Energy Efficiency in Manufacturing Systems (2012) PDF

Sustainable Production, Life Cycle Engineering
and Management
Series Editors
Prof. Christoph Herrmann
Institut fr Werkzeugmaschinen und
Fertigungstechnik
Technische Universitt Braunschweig
Braunschweig
Germany
E-mail: [email protected]
Prof. Sami Kara

School of Mechanical & Manufacturing
Engineering
The University of New South Wales
Sydney
Australia
E-mail: [email protected]
Joint German-Australian Research Group Sustainable Manufacturing and Life

Cycle Management, www.sustainable-manufacturing.com
For further volumes:

http://www.springer.com/series/10615
Sustainable Production, Life Cycle Engineering

and Management
Modern production enables a high standard of living worldwide through products and services.
Global responsibility requires a comprehensive integration of sustainable development fostered by
new paradigms, innovative technologies, methods and tools as well as business models. Minimizing material and energy usage, adapting material and energy ows to better t natural process
capacities, and changing consumption behaviour are important aspects of future production. A life
cycle perspective and an integrated economic, ecological and social evaluation are essential requirements in management and engineering. This series will focus on the issues and latest developments
towards sustainability in production based on life cycle thinking.
Sebastian Thiede
Energy Efciency in
Manufacturing Systems
ABC
Author
Dr.-Ing. Dipl.-Wirtsch.-Ing. Sebastian Thiede
Institut fr Werkzeugmaschinen und Fertigungstechnik
Braunschweig
Germany
ISSN 2194-0541
e-ISSN 2194-055X
ISBN 978-3-642-25913-5
e-ISBN 978-3-642-25914-2
DOI 10.1007/978-3-642-25914-2
Springer Heidelberg New York Dordrecht London
Library of Congress Control Number: 2012935578
c Springer-Verlag Berlin Heidelberg 2012

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microlms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed. Exempted from this legal reservation are brief excerpts in connection
with reviews or scholarly analysis or material supplied specically for the purpose of being entered and
executed on a computer system, for exclusive use by the purchaser of the work. Duplication of this publication or parts thereof is permitted only under the provisions of the Copyright Law of the Publishers
location, in its current version, and permission for use must always be obtained from Springer. Permissions for use may be obtained through RightsLink at the Copyright Clearance Center. Violations are liable
to prosecution under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specic statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
While the advice and information in this book are believed to be true and accurate at the date of publication, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors
or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the
material contained herein.
Printed on acid-free paper
Springer is part of Springer Science+Business Media (www.springer.com)
Foreword
Due to the increased economic and environmental concerns, a systematic

consideration of energy and resource consumption is of increasing importance in
manufacturing. A realistic and goal-driven analysis and derivation of efficiency
potentials demands a holistic system perspective in order to balance conflicting
goals and/or to avoid problem shifting. This involves an extended process
understanding with all relevant input and output flows and their realistic
consumption/emission behavior as well as the necessary consideration of
interactions with technical building services. In the field of energy and resource
efficiency diverse fields of action need to be distinguished. This could be
achieved based on single or continuous data measuring, modeling of energy and
resource flows and their interactions as well as appropriate methods for evaluating
and predicting machine behaviors. The ultimate objective is to integrate energy
and resource oriented variables with the traditional performance indicators
(e.g. cost, quality and time) into the decision system of manufacturing companies.
Measures on process and machine level are the first important steps for
increasing energy efficiency. However, the consumption of energy and resources
and the associated emission of technical equipments are not static but depending on
the specific state of operation. On a factory level which includes coupled
interaction of consumers and emitters - individual consumption and emission
profiles of processes and process chains lead to certain cumulative profiles for the
system as a whole. Thus, in-depth investigation of these consumption and emission
profiles on a factory level leads to additional potentials for improving energy
efficiency. Due to the dynamic interdependencies within the system, there is a
strong demand for a generic energy flow oriented manufacturing simulation
environment which would contribute towards improving energy efficiency in
manufacturing. The work of Dr. Thiede directly addresses this important topic.
With this published work as well as with his active and on-going role, Mr. Thiede
has strongly contributed to the development of the Joint German-Australian
Research Group Sustainable Manufacturing and Life Cycle Management
(www.sustainable-manufacturing.com). We are looking forward to continuing our
work with Dr. Thiede in future.
Prof. Christoph Herrmann

Prof. Sami Kara

The University of New South Wales
Acknowledgment
This book was written in context of my work within the Product- and Life-CycleManagement Research Group of the Institute of Machine Tools and Production
Technology (IWF) at Technische Universitt Braunschweig. Special thanks go to
apl. Prof. Dr.-Ing. Christoph Herrmann as head of the research group for his
support of this book as well as the opportunities, freedom and the excellent
collaboration I could enjoy while working in the institute.
Furthermore I would like to thank Assoc. Prof. Sami Kara from the Life Cycle
Engineering and Management Group of the University of New South Wales
(UNSW) in Sydney, Australia, for the fruitful cooperation in context of the Joint
German-Australian Research Group Sustainable Manufacturing and Life Cycle
Management - specifically during my own research stays at the UNSW. My thanks
also go to Prof. Dr.-Ing. Prof. h.c. Klaus Dilger and Prof. Dr.-Ing. Thomas Vietor for
their contributions which enable the creation of this book.
Big thanks also to all my colleagues in the institute and specifically to those of
the Product- and Life-Cycle-Management Research Group. Dear colleagues, thank
you very much for the excellent teamwork with many fruitful and nice discussions
and experiences which form the positive atmosphere of our team. In particular, I
would like to thank Dr.-Ing. Tobias Luger and Dipl.-Wirtsch.-Ing. Tim Heinemann
for reviewing the book and their constructive criticism.
Lovely thanks go to my fiance Jule Schfer for her understanding and support
specifically within the last intensive months when finalizing this book. I thank
Janne Schfer for proofreading. Last but not least, I would like to thank my parents
- Annerose and Friedrich-Wilhelm Thiede - for all the freedom and support I got
over all the years.
Braunschweig
March 2011
Sebastian Thiede
Contents
List of Figures ..................................................................................................... XI

List of Tables ................................................................................................... XVII
List of Symbols and Abbreviations ................................................................ XIX
1
Introduction .................................................................................................... 1
1.1 Sustainability as New Paradigm in Manufacturing................................... 1
1.2 Motivation ................................................................................................ 4
1.3 Objectives and Work Structure ................................................................. 6
Theoretical Background ................................................................................. 9

2.1 Production and Production Management .................................................. 9
2.2 Energy and Energy Supply ..................................................................... 12
2.3 Energy Consumption in Manufacturing.................................................. 16
2.3.1 Forms of Energy Consumption in Manufacturing ........................ 16
2.3.2 Consumers of Energy ................................................................... 19
2.3.3 Energy Consumption Behaviour of Production Machines ........... 21
2.4 Description of Selected Relevant Energy Flows in Manufacturing ........ 23
2.4.1 Electricity ..................................................................................... 23
2.4.2 Compressed Air Generation ......................................................... 25
2.4.3 Steam Generation ......................................................................... 28
2.5 Energy Efficiency in Manufacturing ...................................................... 30
2.5.1 Definition ..................................................................................... 30
2.5.2 Potentials and Fields of Action..................................................... 31
Derivation of Requirements and Methodological Approach ................... 35

3.1 Requirements from Industrial/Business Perspective .............................. 35
3.2 Requirements from Scientific/Technical Perspective ............................ 37
3.3 Research and Methodological Approach ............................................... 41
3.4 Simulation Background ......................................................................... 45
State of Research.......................................................................................... 51
4.1 Background for Selection and Evaluation of Existing Approaches ....... 51
4.2 Evaluation of Relevant Research Approaches ....................................... 57
4.3 Discussion and Comparison................................................................... 82
4.4 Derivation of Research Demand ............................................................ 86
Contents
Concept Development .................................................................................. 89

5.1 Synthesis of Requirements into Concept Specifications ........................ 89
5.2 Abstraction of Conceptual Framework .................................................. 94
5.3 Description of Simulation Approach ..................................................... 97
5.3.1 Implementation and General Functional Principle ..................... 97
5.3.2 Process Module ........................................................................ 100
5.3.3 TBS Module Compressed Air ............................................... 108
5.3.4 TBS Module Steam Generation............................................. 114
5.3.5 PPC Module ............................................................................. 117
5.3.6 Evaluation/Visualisation (EV) Module .................................... 119
5.3.7 Main Level MS Module ........................................................ 127
5.4 Application Cycle ................................................................................ 129
5.4.1 Application Cycle Synthesis .................................................... 130
5.4.2 Step 1: Objective and System Definition ................................. 132
5.4.3 Step 2: Total Energy Consumption and Contract Analysis ...... 133
5.4.4 Step 3: Identification of Energy Consumers ............................. 135
5.4.5 Step 4: Data Metering and Processing...................................... 137
5.4.6 Step 5: Modelling ..................................................................... 139
5.4.7 Step 6: Validation ..................................................................... 140
5.4.8 Step 7: Scenario Building ......................................................... 141
5.4.9 Step 8: Simulation Runs ........................................................... 141
5.4.10 Step 9: Evaluation .................................................................. 142
5.4.11 Step 10: Implementation......................................................... 144
Application of Concept .............................................................................. 145

6.1 Aluminium Die Casting ....................................................................... 145
6.2 Weaving Mill ....................................................................................... 153
6.3 PCB Assembly ..................................................................................... 161
6.4 Application in Education of Production Engineers .............................. 168
Summary and Outlook .............................................................................. 171

7.1 Summary .............................................................................................. 171
7.2 Concept Evaluation .............................................................................. 172
7.3 Outlook ................................................................................................ 175
References ......................................................................................................... 179

Own References ................................................................................................ 191
Appendix ........................................................................................................... 195
Index .................................................................................................................. 197
List of Figures
Fig. 1: Drivers for sustainability in manufacturing companies (adapted from

Fichter, 2005) .............................................................................................. 1
Fig. 2: Framework for Sustainable Manufacturing (Herrmann, 2009;
Herrmann et al., 2008a). ............................................................................. 2
Fig. 3: Strategies for a sustainable development (Schmidt, 2007). ........................ 3
Fig. 4: Electricity consumption and CO2 emissions related for the case of
Germany (BMWi, 2011). ............................................................................ 5
Fig. 5: Development of energy prices in Germany (compared to progression
of standard living costs) (BMWi, 2011). .................................................... 6
Fig. 6: Hierarchy of objectives and related structure of work. ............................... 7
Fig. 7: Production as Transformation from Inputs into Outputs
(Westkmper, 2005; DIN 8580).................................................................10
Fig. 8: Levels of abstractions in production/manufacturing (Herrmann et al.,
2007b based on Barbian, 2005)..................................................................11
Fig. 9: Classification of manufacturing systems (e.g.Dyckhoff und Spengler,
2010; Schuh, 2006; Westkmper, 2005). ...................................................11
Fig. 10: Control loop of production management (Dyckhoff und Spengler,
2010; Dyckhoff, 1994). ............................................................................12
Fig. 11: Conversion between popular energy units (Dehli, 1998). ........................13
Fig. 12: Efficiency of selected energy conversion processes
(Mller et al., 2009). ................................................................................14
Fig. 13: Energy supply chain (Engelmann, 2009). ................................................15
Fig. 14: Energy flow diagram for Scotland (Scottish government, 2006). ............15
Fig. 15: Electricity net generation 2008 by type and country (top 20 countries)
(EIA, 2009). .............................................................................................16
Fig. 16: Estimation of costs and CO2 emission related to energy consumption
of German manufacturing companies. .....................................................18
Fig. 17: Internal energy consumers and flows in a manufacturing company
(Schmid, 2008).........................................................................................19
Fig. 18: Simplified structure of energy (here: electricity) consumers in a factory
(Westerkamp, 2008).................................................................................20
XII
List of Figures
Fig. 19: Energy used as a function of material removal rate for a 3-axis CNC
milling machine (left, from Gutowski et al., 2006) and electrical energy
consumption of a grinding process (excluding filter system)
(Herrmann et al., 2008b). .........................................................................21
Fig. 20: General structure of electricity supply system (Schufft, 2007). ...............23
Fig. 21: Example of electricity cost composition and sample daily electrical
load profile (own investigation based on actual data from company). .....24
Fig. 22: Losses during the generation of compressed air depicted as
Sankey-diagram (Gauchel, 2006).............................................................27
Fig. 23: specific compressor power demand in kW for generating for one
m/min compressed air depending on nominal system pressure
(Gloor, 2000). ..........................................................................................28
Fig. 24: System for steam generation and distribution (Spirax Sarco, 2006;
Einstein et al., 2001). ...............................................................................29
Fig. 25: Variables to influence the energy efficiency of production machines
(Mller et al., 2009). ................................................................................32
Fig. 26: Measures for influencing energy demand from factory perspective
(Gesellschaft Energietechnik, 1998). .......................................................33
Fig. 27: Influence of PPC on energy demand (Rager, 2008).................................34
Fig. 28: Integrated process model (based on Schultz, 2002). ................................38
Fig. 29: Holistic definition of factory (own illustration, first presented in
Hesselbach et al., 2008b). ........................................................................39
Fig. 30: Steam demand of one and several machines. ...........................................40
Fig. 31: Static ex-post calculation of electricity consumption and comparison
to actual values (left: daily profile, right: monthly values). .....................42
Fig. 32: Example of discrete (left) and continuous (right) state variable
(Banks, 2010). ..........................................................................................46
Fig. 33: Overview simulation paradigms (Borshchev und Filippov, 2004). .........47
Fig. 34: Steps in a simulation study (Banks, 2010). ..............................................48
Fig. 35: Techniques for Verification and Validation and their subjectivity
(Rabe et al., 2008). ...................................................................................49
Fig. 36: Methodology for deriving requirements and criteria for the solution
approach. ..................................................................................................53
Fig. 37: Simplified analysis flow chart of SIMTER approach
(Heilala et al., 2008).................................................................................59
Fig. 38: The Embodied Product Energy framework for modelling energy flows
during manufacture (Rahimifard et al., 2010). .........................................61
Fig. 39: Planning methodology based on energy blocks and related interface to
simulation software (Chiotellis et al., 2009). ...........................................64
Fig. 40: Conceptual framework of simulation approach based on
(Junge, 2007). ..........................................................................................66
List of Figures
XIII
Fig. 41: Conceptual framework of ENOPA coupled simulation approach

(Hesselbach et al., 2008b). ...................................................................... 68
Fig. 42: High accuracy modelling of aggregate systems referring to
(Dietmair and Verl, 2009). ...................................................................... 76
Fig. 43: Linking a Discrete Event Inventory Simulation to a Material Network
(Wohlgemuth et al., 2006). ..................................................................... 79
Fig. 44: State of research - degree of fulfilment regarding identified criteria
towards energy oriented simulation. ....................................................... 84
Fig. 45: Identified paradigms for simulating energy flows in manufacturing
systems based on discrete event simulation (DES). ................................ 85
Fig. 46: Criteria fulfilment of energy flow simulation paradigms. ....................... 86
Fig. 47: Classification of proposed concept in factory life cycle according to
(Schenk, 2004). ....................................................................................... 89
Fig. 48: Mapping of criteria and specific characteristics of the proposed
solution.................................................................................................... 91
Fig. 49: Contribution of Simulation Modules within Control Loop of
Production Management. ........................................................................ 94
Fig. 50: Simulation based interaction of manufacturing system and
technical building services. ..................................................................... 96
Fig. 51: Conceptual Framework of the proposed simulation approach. ................97
Fig. 52: Practical implementation and user interactions with developed
energy oriented manufacturing system simulation environment............. 98
Fig. 53: Description of standardised illustration for modules. ............................. 99
Fig. 54: Underlying state chart logic of process module and connected
modelling of (e.g. energy) consumption of machines. .......................... 100
Fig. 55: Weibull function with different shape parameters b
(Bertsche, 2004). ................................................................................... 102
Fig. 56: Constituting factors of Process Module. ............................................... 103
Fig. 57: Screenshot of graphical depiction of process module in simulation. .... 106
Fig. 58: Results of verification run for process module. .................................... 107
Fig. 59: Integrated control schemes for compressors (Bierbaum und
Htter, 2004). ........................................................................................ 109
Fig. 60: State based control of compressor in compressed air module............... 109
Fig. 61: Inputs, Outputs and Parameters of the Compressed Air Module. ......... 110
Fig. 62: Overview of relevant compressor state variables (screenshot from
GUI of compressed air module). ........................................................... 112
Fig. 63: Allowed switching operations for compressors (Mller et al., 2009). .. 113
Fig. 64: Verification study for compressed air module. ..................................... 114
Fig. 65: Abstraction of steam supply system as underlying model logic. .......... 114
Fig. 66: Inputs, Outputs and Parameters of the Steam Module. ......................... 115
XIV
List of Figures
Fig. 67: Verification results for steam module. ...................................................117

