Energy 93 (2015) 1964e1970

Contents lists available at ScienceDirect

Energy
journal homepage: www.elsevier.com/locate/energy

Methods for evaluation of energy efficiency of machine tools

Timo Schudeleit*, Simon Züst, Konrad Wegener
Institute of Machine Tools and Manufacturing (IWF), ETH Zurich, Technoparkstrasse 1, 8005 Zürich, Switzerland

a r t i c l e i n f o a b s t r a c t

Article history: Energy efficiency of machine tools proves to be an ongoing challenge to manufacturing industries as a
Received 11 May 2015 number of international initiatives shows. The first part of the ISO 14955 series focusses on the basic
Received in revised form understanding, power metering and energy efficient design of machine tool. The ISO standardization
27 August 2015
body (ISO/TC 39 WG 12) is currently working on the second part of the ISO 14955 series, which aims at
Accepted 19 October 2015
defining of a standardized test method. However, a method meant for standardization could not been
Available online 19 November 2015
identified yet, due to the versatile advantages and disadvantages of the different test methods.
In order to find the most feasible test method for standardization, four general energy efficiency test
Keywords:
Energy efficiency
methods are described and compared in a state-of-the-art review. The test methods are then evaluated
Machine tools against seven key characteristic criteria using the Analytic Hierarchy Process (AHP), a structured multiple
Sustainable manufacturing criteria decision-making technique. The criteria selection and judgement of their relative importance has
been carried out in collaboration with experts from the machine tool industry and research institutes.
Hence, weight factors are derived and the best suited test method for both industrial application and
standardization is identified. The validity of the evaluation results is proven using the geometric con-
sistency method.
© 2015 Elsevier Ltd. All rights reserved.

1. Introduction correlation between increasing product complexity and the rising

difficulty to define a reference as challenge towards an energy
The European Commission launched the eco-design directive efficiency label. According to Saidur [5], a unified and standard-
2009/125/EC [1] in order to reduce the energy consumption of ized test procedure is mandatory for the development of an en-
energy-related products (ErP) by paving the way for the energy ergy efficiency label However, for testing machine tools regarding
efficiency labelling directive 2010/30/EC [2]. Since then, numerous energy efficiency a variety of different test methods exists, each
products that cause a substantial amount of energy consumption with its individual advantages and disadvantages. The ISO stan-
have been labelled. Significant achievements towards the European dardization body (ISO/TC 39 WG 12) is facing the challenge to
target (2012/27/EU [3]) can be confirmed. decide on which of four practically proven test methods to
However, governments and international bodies fail to imple- incorporate.
ment a likewise labelling scheme to industries, even though they This paper aims at the comparison of the four methods for
contribute to 42.6% of the worldwide electricity consumption, testing machine tools' energy efficiency using a multiple criteria
which was postulated by Ref. [3]. Hence, energy efficiency in in- decision-making technique. Hence, the most promising candidate
dustries turns out to be a crucial but missing building block for the can be identified and recommended for incorporation into the ISO
achievement of the before stated energy efficiency target in the 14955 series, which builds the cornerstone of the not yet devel-
European Union. opment labelling standard for machine tools.
Machine tools are incomparably more complex and more
versatile in their use than household goods such as refrigerators,
2. State of the art
dishwashers or televisions. Herrmann et al. [4] highlight the

The DIN EN ISO 14020:2000 [6] addresses the labelling of

products in general. Furthermore, the ISO 1402X series distin-
* Corresponding author. Tel.: þ41 44 63-30804; fax: þ41 44 63 21125. guishes three types of environmental labelling applicable in
E-mail address: [email protected] (T. Schudeleit). dependence on the product type:

http://dx.doi.org/10.1016/j.energy.2015.10.074
0360-5442/© 2015 Elsevier Ltd. All rights reserved.
 T. Schudeleit et al. / Energy 93 (2015) 1964e1970 1965

