (2009) Total Energy Efficiency Management. Elektronik Ecode-
[60] Foss R (2002) Managing Compressed Air Energy Part I: Demand Side Issues. sign Congress, Munich.
Maintenance Technology. online http://www.mt-online.com/component/con- [93] ICT Consultation Group. (2009) ICT and Energy Efficiency – The Case for
tent/article/134-august2001/622-managing-compressed-air-energy-part-i- Manufacturing, European Commission. p. 20.
demand-side-issues.html?directory=90 (accessed 26.05.06). [94] ICT Consultation Group. (2009) ICT and Energy Efficiency – The Case for
[61] Frischknecht, R., Jungbluth, N., Althaus, H.J., Doka, G., Dones, R., Hischier, R., Manufacturing, European Commission. p. 27.
Hellweg, S., Nemecek, T., Rebitzer, G., Spielmann, M., 2007, Overview and [95] Ingarao G, Di Lorenzo R, Micari F (2011) Sustainability Issues in Sheet Metal
Methodology, Ecoinvent Report No. 1, Swiss Centre for LCI, Dübendorf. Forming Processes: An Overview. Journal of Cleaner Production 19(4):
[62] German Federal Ministry of Economics and Technology. (2009) Energy 337–347.
Statistics. [96] International Labor Organisation (ILO). (2011) Global Employment Trends
[63] Gibovic D, de Ciurana J (2008) Introduction and Applications of DIN 8580, 2011: The Challenge of a Jobs Recovery, International Labour Office, Geneva.
University of Girona. ISBN: 978-84-96742-73-4. ISBN: 9789221245469.
[64] Global Steel Life Cycle Inventory (LCI). (2010) The World Steel Association. [97] ISO 14001:2004, Environmental Management Systems – Requirements with
online http://www.worldsteel.org/steel-by-topic/life-cycle-assessment.html Guidance for Use, International Organization for Standardization.
(accessed 10.01.12). [98] ISO/CD 14955-1, Environmental Evaluation of Machine Tools, International
[65] Gomes JO, Gomes VEO, Fracaro de Souza J, Kawachi EI (2011) Saving Potential Organization for Standardization, Standard under development.
of Water for Foundry Sand Using Treated Coolant Water. 18th CIRP LCE [99] Jain V, Wadhwa S, Deshmukh SG (2007) Supplier Selection Using Fuzzy
Conference, Braunschweig, 407–412. Association Rules Mining Approach. International Journal of Production
[66] Güley V, Khalifa B, Tekkaya AE (2011) The Effect of Extrusion Ratio and Research 45(6):1323–1353.
Material Flow on the Mechanical Properties of Aluminium Profiles Solid State [100] Jeswiet J, Astle B, Gawal D, Diarra D (2010) Energy Consumption and CO2
Recycled form 6060 Aluminium Alloy Chips. 14th ESAFORM Conference on Emissions for Manufacturing Compressed Air. Trans NAMRI/SME 38:767–773.
Material Forming, AIP Conference Proceedings 1353, 1609–1614. [101] Johansson B, Kacker R, Kessel R, McLean C, Sriram R (2009) Utilizing Combi-
[67] Groscurth HM, Kümmel R, van Gool W (1989) Thermodynamic Limits to natorial Testing on Discrete Event Simulation Models for Sustainable Man-
Energy Optimization. Energy 14(5):241–258. ufacturing. ASME Conference 2:1095–1101.
[68] Gutowski TG, Murphy C, Allen D, Bauer D, Bras B, Piwonka T, Sheng P, [102] Junge, M., 2007, Simulationsgestützte Entwicklung und Optimierung einer
Sutherland J, Thurston D, Wolff E (2005) Environmentally Benign Manufac- energieeffizienten Produktionssteuerung, Dissertation, Universität Kassel.
turing: Observations from Japan, Europe and the US. Journal of Cleaner [103] Kara S, Ibbotson S (2011) Embodied Energy of Manufacturing Supply Chains.
Production 13:1–17. CIRP Journal of Manufacturing Science and Technology 4(3):317–323.
[69] Gutowski TG, Branham MS, Dahmus JB, Jones AJ, Thiriez A, Sekulic DP (2009) [104] Kara S, Manmek S, Herrmann C (2010) Global Manufacturing & the Embodied
Thermodynamic Analysis of Resources Used in Manufacturing Processes. Energy of Products. CIRP Annals – Manufacturing Technology 59(1):29–32.