Fig. 68: Inputs, Outputs and Parameters of PPC Module. ...................................118
Fig. 69: Input parameters of PPC module. ..........................................................119
Fig. 70: Inputs, Outputs and Parameters of the EV Module. ...............................120
Fig. 71: Screenshot of simulation environment with sample model and
diagrams/key figures for evaluation. ......................................................122
Fig. 72: Necessary sample size depending on effect size, statistical power and
error rate (calculated according to Soper, 2011). ...................................125
Fig. 73: Selected statistical key figures for a normal distribution
(e.g. Black, 2008; Anderson, 2002). ......................................................126
Fig. 74: Example for Sankey diagram for the case of a steam plant
(Sankey, 1898 also shown in Schmidt, 2008a). .....................................127
Fig. 75: Inputs, Outputs and Parameters of the MS Module. ..............................128
Fig. 76: Verification results for MS, EV and PPC module. ................................129
Fig. 77: Synthesis of proposed application cycle. ...............................................131
Fig. 78: Matrix for means to influence electricity costs. .....................................134
Fig. 79: Example load profile of manufacturing company. .................................135
Fig. 80: Example for estimation of electricity consumption with pareto
analysis...................................................................................................136
Fig. 81: Energy portfolio as tool for classifying energy consumers. ...................137
Fig. 82: Influence of different sampling rates on accuracy of energy
consumption patterns. ............................................................................138
Fig. 83: Decision tree for level of detail while modelling. ..................................140
Fig. 84: Sample evaluation of simulation results. ...............................................142
Fig. 85: Graphical representation of simulation results. ......................................143
Fig. 86: Structure of considered manufacturing system. .....................................146
Fig. 87: Simulation model for Aluminium die casting case (results based on
scenario A). ............................................................................................147
Fig. 88: Results of simulation run. ......................................................................148
Fig. 89: Results of parameter variation experiment for batch size of blasting
process. ..................................................................................................151
Fig. 90: Results of probabilistic simulation runs. ................................................152
Fig. 91: Energy consumption analysis for weaving mill case. ............................153
Fig. 92: Prioritisation of electricity consumers for weaving mill case. ...............154
Fig. 93: Energy measurements and modelling of weaving machines..................155
Fig. 94: Validation results for weaving mill case. ...............................................156
Fig. 95: Simulated load curves and automatically generated Sankey
diagram of simulated energy flows (base run, in kW). ..........................157
Fig. 96: Impact of changing speed of weaving machines. ...................................159
List of Figures
XV
Fig. 97: Electrical power demand of PCB assembling company. .......................162

Fig. 98: Energy portfolio of PCB assembling company. .....................................163
Fig. 99: Example measurement result of reflow oven and cumulated maximum
power demand in 15 minute interval for main consumers. ....................165
Fig. 100: Simulated electrical load profile for PCB case (second based values
converted to 15min interval) and consumption composition for
scenario A (base scenario). ..................................................................167
Fig. 101: Selected simulated electrical load profiles. ..........................................167
Fig. 102: Screenshot of Java-applet for energy oriented manufacturing system
simulation for educational purposes.....................................................169
Fig. 103: Comparison of proposed simulation based concept with state of
research................................................................................................174
List of Tables
Table 1: Energy consumption for German producing industry with respect to

energy forms and sources (based on data from 2002, in Petajoule). .......17
Table 2: Energy consumption of manufacturing companies and related costs
and CO2 emissions (for Germany) ......................................................... 18
Table 3: Evaluation of general methodological approaches based on
identified requirements (ranking for each requirement from first to
fourth place). .......................................................................................... 44
Table 4: Criteria for evaluation of research approaches ...................................... 56
Table 5: Evaluation of SIMTER approach developed by Heilala et al. ............... 57
Table 6: Evaluation of approach developed by Rahimifard ................................. 60
Table 7: Evaluation of approach developed by Solding et al ............................... 62
Table 8: Evaluation of approach developed by Weinert et al .............................. 65
Table 9: Evaluation of approach developed by Junge ......................................... 67
Table 10: Evaluation of EnoPA approach developed by Hesselbach et al........... 69
Table 11: Evaluation of approach developed by Fraunhofer IPA ........................ 71
Table 12: Evaluation of approach developed by Lfgren .................................... 73
Table 13: Evaluation of approach developed by Johannsson et al ....................... 74
Table 14: Evaluation of approach developed by Dietmair and Verl .................... 76
Table 15: Evaluation of approach developed by Wohlgemuth et al. ................... 79
Table 16: Evaluation of approach developed by Siemens. .................................. 81
Table 17: Comparison of evaluation results. ....................................................... 83
Table 18: Parameter list of process module. ...................................................... 104
Table 19: Parameter list of compressed air module
(n: number of compressor). ................................................................ 111
Table 20: Parameter list of steam module. ......................................................... 116
XVIII
List of Tables
Table 21: Parameter lists of EV module. ............................................................121

Table 22: Results of simulation runs for aluminium die casting case. ................150
Table 23: Results of simulation runs for weaving mill case. ..............................159
Table 24: Simulation results overview for PCB company case. .........................166
Table 25: Evaluation of proposed simulation approach. .....................................173
List of Symbols and Abbreviations
Symbols
Symbol
typ. Unit
Description
a
b
cm
d
TFW
TC
E
E0
eF
ES
F
Fm
f(t)
F(t)
H
hS
hW
k
hours
scale parameter of Weibull function

shape parameter of Weibull function
mass specific heat capacity (e.g. water 4.187)
effect size / Cohens d
temperature difference freshwater - steam
temperature difference condensate - steam
energy (with certain indices)
constant energy demand of machine
constant machine factor
energy demand for steam generation
fuel quantity
manufacturing parameters (e.g. load)
failure probability density function
failures probability
heat/calorific value
specific enthalpy steam
specific enthalpy water / heat of evaporation
machine constant
kJ/kg K
K
K
kWh
kWh
kW
m/s, kg/s
kJ/kg, kJ/m3
kJ/kg
kJ/kg
kg/h
fuel consumption

MTTF
MTTR
n
nFW
nC
nproduction
n
O
P
Pstate
kg/h
hours
hours
bar
steam output
Mean time to failure
Mean time to repair
factor of gamma function
share of fresh water for water supply (0..1)
share of condensate for water supply (0..1)
production quantity
sample size of simulation experiments
operation (with indices)
power
power demand for states (e.g. machine - idle,
process)
compressed air pressure
pieces
runs
W
W
XX
kW
heat input / combustion capacity
s1..n
t
tstate
T
B
kW
sec
sec
C, K
%
boiler output / boiler capacity

variance of data set 1..n
time
duration of states (e.g. idle, process)
temperature
boiler efficiency
m/h
m
fuel consumption
compressed air system volume
W
x
m/sec
J, Ws
-
material processing rate

work
values for data set (e.g. output data)
Abbreviations
AE
ANN
BTU
CA
CBN
CHP
CNC
DCM
DE
DES
EMIS
EnMS
EPE
ERP
EU
EV
FEM
GHG
CIRP
ICT
IE
ISO
IWF
LCA
LCC
LCI
MCDM
MLE
MRR
MS
MTTR
MTTF
OR
PCB
PDCA
PLM
PM
PPC
VSM
SMD
SME
Auxiliary Energy
Artificial Neural Networks
British Thermal Unit
Compressed Air
Cubic Boron Nitride
Combined Heat and Power (Cycle)
Computerised Numerical Control
Die Casting Machine
Direct Energy
Discrete Event Simulation
Energy Management Information System
Energy Management System
Embodied Product Energy
Enterprise Resource Planning
European Union
Evaluation and Visualisation (module)
Finite Element Method
Green House Gas
College International pour la Recherche en Productique/
The International Academy for Production Engineering
Information and Communication Technology
Indirect Energy
International Organisation for Standardisation
Institute of Machine Tools and Production Technology,
TU Braunschweig
Life Cycle Assessment
Life Cycle Costing
Life Cycle Inventory
Multi Criteria Decision Making
Maximum Likelihood Estimation
Median Rank Regression
Manufacturing/Main System (module)
Mean Time to Repair
Mean Time to Failure
Operations Research
Printed Circuit Boards
Plan Do Check Act
Product Lifecycle Management
Process Module
Production Planning and Control
Value Stream Mapping
Surface-Mounted Device
Small and Medium sized enterprises
XXI
XXII
STD
TBS
TE
TEEM
THT
Standard Deviation
Technical Building Services
Theoretical Energy
Total Energy Efficiency Management
Through Hole Technology
Chapter 1
Introduction
1.1 Sustainability as New Paradigm in Manufacturing

Nowadays manufacturing companies are facing diverse economic (e.g. shorter
product life cycles, rising product variant diversity, increasing production volume
fluctuations, rapid changing technologies, financial crisis) but also enormous
environmental (e.g. climate change, resource depletion) and social challenges
(e.g. aging personnel).
Especially the attention to environmental aspects like global warming or
resource depletion is accelerating and different drivers are exerting pressure on
companies (Figure 1). It is more and more an issue addressed in politics (e.g. EU
2020 climate goals) and rising public awareness - potentially resulting in
challenging consequences on the corporate image - can be observed. In addition,
drivers like increasing energy and raw material prices, the potential lack of
critical resources, necessary investments for environmental sound technologies,
and penalties for lacking compliance with environmental regulations as well as
regulative incentives or the introduction of CO2 certificates are issues that directly
connect environmental driven issues to business objectives of a company.
Regulative Pull
Regulative Push
(e.g. restricted emissions)
(e.g. research funding,

incentives)
Society Push
Vision Pull
(e.g. Global Warming

Discussion, NGO)
(e.g. self commitment,

cooperate mission)
COMPANIES
Technology Push
(e.g. efficient electric drives)
Market Pull
(e.g. customer requirements,changing demand,
cost and resource
competition)
Fig. 1 Drivers for sustainability in manufacturing companies (adapted from Fichter, 2005)
S. Thiede: Energy Efficiency in Manufacturing Systems, SPLCEM, pp. 18.

Springer-Verlag Berlin Heidelberg 2012
springerlink.com
1 Introduction
Therefore, besides traditional economical production objectives (e.g. cost, time,

quality), environmental driven objectives (e.g. low CO2 emissions) have become
strategically relevant for manufacturing companies. Altogether, it is necessary to
strive for harmonising the requirements of a sustainable development with the needs
of manufacturing (Brundtland Commission, 1987). Manufacturing processes play an
essential role regarding economic success and environmental impact. Production
processes consume raw materials and transform them into products and wanted or
unwanted by-products using energy as input. While one part of the resources is used
for creating value and embodied into the form and composition of products, another
part is wasted in terms of losses, heat and emissions. Manufacturing systems
predominantly influence the environmental outcome and therefore represent the
major potential to minimise the environmental performance of a company
(Warnecke et al., 1998). Thus, designing and improving manufacturing systems
while advantageously integrating economic, ecological and social goals becomes an
essential strategic objective of manufacturing companies nowadays (Herrmann,
2009; European Commission, 2006; Schultz, 2002). It is clear, that an isolated
consideration of traditional economic variables is not sufficient anymore. In fact,
Sustainable Manufacturing is the new necessary paradigm for manufacturing
companies which involves the integration of all relevant dimensions for all
technological and organisational measures within the normative, strategic and
operative production management (Figure 2).
Dimensions of Sustainability
economical
Process
Layer
Technology
operative
Factory
Layer
r
fo
s lity
e
gi bi
te ina
a
r a
St ust
S
Efficiency
Sufficiency
Consistency
Company
Layer
social
Organisation
strategic
normative
Network
Layer
environmental
Fig. 2 Framework for Sustainable Manufacturing (Herrmann, 2009; Herrmann et al., 2008a)
1.1 Sustainability as New Paradigm in Manufacturing
prod
uctiv
ity
Economic perspective
p >
>
Therefore, all technological and organisational measures within manufacturing

companies have to be evaluated based on a comprehensive set of criteria
nowadays which involves the integration of the economic, environmental and also
the social perspective (known as the triple bottom line). As a holistic approach
which strives to avoid problem shifting within manufacturing companies, their
supply chain and life cycle phases, this involves the consideration of all basic
strategies of sustainability on different layers beginning from the single
(production) process, process chains on a factory layer, strategic decisions on a
company layer or activities in closed looped supply chains like utilising Re-Xoptions, such as remanufacturing or refurbishment (network layer) (Herrmann,
2009). In Figure 3 the strategies of a sustainable development are depicted based
on the coherence of economic and environmental impact. While efficiency strives
to minimise the material and energy usage in all life cycle phases by increasing
resource productivity, sufficiency demands a change in the behaviour of usage and
consumption. The third strategy of sustainability is consistency, which can be
defined as the adaptation of material and energy flows to fit adequately to
biological process capacities (Dyckhoff and Souren, 2008; Herrmann et al., 2007a;
Herrmann, 2009).
vit
cti
u
d
pro
>
yp
v
ucti
prod
scope of possible
technical solutions
=
ity p
s t.
con
c
a b: efficiency
a c: sufficiency
a d: consistency
preferred
acceptable
unacceptable
Use of resources / environmental impact

Fig. 3 Strategies for a sustainable development (Schmidt, 2007)
As shown in Figure 3 the sufficiency strategy may involve the conscious

reduction of (economic) growth, which impedes a broad application in companies.
Significant improvements in sustainability can be achieved by the preferential
application of the strategy consistency, since this forces the substitution of
processes, with which the potential environmental impacts are minimised and
1 Introduction
harmful materials are avoided. While having significant improvement potential

this strategy involves certain development and implementation efforts in terms of
time and costs. Up to now typically from an economic (cost) perspective,
efficiency as improvement of the output to input ratio is an established strategy in
companies already. It also bears significant potential in terms of environmental
improvement and enables a decoupling of economic growth and related
environmental impact. However, the application of the strategy efficiency can
result in rebound effects, which have to be taken into account. Therefore, in order
to consider all interdependencies in advance to the implementation of strategies,
a holistic perspective on the considered system as well as an appropriate
methodology is of importance.
1.2 Motivation
Within the broad paradigm of sustainable manufacturing, the issue of energy
efficiency will be addressed specifically in this book. It focuses on increasing the
efficiency of energy flows in manufacturing companies with certain impact on
both economic as well as environmental target variables. This automatically
includes an improvement of resource efficiency as well since these energy flows
are typically directly or indirectly connected with the depletion of critical
resources (e.g. oil, gas, coal).
The topic energy efficiency in manufacturing is of major relevance from a
national as well as a single company perspective. On a national scale, industry is a
major consumer of energy e.g. German industry is responsible for 42% of the
national electricity and 35% of the national gas consumption (BMWi, 2011).
Considering energy consumption has a very strong relevance from both economic
as well as environmental perspective. On the one hand the energy supply is
directly connected with ecological impacts, e.g.:
Green house gas (GHG) emissions with significant contribution to global

warming. As an example, only through energy demand industry is
responsible for approx. 28% of CO2 emissions (plus approx. 9% through
direct industrial emissions, see Figure 4) in Germany (BMWi, 2011).
Depletion of diverse non-renewable resources (e.g. oil, gas, coal) with
possible lack of these resources in the future - based on currently known
securely mineable deposits and demand the statically estimated supply
range is approx. 40 (oil) respectively 60 (gas) years (BMWi, 2011).
Risks and consequences of using nuclear power plants for electricity
generation such as possible hazardous accidents with nuclear pollution
and problem of radioactive waste disposal.
Land use and harm to landscape and biodiversity through e.g. mining of
coal, oil or uranium or installation of e.g. wind energy equipment.
1.2 Motivation
Energy related
(without
industry, e.g.
households,
transportation);
59%
Industrial
processes
(direct
emissions); 9%
Energy
related (for
industrial
purposes);
28%
Land
use/forestry;
3%
Fig. 4 Composition of CO2 emissions for Germany (BMWi, 2011)
On the other hand, energy consumption also has a very strong economic
dimension. Energy prices for electricity, gas and oil are disproportionately and
steadily increasing in the last years (Figure 5). As a result, energy costs can make
up a very relevant share on total costs of manufacturing companies today. Studies
estimate that energy costs may sum up to 20% on total costs (in some branches)
the average for manufacturing companies is approx. 6% nowadays (Thamling et
al., 2010; IHK 2009). An increase to an average share of approx. 8% is expected
until 2013 (IHK 2009).
Recent studies driven from research as well as industrial practice also underline
the importance of energy efficiency in manufacturing. In an industry survey with
SME (small and medium sized enterprises) approx. 70% named energy efficiency
as an important topic. The main motivation is clearly to decrease energy costs
whereas also the contribution towards environmental protection is an important
reason (Thamling et al., 2010). However, studies also underline the unemployed
potential regarding energy efficiency in manufacturing as well as obstacles which
impede an identification and broad applicability of improvement measures in
practice (Schrter et al., 2009). Obviously there is a strong need of appropriate
methods and tools to support fostering energy efficiency in manufacturing
companies.
1 Introduction
250%
Price index (2000=100%)
200%
150%
100%
50%
gas (households)
gas (industry)
oil (industry)
electricity (households)
electricity (instrustry)
living costs
0%
year
Fig. 5 Development of energy prices in Germany (compared to progression of standard
living costs) (BMWi, 2011)
1.3 Objectives and Work Structure

Against the described background as main objective this book aims at
Strongly contributing towards the improvement of energy efficiency in
manufacturing.
The structure of the book is shown in Figure 6. Following this introduction the
necessary technical background in context of manufacturing and related energy
consumption will be given (Chapter 2). Based on this as well as industrial
experiences, diverse requirements will be derived which serve as background
for reasoning the methodological approach taken here (Chapter 3). These
methodological considerations formulate the objective of
Developing an energy flow oriented manufacturing system simulation
approach.
1.3 Objectives and Work Structure
Hierarchy of objectives
Work structure/chapters
1
Introduction
Contribution towards energy and

resource efficiency
2
Development of energy flow oriented
manufacturing system simulation
Theoretical Background
Embedded in guided
methodlogy with multidimensional evlaluation
Easy to use, also for SME
all relevant energy flows

and their
interdependencies
highly flexible, generic

solution
specific means and characteristics to

address objectives
Derivation of requirements and solution

approach
4
State of research
5
Concept Development
6
Concept Application
7
Summary and Outlook
Fig. 6 Hierarchy of objectives and related structure of the book
In the next step, the more general requirements are broken down to very
specific criteria afterwards. With that, relevant available research approaches are
being analysed and evaluated in detail in order to derive necessary further research
demand (Chapter 4). Based on this detailed analysis, further specific objectives
can be identified.
The aim is to develop an energy flow oriented manufacturing system simulation
approach which
is not related or restricted to a specific case but generic in nature and

applicable to manifold production situations in the sense of a generic
simulation environment.
explicitly pursues a holistic perspective including all relevant energy flows
as well as their interdependencies.
is also applicable for small and medium sized enterprises typically facing
obstacles towards energy efficiency measures and usage of simulation.
is embedded in a guided methodology for goal-oriented identification and
realistic as well as multi-dimensional evaluation of improvement measures
in all relevant fields of actions.
1 Introduction
All these considerations are incorporated in an own innovative solution

approach, which is developed and explained in detail in Chapter 5. Finally, the
flexible applicability and potentials of the approach are shown in four different
case studies (Chapter 6) before closing the book with a summary, concept
evaluation and an outlook (Chapter 7).
Chapter 2
Theoretical Background
Against the background of the scope and objectives of the planned research work,
the following chapter will provide the necessary theoretical background. First of
all the basics of manufacturing and energy consumption will be presented.
Following this, the state of art regarding energy efficiency measures in
manufacturing is described which serves as base for deriving requirements and
potentials for further research demand.
2.1 Production and Production Management