 DIN EN ISO 14021:2012 [7] e Environmental labels and decla- machine tools with similar specifications. Above and beyond, non-
rations e Self-declared environmental claims (Type II environ- productive machine states e such as standby and ready times e are
mental labelling neglected, which make up a major share of the machine times in
 DIN EN ISO 14024:2001 [8] e Environmental labels and decla- industrial use. Moreover, the energy consumption of a machine tool
rations e Type I environmental labelling e Principles and strongly depends on the representative work piece design, which
procedures needs to be agreed on. The method is preferably applicable during
 DIN EN ISO 14025:2011 [9] e Environmental labels and decla- configuration or use phase of the machine tool.
rations e Type III environmental declarations e Principles and
procedures
2.2. Reference process method
The European Association of the Machine Tool Industries
(CECIMO) [10] initiated the rollout a self-declaration initiative, The reference process method contains the measurement of a
which refers to environmental labelling type III. This approach machine tool's energy consumption for a predefined utilization
outsources the assessment to the machine tool builders and re- profile (e.g. defined periods of time in off, standby, ready and
places the check by reporting to the European Commission. processing state), see Fig. 2. Kaufeld [17] introduced an energy ef-
All labelling types require a test procedure on machine tool level ficiency indicator for comparison of machine tools based on a
that need to be standardized. The ISO standardization body (ISO/TC reference process. Giacone and Manco
[18] determined the energy
39 WG 12) initiated the ISO 14995 series to overcome this short- efficiency of a reference process. Peng et al. [19] and Schlosser et al.
coming. The ISO 14995-1 [4] deals with how to measure the energy [20] and Kuhrke [21] chose a static approach by defining energy
consumption of machine tools and to derive design measures for blocks for each operational state that are summed in order to
the improvement of the energetic performance of a machine tool. approximate the total energy consumption of a reference process.
The second part of the ISO 14995 series shifts the focus to the Main advantage of this method compared to the reference part
machine tool utilization and integration into the factory environ- method is the consideration of all main operational states (off,
ment. The former includes the recommendation of a test method to standby, ready and processing) according to their time shares
emulate the use in industrial practice. However, a unified and during a representative workday. This leads to a more realistic use
generic applicable energy efficiency test method has not been scenario, which results in a more comprehensive analysis. How-
standardized yet. Wegener et al. [11] distinguish four alternatives ever, the testing is more time consuming and the drawback of a
(A) of machine tool energy efficiency test methods that shall be limited comparability of different machine tools remains. The ma-
described in more detail in the following: chine tool energy consumption in practice strongly depends on the
scale of production (e.g. small, medium or large scale production)
 Reference Part Method (A1) and also the size of the parts for which the machine is foreseen,
 Reference Process Method (A2) why the method is mainly applied during use phase of a machine
 Specific Energy Consumption Method (A3) tool. Hence, the challenge is to agree on one or more representative
 Component Benchmark Method (A4) reference process.

2.3. Specific energy consumption method

2.1. Reference part method
The specific energy consumption method is a mapping approach
The reference part method refers to the energy consumption that aims at the empirical correlation between the machine tool
needed by a machine tool in order to manufacture a predefined power consumption and a physical output variable, e.g. material
work piece, see Fig. 1. The Japanese Standards Association published removal rate for a grinding machine or cutting speed for a laser
the series JIS TS B 0024 on energy efficiency evaluation using machine, see Fig. 3. This method has been applied to various cases
reference parts [12e15]. A recent example of a work piece used by such as to a turning machine by Li et al. [22] or a milling machine by
the grinding machine manufacturer Fritz Studer AG is given in Gutowski et al. [23].
Gontarz et al. [16]. The specific energy consumption method allows a direct com-
Advantages of the reference part method are the testing close to parison of similar machine tools over the full scale of processing
the manufacturing result in practice, the short test time and from slowest to fastest physical output. However, the mapping is
simplicity of the test method. The major disadvantage is that the considered time intensive and neglects all other operational states
efficiency rating of the machine tool depends on the predefined than the processing state. Additionally, the correlation between
work piece. However, machines are built and optimized for a spe- specific energy consumption and physical output strongly depends
cific application or product family that cannot be comprised by a on the work piece properties and the tool (e.g. grinding wheel or
single predefined work piece. Hence, the definition of a test piece cutting wedge), upon which it is needed to be agreed on for
and the comparison of regarding energy efficiency are limited to standardization.

Test part Manufacturing Power profile

Tool
Power

Machine

Time

Fig. 1. Illustration of the reference part method.