Environmental Science and Technology 43(5):1584–1590. [105] Kara S, Bogdanski G, Li W (2011) Electricity Metering and Monitoring in
[70] Haapala KR, Rivera JL, Sutherland JW (2008) Application of Life Cycle Assess- Manufacturing Systems. 18th CIRP LCE Conference, Braunschweig, 1–10.
ment Tools to Sustainable Product Design and Manufacturing. International [106] Kara S, Li W (2011) Unit Process Energy Consumption Models for Material
Journal of Innovative Computing Information and Control Special Issue on Recent Removal Processes. CIRP Annals – Manufacturing Technology 60(1):37–40.
Advances in Flexible Automation 4(3):577–591. [107] Kaya AD, Phelan P, Chau D, Sarac HI (2002) Energy Conservation in Com-
[71] Haapala KR, Rivera JL, Sutherland JW (2009) Reducing Environmental pressed Air Systems. International Journal of Energy Research 26(9):
Impacts of Steel Product Manufacturing. Transactions of NAMRI/SME 837–849.
37:419–426. [108] Kellens K, Dewulf W, Duflou JR (2011) Environmental Analysis of the Air
[72] Haapala, K.R., Catalina, A.V., Johnson, M.L., Sutherland, J.W., Development and Bending Process, ESAFORM 2011. AIP Conference Proceedings 1353:
Application of Models for Steelmaking and Casting Environmental Perfor- 1650–1655.
mance, CIRP Journal of Manufacturing Science and Engineering. [109] Kellens K, Renaldi Dewulf W, Duflou JR (2011) Preliminary Environmental
[73] Hannon B (1973) The Structure of Ecosystems. Journal of Theoretical Biology Assessment of Electrical Discharge Machining. 18th CIRP LCE Conference,
41(3):535–546. Braunschweig, 377–382.
[74] Harms R, Fleschutz T, Seliger G (2010) Life Cycle Management of Production [110] Kellens K, Yasa E, Renaldi Dewulf W, Kruth JP, Duflou JR (2011) Energy and
Facilities Using Semantic Web Technologies. CIRP Annals – Manufacturing Resource Efficiency of SLS/SLM Processes. SFF Symposium, Texas, 1–16.
Technology 59(1):45–48. [111] Kellens K, Dewulf W, Overcash M, Hauschild M, Duflou JR (2012) Methodol-
[75] Hauschild M, Jeswiet J, Alting L (2005) From Life Cycle Assessment to ogy for Systematic Analysis and Improvement of Manufacturing Unit Process
Sustainable Production: Status and Perspectives. CIRP Annals – Manufacturing Life Cycle Inventory, Part 1: Methodology Description. International Journal of
Technology 54(2):535–555. Life Cycle Assessment 17(1):69–78.
[76] Hawdon D, Pearson P (1995) Input–Output Simulations of Energy, Environ- [112] Klocke F, Schlosser R, Sangermann H (2011) Evaluation of the Energy Con-
ment, Economy Interactions in the UK. Energy Economics 17(1):73–86. sumption of a Directed Lubricoolant Supply with Variable Pressures and Flow
[77] Hayakawa N, Wakazono Y, Kato T, Suzuoki Y, Kaya Y (1999) Minimizing Rates in Cutting Processes. 9th Global Conference on Sustainable Manufactur-
Energy Consumption in Industry by Cascade Use of waste Energy, Energy ing, Saint Petersburg, 208–214.
Conversion. IEEE Transactions 14(3):795–801. [113] Konijn P, de Boer S, van Dalen J (1997) Input–Output Analysis of Material
[78] Hedman BA, Brown HL, Hamel BB (1980) Second Law Analysis of Industrial Flows with Application to Iron, Steel and Zinc. Structural Change and Economic
Processes. Energy 5(8–9):931–936. Dynamics 8(1):129–153.
[79] Heilala J, Vatanen S, Tonteri H, Montonen J, Lind S, Johansson B, Stahre J [114] Kotas TJ (1995) The Exergy Method of Thermal Plant Analysis, reprint ed.
(2008) Simulation-based Sustainable Manufacturing System Design. Winter Krieger, Malabar. ISBN-10: 0894649418.