In the field of production engineering and management a wide range of different
terms and synonyms are not always consistently - used in different disciplines in
research and industrial practice. In order to ensure a necessary and mutual
understanding basic definitions and the connected theoretical background will be
given as base for this book. As far as possible, the terminology will reflect the
glossary/dictionary of the CIRP, The International Academy for Production
Engineering (C. I. R. P., 2008; C. I. R. P., 2004a, C. I. R. P., 2004b).
As a very general term, Operations Management deals with the design and
management of products, processes, services and supply chains. It considers the
acquisition, development, and utilisation of resources which companies transform
into the goods and services their clients want (Massachusetts Institute of
Technology (MIT), 2010). Whereas this definition is relatively broad and includes
all types of transformation and value creation in a company, production as a part
of it is focusing on physical transformation into tangible results. Production can
be defined as a combination of production factors such as labour, material and
technical equipment for the purpose of value creation in form of products
(Gutenberg, 1983). Still the term production is relatively broad in nature and can
also be applied for other areas like the agricultural sector or service industry
(intangible products), which are not the main focus of this book. Thus, the term
Manufacturing is also used which is more specifically the business or industry
of producing goods in large quantities in factories [] (Oxford University Press,
2011). In literature, there is a certain inconsistency regarding the usage of those
terms; in this book both expressions are used while production is larger and
includes manufacturing but not vice versa.
springerlink.com
10
2 Theoretical Background
Figure 7 underlines the understanding of production in context of

manufacturing as transformation of inputs like
raw material (e.g. steel),

auxiliary and operating material (e.g. coolants, paint, screws),
energy (e.g. electricity),
labour/personnel (e.g. for operating and maintaining the machine),
technical equipment for main production process and supporting processes
(e.g. transport, storage, measuring),
information,
into wanted (valuable products) and unwanted (scrap, waste, exhaust heat/air)
outputs (Westkmper, 2005; Schenk, 2004). It also shows a possible classification
of production related transformation processes based on German standard DIN
8580. The actual embodiment of production processes is typically called
Production Engineering.
auxiliary materials/supplies
Information
energy
Input / Initial state
Transformation
Output / Final state
Manufacturing method
personnel/workforce
raw materials
[DIN 8580]
products
Master
forming
Metal forming
Separating
Dividing
DIN 8588
Geometrically
defined
machining DIN
8589
Geometrically
undefined
machining DIN
8589
Joining
Coating
Material
property
changing
process(es)
equipment
(for manufacturing, measuring,
transportation, storage)
manufacturing system
Abrasive
machining DIN
8590
Disassembling
DIN 8591
Cleaning
DIN 8592
heat
information
scrap, Waste
Fig. 7 Production as Transformation from Inputs into Outputs (Westkmper, 2005; DIN 8580)
Like any other process, a production process is a set of interrelated activities

[value creating and supporting activities like transformation, combination,
transport, control, measure or storage (Barbian, 2005)] which transforms inputs
into outputs whereas the inputs to a process are generally outputs of other
processes (DIN 9000). Complex technical products are typically made in multistep production process chains as logically linked sequence of successive or
parallel single processes (and associated activities) over time with one common
goal namely to bring out a defined output (one or several final products) at the
very end (e.g. Arnold, 2002). These processes and process chains involve
technical equipment and personnel, which form manufacturing systems as specific
designated areas for production and, on a higher level of aggregation, factories
(Figure 8).
In this context manufacturing systems can be classified according to different
criteria, which specify the properties of the specific system (Figure 9).
2.1 Production and Production Management
11
factory level
order
Customer order
products
raw material
waste
resources
(manufacturing) system level
products
raw material
waste
production plan
emission
process/machine level
resources
material
information
feed
in
control
transformation
measure
combination
storage
transport
feed
out
parts
waste
emission
Fig. 8 Levels of abstractions in production/manufacturing (Herrmann et al., 2007b based on

Barbian, 2005)
material flow
repetition
Diverge
Converge
Rearrange
Continuous
Single production individual

products, uniquely produced
(e.g. Ships)
Serial production Limited
number of a product type (e.g.
furniture).
Batch production temporary
produced of large amounts of
one product type (e.g.
screws).
Mass production Open-end
production of a large number
of pieces (electronic parts,
automobile industry).
spatial alignment
workshop production; several
machines with the same function to
realise one production step (turning
centre, grinding centre, etc.).
production cells; different machines
to produce a product in one spot
with a manual production and
material flow.
flexible manufacturing systems ;
spatial aggregation similar to the
production cells but an automated
production and material flow.
continuous flow production; linking
of working stations through a
conveyor belt with synchronous
material flow.
transfer line; linking of working
stations through an conveyor belt
with asynchronous material flow.
Fig. 9 Classification of manufacturing systems (e.g. Dyckhoff and Spengler, 2010; Schuh,
2006; Westkmper, 2005)
By common definition Production Management is responsible for planning

and controlling production in order to produce the right product in terms of type
and quantity, in the right quality, at the right time and, for acceptable costs. (e.g.
Westkmper, 2005) Figure 10 shows the connected control loop of production
management. As also mentioned in the definition, main reference input variables
of production management typically refer to costs, time (e.g. reliability, speed)
and quality targets (e.g. Bickford et al., 1996). Production management can
12
reference input variable(s)
(e.g. time, cost, quality)
planning and control

coordination
inf ormation
Defining
Measuring
production management
> job order planning

> resource allocation
> work sequences
reference
(actuating variable)
> order fulfillment

> utilisation
> stock
> throughput time
actual state
(feedback)
Input
Output
disturbance variable(s)
Fig. 10 Control loop of production management (Dyckhoff and Spengler, 2010; Dyckhoff,
1994)
influence the manufacturing system through actuating variables on a strategic (e.g.

production structure/layout) and operative (e.g. job order planning, resource
allocation) layer. Feedback variables (e.g. utilisation, throughput times) enable the
comparison of the reference with the actual state, which might differ due to
disturbance variables acting on the manufacturing system. The control loop is
closed through adjusting actuating variables in order to meet the objectives
management (Dyckhoff and Spengler, 2010).
2.2 Energy and Energy Supply

By popular definition energy is the capacity to do work (e.g. McKinney et al.,
2007) respectively the inherent ability of a system to generate external impact
(e.g. Planck and Psler, 1964) therefore it is necessary to execute any kind of
designated tasks. Energy (E) is a state variable connected with Work (W) as
process variable, which describes the energetic difference when a system changes
from one state to the other. Power (P) is the rate of energy usage related to a
period of time (t).
( 1)

(2)
13
The standard unit (derived from SI-units) for energy is Joule [J], for power it is
Watt [W].

(3)
However, for different areas of application diverse units for energy can be found.
The conversion between different energy units is shown in Figure 11.
litre
oil
kcal
x 8600
x 7000
BTU
x4
x 240
x 1,23
kg
x 860
coal
equivalents
MJ
x 29,3
x 3,6
x 1,1
x 106
x 8,14
m3
gas
kW h
x 3,6*106
Ws
J
(Joule)
Nm
Fig. 11 Conversion between popular energy units (Dehli, 1998)
Fundamental physics distinguish between only two basic types of energy:

potential (stored) and kinetic (working) energy (Viegas, 2005; EIA, 2009).
However, when going into more details with mechanical, thermal, chemical,
electric, electromagnetic and nuclear energy more forms can be differentiated (e.g.
EIA, 2009). Conversion between different energy forms is basically possible.
Referring to the two basic laws of thermodynamics within a system the sum of
energy stays constant but every conversion is connected with losses because not
the whole amount of energy ends in the designated form. In this context energy is
considered as sum of exergy and anergy: exergy is the usable part of energy of a
system, which is being converted from one energy form to the other. Anergy is
energy which cannot be further utilised and is referred to as loss (typically in form
of heat) (Mller et al., 2009). A system strives towards a share of exergy of zero,
which means that it is in equilibrium and no further work can be done.
14
To enable (technical) application of energy, conversion in between different

forms is inevitable. Figure 12 shows selected conversions with connected
efficiencies resulting in certain losses.
transformation from
in
through
efficiency
transformator
electric drive
electric heating
battery, electrolysis
light bulb
fluorescent lamp
laser
95%
95%
100%
70%
5%
20%
up to 35%
mechanical electrical
mechanical
thermal
generator
gearbox
mech. brake
95%
99%
100%
thermal
thermocouple
diesel engine
otto engine
heat exchanger
5%
35%
25%
90%
Battery
fuel cell
coal heating
5%
35%
25%
electrical
electrical
mechanical
thermal
chemical
radiation
electrical
mechanical
thermal
chemical
electrical
thermal
Fig. 12 Efficiency of selected energy conversion processes (Mller et al., 2009)
In an energy supply chain different energy carriers are of importance (Figure

13). In nature, primary energy - without any conversion so far - can be found in
form of e.g. oil, gas, coal (chemical energy) or renewable source (e.g.
radiation/solar energy). These primary energy carriers are being converted to
secondary energy (e.g. electricity, heating/fuel oil) and transferred to the
designated destination. Further conversions into the targeted/useful form of energy
(e.g. compressed air, heat/cold) might be necessary in order to fulfil the designated
function (e.g. enable rotation of drives, movement of actuators, heating up space)
(Brettar, 1988; VDI, 2003). Against the background of the physical coherences as
described before, this whole supply chain involves losses from conversions itself
and transmission as well as inappropriate control and usage (e.g. leakages). For
example, in Europe (on average) electricity has a primary energy factor of about
3.3 - that means for each kWh of electricity 3.3 kWh of primary energy need to be
deployed (ISO EN15603).
Figure 14 shows the energy flows from a nations perspective, in this case
Scotland. It reveals a typical mix of energy sources for electricity generation and
the significant amount of energy, which is involved as well as the main consumers
of different forms of energy.
type of energy
(exergy) losses
description
examples
primary energy
15
secondary
energy
transformation
losses
use energy
transportation
losses
net/effective
energy
control-/
distribution losses
energy
services
usage losses
natural
resources
usable form
place of usage
directly
required form
impact on
environment
wind
sun radiation
oil, natural gas
electricity
gas
fuel oil
electricity
gas
fuel oil
electricity
compressed air
heat
running motor
running pump
heated space
Fig. 13 Energy supply chain (Engelmann, 2009)
Fig. 14 Energy flow diagram for Scotland (Scottish government, 2006)
The supply with energy is directly connected with environmental impacts. On

the one hand energy consumption involves the depletion of diverse non-renewable
resources (e.g. oil, gas, coal). Besides issues related to the actual exploration of
these resources (e.g. mining), this is a challenge in the longer-term perspective:
based on currently known securely mineable deposits and demand the statically
estimated supply range is approx. 40 (oil) respectively 60 (gas) years (BMWi,
2011). On the other hand, the generation and usage of energy through burning
coal, gas or oil results in green house gas (GHG) emissions with significant
contribution to global warming. GHG emissions from electricity usage directly
depend on the actual mix of energy sources for generation, which strongly differs
between countries. Generally, three different energy sources can be distinguished:
Conventional thermal energy generation by incineration of non-renewable

resources such as coal or gas.
16
Nuclear power generation.

Energy generation from renewable resources, such as wind, water or solar
power.
Figure 15 shows the energy mix composition for the electricity net generation in
different countries worldwide. Significant differences can be observed between
countries largely depending on conventional thermal energy generation with high
specific GHG emissions, such as Australia (0.924 kg CO2/kWh electricity, EIA,
2009) or Saudi Arabia (0.816 kg CO2/kWh electricity), and countries mainly relying
on renewable energy sources like Brazil (0.093 kg CO2/kWh electricity) or Norway
(0.005 kg CO2/kWh electricity). Thus, energy consumption in specific countries is
associated with a specific environmental impact depending on the sources.
share of sources for electricity generation
100%
3%
10% 12%
90%
24%
14%
80%
70%
60%
61%
67% 68%
71%
81%
61%
64%
55%
76%
80% 81%
82%
93%
50%
62%
77%
30%
9%
42%
34% 6%
17% 24%
15%
2%
Nuclear
17%
2%
96%
100%
21%
17%
20%
10% 20%
100%
1%
16%
9%
82%
85%
40%
0%
0%
24%
15%
14%
3%
Renewables
20%
20% 19%
0%
18%
7%
4%
0%
0% 0%
5%
0% 0%
Conventional Thermal
Fig. 15 Electricity net generation 2008 by type and country (top 20 countries) (EIA, 2009)
2.3 Energy Consumption in Manufacturing

2.3.1 Forms of Energy Consumption in Manufacturing
As described before, manufacturing processes require a significant amount of
resources and energy whereas one part of the input is used for creating value,
another part is wasted in terms of losses. Hence, it is involving relevant (and to a
certain extend unavoidable) environmental impact through energy consumption
with related resource depletion and GHG emissions. Table 1 shows the necessary
forms of energy for industrial (manufacturing) purposes in the case of Germany.
17
According to that, (space and process) heat and mechanical energy are mainly
needed (Seefeldt and Wnsch, 2007) which are getting converted from energy
sources like electricity (electrical energy), gas, oil or coal (chemical energy). The
study also underlines that the actual composition of energy form and sources
differs significantly between different branches. Whereas coal is mainly used in
metal founding, cement or chemical industry (almost 90% of coal is used by these
branches), oil and especially electricity as well as gas are far more common
through all other industries. In machinery and automotive industry for example,
electricity counts up for over 50% of total energy consumption (Seefeldt and
Wnsch, 2007).
Table 1 Energy consumption for German producing industry with respect to energy forms
and sources (based on data from 2002, in Petajoule)
space
process
mechanical
heat
heat
energy
lighting
total
total
345.6
1589.3
522.3
72.1
2529.3
electricity
21.8
234.8
490.4
72.1
819.1
gas
179.6
792.1
2.0
0.0
973.7
oil
97.7
129.6
3.9
0.0
231.2
coal
10.3
397.5
0.0
0.0
407.7
district heat
27.8
27.9
0.0
0.0
55.7
renewable
8.4
7.4
0.0
0.0
15.7
fuel
0.0
0.0
26.1
0.0
26.1
On a national scale, industry is one of the major consumers of natural gas as

primary energy carrier, e.g. in Germany the share is 36% (BMWi, 2011).
Additionally, industry consumes the major share of electricity which is a
secondary energy carrier and is produced using primary sources including
significant losses. In Germany, industry is responsible for the consumption of 47%
of the national electricity (BMWi, 2011). As mentioned above, energy
consumption has a very strong relevance from both an economic as well as an
environmental perspective. Thereby the pure energetic view as shown in Table 1
is only one perspective; whereas striving towards sustainability in manufacturing
demands a more detailed analysis of connected economic as well as environmental
impacts (here depicted with related CO2 emissions). Therefore (based on the data
from Seefeldt and Wnsch, 2007) Table 2 and related Figure 16 show the
estimated energy costs and CO2 emissions of the German manufacturing industry
for the main energy sources.
The calculation is based on the average energy prices for the considered years and
the emitted CO2 for either generating electricity (energy source mix for Germany) or
directly burning oil, gas or coal. The calculations underline the major importance of
18
Table 2 Energy consumption of manufacturing companies with related costs and CO2
emissions (for Germany)
electricity
gas
oil
coal
total
energy
consumption
[in PJ]
819,1
973,7
231,2
407,8
2431,8
energy costs
(2000)
[in ]
10.012.650.793
4.577.253.331
1.055.855.319
586.200.977
16.231.960.420
energy costs
(2008)
[in ]
20.073.221.336
9.094.440.438
2.204.659.000
1.566.545.164
32.938.865.938
related CO2
emissions
[in t]
130.933.135
38.745.623
10.395.556
37.185.949
217.260.264
120%
coal
100%
17%
80%
60%
10%
oil
gas
5%
7%
28%
electricity
17%
5%
18%
40%
40%
20%
61%
60%
cost perspective
(2008)
CO2 emissions
34%
0%
consumption in PJ
Fig. 16 Estimation of costs and CO2 emission related to energy consumption of German
manufacturing companies
considering electricity in comparison to primary energy sources (due to upstream

supply chain). Only through its electricity consumption, industry is responsible for
approx. 18% of CO2 emissions (plus approx. 20% through direct industrial
emissions) in Germany (BMWi, 2011). Furthermore, the calculation stresses the
very strong economic relevance of industrial energy consumption. Energy prices for
electricity, gas and oil have been steadily increasing for the last couple of years
(BMWi, 2011). As shown in Table 2, energy costs for manufacturing companies
have been more than doubled from the year 2000 to 2008.
19
2.3.2 Consumers of Energy