1966 T. Schudeleit et al. / Energy 93 (2015) 1964e1970

Test part(s) & schedule Manufacturing Power profile(s)

Tool

Power
Machine
Schedule

Time

Fig. 2. Illustration of the reference process method.

Test part(s) Manufacturing SEC profile

Tool

SEC
Machine

MRR

Fig. 3. Illustration of the specific energy consumption method.

2.4. Component benchmark method All in all, one major challenge needs to be overcome for the
possible standardization of this method: the “degree of need”
The component benchmark method is a bottomeup test and needs to be determined, since using efficient components does not
evaluation method. The machine tool is subdivided into compo- necessarily mean that the superordinate system (the machine tool)
nents, which are evaluated and compared according to their en- is energy efficient.
ergy efficiency, see Fig. 4. The component benchmark method
bases on following simply stated concept for machine tool
building: Energy efficient components combined in a need- 2.5. Comparative method overview and summary
oriented way lead to an energy efficient machine tool design.
Approaches to improve the energy efficiency of machine tool The four presented alternatives for testing the energy efficiency
components are presented in literature, e.g. for pneumatic com- of machine tools enable different advantages and comparison op-
ponents [14], drives of auxiliary components [15] or the spindle tions. Assessments of the alternatives performances require quali-
drive gearbox design [16]. Draganescu et al. [24] developed a fication criteria. Table 1 lists a comparative overview of the
method to map the efficiency of a spindle of a milling machine. A beforehand described energy efficiency test methods.
power mapping of the complete assemblage of a machine tool has In conclusion, each method has its particular advantages and
not been performed yet and the evaluation of need-based linking of disadvantages for application in practice. It is unclear which of the
components is insufficiently covered by today's research. test procedures the best compromise is. Hence, an assessment is
The component benchmark method neither requires a work piece required to select the optimal alternative regarding the re-
nor a tool specification. The consideration of the component inter- quirements of the customer (industry). This is addressed by the
linkage allows the assessment of the machine tool design and indi- following analysis and evaluation procedure.
rectly of the operational states. The method is therefore well suited for
application during design stage and the comparison of different ma-
chines based on their components. Moreover, conclusions regarding 3. Method
oversizing of components can be drawn, if tool and work piece
respectively the process is specified. However, it is a time consuming This issue shall be addressed using a structured multiple criteria
method since multiple components need to be measured and decision-making technique called Analytic Hierarchy Process
compared. Additionally, the imitation of a realistic use scenario is not (AHP), which has been developed by Saaty [25].The system con-
part of the test method, but need to be settled somehow. tains of three linked levels from top to bottom:

Component power
Load profile(s) Components
profiles
C1: Cooling System
Power
Load

C2: Spindle

C3: Axis
Time Time

Fig. 4. Illustration of the component benchmark method.

 T. Schudeleit et al. / Energy 93 (2015) 1964e1970 1967

Table 1
Comparative overview of energy efficiency test methods.

Criterion Reference part Reference process Specific energy consumption Component benchmark

C1: Time required Low Medium Medium-high High

C2: Simplicity of the test method Low complexity Low complexity Medium complexity High complexity
C3: Comparison of machine tools Yes, with very large Yes, with larger Yes, with few restrictions Yes, from component
restrictions restrictions level
C4: Independence on work piece and tool No No Partly Yes
C5: Imitation of a realistic use scenario Yes, with few restrictions Yes Yes, with larger restrictions No
C6: Implementation in the early development No, configuration or use phase No, mainly use phase Yes, late design phase Yes, design phase
stage (prototyping)
C7: Evaluation of main operating states Partly, processing and ready Yes, all No, only processing No, only indirectly

 Level 1: Goal Entries on the diagonal refer to an item compared to itself.