Simulation Conference 1922–1930. [115] Kurz HD, Salvadori N (2000) Classical Roots of Input–Output Analysis a Short
[80] Hendricks T, Choate WT (2006) Engineering Scoping Study of Thermoelectric Account of its Long Prehistory. Economic Systems Research 12(2):153–180.
Generator Systems for Industrial Waste Heat Recovery, Industrial Technologies [116] Lai J, Harjati A, McGinnis L, Zhou C, Guldberg T (2008) An Economic and
Program, U.S. Department of Energy. Environmental Framework for Analyzing Globally Sourced Auto Parts Packa-
[81] Herrmann C, Hesselbach J, Bock R, Dettmer T (2007) Coolants Made of Native ging System. Journal of Cleaner Production 16(15):1632–1646.
Ester – Technical, Ecological and Cost Assessment from a Life Cycle Perspec- [117] Larjola J (1995) Electricity from Industrial Waste Heat Using High-Speed
tive. 14th CIRP LCE Conference, Tokyo, 299–303. Organic Rankine Cycle (ORC). International Journal of Production Economics
[82] Hermann C, Thiede S (2009) Process Chain Simulation to Foster Energy 41(1–3):227–235.
Efficiency in Manufacturing. CIRP Journal of Manufacturing Science and Tech- [118] Lave LB, Cobas-Flores E, Hendrickson CT, McMichael FC (1995) Using Input–
nology 1(4):221–229. Output Analysis to Estimate Economy-Wide Discharges. Environmental
[83] Herrmann C, Kara S, Thiede S, Luger T (2010) Energy Efficiency in Manufac- Science & Technology 29(9):420–426.
turing – Perspectives from Germany and Australia. 17th CIRP LCE Conference, [119] Leão F, Pashby I (2004) A Review on the Use of Environmentally-Friendly
Hefei, 23–28. Dielectric Fluids in Electrical Discharge Machining. Journal of Materials
[84] Herrmann C, Zein A, Winter M (2010) Procedures and Tools for Metering Processing Technology 149(1–3):341–346.
Energy Consumption of Machine Tools. Manufacturing 2010, Poznan. [120] Leontief W (1936) Quantitative Input–Output Relations in the Economic
[85] Herrmann C, Zein A (2010) Polymer Water as Optimal Cutting Fluid – System of the United States. Review of Economics and Statistics 18(3):
Technical Analysis. 43rd CIRP Conference on Manufacturing Systems, Vienna. 105–125.
[86] Herrmann C, Thiede S, Kara S, Hesselbach J (2011) Energy Oriented Simula- [121] Li W, Zein A, Kara S, Herrmann C (2011) An Investigation into Fixed Energy
tion of Manufacturing Systems – Concept and Application. CIRP Annals – Consumption of Machine Tools. 18th CIRP LCE Conference, Braunschweig,
Manufacturing Technology 60(1):45–48. 268–273.
[87] Herrmann C, Suh S, Bogandanski G, Zein A, Cha J, Um J, Guzman A (2011) [122] Li Y, Hewitt CN (2008) The Effect of Trade between China and the UK on
Context-aware Analysis Approach to Enhance Industrial Smart Metering. National and Global Carbon dioxide Emissions. Energy Policy 36:1907–1914.
18th CIRP LCE Conference, Braunschweig, 323–328. [123] Liker J (2003) The Toyota Way: 14 Management Principles from the World’s
[88] Herrmann C, Zein A, Wits W, Van Houten F (2011) Visualization of Environ- Greatest Manufacturer, McGraw-Hill. ISBN: 0-07-139231-9.
mental Impacts for Manufacturing Processes using Virtual Reality. CIRP [124] Lin Y, Zhang Z, Wu F, Deng N (2004) Development of Ecological Industrial
Conference on Manufacturing Systems, Madison. Parks in China. Fresenius Environment Bulletin 13(7):600–607.
608 J.R. Duflou et al. / CIRP Annals - Manufacturing Technology 61 (2012) 587–609

[125] Löfgren, B., 2009, Capturing the Life Cycle Environmental Performance of a [158] Quadriguasi J, Walther G, Bloemhof J, Van Nunen JAEE, Spengler T (2009) A
Company’s Manufacturing System, Ph.D. thesis, Chalmers University of Methodology for Assessing Eco-Efficiency in Logistics Networks. European
Technology. Journal of Operational Research 193(3):670–682.