Table 2 already gave an overview over main energy forms needed in industry.
Altogether the most typical energy conversions are from gas to process heat and
from electricity to mechanical energy. Due to their relevance in general and for
this book in particular, selected energy flows from these categories will be
presented in more detail in chapter 2.4. For a deeper insight regarding the
coherences in a manufacturing company, Figure 17 shows internal energy flows
with respect to different consumers and energy carriers. The figure underlines the
manifold technologies, which are involved to keep a factory operating whereas the
actual embodiment evidently depends on the specific case. On average, space and
process heat sum up to a major share on total energy consumption (in PJ or kWh)
in industry, altogether approx. 75%. However, this consumption mostly bases on
gas, coal or oil and is also branch specific. As shown above, electricity is of
specific relevance due to its cost as well as environmental impact and the broad
range of application. Therefore, Figure 17 shows typical users of electricity in
industry. It is mainly used to run electric drives to generate mechanical energy.
Typical applications are pumps, air conditioning (chill generation, ventilation),
compressed air generation and of course the actual movement and processing of
production machines (e.g. spindle motor, conveyor belt drive). Furthermore
electricity is necessary to operate lighting as well as information and
communication technologies (ICT) (Schmid and Layer, 2003).
This consideration focuses on cross-sectional technologies with broad relevance
for all industries to give a general overview from an energetic perspective. In the
district heat
waste
materials
fossil fuels
heat recovery
electricity
generation
space heat
combined heat
and power
plant
process
heat
building
refrigerating
plant
cooling
energy
compressed air
system
compressed
air
electrical
drives
mechanical
energy
lighting
light
ICT
communications
waste
renewables
steam and hot
water supply
electricity
electricity
losses
Fig. 17 Internal energy consumers and flows in a manufacturing company (Schmid, 2008)
20
specific case, these technologies are applied in very complex production

environments and embodied in specific machines. Energy consumption takes place
on diverse levels of consideration: it can be distinguished between processes and
machines for actual value creation and energy consuming equipment for diverse
supporting activities (e.g. coolant treatment in machining Bode, 2007) including
building shell and technical infrastructure (Schenk, 2004; Clarke et al., 2008). In this
context, the term (technical) building services (TBS) is often used. TBS are
responsible for essential tasks like heating and cooling (e.g. space and process heat),
ventilation and air conditioning (e.g. exhaust air purification, air technology), power
engineering (e.g. energy supply, lighting), or water/media supply and treatment.
Hence they provide the needed production environment and necessary process
energy in different forms as well as process-related media like water. (Hall and
Greeno, 2009; Chadderton, 2004). Altogether, referring to a European study, 3540% of industrys energy consumption is caused by TBS (Eichhammer et al., 1996).
Altogether, an example breakdown of different energy consumers in a factory is
shown in Figure 18.
Fig. 18 Simplified structure of energy (here: electricity) consumers in a factory (Westerkamp,

2008)
21
2.3.3 Energy Consumption Behaviour of Production Machines

In addition to the general overview of energy consumption in manufacturing
companies, the analysis of the consumption behaviour of production machines is
necessary. As diverse studies for different types of production machines show,
their energy consumption is usually not constant over time but rather highly
dynamic depending on the production process and the actual state of the machine.
Machines consist of several energy consuming components (e.g. electric drives)
that generate a specific energy load profile when producing (Eckebrecht, 2000;
Gutowski et al., 2006; Binding, 1988). This typically applies to electricity, but is
also true for other forms of energy like compressed air, process heat or gas since
their consumption naturally also differs depending on process and machine states.
As example, Figure 19 shows an electrical load profile for the case of a grinding
machine.
Internal cylindrical grinding
Q'w = 1,5 mmmm-1s-1
Grinding wheel: CBN
V'w = 200 mmmm-1
Workpiece: 100Cr6 (62HRC)
vc = 60 ms-1
12
Power [kW]
10
8
6
4
process power
basic power
0
0
Machine
startup
50
Exhaust air
system
startup
100
Spindle
startup
150
200
Machining
250
Time
[s]
300
Spindle and
air system
stopped
Fig. 19 Energy used as a function of material removal rate for a 3-axis CNC milling
machine (left, from Gutowski et al., 2006) and electrical energy consumption of a grinding
process (excluding filter system) (Herrmann et al., 2008b)
In general, different typical main states of a machine can be distinguished,

whereas, depending on the specific machine, a more detailed differentiation or
combination of states is possible (e.g. Binding, 1988; Dietmair and Verl, 2008;
Dahmus and Gutowski, 2004; Devoldere et al., 2007):
Off: main switch off, no energy consumption

Start-up: many machines consist of distinctive start-up phases, with energy
demand peaks caused by switching on certain components, heating-up phases
etc.
Idle: typically relatively constant energy consumption after main supporting
components completed start-up and machine is ready for production.
Run-time/ready for machining: positioning and loading straight before
actual processing (e.g. movement of spindle in position towards workpiece
but without material removal)
Operation: actual production process takes place, physically necessary
energy to fulfil production task (e.g. remove material)
22
In general, energy profiles can be subdivided into constant and variable energy
consumption (Figure 19, Gutowski et al., 2006). The constant energy consumption
includes the energy requirements of machine components like control units,
pumps (e.g. oil pressure, coolant) or coolers, which enable an operating state. The
variable energy consumption of a production machine enfolds the required energy
for tool handling, positioning and the actual operation (e.g. cutting). Studies have
shown that machine tools with increasing levels of automation reveal higher
constant energy consumptions resulting from the amount of additional integrated
machine components. The energy consumption E is therefore not only determined
by the cutting operation, but may be dominated by the basic power consuming
components (Klocke et al., 2010; Dahmus and Gutowski, 2004). Altogether, the
energy consumption of the machine as a whole (E) depends on the design and
control schemes of the machine as well as actual process parameters. A simplified
linear consumption equation can generally be formulated as (Wohinz and Moor,
1989):

(4)
In this case E0 is the constant energy demand (typically idle demand) whereas the
term eF*Fm describes the variable demand depending on constant factors reflecting
machine characteristics (eF) and the actual manufacturing parameters (Fm, e.g.
load, speed, production volume). Similar to this equation, Gutowski et al. specify
for the case of electricity demand of a machine tool (Gutowski et al., 2006):

(5)
P [in kW] is the total power consisting of Pidle (idle power), a machine specific
constant k and the rate of material processing in cm3/sec.
A certain period of production time typically involves both times of operation
and machine idleness. Having in mind equation (1) and (4)/(5) the following
equation for the energy consumption (of e.g. one shift) can be formulated with the
state based power demand P, the cycle and total time (t) and the production
volume n (Engelmann, 2009):
(6)
The equation reveals the energy demand when not producing (often equal to the
constant power portion) in combination with the actual idle time, both can have
significant impact on the overall energy consumption of a machine. Devoldere et
al. conducted time studies for different types of production machines in industry
cases, which underlined this relevance of non-productive energy consumption.
This is specifically true if a major share of process-related inevitable handling and
measuring activities is involved (e.g. in bending) (Devoldere et al., 2007;
Devoldere et al., 2008). As a consequence, the energy consumption during nonproduction time should be reduced through organisational (e.g. influencing worker
behaviour, batch processing) and technical (e.g. standby-mode) measures.
Additionally, not only normative values but also state related energy consumption
values should be considered when acquiring new production machines (Kuhrke
et al., 2010).
2.4 Description of Selected Relevant Energy Flows in Manufacturing
23
2.4 Description of Selected Relevant Energy Flows in

Manufacturing
The prior description revealed the relevance of different energy flows in
manufacturing companies. Due to their relevance for this book, selected forms of
energy will be presented in more detail. This will allow a deeper understanding of
specific characteristics and provide the necessary theoretical background information
for latter considerations.
2.4.1 Electricity
Electrical energy or just electricity - is of major relevance in industrial practice
due to some significant advantages in comparison with other energy carriers
(Schufft, 2007; Ridder, 2003). Electricity,
can basically be generated from any primary energy carrier

has relatively few losses in distribution
can be measured easily and also used for control purposes
is completely convertible without any residues
has a significant relevance in information and communication technologies (ICT)
However, three major disadvantages need to be stated as well (Schufft, 2007):
Electricity is basically not storable.

Electricity has to be generated in conversion processes with relatively low
efficiencies.
Extensive facilities are necessary for electricity generation and distribution.
As described above, electricity is being generated in power plants through using

coal, gas, oil, nuclear power or renewable sources such as wind, solar power or
hydropower. In the electric grid, alternating current is used because of the
advantages in comparison with direct current. Distribution involves diverse current
levels, which are obtained through different transformation steps. Figure 20 shows
the structure of an electricity supply system with generation and distribution to
consumers, which differ in terms of their connection due to their power demand.
Manufacturing companies are typically connected to the high or mean voltage grid,
in smaller cases (e.g. SME) low voltage might be sufficient as well.
large industry
110kV
220kV
power
plants
230/400V
residential areas
public buildings
commercial areas
20kV
230/400V
110kV
railway transportation
Fig. 20 General structure of electricity supply system (Schufft, 2007)
small companies
agriculture
private houses
24
Pricing
The calculation of electricity costs in manufacturing companies is very complex
and is influenced by diverse variables. Additionally there are typically no standard
contracts but rather very specific conditions for each company as a result of
individual negotiations between the supplier and customer (Specht, 2005). Based
on analyses of different electricity supply contracts, this section just gives a brief
summary of relevant cost aspects and calculation procedures. Three different main
cost portions can be distinguished:
costs for electrical work (

costs for electrical power (
fixed and variable standard costs (
Therewith, the composition of total electricity costs ( sums

up to

(7)
In the following, these cost portions will be described briefly. The base for
electricity cost calculation is a time based electrical load curve as exemplarily
shown in Figure 21. The measuring respectively relevant billing interval is
typically 15 minutes the measured electrical power demand (which be
technically available with significantly higher sampling rate) is being averaged
over this period (Mller, 2001). This results in 96 electrical power values for one
day (24 hours).
250,000
standard cost portions
base
peak
base
200,000
power [kW]
150,000
costs for electrical work

(power multiplied with time)
100,000
costs for power demand

(highest 15min power value(s)
in specified period)
50,000
0,000
00:00
01:00
02:00
03:00
04:00
05:00
06:00
07:00
08:00
09:00
10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00
20:00
21:00
22:00
23:00
energy cost composition [%]
(e.g. network access, taxes, fees)
time [h]
Fig. 21 Example of electricity cost composition and sample daily electrical load profile
(own investigation based on actual data from company)
Costs for Electrical Work

Mathematically, the electrical work is the area beneath and so the integral of
the electrical load curve. Based on equation (1) and a given price rate ( per kWh)
the costs of electrical work can be calculated with:
( 8)
25
In energy economics there is a differentiation between so called peak and base

times. Against the background of typical market behaviour influenced by supply
and demand, energy suppliers may charge a higher price per kWh during the
day (peak, typically between 6am and 8pm) because of higher general demand. In
base times (evening/night) this rate can be significantly lower because at this time
energy suppliers want to utilise their power plants as well. However, as mentioned
above, this is just a general framework and the specific solution for an individual
company is the subject of individual negotiation this might also mean that there
is a single cost rate based on a mixed price model.
Costs for Electrical Power
In contrast to private household billing, industrial companies as relatively large
consumers are typically also charged based on electrical power demand. Energy
suppliers argue that power demand stresses the electric grid as a whole and
guaranteeing capacity and reliability of the system involves significant effort. This
cost portion can make up a very significant share on total electricity costs,
especially in SME with volatile electricity consumption patterns. For the
calculation of this cost portion the highest 15min-value of electrical power
demand for a defined period of time (e.g. for a month) is being multiplied with the
specific demand related price rate (/kW)

(9)
Again, the actual calculation scheme and price rate is highly individual: for
example, the highest electrical power value for a month could be the base of
calculation. However, also cost calculations which take e.g. the three highest
values over a whole year into account can be found in industrial practice.
Fixed and Variable Standard Costs
This electricity cost portion should subsume the manifold additional cost
elements, which are being charged, connected either to the electrical work or just a
fixed rate.

(10)
Most of these issues are regulated by the government, such as taxes, fees for
network access or country specific dues (e.g. in Germany for supporting the
introduction of renewable energies) and cannot be negotiated. Additionally there
are standard charges from the supplier for e.g. measurement or billing.
2.4.2 Compressed Air Generation

For industrial purposes compressed air offers some important advantages compared to
e.g. hydraulic, electrical or mechanical systems (Bierbaum and Htter, 2004). For
instance, compressed air is:
26
easy to store and transport (also possible over longer distances, no

recirculation necessary)
clean and dry medium
safe to operate
components of the system are often less expensive, more robust/simple,
lighter and easier to maintain (e.g. no exchange of fluids necessary)
fast medium
overload normally causes no problem
parameters infinitely variable
In industry, compressed air is typical used for e.g. clamping (e.g. fixation of parts
in automated systems), transportation (e.g. bulk goods in pipes), pneumatic drives
(e.g. automatic screwing), spraying (e.g. painting, blasting), blowing (e.g. glass
industry, cleaning of parts), testing (e.g. pressure checks) or control tasks
(Bierbaum and Htter, 2004). The generation of compressed air is a conversion of
electrical energy to mechanical energy. Physically, the compressed air system is
determined by three main variables, temperature (T), pressure (p) and volume (V),
which are thermodynamically directly connected with each other (Bierbaum and
Htter, 2004):
(11)
Assuming one of the variables as being constant, the two others are changing
proportionally to each other, e.g. (for constant temperature)

(12)
The conversion process takes place in one or several compressors. Diverse kinds
of compressors are available which differ in terms of functional principle (e.g.
rotary and reciprocating compressors), energy consumption, compressed air
volume rate etc, but also regarding control schemes (Bierbaum and Htter, 2004;
Ruppelt, 2003). A comprehensive European market study revealed that so-called
screw compressors dominate the market (approx. 75% market share) because of
their reliability, simplicity and relatively low costs (Radgen and Blaustein,
2001). The same study also gives an impression of the immense relevance of
compressed air generation in terms of energy consumption. About 10% of total
industrial electricity consumption is caused by generation of compressed air
(which means 80 TWh or 55 million tons CO2). The average compressor has a
demand of 71 kW and runs 3,500 hours per year (Radgen and Blaustein, 2001).
The compressor(s) are just one part of a whole compressed air system which
consists of several other centrally controlled components for air supply and
treatment (filter, dehumidification/drying), distribution (e.g. pipe network, puffer
tanks) as well as air usage (e.g. production machines or manual devices such as
blow guns) (Radgen and Blaustein, 2001; Ruppelt, 2003; Mller et al., 2009).
As one big disadvantage compressed air usage is often connected with very
high system losses. Studies show that not even 10% of inserted energy ends up as
usable mechanical energy at the end-use device (Figure 22). As a result
compressed air is actually one of the most expensive forms of energy in industry
27
Fig. 22 Losses during the generation of compressed air depicted as Sankey-diagram (Gauchel,
2006)
(Gauchel, 2006; Gloor, 2000). As Figure 22 underlines, losses occur in all

components of the system. The main amount of energy (>90%, Bierbaum and
Htter, 2004) gets lost as heat at the compressor (typically there are also leakage
losses in the system). As orientation for evaluating the performance, Figure 23
shows typical specific compressor power demand for generating a compressed
airflow of one m3/min with respect to the nominal pressure of the system (typical
range for regular industrial application) whereas the lower boundary is a loss-free
isothermal compression (Gloor, 2000). Having in mind the broad application
fields in industry underlines the huge additional energy demand caused by
compressed air (e.g. a simple impact wrench needs approx. 0.2-1 m/min; a small
sandblasting workplace alone needs at least 1 m/min larger workplaces reach
3-5 m/min) (Bierbaum and Htter, 2004; IES Industrie Engineering Service
GmbH, 2009) as well as the significant losses which are involved (even in a
good case).
Literature mentions different areas for improvement of compressed air generation,
e.g. (Radgen and Blaustein, 2001; Gloor, 2000; Saidur et al., 2010):
improvement of drives in compressors

optimal choice of type and dimensioning of compressor (over dimensioning
as a problem)
recuperating waste heat
decreasing system volume (e.g. block pipes that are not in use)
reducing nominal system air pressure
network design (e.g. pipes, avoiding filtering/flow restrictions)
28
optimising control
avoiding air leaks
improved air treatment
Studies reveal that these opportunities are not seized yet; saving potentials are
estimated with 5-50% (average approx. 33%) for the next 15 years (Radgen and
Blaustein, 2001).
power demand [kW] per [m/min]

not favorable
favorable
boundary
nominal air pressure

Fig. 23 Specific compressor power demand in kW for generating for one m/min compressed
air depending on nominal system pressure (Gloor, 2000)
2.4.3 Steam Generation

Steam is another important form of energy and widely used in different industries
for diverse purposes besides its major role of generating electricity in power
plants. For example, 37% of all fossil fuel burned in US industry are used to
generate steam (Einstein et al., 2001). Typical producing industries with heavy
steam usage are, for instance, the food, paper, pharmaceutical, chemical, textile,
petroleum refining and primary metal industry (Spirax-Sarco Limited, 2005;
Einstein et al., 2001). Steam is a relatively efficient way of generating and
distributing heat which is needed in order to conduct certain production processes
(e.g. sterilising, cleaning, drying, distillation, cooking) or provide space heating.
Advantages of this form of energy are the high heat content and heat transfer
capabilities, safety, cleanness and flexibility (Wagner, 2010; Spirax-Sarco
Limited, 2005). The steam system consists of diverse components water
treatment, boiler, pipes, consumers, steam traps and condensate loop - which are
depicted in Figure 24 (Spirax Sarco, 2006). Steam with predetermined pressure
and temperature is generated through vaporising water through burning fuel
(Rajan, 2008). The steam is distributed to its consumers where heat transfer takes
29
steam
fresh
water tank/freshwater
treatment
steam
using unit
steam
using unit
steam trap
steam trap
economizer
preheated
water
steam trap
flue
gas
boiler
pump
condensate
burner
fuel (e.g. gas)
Fig. 24 System for steam generation and distribution (Spirax Sarco, 2006; Einstein et al.,
2001)
place and the steam cools down to change the aggregate state back to liquid
(condensation). This condensate is flowing back into the boiler to close the cycle
and be vaporised again.
The steam generation process is the core element of the steam system. Within
the steam raising unit, pre-treated (to avoid impurities, which may harm the
system) water is brought to its boiling temperature before being vaporised in the
boiler. In superheaters the steam can be further heated in order to achieve the
designated temperature and pressure. The coherences between the determining
variables (e.g. temperatures, pressure, enthalpy) in steam generation can be found
in so called steam tables (e.g. Bge, 2009; Babcock and Wilcox Company, 2010).
The necessary input energy for vaporisation comes from burning fossil fuels or
biomass/waste used as fuel (Einstein et al., 2001) which results in flues emitted to
the environment. Diverse types of steam raising units are available in industrial
practice; an overview can be found in e.g. (Effenberger, 2000). They differ in
terms of functional principle, possible volume rate and characteristics of the steam
and efficiency which is defined as (in stationary operation and full load)
(Recknagel et al., 2010):

(13)
The boiler capacity is the decisive value to estimate the possible steam generation
rate. The equation shows that the efficiency is determined by the supply rate and
calorific value of the utilised fuel. The necessary amount of fuel ( ) with respect
to the volume rate of steam ( ) as well as the specific enthalpy of steam and
water ( ) can be calculated with (Saacke, 2009):

(14)
Modern steam boilers achieve an efficiency of around 80-85%; main losses occur
through heat loss in flue (approx. 75% of heat losses), radiation on surfaces and
blow-down (Dalzell, 2000; Spirax-Sarco Limited, 2005). In addition to the boiler
30
losses, the distribution system incorporates further losses of 5-30% (Dalzell,

2000). Studies reveal an economic potential of 18-20% for the improvement of
energy consumption in steam systems (Einstein et al., 2001). In the case of the
boiler, relevant leverages are an improved process control, the reduction of flue
gas quantity and excess air (for combustion), improved insulation, boiler
maintenance, blow-down steam and flue gas heat recovery (through economisers)
as well as the usage of alternative fuels. For the heat distribution, improving and
maintaining insulation and steam traps, the avoidance of leaks and circuitry of
condensate are important issues (Einstein et al., 2001).
2.5 Energy Efficiency in Manufacturing

2.5.1 Definition
As mentioned above, efficiency is one of the three main strategies towards a
Sustainable Development as promoted by e.g. the Brundtland Report (Brundtland
Commission, 1987). In general, efficiency is defined as ratio of any output to the
necessary input of a system (DIN 9000). For the case of energy it can be
formulated as (EuP Directive 2005/32/EC, 2005):

(15)
This term is very generic and therefore energy efficiency can mean different
things at different times and in different places or circumstances (European
Commission, 2009). In technical systems (e.g. energy transformations in general
in Chapter 2.2, see steam generation in Chapter 2.4.3) the term efficiency is often
used thermodynamically as the ratio of output energy to input energy in order to
assess the quality of energy conversions (Patterson, 1996).