 Level 2: Criteria Hence, these entries are one. The pairwise comparison leads to a
 Level 3: Alternatives positive and symmetric property of the matrix

Each alternative is linked to each criterion by their

degree of fulfilment and each criterion is linked to the goal by
their degree of importance. The entire system is depicted in Fig. 5.
The individual priorities describe the degree of importance of (2)
each criterion with respect to the goal. The local priorities express
the degree of fulfilment of each alternative evaluated against each
criterion. Both types are determined by pairwise comparison and
statistically aggregated in the same manner. Therefore, in the Saaty [27] suggests a nine-step scale from 1 to 9 as fundamental
following will be referred to both as priorities that consist of a set of scale of intensity values for the pairwise comparison.
items. Each set of weighting factors needs to meet a consistency
criterion to ensure the validity of the results. Finally, each alter-
native's contribution to the goal achievement can be calculated,
which allows setting up a ranking.
3.2. Priority aggregation and normalization
3.1. Determination of priorities ½k ½k
Ao ¼ ðaijo Þ is defined as the judgement matrix gained from the
The priorities are determined using pairwise comparison of n k-th participant of the group of participants G by comparing n el-
items, for which (n2n)/2 subsequent comparisons are required. ements (i,j ;¼ ;1,…,n) for the alternative (o ¼ 1,…,p).
The pairwise comparisons ensure that the relations between each The normalization is performed on the basis of Escrobar et al. [28]
½k
item of the hierarchical structure can be established and lead to a and modified accordingly. The normalized priority wi o of ;the
dominance matrix from which ratio scales of priorities are calcu- ;item i for the participant k(k ¼ 1,…,m) using the geometric mean
lated in the form of principal eigenvectors [26]. The pairwise method and equal item weighting (bi ¼ 1/n) is derived by
judgement of the set of items I1,I2,…,In with the number of items
11=n
n2N leads to the dominance matrix Y
n
½k
aijo A !
½k j¼1 ½k
X
n
½k
wio ¼ 11=n ; with wio > 0; wio ¼ 1 (3)
Pn Y
n
½k
i¼1
i¼1 aijo A
(1) j¼1

The normalized priorities by participant are likewise aggregated

to normalized group priorities using equal participant weighting
(bk ¼ 1/m) by

Level 1: Goal

Individual priorities
“degree of importance”

Level 2: Criteria

Local priorities
“degree of fulfillment”

Level 3: Alternatives

Fig. 5. System structure of the AHP-method.

1968 T. Schudeleit et al. / Energy 93 (2015) 1964e1970

Table 2 Table 3
Maximal GCI level to ensure consistency [29,30]. Normalized local priorities.

n 3 4 >4 C1 C2 C3 C4 C5 C6 C7
GCImax(CR  0.1) 0.31 0.35 0.37
A1 32.23% 35.91% 9.29% 5.04% 20.11% 6.20% 21.22%
A2 7.08% 42.73% 6.74% 16.62% 67.11% 6.45% 67.27%
A3 54.39% 15.37% 61.65% 12.18% 6.51% 19.34% 5.76%
A4 6.29% 5.99% 22.32% 66.16% 6.27% 68.01% 5.76%
!1=m
Y
m
½k
wio !
k¼1
X
n
wG
io ¼ !1=m ; with wG
io > 0; wG
io ¼1 (4) Table 4
Pn Y
m
½k i¼1
GCIs of local priorities.

i¼1 wio Criterion GCI GCImax(n ¼ 7)

k¼1
C1 0.041 0.35
The normalized group priorities are used to assemble the C2 0.015 0.35
normalized local priorities matrix (n  p) and the normalized in- C3 0.039 0.35
C4 0.054 0.35
dividual priorities vector(n  1).
C5 0.035 0.35
C6 0.046 0.35
C7 0.060 0.35
3.3. Consistency evaluation

The consistency evaluation ensures that the answers given are The respective GCI has to be below the specified level in order to
consistent in themselves. For instance, if A is much higher rated fulfil the consistency condition sufficiently. If the necessary con-
than B and B is much higher rated than C, C consequently needs to sistency condition is not fulfilled, redesign of the study and/or a
be rated much lower than A in order to ensure consistency. Thus, repetition of the judgement need to be carried out.
randomness of answers can be detected, misconceptions can
directly be determined and the evaluation validity be measured. 3.4. Synthesizing: calculation of the degree of goal fulfilment
Here, the consistency is measured using the geometric consistency
index (GCI) according to Aguaro n and Moreno-Jime nez [29]. The By taking into account each alternative's degree of criteria
GCI for the answers of each participant (k ¼ 1,…,m) is derived by fulfilment and each criterion's degree of importance for the defined
0 1 goal, the share of each alternative to achieve the goal is calculated
P ½k w
½k
by multiplication of the normalized local priorities matrix with the
log2 @aij j½k A normalized individual priorities vector
w
i<j i