[126] Logožar K, Radonjič G, Bastič M (2006) Incorporation of Reverse Logistics [159] Quesnay F (1894) Tableau Economique (1758), first reprint edition. British
Model into In-Plant Recycling Process: A Case of Aluminium Industry. Economic Association, London.
Resources Conservation and Recycling 49(1):49–67. [160] Quinkertz R, Rombach G, Liebig D (2001) A scenario to Optimize the Energy
[127] Lowe EA, Moran SR, Holmes DB (1995) Field Book for the Development of Eco- Demand of Aluminium Production Depending on the Recycling Quota,
Industrial Parks, Draft Report, Indigo Dev. Co, Oakland, CA. (quoted from Resources. Conservation and Recycling 33(3):217–234.
Chertow 2000). [161] Rahimifard S, Seow Y, Childs T (2010) Minimising Embodied Product Energy
[128] Mani M, Lyons KW, Rachuri S, Subrahmanian E, Ameta G, Sriram RD, Feng SC to Support Energy Efficient Manufacturing. CIRP Annals – Manufacturing
(2008) Introducing Sustainability Early Into Manufacturing Process Planning. Technology 59(1):25–28.
ASME 2008, Evanston, IL, 319–326. [162] Radgen P, Lucas K (1996) Energy System Analysis of a Fertilizer Complex –
[129] Mattila T, Pakarinen S, Sokka L (2010) Quantifying the Total Environmental Pinch Analysis vs. Exergy Analysis. Chemical Engineering & Technology
Impacts of an Industrial Symbiosis – A Comparison of Process-, Hybrid and 19(2):192–195.
Input–Output Life Cycle Assessment. Environmental Science and Technology [163] Rager, M., 2008, Energieorientierte Produktionsplanung – Analyse, Konzep-
44(11):4309–4314. tion und Umsetzung, Dissertation, Universität Augsburg.
[130] Mentzer JT, DeWitt W, Keebler JS, Min S, Nix NW, Smith CD, Zacharia ZG [164] Reich-Weiser C, Dornfeld D (2009) A Discussion of Greenhouse Gas Emission
(2001) Defining Supply Chain Management. Journal of Business Logistics Tradeoffs and Water Scarcity Within the Supply Chain. Journal of Manufac-
22(2):1–26. turing Systems 28(1):23–27.
[131] Meyer, L., 2006, Exergiebasierte Untersuchung der Entstehung von Umwelt- [165] Reich-Weiser C, Vijayaraghavan A, Dornfeld D (2010) Appropriate Use of
belastungen in Energieumwandlungsprozessen auf Komponentenebene: Green Manufacturing Frameworks. 17th CIRP LCE Conference, Heifei,
Exergoökologische Analyse, Dissertation Universität Darmstadt (in German). 196–201.
[132] Miller RE, Blair PD (1985) Input–Output Analysis: Foundations and Extensions, [166] Reinhardt G, Reinhardt S, Föckerer T, Zäh M (2011) Comparison of the
Prentice-Hall, Englewood Cliffs, NJ. Resource Efficiency of Alternative Process Chains for Surface Hardening.
[133] Mognol P, Lepicart D, Perry N (2006) Rapid Prototyping: Energy and Envir- 18th CIRP LCE Conference, Braunschweig, 311–316.
onment in the Spotlight. Rapid Prototyping Journal 12(1):26–34. [167] Renaldi R, Kellens K, Dewulf W, Duflou J (2011) Exergy Efficiency Definitions
[134] Moran MJ, Sciubba E (1994) Exergy Analysis: Principles and Practice. Journal for Manufacturing Processes. 18th CIRP LCE Conference, Braunschweig,
of Engineering for Gas Turbines and Power 116(2):285–290. 329–334.
[135] Mori M (2010) Power Consumption Reduction of Machine Tools. Presentation [168] Risi JD (1995) Energy Savings with Compressed air. Journal of the Association
at EERE Working Group Meeting, CIRP General Assembly, Pisa. of Energy Engineers 92(6):49–58.
[136] Mori M, Fujishima M, Inamusa Y, Oda Y (2011) A Study on Energy Efficiency [169] Rother M, Shook J (2003) Learning to See – Value Stream Mapping to Create
Improvement for Machine Tools. CIRP Annals – Manufacturing Technology Value and Eliminate Muda, The Lean Enterprise Institute, Cambridge, MA.
60(1):145–148. ISBN: 0-9667843-0-8.