(16)
In the context of energy efficiency in manufacturing, a physical-thermodynamic

meaning is typically applied, which means optimising the ratio of the production
output (e.g. in terms of quantities with defined quality) to the total energy input
(electricity, gas, oil) (Freeman et al., 1996; Quadriguasi et al., 2009).

(17)
However, this is still a general and simplified equation. Both determining variables
are not easy and generally definable and are subject of discussion in literature
(Patterson, 1996). To correctly evaluate the energy efficiency of manufacturing
systems the consistent definition of the system boundaries (e.g. machine, factory,
country) as well as of the different possible input and output variables (respectively
their units) is critical for application. This is specifically true since just one
efficiency figure has a limited meaning it is important to have a comparison of
values in order to generate meaningful information about the state of a
manufacturing system (Engelmann, 2009). For example, production output could
31
be measured in terms of produced quantities (unit: pieces), masses (unit:

kilograms/kg) or generated revenue (unit: Euro/). Energy input can be expressed,
for instance, by energetic value (e.g. in kWh), energy costs (unit: Euro/) or
environmental impact (e.g. induced CO2 emissions) (Verfaillie and Bidwell, 2000).
The energy efficiency ratio should be as low as possible whereas physical
boundaries set the lower limit. The following equation shows a more specific
energy efficiency definition for the case of a production machine (or system of
machines) powered by electricity for a certain period of time.

(18)
One strategy towards energy efficiency is reducing energy consumption (through

less power demand and/or less necessary time) while keeping output (at least) at
the same level. However, the ratio clearly shows that increasing the output with no
or proportionately less increase of the input energy is also a legitimate way for
improvement.
2.5.2 Potentials and Fields of Action

Different studies reveal the significant improvement potential within industry. The
study Energy Efficiency in Manufacturing carried out by Fraunhofer Gesellschaft
underlines the relevance of production processes in single companies as well as on a
global base and highlights the major potential of increased production process
efficiency to optimise the environmental as well as economic performance
(Fraunhofer IWU, 2008). The study Energy Efficiency in Manufacturing: The Role
of ICT claims saving potentials of 10-40% in manufacturing and stresses the
importance of ICT (Information and Communication Technologies) as an enabler
for energy efficiency (European Commission, 2008). A comprehensive study carried
out in Germany reveals similar significant potentials in the manufacturing industry
regarding e.g. the efficient usage of energy (Seefeldt and Wnsch, 2007).
Altogether, based on the available technology in 2002 and depending on the field of
action, a saving potential of 10-30% on energy consumption was identified. From
todays perspective the potential is likely to be even higher. These rather research
driven estimations of potentials are also confirmed through studies in industry. A
broad survey on different companies revealed an average estimated potential of
energy efficiency of 15% (Schrter et al., 2009).
Some general fields of action to increase the energy efficiency in
manufacturing can be distinguished and will be briefly described in the following
section.
Production Machine Level
Figure 25 shows the possible approaches to increase the energy efficiency on
production machine layer. Relevant influencing variables are:
The design of the machine which includes the general structure/functional

principle (including definition of necessary energy inputs for different
32
purposes), appropriate dimensioning and the selection of energy efficient

components (e.g. efficient electric drives for spindle, pumps etc) (Kuhrke et
al., 2010; Devoldere et al., 2007).
Machine control offers saving opportunities through intelligent logic
interaction of all machine elements (Dietmair and Verl, 2010; Cannata et al.,
2010), e.g. intelligent shutdown (in idle mode).
Process parameters can have significant influence, e.g. increasing the speed
of processing often leads to higher energy consumption (Lanz et al., 2010).
However, energy efficiency as key figure is important to consider since this
increase does not have to be proportional. As mentioned in 2.3.3, the fixed
energy consumption share is often relatively high; processing faster would
therefore reduce the specific energy per part (Klocke et al., 2010; Boos and
Kuhlmann, 2010).
further use of occuring
energy
energy
recovery
substitution of
energy carriers
Eoperation bzw.
auxiliaries
Input
execution
production machine
Efficiency
Losses
Necessary energy
product design
dimensioning
control
Eloss
Output
products
materials
process time t
Fig. 25 Variables to influence the energy efficiency of production machines (Mller et al.,
2009)
Production Planning and Control (PPC)

As mentioned before, the energy demand of a factory is determined by diverse
consumers and adds up to a specific load profile of the different energy carriers.
Figure 26 shows general possibilities to influence the cumulative load curve on
factory layer (Gesellschaft Energietechnik, 1998). In this context, the avoidance of
energy consumption peaks is an important aspect for manufacturing companies in
order to avoid peak surcharges (e.g. electricity billing). Another option is the
shifting of consumption to daytimes with less expensive energy price rates (e.g.
base time at night) (Tnsing, 1996). These options are certainly focusing less on
the decrease of energy consumption but increasing the energy cost efficiency.
However, technical and environmental targets are also relevant. On the one hand,
peak
reduction
utilising load
troughs
33
shifting
load
long term
reduction
Fig. 26 Measures for influencing energy demand from factory perspective (Gesellschaft
Energietechnik, 1998)
a balanced consumption behaviour supports continuous energy and media supply

in favourable working points and therefore less capacity has to be provided.
Altogether, this can result in less investment, less operating costs and also less
energy consumption with its related environmental impact (e.g. dimensioning of
compressed air system or electricity grid). Actual means for influencing the
consumption behaviour are technical measures like energy/load management
systems (e.g. Gesellschaft Energietechnik, 1998) or rather organisational measures
like an energy-aware production planning and control (PPC, e.g. scheduling, lot
sizes)(Schieferdecker, 2006; Bonneschky, 2002). Energy aware PPC in particular
also allows energy consumption reduction through optimal utilisation of
equipment and avoiding energy waste in idling machines. The impact of shifted
production order sequences on energy demand is exemplarily shown in Figure 27
(Rager, 2008).
Supporting Processes Including Technical Building Services (TBS)
In the context of technical building services, the technical configuration of the
equipment (e.g. dimensioning, materials, usage of energy efficient components,
detail design of tubes etc), an efficient process control (e.g. high utilisation avoid
standby, continuous runs versus start-stop-operation, processes at favourable
working point) as well as the avoidance of losses (e.g. leakage, lacking insulation)
are main levers to improve e.g. energy efficiency. Simulation-based approaches
are available to support the development of different measures (Chow, 1996;
Andreassi et al., 2009; Wischhusen et al., 2003; Rebhan, 2002; Kircher et al.,
2010). Some specific fields of action for the case of compressed air supply were
already presented in Chapter 2.4.2, for steam in Chapter 2.4.3.
34
energy
demand
O21
O22
O11
O23
O14
O12
O13
time
energy
demand
O23
O21
O11
O22
O12
O2
O14
O13
time
Oi
operation j of production order i

energy demand of production order J1
cumulative energy demand

r
planned energy demand
der
energy demand of production order J2
Fig. 27 Influence of PPC on energy demand (Rager, 2008)
Building Shell
In civil engineering the improvement of (energy) efficiency of the building itself is
an important topic for many years, and therefore energy efficiency measurements
(and labelling) are commonly used. Typical approaches focus on adequate design
of building elements like insulation, roofing, walls, windows, slabs and
foundations (Energy Star, 2010). The consideration of airflows within the building
is a major element which is supported by diverse available simulation approaches
(e.g. TRNSYS software tool for simulation air/heat flows in buildings). A
comprehensive overview of available tools can be found at (U.S. Department of
Energy, 2010).
Chapter 3
Derivation of Requirements and Methodological

Approach
The following section will first focus on the derivation of requirements for further
support towards energy efficiency in manufacturing. This discussion is based on
the theoretical background but also the industrial demand. These requirements
serve as input for the later selection of the methodological solution approach
being used.
3.1 Requirements from Industrial/Business Perspective

As underlined in the previous chapter; significant potentials to improve the energy
efficiency in manufacturing companies are basically available. However, diverse
studies based on comprehensive surveys also highlight that there are - specifically
for small and medium enterprises (SME) - relevant obstacles which impede a
broad implementation of promising measures despite the general willingness of
many companies (Sorrell et al., 2000; Thollander, 2009; Thamling et al., 2010;
Schmid, 2008). These issues should be consciously addressed and serve as
requirements for a new solution approach. Based on diverse studies/references the
following main obstacles (O) can be identified (Sorrell et al., 2000; Thollander,
2009; Gesellschaft Energietechnik, 1998; Beyene, 2005; Kaiser and Starzer, 1999;
Offner, 2001; Schmid, 2008; Brggemann, 2005; Thamling et al., 2010; Stern,
1984; Jaffe and Stavins, 1994).
O1: Due to necessary investment for e.g. new technologies and the relatively
low level of energy prices, energy efficiency measures may just sum up to
small cost savings with long amortisation time - which make those
measures not attractive to implement. Measures are often too selective and
restricted to certain areas larger approaches with more leverage are missing.
O2: Companies often face lacking access to necessary capital for
implementing energy efficiency measures. Capital is a scarce resource if
available, it is typically needed for investment in other areas of production.
O3: There are no resources for energy efficiency measures personnel and
available time are normally allocated to other tasks in order to keep value
creation running.

springerlink.com
36
3 Derivation of Requirements and Methodological Approach
O4: Lacking organisational responsibility impedes a systematic consideration

of energy consumption and efficiency.
O5: There is very restricted transparency regarding the energy
consumption in the company. Due to the lack of data, main consumers are
rarely known and systematic improvement not possible. Additionally, the
topic as a whole is too complex or assumed to be too complex.
O6: Companies lack the necessary detailed knowledge of energy efficiency
measures including an overview over available technologies.
O7: Altogether, production objectives have often the highest priority whereas
energy consumption is of lower priority. There is a concern about potential
negative influences on production that have to be addressed.
O8: While diverse means for improving the energy efficiency are identified
and described in literature, not every measure may be transferable and
suitable for the specific case of a company. Hence, transferability
respectively verifiability is an important issue which would allow to
estimate the impact of certain measures for the actual case.
Some obstacles are more an issue of politics (e.g. O2 is addressed through support
programs, incentives towards energy efficiency measures) and also market
behaviour (e.g. O1 higher energy prices will automatically drive implementation).
Evidently, keeping up value and creating production is of main interest. In this
context, strategic and operative production management plays a key role as
important interface to energy efficiency related concepts. Besides the determination
of production activities to generate valuable outcome, production management also
heavily influences the energy consumption of the factory system when planning
(e.g. structuring production, selection and dimensioning of technical equipment) and
operating (e.g. scheduling) the facility. However, studies underline that compared
to other variables - energy efficiency is still a minor relevant target objective even
though specifically system planning has significant influence on energy
consumption behaviour in later use phases (Engelmann, 2009).
Against the background of the mentioned obstacles and the relevance of
production management, the following requirements (R) can be derived in order to
support the implementation of energy efficiency measures in industry:
R1: Considering energy efficiency should not be an additional burden in daily

work. No add-on but an integrative approach is needed which delivers
simultaneous perspectives on all time, cost, quality and energy related
objectives. This avoids too much additional effort, supports thinking in cause
and effect chains, enables the identification of impacts on production and
automatically draws attention to the topic in daily work (addresses O3, O4,
O5 and O7).
R2: Complex tasks, simple tools considering energy flows in manufacturing
is very complex. However, while keeping this complexity in the background this
task must be broken down to simple, controllable solutions. This enables
operation in practice without too much effort while coping with the complexity
of the issue (addresses O3 and O5).
3.2 Requirements from Scientific/Technical Perspective
37
R3: There should be appropriate means for easy identification of main

consumption drivers. This allows concentrating on promising fields of
action with large effects, serves as filter for target oriented knowledge
acquisition (e.g. deep analysis of available technology) and reduces
unnecessary efforts (addresses O1, O3, O5 and O6).
R4: It is necessary to be able to evaluate energy efficiency measures on
different levels within one solution, including their interdependencies. For
example, it should be possible to assess and compare both the impact of
smaller improvements (e.g. installing efficient drives in one machine) but also
of more complex solutions (e.g. energy aware production control). This
allows prioritising possible measures (for sequencing implementation) and
fosters systematic improvement (addresses O1, partly O2, O5 and O7, O8).
R5: Finally, the realistic and reliable calculation of key figures is a major
issue, because they serve as base for implementation decisions. This includes
the realistic calculation of actual (energy) costs and the relevant production
performance criteria (e.g. output, throughput time, utilisation) as well as
the necessary consideration of uncertainty and risk (addresses O1, O5
and O7).

In addition to requirements from industry as potential users, there are further
aspects from scientific respectively technical perspective which have to be
addressed by an appropriate solution approach. Section 2 gave an overview on
aspects that have to be considered in context of energy efficiency in
manufacturing. It revealed the enormous complexity of the topic, which is mainly
determined by (Schmid, 2008):
the multitude of elements (each single light bulb or electric drive influences
the energy consumption)
the diversity of elements (entirely different technologies and disciplines
involved)
quantity and diversity of element interactions in context of energy
consumption
the time based volatility of energy consumption that is based on seasons,
intensity of usage etc
A realistic and goal-oriented identification of improvement potentials and actual

measures without the danger of problem shifting, demands a comprehensive
system-oriented view. Against this background, further critical requirements (R)
towards a holistic comprehension of the factory system can be derived (first
published in Herrmann et al., 2010a).
R6 - Extended process comprehension: All relevant input and output flows

of production processes must be explicitly considered in order to avoid
focusing on less relevant issues (while neglecting major challenges) and local
38
optimisation with problem shifting. This includes all energetic (e.g.

compressed air, electrical power, waste heat) and material (e.g. auxiliary
materials as cooling lubricants) flows, which lead either directly or indirectly
to additional energy and/or resource consumption. Against this background
Figure 28 shows an extended process model, which integrates previously
separately considered ecological and economic process perspectives (Schultz,
2002; Herrmann and Thiede, 2009a).
Ecological Process Model
Economic Process Model
Input
Output
Energy
Electric energy
compressed Air
Energy Emission
Waste heat
Waste air
Input
Process
Material
Raw material
Auxiliary material
Output
Products
Resources
Energy
Material
Personnel
Material
Emission
Waste material
Special waste
Information
Process
Scrap
Waste
Integrated Process Model