GCI ½k ¼ 2 ; i; j ¼ 1; …; n (5)

ðn  1Þðn  2Þ

and for the entire group by (8)

!
P wG
log2 aG j
ij wG
i<j i
GCI G ¼ 2 ; i; j ¼ 1; :::; n (6)
ðn  1Þðn  2Þ The normalized global priorities vector summarizes each alter-
native's share to achieve the goal. The alternatives can be ranked
A specific level of consistency is required to ensure the validity according to their degree of achievement in order to identify the
of the survey. Saaty [30] introduced a consistency condition in most recommendable alternative to accomplish the goal.
dependency on the number of elements
4. Application
CI
CR ¼  0:1 (7)
RIðnÞ 4.1. System definition and methodology

with the consistency ratio CR, the consistency index CI and the The system definition contains the goal statement, description
random index RI. This, combined with the geometric consistency of the alternatives and the criteria selection. The goal is to find an
 n and Moreno-Jime
condition stated by Aguaro nez [29] leads to the applicable test procedure for energy efficiency evaluation of ma-
consistency requirement for the maximum allowed GCI as listed in chine tools for standardization. Both the description of the alter-
Table 2. natives and the criteria statement can be found in Section 2.

Calculation and
Data acquisition Synthesizing
evaluation

Workshop Local priorities

Global priorities
Interview Individual priorities

Fig. 6. Application methodology of the AHP-method.

 T. Schudeleit et al. / Energy 93 (2015) 1964e1970 1969

Table 5 Table 6
Normalized individual priorities. GCIs of individual priorities.

Criterion Goal Participant k GCI[k] GCImax(n ¼ 7)

C1 5.04% 1 0.169 0.37

C2 10.39% 2 0.350 0.37
C3 15.06% 3 0.165 0.37
C4 11.08% 4 0.144 0.37
C5 24.58% 5 0.089 0.37
C6 17.75% 6 0.057 0.37
C7 16.10% 7 0.132 0.37
8 0.108 0.37
9 0.099 0.37
Combined 0.014 0.37
The calculation of the local and individual priorities requires
surveying participants. The above stated methodology has been
applied in practice in collaboration with nine participants from in-
followed by the component benchmark method (A4) with 26.2%
dustries (Agathon AG, Fritz Studer AG, Bystronic Laser AG, Starrag
and the specific energy consumption method (A3) 20.9%. Last
Group, Trumpf Maschinen AG), research institutes (inspire AG and ETH
ranked is the reference part method (A1) with 16.8%. Fig. 7 illus-
Zurich) and members of the standardization body ISO/TC 39 WG 12,
trates the study results.
who are experts in the field of both energy efficiency of machine tools
and standardization. The application methodology is depicted in Fig. 6.
5. Discussion and outlook
The local priorities are determined using the pairwise comparison
technique in a workshop. The workshop has been chosen due to the
The consistency evaluation proved the validity of the study re-
complexity of the alternatives. A presentation of the test methods and a
sults. The study led to significantly different normalized global
subsequent discussion ensured an in-depth understanding and paved
priorities and a therefore distinctive ranking of the different test
the way for a consensus finding. The individual priorities are derived
methods. In brief, none of the test methods shows a balanced score
from individual interviews. The data acquisition was followed by the
of strengths and criteria importance, which clearly underlines the
priorities calculation and the consistency evaluation. Finally, the syn-
complexity of the decision the ISO standardization body (ISO/TC 39
thesizing took place in order to calculate the global priorities.
WG 12) is facing. The reference part method (A1) shows an even
contribution of the criteria C2, C5 and C7 to the overall score. The
4.2. Calculation and consistency evaluation of local priorities same criteria make up the main contribution of the overall score,
which refers to the match of the test method's strengths and the
Table 3 lists the normalized local priorities of the alternative high importance of simplicity (C2), imitation of a realistic use
with respect to the criteria. scenario (C5) and the evaluation of main operational states (C7).
The number of alternatives is four (n ¼ 4), which leads to the The same is true for the reference process method (A2) to an even
consistency condition GCImax(n ¼ 4)  0.35 according to Table 2. greater extent. In comparison of only A1 and A2, the latter method
Table 4 contains the GCIs of the local priorities. The consistency is the clearly favourable one.
condition is fulfilled for all seven criteria. For the specific process energy method (A3) a different charac-
teristic can be observed. The match of criteria fulfilment by the test
4.3. Calculation and consistency evaluation of individual priorities method and criteria importance is highest for the comparison of
machine tools (C3) followed by the implementation in the early
Table 5 lists the normalized individual priorities of each crite- development stage (C6) and time spent (C1). For comparing ma-
rion to contribute to the beforehand stated goal. chine tools regarding energy efficiency, A3 is clearly the method of
Table 6 lists the GCIs of the individual priorities both by choice, followed by the component benchmark method (A4).
participant and combined for the entire group. In any case, the However, the major match of strengths and importance is embodied
consistency condition is fulfilled. The number for alternatives is by the methods independence on work piece and tool (C4) and the
seven (n ¼ 7), which leads to the consistency condition
GCI(n  4)  0.37 from Table 2.