[137] Morrow W, Kim H, Mazumder J, Skerlos S (2007) Environmental Aspects of [170] Saidur R, Rahim NA, Hasanuzzaman M (2010) A Review on Compressed-air
Laser-Based and Conventional Tool and Die Manufacturing. Journal of Cleaner Energy Use And Energy Savings. Renewable and Sustainable Energy Reviews
Production 15(10):932–943. 14(4):1135–1153.
[138] Mosher GE (1994) Calculating Emission Factors for Pouring, Cooling and [171] Saiganesh S, Sekulic DP (2010) Balancing Material and Exergy Flows for a PCB
Shakeout. Modern Casting 84(10):28–31. Soldering process: Method and a Case Study. IEEE International Symposium on
[139] Munoz AA, Sheng P (1995) An Analytical Approach for Determining the Sustainable Systems and Technology, Arlington.
Environmental Impact of Machining Processes. Journal of Materials Processing [172] Santos JP, Oliveira M, Almeida FG, Pereira JP, Reis A (2011) Improving the
Technology 53(3–4):736–758. Environmental Performance of Machine Tools: Influence of Technology and
[140] Neugebauer R, Wertheim R, Hochmuth C, Schmidt G, Dix M (2010) Modelling Throughput on the Electrical Energy Consumption of a Press-brake. Journal of
of Energy and Resource-Efficient Machining. Proceedings of the 4th CIRP Cleaner Production 19(4):356–364.
Conference on High Performance Cutting, Gifu, 295–300. [173] Schieferdecker, B.. (2006) Energiemanagement-Tools – Anwendung im Indus-
[141] Neugebauer R, Schieck F, Göschel A, Schönherr J (2011) Heuristic-based trieunternehmen, Springer, Berlin, Heidelberg. ISBN-10: 3540254781.
Evaluation of Energy Flows in Press Hardening Process Chains. Key Engineer- [174] Schipper M (2006) Energy-Related Carbon Dioxide Emissions in U.S. Manufac-
ing Materials 473:816–823. turing, U.S. Energy Information Administration. DOE/EIA-0573(2005), online
[142] Neugebauer R, Putz M, Böhme J, Todtermuschke M, Pfeifer M (2011) New http://www.eia.doe.gov/oiaf/1605/ggrpt/pdf/industry_mecs.pdf (last
aspects of Energy Consumption Analysis in Assembly Processes and Equip- accessed 05.07.11).
ment. 9th Global Conference on Sustainable Manufacturing, Saint Petersburg, [175] Schlosser R, Klocke F, Döbbeler B, Riemer B, Hameyer K, Herold T, Zimmer-
203–207. mann W, Nuding O, Schindler B, Niemczyk M (2011) Assessment of Energy
[143] New Zealand Business Council (NZBC) for Sustainable Development. (2003) and Resource Consumption of Processes and Process Chains within the
Business Guide to a Sustainable Supply Chain – A Practical Guide. Automotive Sector. 18th CIRP LCE Conference, Braunschweig, 45–50.
[144] Newman ST, Nassehi A, Imani-Asrai R (2010) Energy Efficient Process Plan- [176] Schmitt R, Bittencourt JL, Bonefeld R (2011) Modeling Machine Tools for Self-
ning for CNC Machining. 1st International Colloquium of the Cluster of Excel- Optimization of Energy Consumption. 18th CIRP LCE Conference, Braunsch-
lence (eniPROD), Chemnitz, 57–79. weig, 253–257.
[145] Noble SB (1978) Material/Energy Accounting and Forecasting Models, Resources, [177] Schwarz EJ, Steininger KW (1997) Implementing Nature’s Lesson: The Indus-
Environment, and Economics, R.U. Ayres, New York, Wiley. trial Recycling Network Enhancing Regional Development. Journal of Cleaner
[146] NREL, U.S. Life-Cycle Inventory Database, online http://www.nrel.gov/lci/ Production 5(1–2):47–56.
database (accessed 10.01.12). [178] Sheng P, Srinivasan M, Kobayashi S (1995) Multi-Objective Process Planning
[147] OECD Fact book. (2009) Economic, Environmental and Social Statistics. ISBN: in Environmentally Conscious Manufacturing: A Feature-Based Approach.
92-64-05604-1. CIRP Annals – Manufacturing Technology 44(1):433–437.