Input
Output
Personnel
Products
Information
Scrap
Energy
Electric energy
compressed Air
Material
Raw material
Auxiliary material
Process
Machine
Quality gate
Buffer
Energy Emission
Waste heat
Waste air
Material
Emission
Waste material
Special waste
Fig. 28 Integrated process model (based on Schultz, 2002)
R7 - Holistic system definition of the factory: Manifold interdependencies

between the constituting elements of a factory also demand for an enhanced
comprehension of the factory system as a whole. This system can be divided
into the three sub-systems, namely production (machines and employees,
coordinated by production planning and control), technical building services
(TBS) and building shell. Altogether, this results in a complex control system
with dynamic interdependencies between different internal and external
influencing variables (Figure 29, Hesselbach et al., 2008b).
39
local
climate
cooling
heating
need for defined

production conditions
(e.g. temperature, moisture, purity)
gas, oil,
electricity
technical
building
services
(TBS)
waste heat
exhaust air
allocation of
media
(e.g. compressed air,
steam, cooling water)
production
machines
(e.g. steam,water)
water
backflow of media
electricity
Fig. 29 Holistic definition of factory (own illustration, first presented in Hesselbach et al., 2008b)
As shown in Figure 29, one major task of technical building services is to

ensure the needed production conditions in terms of temperature, moisture and
purity through cooling / heating and conditioning of the air. The essential
influencing variables are the local climate at the production site (seasonal
influences) and the exhaust air and the waste heat that is primarily emitted by
production machines but also by other production factors like transportation
equipment or even personnel. Furthermore, production machines need energy
(mostly electricity) and diverse media like compressed air, steam or cooling
water to fulfil their designated processes. Technical building services are also
responsible for the supply of these essential media whereas this involves their
generation, their circuitry as well as the required conditioning (e.g.
temperatures, pressures, purity). Altogether an evaluation of a factorys energy
consumption must consider all non-regenerative energy flows that are
externally supplied (e.g. electrical power, oil, gas) for running technical
equipment for both production and TBS. Evidently, these sub-systems strongly
interact with each other - they cannot be seen independent from each other
when pursuing a holistic approach towards energy efficiency in manufacturing.
R8 - Dynamics of consumption and emission behaviour on machine and
factory level: As shown before, all relevant input and output flows are typically
not static values but highly dynamic depending on the operating conditions of
40
the processes and the machines. These profiles add up to cumulative load
profiles on the factory level. In the end these dynamic cumulative load profiles
(e.g. process heat demand, compressed air demand, heat flow into the factory
building, electrical power demand) are decisive for design and control of the
technical equipment (e.g. dimensioning of compressed air system) as well as for
billing (e.g. energy supplier). However, the specific consumption and emission
patterns of machines and the system as a whole are rarely available. For
machines, typically just single nominal values are available which describe a
maximum demand and inherently incorporate certain safety factors. Hence,
these values neither reflect magnitude nor state based dynamics of consumption
sufficiently. Even though, nominal values combined with demand/simultaneity
factors are often applied in order to dimension infrastructure (e.g. electric grid,
compressed air system, air conditioning, boilers) on factory level. Due to
lacking transparency of consumption patterns these factors are typically also
just rough estimations. Altogether, the insufficient data base in connection with
the application of unrealistic (while unknown) demand factors plus additional
safety factors lead to an unfavourable dimensioning of technical equipment,
which results in unnecessary energy consumption (Mller et al., 2009). As an
example, Figure 30 shows the superposition of single machine consumption
patterns to a cumulative load curve on factory layer. In this case steam is
considered that has to be provided through boilers. Taking into account the
considerations above, the alternative cumulative profiles will result in different
necessary dimensioning of the boilers as well as different energy costs (e.g.
through consumption peaks and different idle consumption).
25000
energy demand [e.g. kW]
20000
18000
14000
15000
power [W]
power consumption [W]
16000
12000
10000
10000
8000
6000
5000
4000
2000
0
14:24
14:52
15:21
15:50
16:19
16:48
17:16
17:45
time
18:14
18:43
time [min]
10
12
0
23
0
34
0
45
0
56
0
67
0
78
0
89
0
10
00
11
10
12
20
13
30
14
40
15
50
16
60
17
70
18
80
19
90
21
00
22
10
23
20
24
30
25
40
26
50
27
60
28
70
29
80
30
90
32
00
33
10
34
20
35
30
36
40
37
50
38
60
39
70
Leistung Sterilisa tionszyklus

20000
time [min]
One machine, several cycles
time [s]
One machine, one cycle
Several production machines, several cycles

70000
70000
60000
60000
50000
50000
40000
40000
30000
30000
20000
20000
10000
10000
0
0
1
13 19 25
31 37
43 49 55 61 67 73 79
85 91 97 103 109 115 121
time [min]
unfavorable
Fig. 30 Steam demand of one and several machines
13 19 25 31 37 43 49
55 61 67 73 79 85
better
time [min]
91 97 103 109 115 121
3.3 Research and Methodological Approach
41
R9 - Thinking in process chains: Final products are usually not the result of
single production processes, but are rather manufactured in several steps on
different production lines in the sense of production process chains. Against
the background of energy and resource efficiency, the process chain has to be
regarded and evaluated as a whole, as it may involve further potentials (e.g.
combination of processes). Moreover, problem shifting might occur while
improving measures in one process can possibly lead to worse performance of
others.
R10 - Life-cycle-oriented perspective: Analogous to the thinking in process
chains, all life phases of products (this includes also all the technical
equipment within the factory itself) have to be considered when it comes to
deriving measures concerning energy efficiency. Thus, the decisive factor for
increasing the energy efficiency of a machine tool, for instance, is less the
improvement of single parameters of a specific process than rather the
development of the machine itself. Moreover, the choice of specific processes
(e.g. joining techniques) has direct effects on the use and disposal phase
which could lead to increased efforts in those phases. The life-cycle-oriented
perspective also stresses the importance of energy aware planning of
production facilities as the energy consumption behaviour is mainly
determined in the planning phase (Herrmann, 2009; Herrmann et al., 2007a).
R11 - Consideration of all sustainability dimensions and integrated
evaluation: In order to deduce advantageous solutions, several relevant target
dimensions have to be considered simultaneously. Going even further than
requirement R5, besides a realistic economic (on the basis of a suitable cost
model which integrates real contract conditions) and technical evaluation (e.g.
effects on product quality), this includes an ecological evaluation (with a
correct balance of the different input and output parameters, e.g.
environmental effects of electricity and gas consumption). Possible conflicts
of goals must be disclosed and decision support to solve them has to be
offered (Herrmann, 2009).

As short summary of the requirements derived above the holistic solution
approach should be an easy to use solution for decision support in strategic and
operative production management under simultaneous consideration of
all relevant energy flows and connected dynamics,

energy driven objectives with conventional time, quality and cost based
targets,
realistic calculation of energy induced costs and environmental impact,
means towards energy efficiency on both single machine as well as factory
system level.
As described before, single machine considerations are necessary but not

sufficient for a holistic perspective on energy efficiency in manufacturing. Hence,
42
the estimation of time based consumption patterns on a factory level is of major

importance since this is the decisive base for
dimensioning infrastructure respectively supporting processes like technical

building services,
realistic evaluation of energy costs, environmental impact and technical
performance,
derivation and assessment of energy efficiency measures going beyond single
machine improvement.
For considering factory system consumption profiles four alternative methodological

approaches could be pursuit: static calculations, fuzzy logic, artificial neural
networks and simulation.
Static Calculations
Through combining available single machines consumption information and
production data, the energy demand of a factory can be calculated statically using
standard office software like MS Excel. Figure 31 shows an analysis for the
electricity consumption of a SME company. The electricity consumption of each
major consumer was measured and ex-post - put together with recorded
production data (e.g. production program, machine runtime). The figure shows the
good approximation in comparison with the electricity consumption as measured
by the energy supplier (for the considered month the variance is even below 1%).
2500
120
100
100
80
80
60
60
40
40
20
20
electricity consumption [kWh]
electricity consumption [kWh]
2000
120
1500
1000
time
1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73 76 79 82 85
88 [hours]
91 94
Real 15 min measurement
MIN prognosis
MAX prognosis
Real 1h average
500
time [day]
0
1
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
total electricity consumption - MIN prognosis
total electricity consumption - MAX prognosis
total electricity consumption - measured
Fig. 31 Static ex-post calculation of electricity consumption and comparison to actual

values (left: daily profile, right: monthly values)
The method increases transparency regarding energy flows and allows a good
first estimation of consumption patterns in factory as well as a simplified
evaluation of selected efficiency measures. However, while being an ex-post
consideration, it obviously relies on the existence of detailed production data for
the considered period of time. Ex-ante analyses based on planned production
program are possible but would hardly achieve the necessary accuracy since
uncertainties and interdependencies cannot be considered. This is specifically true
for more complex cases with more consumers and energy flows (and more
complex efficiency measures), which would also result in significant higher
demand for calculation.
43
Fuzzy Logic
Lau et al. presented a fuzzy logic approach (Yager and Zadeh, 1992) to forecast
energy consumption change in a manufacturing company (Lau et al., 2007). It
aims at predicting the amount of energy consumption change compared to a
reference energy consumption level that is based on selected production input
variables. This shall help to e.g. consider whether the necessary power can be
supplied. In their case study, Lau et al. applied their methodology to a clothing
manufacturing plant. Based on the input variables (recorded for a certain period of
time) daily total mass of finished products, total labour hours of operators in the
plant in one day and the total running time of equipment in the plant they predict
an additional demand on energy of 2.64% for this period compared to the
reference consumption (Lau et al., 2007).
The fuzzy logic approach is based on a black-box-oriented perspective of the
manufacturing system without detailed consideration of actual single consumers
and their interdependencies. Furthermore, the method can provide the energy
demand but not the necessary time based load profiles. These factors combined
with the specific expertise needed for application restrict the applicability in
context of the aforementioned requirements.
Artificial Neural Networks (ANN)
Some authors propose the usage of Artificial Neural Networks (ANN) to predict
the consumption profile of a company (Fiedler et al., 2007; Lang and Hesselbach,
2009; Hufendiek and Kaltschmitt, 1998). Measured consumption data is used to
train the network and the impact of selected influencing parameters can also be
considered. A well trained ANN allows a good estimation of the time based
consumption profile of e.g. electricity or gas (Fiedler et al., 2007). However,
similar to the fuzzy logic approach the manufacturing system is considered as
black-box being determined by selected influencing variables. It is not possible to
break consumption down to single consumers. Additionally, ANN is a relatively
complex method, which requires expertise and time for network setup and
training.
Manufacturing System Simulation
Simulation is an established method within the planning phase of manufacturing
systems and diverse mature software solutions (e.g. Plant Simulation,
Delmia) are commercially available. However, energy related aspects are so far
not being considered in these tools. In research, first approaches can be found that
aim at augmenting material flow simulation with energy consumption data.
Furthermore, detailed simulation models for e.g. technical building services or the
building itself are available. In general these approaches are basically enabled
to estimate the cumulative energy demand while keeping the necessary level of
detail regarding cause and effect relations. Whereas being able to consider
interdependencies and uncertainties (e.g. waiting times, failures) also a relatively
reliable ex-ante prediction of energy demand might be possible. However, also
44
simulation is more an expert tool and involves costs (licence) and time for model
setup, validation/verification and simulation studies.
Table 3 shows a simplified evaluation of the four alternative methodological
approaches based on a ranking with respect to the possible fulfilment of the identified
requirements. A high ranking does not mean that all requirements are already fulfilled
in those approaches it simply indicates that it would be possible to address them
properly from a methodological perspective (and compared to the alternatives).
Manufacturing simulation clearly scores highest in almost all categories. It would
allow addressing all requirements appropriately. Static calculation scores second but
has major disadvantages when it comes to more complex energy efficiency analyses.
ANN offers some advantages; at least once a well-trained network is available. It
does however not allow deeper looks into the factory, which limits fields of
application. The applicability of fuzzy logic is even more limited since not even a
cumulative load profile can be predicted. Besides that it is a rather complex approach
with limited usability in daily industrial practice.
Table 3 Evaluation of general methodological approaches based on identified requirements
(ranking for each requirement from first to fourth place)
Static calculation
Fuzzy Logic
Artificial Neural
Networks (ANN)
Manufact. System
Simulation
R 1: no add-on, integrative
approach
R2: simple, controllable

solution
R3: identification of main

consumption drivers.
R4: evaluation of energy

efficiency measures on
different levels
R5: realistic and reliable

calculation of key figures
R6 - extended process
comprehension:
R7 - holistic system
definition of the factory
R8 - dynamics of
consumption behaviour
R9 - Thinking in process
chains
R10 - Life-cycle-oriented
perspective
R11 integrated
consideration of all
sustainability dimensions
2.0
3.4
2.7
1.1
Mean Average Ranking
As a consequence, manufacturing system simulation is selected as methodological

approach for the later concept. In the following some theoretical background will be
provided.
3.4 Simulation Background
45

Simulation in general can be defined as imitation of the operation of a real-world
process or system over time (Banks, 2010, similar definitions can be found in e.g.
VDI, 2007). In this context, a system is defined to be a collection of entities, e.g.
people or machines that act and interact together toward the accomplishment of
some logical end (Law, 2007 referring to Schmidt, 1970). Thereby, a system is
determined by its structure (the inner constitution of a system) and behaviour (its
outer manifestation) (Zeigler et al., 2000). The behaviour of a system determines
the relationship between its inputs and outputs. The internal structure includes the
systems states and state transition mechanisms (Zeigler et al., 2000). For
simulation, an appropriate model is necessary, which is a representation of a
system for the purpose of studying that system (Banks, 2010). In contrast to a
physical model, a mathematical model is meant in this context (Law, 2007).
Simulation is an established method for years in research and industry. Several
examples of diverse applications can be found in (Engelhardt-Nowitzki et al., 2008;
Bayer, 2003; Krenn, 2007); whereas a comprehensive overview over recent
simulation applications provides (Jahangirian et al., 2010). Simulation offers some
important advantages (summary based on Banks, 2010; Pegden et al., 1995; Law,
2007; Chung, 2004):
Testing alternatives (e.g. policies, procedures, system design and control

strategies) without disturbing the real system
Testing of new system elements without necessity to acquire them (e.g.
transportation systems, layout)
Time can be compressed or expanded, which allows detailed analysis of
interactions and/or long-time studies
Simulation can provide deep understanding of the system with respect to its
cause and effect relations (interaction and importance of variables). It can help
in understanding how the system operates rather than how individuals think the
system operates (Banks, 2010) and can make it visible to all stakeholders.
A detailed analysis of e.g. hypotheses, certain phenomena or new design
concepts is possible. This includes bottleneck analysis and answering "What
if? questions in order to improve the system.
However, simulation also involves some drawbacks which have to be addressed

carefully when conducting studies (e.g. Banks, 2010; Chung, 2004).
Simulation requires expertise and each model is influenced by the individual

who created it.
Simulation results can be difficult to interpret and their quality directly
depends on the quality of input variables. The selection of appropriate
performance criteria is critical and evaluation may involve statistical analyses.
Simulation requires significant effort in terms of time and costs. For each case
it should be carefully considered whether other (analytical) solutions might be
possible or preferable.
46
Head of water behind the dam
Number of customers waiting in line

or being served
Simulation approaches can be distinguished with respect to different criteria (e.g.

Law, 2007). For example, static simulation models representing a system at a
certain point of time (e.g. Monte-Carlo Simulation) whereas dynamic models are
considering the evolution over a period of time. Deterministic models do not
involve stochastic effects and lead to a distinct, repeatable result. In contrast to
that probabilistic models involve one or several random variables (described
through certain distribution function) and different runs will have different results.
Therefore statistical treatment is an important issue.
Furthermore, a major distinction can be made between discrete and continuous
simulation models. In discrete systems state variable(s) change only at a discrete
set of points of time (Banks, 2010) whereas in continuous systems the state
variable(s) change continuously over time (Banks, 2010; Cassandras and
Lafortune, 2008). This important differentiation is shown in Figure 32. Nowadays,
also so called combined discrete-continuous or hybrid-simulation approaches
exist, which are able to combine both perspectives. Thereby, discrete events may
change the state of a continuous variable or (vice versa) a continuous variable
which achieves a threshold triggers a discrete event (Pritsker, 1995).
3
2
1
Time [e.g. min]
Time [e.g. min]
Fig. 32 Example of discrete (left) and continuous (right) state variable (Banks, 2010)
Based on these classifications four main simulation paradigms can be

distinguished, which are displayed in Figure 33. Due to their different fields of
application and logic, these paradigms were used by different research disciplines
and implemented using different software tools (Borshchev and Filippov, 2004).
Dynamic Systems Simulation is typically based on state variables and

algebraic equations (often depicted through block diagrams), which describe
the behaviour of physical systems. The standard tool for this application is
MATLAB (www.mathworks.com/products/matlab).
System Dynamics Simulation was invented by J.W. Forrester in the 1950s
and describes the system with balancing or reinforcing feedback loops
modelled as stock and flows (Forrester, 1969). It typically addresses problems
on a higher level of abstraction such us whole ecological or economic systems
(e.g. market behaviour, Sterman, 2000). An established tool for system
dynamics modelling and simulation is Vensim (www.vensim.com).

High Abstraction
Less Details
Macro Level
Strategic Level
47
Aggregates, Global Causal Dependencies, Feedback Dynamics,...
Agent Based
(AB)
Active objects
Middle Abstraction
Medium Details
Meso Level
Tactical Level
"Discrete
Event" (DE)
System Dynamics (SD)

Levels (aggregates)
Stock-and-Flow diagrams
Feedback loops
Individual
behaviour rules
Direct or indirect
Entities (passive
dobjects)
Flowcharts and/or
f f interaction
Environment
models
dtransport networks
Dynamic Systems (DS)

Physical state variables
Block diagrams and/or
Resources
Low Abstraction
More Details
Micro Level
Operational Level
algebraic-dif f erential equations

Mainly discrete
Mainly continous
Individual objects, exact sizes, distances, velocities, timings, ...
Fig. 33 Overview simulation paradigms (Borshchev and Filippov, 2004)
Discrete Event Simulation is widely used in research and industrial practice

for diverse purposes. Typically, passive entities (which represent e.g.
products, people, messages) travel through a system (e.g. blocks in a
flowchart) where they trigger certain actions and being e.g. delayed,
processed, or combined (Borshchev and Filippov, 2004). An established
application is the simulation of material flows in manufacturing systems.
Therefore diverse specialised software tools such as Plant Simulation
(www.plant-simulation.com) or Delmia (www.3ds.com/products/delmia)
are available.
As relatively new paradigm Agent Based Simulation focuses on
decentralised modelling of individual object behaviour in a defined
environment via state charts (Borshchev and Filippov, 2004). Each object acts
individually based on his inherent logic and interacts with other objects.
Altogether this leads to the behaviour/state of the considered system as a
whole (e.g. transportation, pedestrians).
Some more recent software tools allow the integrative application of diverse
paradigms, e.g. for combined discrete-continuous simulation models. Examples are
Arena (www.arenasimulation.com, Kelton et al., 2010), Extend (www.extendsim.
com) or specifically AnyLogic (www.xjtek.com/anylogic, Borshchev and Filippov,
2004).
Independent from the underlying paradigm some typical phases for conducting
simulation studies can be distinguished (Banks, 2010; Law, 2007; Wenzel et al.,
2008; VDI, 2007). An example is shown in Figure 34.
48
Problem
formulation
Setting of
objectives
and overall
project plan
Model
conceptualisation
5
No
Model
translation
6
Verified?
Yes
7
No
No
Validated?
Yes
8
Yes
Experimental
design
Production runs
and analysis
10
Yes
More runs?
No
11
Documentation
and reporting
12
Implementation
Fig. 34 Steps in a simulation study (Banks, 2010)
Data
collection
49
Verification and validation play an important role for reliable and realistic
simulation results. Verification stands for the formal proof that the model itself is
correct and its behaviour is consistent with other models (e.g. description model)
(Rabe et al., 2008). Validation is the assessment whether the model behaves with
satisfactory accuracy consistent with the study objectives (Banks, 2010).
Practically, this could be the comparison of a simulation run with the comparable
situation of the real system whereas the results should be consistent. A collection
of available methods for validation and verification (differing in terms of
applicability and subjectivity) can be found in (Rabe et al., 2008); an overview is
shown in Figure 35.
degree of subjectivity
very high
low
Animation
Review
Desk Checking
Structured Walkthrough
Submodel Testing
Face Validity
Monitoring, Operational Graphics
Trace Analysis
Turing Test
Sensitivity Analysis
Comparison to other Models
Fixed Value Test
Extreme-Condition Test
Cause-Effect Graph
Predictive Validation
Historical Data Validation
Dimensional Consistency Test
Event Validity Test
Statistical Techniques
Internal Validity Test
Fig. 35 Techniques for Verification and Validation and their subjectivity (Rabe et al., 2008)
Chapter 4
State of Research
The latter section revealed that simulation techniques are necessary to be able to
realistically consider energy efficiency in manufacturing systems. There is a need
to combine manufacturing system respectively factory level oriented simulation
approaches with realistic representation of energy flows. Against this background
the following section analyses and evaluates existing research approaches in this
area with specific criteria derived from aforementioned requirements. Based on
this detailed consideration further research demand will be deduced.
4.1 Background for Selection and Evaluation of Existing