4.4. Synthesizing and evaluation results

The matrix-vector multiplication of the normalized local prior-

ities matrix and the normalized individual priorities vector leads to
the normalized global priorities vector

(9)
A ranking can be derived from the global priorities vector. First
ranked is the reference process method (A2) with 36.1%, which is Fig. 7. Study results.
1970 T. Schudeleit et al. / Energy 93 (2015) 1964e1970

implementation in the early development stage (C6). For the latter [10] CECIMO. Concept description for CECIMO's self-regulatory initiative (SRI) for
the sector specific implementation of the directive 2005/32/EC (EuP direc-
criterion, A4 outperforms all other alternatives.
tive). 2009 [accessed: 12.04.2014] Available, http://www.ecodesign-info.eu/
The reference process method turned out to be the first ranked documents/Machine_tools_VA_20Oct09.pdf.
test procedure due to its highest total score and has already been [11] Wegener K, Weiss L, Gontarz A. Methods and tools for evaluation of energy
recommended as method for machine tool energy efficiency eval- efficiency in production. In: International chemnitz manufacturing collo-
quium ICMC 2012, 2nd international colloquium of the cluster of excellence
uation to the ISO standardization body (ISO/TC 39 WG 12). How- eniPROD, Auerbach; 2012. p. 593e614.
ever, the reference process method's application during design [12] JIS TS B 0024-1:2010. Machine tools e test methods for electric power con-
phase is very restricted compared to the second ranked component sumption e part 1: machining centres. Japanese Standards Association; 2010.
[13] JIS TS B 0024-2:2010. Machine tools e test methods for electric power con-
benchmark method, which is explicitly recommended for machine sumption e part 2: numerically controlled turning machines and turning
tool design purpose. The findings of the study will be integrated centres. Japanese Standards Association; 2010.
into the ISO 14995 series in future work. [14] JIS TS B 0024-3:2010. Machine tools e test Methods for electric power con-
sumption e part 3: horizontal grinding wheel spindle and reciprocating table
type surface grinding machines. Japanese Standards Association; 2010.
Acknowledgements [15] JIS TS B 0024-4:2010. Machine tools e test methods for electric power con-
sumption e part 4: cylindrical grinding machines. Japanese Standards Asso-
ciation; 2010.
We gratefully appreciate our sincere thanks to the experts from [16] Gontarz A, Schudeleit T, Wegener K. Framework of a machine tool config-
Agathon AG, Fritz Studer AG, Bystronic Laser AG, Starrag Group and urator for energy efficiency. Procedia CIRP 2015;26:706e11.
Trumpf Maschinen AG as well as inspire AG, IWF of ETH Zurich and [17] Kaufeld M. Energieeffizienz-Eine Mo € glichkeit des Maschinenvergleichs.
Werkstatt und Betrieb 2011;144:60.
members of the standardization body ISO/TC 39 for their cooper-
S. Energy efficiency measurement in industrial processes.
[18] Giacone E, Manco
ation and valuable contributions. Energy 2012;38:331e45.
[19] Peng T, Xu X, Wang L. A novel energy demand modelling approach for CNC
References machining based on function blocks. J Manuf Syst 2014;33:196e208.
[20] Schlosser R, Klocke F, Do € bbeler B, Riemer B, Hameyer K, Herold T, et al.
Assessment of energy and resource consumption of processes and process
[1] European Commission. Directive 2009/125/EC of the European parliament and chains within the automotive sector. In: Hesselbach J, Herrmann C, editors.