[148] Oliveira JFG, Alves SM (2006) Development of Environmentally Friendly [179] Shi CWP, Rugrungruang F, Yeo Z, Gwee KHK, Ng R, Song B (2011) Identifying
Fluid for CBN Grinding. CIRP Annals – Manufacturing Technology 55(1): Carbon Footprint Reduction Opportunities through Energy Measurements in
343–346. Sheet Metal Part Manufacturing. 18th CIRP LCE Conference, Braunschweig,
[149] Oliveira M, Santos J, Almeida F, Reis A, Pereira JP, Rocha A (2011) Impact of 389–394.
Laser-based Technologies in the Energy-consumption of Metal Cutters: [180] Shui B, Harriss R (2005) The Role of CO2 Embodiment in US–China Trade.
Comparison between Commercially Available Systems. Key Engineering Energy Policy 34(18):4063–4068.
Materials 473:809–815. [181] Siemens AG (2010) Simulation of the Energy Consumption of Conveyor Lines
[150] Overcash M, Twomey J, Isaacs J (2009) Manufacturing Unit Process Life Cycle with Tecnomatix Plant Simulation. online www.slideshare.net/SiemensPLM
Inventories (UPLCI). Permis’09, Gaithersburg, 30–31. (accessed 10.01.12).
[151] Pantke K, Güley V, Khalifa B, Heilmann M, Biermann D, Tekkaya EA (2010) [182] Sirkin T, ten Houten M (1994) The Cascade Chain: A Theory and Tool for
Aluminium Scrap Recycling by Hot Extrusion without Melting Process. 12th Achieving Resource Sustainability with Applications for Product Design.
International Conference on Aluminium Alloys, Yokohama, 242–247. Resources Conservation and Recycling 10(3):213–276.
[152] Pecas P, Ribeiro I, Folgado R, Henriques E (2009) A Life Cycle Engineering [183] Sokka L, Pakarinen S, Melanen M (2011) Industrial Symbiosis Contributing to
Model for Technology Selection: A Case Study on Plastic Injection Moulds for More Sustainable Energy Use – An Example From the Forest Industry in
Low Production Volumes. Journal of Cleaner Production 17(9):846–856. Kymenlaakso, Finland. Journal of Cleaner Production 19(4):285–293.
[153] Pechmann A, Schöler I (2011) Optimizing Energy Costs by Intelligent Pro- [184] Solding P, Petku D, Mardan N (2009) Using Simulation for More Sustainable
duction Scheduling. 18th CIRP LCE Conference, Braunschweig, 293–298. Production Systems – Methodologies and Case Studies. International Journal
[154] Pehnt, M., Bödeker, J., Arens, M., 2010, Die Nutzung industrieller Abwärme – of Sustainable Engineering 2(2):111–122.
technisch-wirtschaftliche Potenziale und energiepolitische Umsetzung, [185] Steiner, R., Frischknecht, R., 2007, Metals Processing and Compressed Air
Research report (in German). Supply, Ecoinvent report nr. 23, Swiss Centre for LCI, Dübendorf.
[155] Pei Y, Shi X, LaLonde A, Wang H, Chen L, Snyder GJ (2011) Convergence of [186] Sternlicht B (1982) Waste Energy Recovery: An Excellent Investment Oppor-
Electronic Bands for High Performance Bulk Thermoelectrics. Nature 473: tunity. Energy Conversion and Management 22(4):361–373.
66–69. [187] Suh S (2005) Theory of Materials and Energy Flow Analysis in Ecology and
[156] Pusavec F, Krajnik P, Kopac J (2010) Transitioning to Sustainable Production – Economics. Ecological Modeling 189(3–4):251–269.
Part 1: Application on Machining Technologies. Journal of Cleaner Production [188] Suh S (2009) Handbook on Input–Output Economics in Industrial Ecology, ECO-
18(2):174–184. Efficiency in Industry and Science, Springer. ISBN: 978-1-4020-4083-2.
[157] Pusavec F, Kramar D, Krajnik P, Kopac J (2010) Transitioning to Sustainable [189] Sutherland JW, Gunter KL (2001) Environmental Attributes of Manufacturing
Production – Part 2: Evaluation of Sustainable Machining Technologies. Processes. Handbook of Environmentally Conscious Manufacturing, Kluwer
Journal of Cleaner Production 18(12):1211–1221. Academic Publishers. pp. 293–316 (Chapter 13), ISBN: 978-0792384496.