Approaches
Procedure and Limitations of Analysis
For the matter of this analysis a deep review of relevant books and research papers
in the field of (e.g. manufacturing) engineering but also adjacent disciplines (e.g.
operations research, computer science) was conducted. Taking into account the
reasoning of the chapters before, necessary limitations regarding the selection of
research approaches can be derived directly:
While being an appropriate and necessary method to capture the dynamics of

the factory system, only research approaches involving the simulation in
(discrete) manufacturing are being considered in detail. To be more
specific, considered approaches should (at least partly) involve discrete event
simulation while this is the most established paradigm for manufacturing
system simulation. This excludes pure mathematical (optimisation) models
(e.g. Larsson and Dahl, 2003; Ferretti et al., 2008) and/or specific approaches
for industries which are not in focus of this book (e.g. agriculture Neumann,
1985, chemical/process engineering Mark et al., 2009).
Numerous approaches for different fields of application can be found which
are focusing on the simulation of single processes or machines (e.g. FEM
finite element simulation of machining processes). However, multimachine manufacturing systems are considered here.
While being of major interest, relevant energy related flows need to be
considered at all and with a certain level of detail. Just naming a
cumulative energy demand on a high level of aggregation (e.g. for calculation

springerlink.com
52
4 State of Research
of overhead costs) is not sufficient. Energy consumption should be broken

down to at least defined production areas if not individual machines with
appropriate time resolution.
A direct relation to and certain focus on manufacturing systems is
necessary. This excludes research work being done to dynamise LCA (Life
Cycle Assessment) with a significantly broader scope. These approaches
certainly include one or even several production process(es) but specific
interactions (e.g. queuing, machine behaviour) are rarely considered (examples
are Gbel and Tillman, 2005; stergren et al., 2008). Additionally this
limitation excludes research that exclusively focuses on the simulation of
technical building services without direct relation to the production environment
(e.g. Kircher et al., 2010; Wischhusen et al., 2003; Andreassi, 2009).
Definition of Criteria
Against the previously discussed theoretical background and connected
requirements for the pursued solution approach, manifold specific criteria were
derived, which will serve as base for evaluating relevant research approaches
(Figure 36).
These criteria are structured in four main areas and described in the following
oriented towards an idealistic degree of compliance.
Consideration of Energy and Resource Flows: Referring to the described
theoretical background, an ideal approach has to consider all relevant energy and
resource flows. This includes all forms of energy including externally supplied
gas, oil or electricity as well as internal energy flows like compressed air, cooling
water or heat. Additionally, raw or auxiliary material consumption might be of
interest as it has direct and indirect influence on energy and resource efficiency of
the manufacturing system. Not all flows are important in every case. However, the
relevance of considered and ignored flows and possible limitations should be
clearly pointed out (criterion completeness of energy and resource flows). Energy
and resource consumption is not static but depending on machines respectively
process states and time. This dynamic consumption behaviour has to be
considered with sufficient accuracy. This is also true for the cumulative load
profiles on manufacturing system respectively factory level, which are important
to e.g. dimension and control technical building services (e.g. cumulative
compressed airflow demand for whole factory) and billing purposes (e.g. 15
minute interval of total electricity consumption as base of energy cost calculation)
(criterion realistic representation of consumption dynamics). Following the
previous criteria, interdependencies within the whole factory system need to be
considered realistically. For example, energy and resource consumption as well as
supply of supporting processes (e.g. technical building services) are directly
depending on the state of the manufacturing system and vice versa. Just naming
e.g. the total electricity consumption of compressors for a month is certainly a
starting point. However, it does not support a detailed consideration of cause and
effect in order to optimise the whole system or realistic allocation of consumption
and costs (criterion interdependencies with technical building services).
4.1 Background for Selection and Evaluation of Existing Approaches
53
Background
Requirements
Criteria
Sources for relevant

input information.
Derivation of general
requirements.
Breakdown to detailed,
assessable criteria.
Obstacles
towards
implementation
of energy
efficiency
measures
Energy and
resource flows
Completeness
Dynamics
TBS
Industry /
business
perspective
Fields of action
(R1R5)
Technological
Organisational
Optimisation
Evaluation
Scientific /
technical
perspective
Implementation
(R6R11)
Selection of
simulation as
methodological
background
requirements
Technical
background in
context of
energy flows in
manufacturing
Economic
Ecological
Technical
Decision support
Uncertainty
Transferability
Effort
Visualisation
Application Cycle
Evaluation of
existing simulation
approaches and
derivation of
research demand
Fig. 36 Methodology for deriving requirements and criteria for the solution approach
Focused Fields of Action for Improvement: As pointed out above, diverse

measures to improve energy efficiency in manufacturing systems are basically
available. Simulation based approaches should be able to derive, analyse and/or
evaluate all kinds of measures on different levels. This includes technological
measures (criterion technological measures) like machine, TBS or manufacturing
system design and control (e.g. impact of new machine concept in context of whole
system, manufacturing system layout/logistics, installed load management to avoid
energy peaks) as well as organisational measures (criterion organisational
measures) like production planning (e.g. sequencing, order/lot sizes), maintenance
strategies (e.g. preventive maintenance) or the management of employee behaviour
(e.g. rules, operating procedures). Ideally, the usage of optimisation methods should
be possible: going beyond conducting simulation experiments with pre-defined
54
4 State of Research
parameter sets this would allow to strive towards not just improved but actually
optimal solutions (criterion possibility of optimisation).
Evaluation Methods: The results of an evaluation are the base for decisions
regarding the implementation of certain measures. By definition, striving towards
more energy efficient solutions requires three dimensions of evaluation. On the
one hand technical aspects have to be considered to ensure that the system is
capable to fulfil its designated technical production task (to deliver a product with
defined quality and at the right time) at all (criterion scale and scope of technical
evaluation). Example key figures in the present context are throughput times,
utilisation/availability/quality rates, which may be available on system level or
even be broken down to single machines and processes. These aspects are closely
connected with the economic evaluation whereas this is a transformation into cost
impact (criterion scale and scope of economic evaluation). Two main issues are of
major relevance in this context: firstly energy and specifically electricity costs
demand complex cost models which respect also peak surcharges and daily time in
order to realistically calculate actual costs. Secondly, for the consideration of
further production cost components (e.g. personnel costs for operation or
maintenance, material costs, downtime/failure costs) it is desirable to put energy
related efforts in an overall costs scheme and avoid negative side effects (e.g. cost
shifting). As third dimension, an ecological evaluation is necessary, which reflects
the environmental burden caused by energy consumption (criterion scale and
scope of ecological evaluation). A wide range regarding the level of detail is
possible, ranging from a conversion of consumption values to related greenhouse
gas emissions (based on local energy source mix) up to detailed product or process
based LCA (Life Cycle Assessment) with consideration of all input and outflows
and diverse impact categories. As pointed out before in decision situations all
evaluation dimensions need to be considered simultaneously, which may lead to
conflicts of goals. To support decision appropriate methods to balance conflicts of
goals and derive the most promising solutions from an integrated perspective is
necessary (criterion decision support). As both are an advantage but also challenge
applying simulation often involves the consideration of stochastic behaviour, e.g.
when considering failure probabilities of machines, or volatile time distributions
of operations or maintenance actions. This issue needs to be addressed to ensure
the quality of results with either assuring that stochastic effects are not significant
in the specific case or providing statistical confidence through sufficient
simulation runs (criterion consideration of uncertainty).
Implementation: Besides rather content driven criteria some aspects regarding
the embodiment of simulation based research approaches (referring to the
challenges connected with simulations studies as described above) need to be
addressed. Transferability is one example which stands for the broad applicability
of the approach for different cases and purposes with reasonable effort (criterion
transferability). This involves several perspectives: on the one hand not only one
specific case shall be considered, e.g. the production of one selected product on a
fixed production line. It should rather be possible to consider multi-product
4.1 Background for Selection and Evaluation of Existing Approaches
55
production with, ideally, flexible production process chains. Furthermore the

solution should not be too company and industry specific but be easily applicable
to other companies of the same or other industries. Somehow related to that issues
like the possible adaption to other questions (e.g. using the same approach to
improve manufacturing system and/or technical building services and/or product
design while determining related environmental impact ) or the necessary
knowledge of users (easy involvement of other people possible?) will also be
evaluated. Another crucial aspect is the effort in terms of time and costs for
modelling and simulation (criterion modelling and simulation effort). Applying
simulation methods demand specific and often expensive software tools as well as
time to model, simulate (computing time) and evaluate the actual case. Being
more visionary than realistic, an ideal approach should be based on free or
reasonable priced standard software with fast simulation routines as well as
appropriate methods to support easy and flexible design of the model (e.g.
standard modules) and integrated/automated evaluation. Additionally,
visualisation is of major importance while it enables to check the state of the
simulation during runtime and supports the understanding of cause and effect
within the system (criterion visualisation). It can be distinguished between the
visualisation of the considered system itself displaying all things happening during
runtime (e.g. material flow, machine states) and also of the relevant target criteria
with proper key figures (and hints for improving the system, e.g. share of standby
consumption) for evaluating the energy efficiency. To tab the full potential of
simulation studies both perspectives should be displayed meaningfully and
simultaneously during runtime. In the worst case, simulation runs are processed in
the background without further visualisation and some results were generated
which may even have to be transformed manually into relevant key figures.
Certainly it should be mentioned that for some purposes (e.g. conducting various
runs for parameter optimisation) a detailed visualisation of the production
environment is not reasonable while unnecessarily leading to longer computing
time. However it should be at least available as an option while strongly
contributing to understand system behaviour. Finally, to support the correct and
target oriented applicability, the simulation approach should ideally be embedded
within an appropriate comprehensive methodology respectively application cycle
(criterion embedment within application cycle). Starting from scratch without any
specific prerequisite therewith, it shall be possible to get to promising measures to
improve energy efficiency based on a verified/validated simulation model with
appropriate scale and scope (e.g. definition of system boundaries and evaluation
criteria, relevant energy and resource flows and possible data sources, level of
detail and accuracy).
Table 4 gives an overview of all criteria with their specific range of value in a
five step evaluation scheme. Each field (except if there is no fulfilment) counts for
a quarter of a point and can sum up to a maximum of one (100% fulfilment) per
criterion.
56
4 State of Research
Table 4 Criteria for evaluation of research approaches
Criteria
Energy and Resource Flows

no flows
one flow
more than one

flow
all internal
Dynamics
single value
one flow state

based
more than one

state based
cumulative
load profile for
one flow
available
TBS
not considered
consumption of consumption of
one TBS
more than one
subsystem
TBS subsystem
considered
considered
interactions of
one TBS
subsystem
considered
Technological
not considered
Organisational
not considered
Optimisation
not considered
Economic
not considered
Ecological
not considered
Technical
not considered
Decision Support
not considered
Uncertainty
not considered
Transferability
specific solution
Effort
time and costs

high
little expertise
necessary
low software
costs
Visualisation
not provided
material flow
visualised in
runtime
results in
runtime
Application Cycle
not provided
comprehensiveness ensured
validity
ensured
Completeness
Fields of action
Machine
TBS
improvement
improvement
PPC focus
(depending on actual strength)
addressed and possible, not
necessarily conducted
Evaluation
conversion in
costs
consideration
of several
energy carriers
conversion to
conversion to
e.g. CO2 one
e.g. CO2
several flows
energy flow
output/production time
considered
comparison/discussion
addressed, not necessarily
actually considered in detail
Implementation
approved for
branch
different
spanning
manufacturing
approach
systems
all external
cumulative
load profiles
for several
flows
available
interactions of
more than one
TBS
subsystem
considered
Prod. System
integrated
improvement
perspective
employee/behaviour focus
(depending on actual strength)
optimisation studies are
conducted
realistic cost
complete cost
model
model with
including
different cost
power peaks
portions
simplified LCA
full LCA
with selected
based on LCI
flows/ impacts
databases
further aspects considered
methods for integrated
evaluation
considered through appropriate
methods
appropriate to
solve other
questions
low modelling
time
automatic
processing of
results at the
end
support for
systematic
improvement
not
necessarily
expert usage
low
simulation
time
meaningful
processing for
decision
support
support for
data
acquisition
4.2 Evaluation of Relevant Research Approaches
57

Based on the previously described limitations 12 relevant research approaches
were identified. They will be presented in the following. Beginning with some
general information and a short description, each approach will be discussed based
on the evaluation criteria described above. The evaluation can obviously just be
based on information given in publicly available sources. This may involve that
certain criteria cannot be assessed due to a lack of information. However, this will
be indicated in the discussion.
Table 5 Evaluation of SIMTER approach developed by Heilala et al.
Main
Author(s)
Institutions
References
Software
tools
Juhani Heilala et al., Bjrn Johannsson et al. (SIMTER

consortium)
VTT Technical Research Centre of Finland, Finland
Product and Production Development, Chalmers
University of Technology, Gothenburg, Sweden
(Heilala et al., 2008; Lind et al., 2009)
3DCreate and 3DRealize of Visual Components, MS
Excel
Completeness
Dynamics
TBS
all external energy sources considered, no internal

flows
state based energy consumption of machines, no
cumulative load profiles
TBS not considered at all
Fields of action
Technological
mainly manufacturing system design is focused
Organisational
not considered
Optimisation
not addressed
Evaluation
Economic
not considered
Ecological
LCA based on LCI data planned; practically mainly

energy related CO2 emission considered
58
4 State of Research
Table 5 (continued)
Technical
Decision
Support
Uncertainty
mainly capacity considered in context of energy

consumption evaluation
Methods for an integrated evaluation of ergonomics,
level of automation and environmental impact, but less
in terms of energy efficiency
not addressed
Implementation
Transferability
Effort
Visualisation
Application
Cycle
different branches and questions can be considered,

unclear whether appropriate for flexible multi-product
production, rather expert solution
if modules in simulator available relatively fast
modelling possible, expensive simulators and expertise
necessary
material flow visualised in 3D, due to model approach
no energy related results during runtime, manual postprocessing of data necessary, no support regarding
meaningful visualisation of results
not provided
The SIMTER project was carried out during 2007 and 2009 as a Finnish-Swedish
cooperation (Lind et al., 2009). The project as a whole aimed at developing an
integrated simulation tool simultaneously analysing the level of automation, the
environmental impact and ergonomics of manufacturing systems as decision support
for manufacturing system design engineers (Heilala et al., 2008). It focuses on the
manufacturing system planning phase before first implementation and start-up. Thus
issues like manufacturing system layout/bottleneck analysis are of major interest e.g.
through varying the number and dimensioning of production equipment (e.g. cycle
times, buffer sizes). Figure 37 shows the analysis flow of the environmental impact
calculation sub-tool which is mainly relevant here.
Basically it is a combination of discrete event simulation and analytic
calculation. The manufacturing system simulation is based on 3DCreate and
3DRealize of Visual Components and generates operation data (e.g. piece count,
material used) and percentages of machine states (on, off, standby, under repair)
as output. With this, after simulation a MS Excel based analytical calculation of
environmental metrics (based on LCI data) is conducted, which includes impacts
through energy and material consumption as well as emissions. This separation of
simulation and evaluation procedure impedes the consideration of time dependent
59
Fig. 37 Simplified analysis flow chart of SIMTER approach (Heilala et al., 2008)
cumulative load profiles and evaluating energy related cause and effect during
runtime. The SIMTER approach does not require specific machine models (only
an abstract simple state representation is necessary) and employs conventional
discrete event manufacturing simulation tools so transferability as well as
modelling effort in terms of time and cost are mainly restricted by properties,
opportunities and limitations (e.g. library of standard machines, licence costs) of
these tools. Case studies in different industries (automotive, food industry)
underline the flexibility of the approach. However, each case considers only single
product production situations with fixed production layout and no detailed
production program (just one product continuously running through). The question
arises whether this approach is also applicable for multi-product production with
more complex and flexible production process chains with certain production
planning and control.
Rahimifard presents a more product oriented approach for modelling embodied
product energy (EPE, in kJ) during manufacturing. The consumption behaviour of
machines and processes over time is not modelled; the energy is calculated per
product based on equations in combination with metered data. Hence, a load
profile of the whole factory system is not available. Rahimifard categorises energy
consumption in manufacturing into two groups which sum up to the embodied
energy of a product (Figure 38). Direct energy (DE) is required to manufacture the
product in a specific process and can be further on, divided in theoretical energy
(TE, physically necessary energy for actual value creation) and auxiliary energy
(AE, supporting activities for the individual machine/process e.g. coolant pump).
With available equations from production engineering TE can be calculated based
on the characteristics of the product. AE needs to be determined through
measurements. Additionally Rahimifard defines the Indirect Energy (IE) that is
necessary to maintain the production environment (lighting, heating or ventilation)
whereas different zones can be distinguished. The allocation of IE to the embodied
product energy is based on the total IE consumption per hour of the specific zone
60
4 State of Research
Table 6 Evaluation of approach developed by Rahimifard
Main
Author(s)
Institutions
References
Software
tools
Shahin Rahimifard
The Centre for Sustainable Manufacturing And
Reuse/Recycling Technologies (SMART),
Loughborough University, United Kingdom
(Rahimifard, 2009; Rahimifard et al., 2010)
Arena from Rockwell Automations
Completeness
Dynamics
TBS
approach speaks of energy (with the unit kilo joule)