of the Council of 21 October 2009 establishing a framework for the setting of Glocalized solutions for sustainability in manufacturing. Springer Berlin/Hei-
ecodesign requirements for energy-related products. 2009 [accessed: delberg; 2011. p. 45e50.
29.04.2015] Available, http://eur-lex.europa.eu/legal-content/EN/TXT/PDF/? [21] Kuhrke B. Methode zur Energie- und Medienbedarfsbewertung spanender
uri¼CELEX:32009L0125&from¼EN. Werkzeug-maschinen [Dissertation]. TU Darmstadt; 2011.
[2] European Union. Directive 2010/31/EU of the European parliament and of the [22] Li W, Kara S. An empirical model for predicting energy consumption of
Council of 19 May 2010 on the energy performance of buildings. 2010 manufacturing processes: a case of turning process. Proc Institution Mech Eng
[accessed: 03.02.2015] Available, http://eur-lex.europa.eu/legal-content/EN/ Part B J Eng Manuf 2011;225:1636e46.
TXT/PDF/?uri¼CELEX:32010L0031&from¼EN. [23] Gutowski T, Dahmus J, Thiriez A, Branham M, Jones A. A thermodynamic
[3] European Union. Directive 2012/27/EU of the european parliament and of the characterization of manufacturing processes. In: Proceedings of the 2007 IEEE
council of 25 October 2012 on energy efficiency, amending directives 2009/ international symposium on electronics & the environment, conference re-
125/EC and 2010/30/EU and repealing directives 2004/8/EC and 2006/32/EC. cord; 2007. p. 137e42.
2012 [accessed: 12.05.2015] Available: http://eur-lex.europa.eu/LexUriServ/ [24] Draganescu F, Gheorghe M, Doicin CV. Models of machine tool efficiency and
LexUriServ.do?uri¼OJ:L:2012:315:0001:0056:EN:PDF. specific consumed energy. J Mater Process Technol 2003;141:9e15.
[4] Herrmann C, Bergmann L, Thiede S, Zein A. Energy labels for production [25] Saaty TL. The analytic hierarchy process: planning, priority setting, resource
machinesean approach to facilitate energy efficiency in production systems. allocation. New York; London: McGraw-Hill International Book Co.; 1980.
In: Proceedings of 40th CIRP international seminar on manufacturing systems [26] Saaty TL, Vargas LG. Models, methods, concepts & applications of the analytic
location, Liverpool, UK; 2007. hierarchy process. 2nd ed. Boston, MA: Springer US; 2012.
[5] Saidur R. A review on electrical motors energy use and energy savings. Renew [27] Saaty TL. Decision making with the analytic hierarchy process. Int J Serv Sci
Sustain Energy Rev 2010;14:877e98. 2008;1:83e98.
[6] DIN EN ISO 14020:2002. Environmental labels and declarations e general [28] Escobar MT, Aguaro n J, Moreno-Jime nez JM. A note on AHP group consistency
principles. 2002. DIN EN ISO. for the row geometric mean priorization procedure. Eur J Operational Res
[7] DIN EN ISO 14021:2012. Environmental labels and declarations e self- 2004;153:318e22.
declared environmental claims. 2012. DIN EN ISO. [29] Aguaro n J, Moreno-Jime nez J María. The geometric consistency index:
[8] DIN EN ISO 14024:2001. Environmental labels and declarations e type I approximated thresholds. Eur J Operational Res 2003;147:137e45.
environmental labelling e principles and procedures. 2001. DIN EN ISO. [30] Saaty RW. The analytic hierarchy process e what it is and how it is used. Math
[9] DIN EN ISO 14025:2011. Environmental labels and declarations e type III Model 1987;9:161e76.
environmental declarations e principles and procedures. 2011. DIN EN ISO.