 J.R. Duflou et al. / CIRP Annals - Manufacturing Technology 61 (2012) 587–609 609

[190] Symbiosis Institute. (2011) Homepage of the Kalundborg Industrial Symbiosis [211] Verl A, Westkamper E, Abele E, Dietmair A, Schlechtendahl J, Friedrich J, Haag
Activities. online www.symbiosis.dk (accessed 27.06.11). H, Schrems S (2011) Architecture for Multilevel Monitoring and Control of
[191] Szargut, J., Morris, D.R., Steward, F.R., 1988, Exergy Analysis of Thermal, Energy Consumption. 18th CIRP LCE Conference, Braunschweig, 45–50.
Chemical, and Metallurgical Processes, Hemisphere, ISBN-10: 0891165746. [212] Vijayaraghavan A, Dornfeld D (2010) Automated Energy Monitoring of
[192] Szargut J, Morris DR (1990) Cumulative Exergy Losses Associated with the Machine Tools. CIRP Annals – Manufacturing Technology 59(1):21–24.
Production of Lead Metal. International Journal of Energy Research 14(6): [213] Wahl, E., Vincke, K., Himmelsbach, M., 2011, Recovery of Energy from a Laser
605–616. Machining System, US Patent Application publication, US2011/0024401.
[193] Tekkaya EA, Schikorra M, Becker D, Biermann D, Hammer N, Pantke K (2009) [214] Wall, G., 1977, Exergy – A Useful Concept within Resource Accounting, Report
Hot Profile Extrusion of AA-6060 Aluminium Chips. Journal of Materials No. 77-42, Institute of Theoretical Physics, Chalmers University of Technol-
Processing Technology 209(7):3343–3350. ogy and University of Göteborg, Göteborg, Sweden.
[194] The Life Cycle Initiative. (2012) United Nations Environmental Programme and [215] Wang J, Yu J, Li S, Xiao G, Chang Q, Biller S (2008) A data enabled operation-
the Society for Environmental Toxicology and Chemistry (UNEP–SETAC). online based simulation for automotive assembly. Asia Simulation Conference – 7th
http://lcinitiative.unep.fr (accessed 10.01.12). International Conference on System Simulation and Scientific Computing, Beij-
[195] Thiede S, Herrmann C, Kara S (2011) State of Research and Innovative ing, 726–731.
Approach for Simulating Energy Flows of Manufacturing Systems. 18th CIRP [216] Weinert K, Inasaki I, Sutherland J, Wakabayashi T (2004) Dry Machining and
LCE Conference, Braunschweig, 347–352. Minimum Quantity Lubrication. CIRP Annals – Manufacturing Technology
[196] Thiede, S., 2012, Energy Efficiency in Manufacturing Systems, Dissertation, 53(2):511–537.
TU, Braunschweig, Springer, ISBN: 978-3-642-25913-5. [217] Weinert N, Chiotellis S, Seliger G (2009) Concept for Energy-Aware Produc-
[197] Thiriez A, Gutowski T (2006) An Environmental Analysis of Injection Mold- tion Planning based on Energy Blocks. 7th Global Conference on Sustainable
ing. IEEE Symposium on Electronics and the Environment, Scottsdale, 195–200. Manufacturing, Chennai, 75–80.
[198] Todd RH, Allen DK, Alting L (1994) Fundamental Principles of Manufacturing [218] Weinert N, Chiotellis S, Seliger G (2011) Methodology for Planning and
Processes, Industrial Press Inc. ISBN: 0-8311-3050-4. Operating Energy-Efficient Production Systems. CIRP Annals – Manufacturing
[199] Tönsing E (1996) Energiemanagement im Betrieb reduziert Kosten. Magazin Technology 60(1):41–44.
Wirtschaft der Industrie- und Handelskammer Region Stuttgart 1:19–20. [219] Weinert, N., 2011, Vorgehensweise für Planung und Betrieb energieeffizien-
[200] Ukidwe, N., 2005, Thermodynamic Input–Output Analysis of Economic and ter Produktionssysteme, Dissertation, TU Berlin, Fraunhofer Verlag.
Ecological Systems for Sustainable Engineering, Ph.D. thesis, Ohio State [220] Wilson SE (2001) Michigan’s Sand Dunes, Pamphlet 7, Michigan Dept. of
University. Environmental Quality, Geological Survey Division, Lansing, MI. online
[201] United Nations Environmental Programme (UNEP), 2011, Towards a Green http://www.michigan.gov/documents/deq/PA07_304669_7.pdf (accessed
Economy: Pathways to Sustainable Development and Poverty Eradication, 10.01.12).
Chapter on Manufacturing, online http://www.unep.org/greeneconomy [221] Winter M, Öhlschläger G, Dettmer T, Ibbotson S, Herrmann C, Kara S (2012)
(accessed 28.07.11). Using Jatropha Oil Based Metalworking Fluids in Machining Processes: A
[202] UPLCI, 2011, Online Database and Taxonomy (Screening Approach), Functional and Ecological Life-Cycle Evaluation. 19th CIRP LCE Conference,
www.wichita.edu/sustainability (accessed 11.09.11). Berkeley.
[203] U.S. Department of Energy (USDOE). (1999) Energy and Environmental Profile [222] Wohlgemuth V, Page B, Kreutzer W (2006) Combining Discrete Event Simu-
of the U.S. Metalcasting Industry, Office of Industrial Technologies. lation and Material Flow Analysis in a Component-Based Approach to
[204] U.S. Energy Information Administration (EIA), 2010, online http://www.eia.- Industrial Environmental Protection. Environmental Modeling & Software
doe.gov (accessed 28.06.11). 21(11):1607–1617.
[205] U.S. Environmental Protection Agency (USEPA). (1998) Profile of the Metal [223] Xue H, Kumar V, Pandit SM, Sutherland JW (2004) An Enhanced Input–
Casting Industry, EPA/310-R-97-004, EPA Office of Compliance Sector Note- Output Model for Material Flow Analysis of Manufacturing Processes, JUSFA,
book Project. Denver.
[206] U.S. Environmental Protection Agency (USEPA). (2002) Beneficial Reuse of [224] Xue H, Kumar V, Sutherland JW (2007) Material Flows and Environmental
Foundry Sand: A Review of State Practices and Regulations, U.S. Environmental Impacts of Manufacturing Systems via Aggregated Input–Output Models.
Protection Agency, Washington, DC. online http://www.epa.gov/sectors/sec- Journal of Cleaner Production 15(13–14):1349–1358.
torinfo/sectorprofiles/metalcasting/reuse.pdf (accessed 10.01.12). [225] Yuan C, Zhang T, Rangarajan A, Dornfeld D, Ziemba B, Whitbeck R (2006) A
[207] U.S. Geological Survey (USGS), 2010, Aluminium Statistics, Historical Statistics Decision-Based Analysis of Compressed Air Usage Patterns in Automotive
for Mineral and Material Commodities in the United States: USGS Data Series Manufacturing. Journal of Manufacturing Systems 25(4):293–300.
140, online http://minerals.usgs.gov/ds/2005/140/ (accessed 05.07.11). [226] Zein A, Li W, Hermann C, Kara S (2011) Energy Efficiency Measures for the
[208] U.S. President’s Council on Sustainable Development (USPCSD). (1996) Sus- Design and Operation of Machine Tools: An Axiomatic Approach. 18th CIRP
tainable America: A New Consensus for Prosperity, Opportunity, and a Healthy LCE Conference, Braunschweig, 274–279.
Environment for the Future, U.S. Government Printing Office, Washington, DC. [227] Zhu Q, Côté RP (2004) Integrating Green Supply Chain Management into an
(quoted from Chertow, 2007). Embryonic Eco-Industrial Development: A Case Study of the Guitang Group.
[209] van Beers D, Corder G, Bossilkov A, van Berkel R (2007) Industrial Symbiosis Journal of Cleaner Production 12(8–10):1025–1035.
in the Australian Minerals Industry: The Cases of Kwinana and Gladstone. [228] Zhu Q, Lowe E, Wei Y, Barnes D (2007) Industrial Symbiosis in China: A Case
Journal of Industrial Ecology 11(1):55–72. Study of the Guitang Group. Journal of Industrial Ecology 11(1):31–42.
[210] Van Berkel R (2010) Quantifying Sustainability Benefits of Industrial Sym- [229] Züst R, Züst S, Studer S (2010) Ecodesign-Potenzialanalyse in der Schweizer
bioses. Journal of Industrial Ecology 14(3):371–373. MEM-Industrie – eine explorative Studie, Bundesamt für Umwelt (BAFU).