in general, specific forms of energy are not mentioned
or distinguished
consumption behaviour of machines is not modelled;
energy is calculated per product based on equations in
combination with metered data
Energy consumption of TBS only considered in
context of providing production conditions, certain
interdependencies and/or the internal supply of
different forms of energy are not addressed
Fields of action
Technological
fields of action are being revealed in general, feedback

to product possible
Organisational
PPC measures slightly addressed (e.g. batch sizes)
Optimisation
not considered
Evaluation
Economic
no specific economic criteria considered
Ecological
no specific ecological criteria considered
Technical
Decision
Support
Uncertainty
embodied product energy (EPE) in general as key

criterion
distinction between EPE portions as base for efficiency
ratios and decisions
not considered
61
Table 6 (continued)
Implementation
Transferability
Effort
Visualisation
Application
Cycle
basically flexible but specific equations and/or metered

data necessary, application to more complex
production scenarios questionable, rather expert
approach
Expensive simulation software as well as significant
time for modelling (e.g. equations for TE) necessary if
library not available
material flows and relevant results basically available
in runtime, within methodological restrictions manual
generation of meaningful charts possible
not provided
divided by the total number of products processed in one hour. Taking this into
consideration the embodied energy over the whole process chains (several
processes in different zones) for products can be calculated. Applicability directly
depends on the availability of appropriate equations and metered data. To facilitate
applicability while including the complexity of the problem, an energy simulation
model was developed (based on simulation tool Arena from Rockwell
Automations) which automatically calculates the energy values. However,
dynamics of energy and resource consumption and also material flow are rarely
considered in the presented cases; as also illustrated in (Rahimifard et al., 2010)
the EPE could even be calculated statically. Certainly, there is potential when
considering more complex cases and e.g. waiting times play a more important
role. Whereas supporting decisions in operations management is possible, the
simulation model shall explicitly help to derive improved design solutions in
terms of design for energy minimisation.
Fig. 38 The Embodied Product Energy framework for modelling energy flows during
manufacture (Rahimifard et al., 2010)
62
4 State of Research
Already in 2005 Solding et al. firstly presented their approach, which focuses on
increasing the energy efficiency of Swedish foundries through applying energy
aspects on discrete event simulation. The authors conducted several case studies of
different foundries while individually extending conventional simulators like
AutoMod or ED to depict the specific characteristics for each case (e.g. objective of
case studies, scope of simulation study, manufacturing system structure, PPC
aspects). In terms of energy carriers electricity is considered exclusively; state based
consumption behaviour of machines is realised through additional programming
within the base simulation tools. The additional overhead energy consumption of
supporting processes is addressed (lighting, ventilation, space heating) and modelled
as continuous consumer or depending on production machine working periods. Later
publications introduced a tool for structured data gathering and transfer to the
simulation environment. Improvement measures focus on production planning and
control. Thereby, besides reducing the energy consumption, the consideration of
energy peaks is of major interest while it may have significant cost impact
especially in energy intensive companies like metal foundries. The evaluation of
peaks is possible as the simulation is able to generate cumulative electricity load
curves (with appropriate time resolution) for the considered manufacturing system
based on single machine consumption behaviour.
Table 7 Evaluation of approach developed by Solding et al.
Main
Author(s)
Institutions
References
Software
tools
Petter Solding, (Damir Petku, Patrik Thollander)

Department of Production, Energy and Environment
Swedish Foundry Association, Jnkping, Sweden
Production and Energy Technology, Swerea
SWECAST AB, Jnkping, Sweden
Vestas Guldsmedshyttan AB, Guldsmedshyttan,
Sweden
(Solding and Petku, 2005; Solding and Thollander,
2006; Solding et al., 2009)
AutoMod from Applied Materials, ED from Incontrol
Enterprise Dynamics
Completeness
Dynamics
TBS
just electricity consumption is considered in

simulation
state based electricity consumption with cumulative
load profile
consumption of different supporting processes may be
included but without detailed consideration of energy
related interdependencies
63
Table 7 (continued)
Fields of action
Technological
manufacturing system improvement partly relevant in

selected case studies
Organisational
strong focus of coordination of processes through PPC
Optimisation
not really addressed, mentioned in outlook of most

recent paper
Evaluation
Economic
diverse issues of electricity pricing are addressed and

seemingly included (not in detail shown in case
studies, maybe due to confidentiality issues)
Ecological
no ecological evaluation
Technical
Decision
Support
Uncertainty
productivity (number of parts) as indicator for

technical performance
technical and energy related issues are discussed, no
actual decision support
not addressed
Implementation
Transferability
Effort
Visualisation
Application
Cycle
quite specific solution for each case (individually

built), high expertise necessary
Expensive simulation software as well as significant
time and expertise for modelling (specifics of each
case need to be addressed) necessary
material flow (apparently) visualised in runtime,
detailed results (including cumulative load profile)
just available after simulation run, no processing for
systematic decision support observable
software aided data collection was introduced which
also supports to ensure the comprehensiveness of the
energy study
Weinert et al. developed a concept for energy-aware production planning and

control (Chiotellis et al., 2009) based on so called energy blocks. They argue that
the specific load profiles of production machines can be separated into distinctive
classes (energy blocks) based on operational states (Figure 39).
64
4 State of Research
Fig. 39 Planning methodology based on energy blocks and related interface to simulation
software (Chiotellis et al., 2009)
The energy consumption of each block (exclusively focusing on electricity) can

be described through mathematical equations. Therewith energy consumption data
is made processable and can be used to model specific production tasks (process
chains) through application and combination of the relevant energy blocks.
Measured and processed energy blocks are stored in a specific database. In a
further step this data was combined with discrete event simulation. As shown in
Figure 39 conventional simulation tools like 3DCreate or emPlant (Plant
simulation) can be used to simulate manufacturing systems and deliver relevant
data like operation states or process times. Within a calculation module this data is
merged with involved energy blocks hence, the calculation of the energy
consumption is enabled.
In 2007, in his dissertation Junge presented a unique approach for considering
energy consumption in manufacturing systems, strongly focusing on the specific
case of injection moulding. The considered manufacturing system is modelled
with Simflex 3D, a material flow simulation software that is an own development
of the involved research institution. State based electricity consumption is directly
handled within the simulation environment. However, in injection moulding
additional, TBS related, energy consumption for air processing (possible
emissions of harmful substances) and heating is very relevant. To tackle this issue
Junge uses TRNSYS, which is an established tool for simulation air/heat flows in
buildings normally applied by civil engineers. Junge couples both material flow
and building simulation through defined interfaces (Figure 40): time respectively
state depending heat and particle emissions of injection moulding machines
simulated in Simflex are being transferred to TRNSYS where the additional
demand of electricity of oil for heating and air conditioning is calculated
dynamically (also including influencing variables like the local climate).

Table 8 Evaluation of approach developed by Weinert et al.
Main
Author(s)
Institutions
References
Software
tools
Nils Weinert, Stylianos Chiotellis et al.

Department of Machine Tools and Factory
Management (IWF), Chair of Assembly Technology
and Factory Management, Technische Universitt
Berlin, Germany
(Weinert et al., 2009; Weinert, 2010; Chiotellis et al.,
2009; Fleschutz et al., 2010)
emPlant/3dCreate
Completeness
Dynamics
TBS
just electricity is considered as of now

state based modelling of electricity consumption,
however no cumulative load profiles
focus on process chain, no consideration of any
supporting processes
Fields of action
Technological
manufacturing system design is mainly addressed
Organisational
PPC issues partly addressed, but not main focus
Optimisation
not considered
Evaluation
Economic
cost calculation (including peak prices) is addressed
Ecological
not considered in original approach by Weinert et al.
Technical
throughput time is considered
Decision
Support
not considered in original approach by Weinert et al.
Uncertainty
not addressed
65
66
4 State of Research
Table 8 (continued)
Implementation
basically branch spanning just depending one data
availability, different studies possible, application to
complex production scenarios unclear
if modules in simulator available relatively fast
modelling possible, expensive simulators and expertise
necessary
material flow shown in runtime but no energy related
results, manual processing of results necessary
Transferability
Effort
Visualisation
Application
Cycle
not provided
optimisation of parameters
(OptiS)
parameter
temperature,
energy demand
production
control
temperature,
energy demand
material flow
simulation
machine
utilisation
order data
results
TCP/IP
TCP/IP
machine
media/energy
consumption
events
material flow model
temperature,
energy demand
TCP/IP
building
simulation
heat losses,
emissions
machine model
building model
Fig. 40 Conceptual framework of simulation approach based on (Junge, 2007)
Diverse scenarios for different production strategies but also for different
locations (with different climate) are considered. The derivation of promising
strategies is supported by a tool for support parameter optimisation (OptiS, own
development). For evaluation, Junge introduces a multi-criteria approach that
includes diverse logistical objectives as well as an extensive cost model going
beyond just energy cost calculation.
Hesselbach et al. introduced a more complex simulation approach which
explicitly considers the interdependencies between production factors and
technical building services. It was developed within the research project ENOPA
(Energy efficiency through optimised coordination of production and TBS). It
aims at an optimised dimensioning and efficient process control of technical

Table 9 Evaluation of approach developed by Junge
Main
Author(s)
Institutions
References
Software
tools
Mark Junge
Fachgebiet Umweltgerechte Produkte and Prozesse
(upp), Universitt Kassel, Germany
(Junge, 2007; Hesselbach and Junge, 2005)
Simflex 3D, TRNSYS, Optis
Completeness
Dynamics
TBS
almost all relevant energy flows considered, except

compressed air
all flows sufficiently dynamically considered,
cumulative profile available
heating/AC is considered in interaction with
Fields of action
Technological
not in focus, no technological changes on machines,

manufacturing system or TBS considered
Organisational
very strong focus on PPC
Optimisation
parameter optimisation studies were conducted

Evaluation
Economic
holistic cost calculation including several cost

components
Ecological
not considered
Technical
diverse logistic/technical key figures (e.g. utilisation,

throughput time, stocks) are considered
Decision
Support
Comparison/discussion but no actual decision support
Uncertainty
is addressed through proving the independence from

stochastic effects
67
68
4 State of Research
Table 9 (continued)
Implementation
Transferability
Effort
Visualisation
Application
Cycle
very specific solution which demands expert

knowledge, considering a new manufacturing system
would need a completely new modelling
two simulation tools involved, extensive time needed
for modelling and simulation
material flow simulation in 3D during runtime, manual
processing for meaningful visualisation necessary
not provided
building services with respect to dynamic influences from production and, vice
versa, an energy efficient production management which includes restrictions and
opportunities given by technical building services. A holistic simulation approach
is used, which integrates and dynamically couples different simulation tools for all
relevant layers of the problem through certain interfaces (Figure 41):
Technical Building
Services
Building
Coupling
Material Flow
HKSim
Production Management
Simflex/3D
TRNSYS
AnyLogic
Fig. 41 Conceptual framework of ENOPA coupled simulation approach (Hesselbach et al.,

2008b)
69
Technical Building Services: the simulation tool HKSim was invented by the
company Imtech Deutschland GmbH & Co KG. It allows simulating
dynamically the generation, circuitry and consumption of energy and media
within technical building services. Thereby HKSim supports the design and
control of these complex systems (Wischhusen et al., 2003).
Building (climate): the (plant) building itself is the place where technical
building services (e.g. air conditioning) and production (e.g. waste heat,
exhaust air) interact, which is specifically important when a defined
production environment is crucial. To simulate the thermal processes
TRNSYS as an established tool in this field is used.
Production Machines / Material Flow: the material flow within the manufacturing
system can be simulated with SIMFLEX/3D, which was developed by the
University of Kassel
Production Management: the software AnyLogic from XJTek as multiparadigm simulation platform is used to simulate the influence and
interdependencies of production management measures (e.g. production
program, production control, production strategies) within the process chain
from a more abstracted perspective. Promising scenarios are being transferred
to the more detailed coupled simulation.
The simulation tools themselves are not new; they are all established solutions
within their specific fields of application. The most important aspect is the shift
away from an isolated usage to an integrated approach by coupling them. This
allows a realistic and holistic modelling of the whole system of production and
technical building services. Under consideration of all relevant interdependencies
and dynamics of the system, different scenarios can be analysed and evaluated
with both economic and ecological criteria. As a result, this enables the derivation
of optimal solutions from a global perspective.
Table 10 Evaluation of EnoPA approach developed by Hesselbach et al.
Main
Author(s)
Institutions
References
Software
tools
Jens Hesselbach, Lars Martin, Christoph Herrmann,

Sebastian Thiede, Rdiger Detzer, Bruno Ldemann
(EnoPA consortium)
Fachgebiet Umweltgerechte Produkte und Prozesse,
Universitt Kassel, Germany
Product- and Life-Cycle-Management Research Group,
Institute of Machine Tools and Production Technology
(IWF), Technische Universitt Braunschweig
Imtech Deutschland GmbH
(Hesselbach et al., 2008a; Hesselbach et al., 2008b)
Simflex 3D, TRNSYS, HKSim, (AnyLogic)
70
4 State of Research
Table 10 (continued)

Completeness
Dynamics
TBS
almost all relevant energy flows considered, except

compressed air
all flows sufficiently dynamically considered,
heating, AC, cooling media as well as steam generation
is considered in interaction with manufacturing
system
Fields of action
Technological
manufacturing system and TBS consideration with

integrated perspective
Organisational
effects of PPC measures are considered
Optimisation
not focused
Evaluation
Economic
realistic cost calculation for different energy carriers
Ecological
conversion to CO2 with distinction of different

countries
Technical
production time/output as main technical variables
Decision
Support
comparison of different goals, discussion of integrative

evaluation
Uncertainty
not addressed
Implementation
Transferability
Effort
Visualisation
Application
Cycle
very specific solution which demands expert

knowledge, considering a new case would need a
completely new modelling
at least three simulation tools involved, extensive time
needed for modelling and simulation
material flow simulation in 3D during runtime, manual
processing for meaningful visualisation necessary
not provided
71
In 2009, Fraunhofer IPA introduced their Total Energy Efficiency Management

(TEEM) concept, which aims to provide support for energy data metering,
visualisation, monitoring, and systematic derivation of improvements. As one
aspect, discrete event simulation is used to evaluate and derive measures towards
energy efficiency. Available literature focus on one specific example to describe
procedure and potentials of TEEM, namely the process chain injection moulding,
painting, drying and assembly for plastics. This process chain was also modelled
in Plant Simulation from Siemens UGS with consideration of main input energy
carriers, electricity, compressed air and process heat. While being relevant in
context of energy consumption for painting and drying also air ventilation was
considered with specific models, which enable to evaluate improvements in this
subsystem. TEEM simulation provides an economic evaluation of improvement
measures including both energy costs (simplified calculation just based on
consumption) and necessary investment in order to calculate amortisation time.
Table 11 Evaluation of approach developed by Fraunhofer IPA
Main
Author(s)
Institutions
References
Software
tools
e.g. Markus Hornberger (TEEM-Total Energy

Efficiency Management)
Fraunhofer IPA, Stuttgart, Germany
(Hornberger, 2009, Hornberger, o. J.; Wahren, o. J.)
Plant Simulation 8.1
Completeness
Dynamics
TBS
relevant energy flows are considered but no details

regarding origin of compressed air and process heat
from energy perspective
single machine as well as cumulative load profiles for
all considered energy flows are obviously available
consideration of air ventilation for selected processes,
no further details regarding generation of compressed
air and process heat
Fields of action
Technological
focus on machine and TBS improvement,

manufacturing system rather fixed
Organisational
PPC is addressed but not in main focus
Optimisation
not addressed
72
4 State of Research
Evaluation
Economic
simplified conversion to energy costs with same price

rate for all energy carriers, investment/amortisation of
measures included
Ecological
not considered
Technical
production time/rate considered
Decision
Support
comparison of different target variables and calculation

of amortisation for but e.g. not consideration of
technical issues in this case
Uncertainty
not addressed
Implementation
Transferability
Effort
Visualisation
Application
Cycle
basically yes, but not proved so far and always quite

specific solution
Expensive licence, time and expertise for specific
model necessary
Very little information available but seemingly
material flow and possible result visualisation in
runtime, manual post-processing for meaningful charts
necessary
Embedded in TEEM method which shall support in
terms of data acquisition and systematic derivation of
improvements
Lfgren published first work on a LCA oriented discrete simulation approach

which concentrates on the complex manufacturing system of one specific Swedish
company (SKF). The idea is to dynamically generate profiles in terms of energy
consumption but also material losses by means of simulation in order to conduct
LCA studies as base for improving the system. Besides manufacturing system
improvement, an important motivation is to show how the environmental outcome
can be influenced by employee behaviour.
Johansson et al. were also members of the aforementioned SIMTER
consortium. While having extended the SIMTER approach in some important
issues the evolved approach will be considered here separately as well. Johansson
et al. focus on the ecological evaluation of manufacturing systems. The functional
principle modelling with conventional simulation software and manual
processing in MS Excel afterwards is quite similar to SIMTER whereas
AutoMod was used here. In further publications they presented different case
73
Table 12 Evaluation of approach developed by Lfgren
Main
Author(s)
Institutions
References
Software
tools
Birger Lfgren
Chalmers University of Technology, Sweden
SKF Sverige AB
(Lfgren, 2009)
?
Completeness
Dynamics
TBS
several important flows are considered, not necessarily

comprehensive
state based modelling for all flows, not clear whether
cumulative profiles are available
consumption of compressed air and fluids are
considered but simply converted to connected
electricity consumption, no interactions
Fields of action
Technological
machine and manufacturing system perspective
Organisational
PPC not in focus, but influence of employee behaviour

is addressed
Optimisation
not considered
Evaluation
Economic
no economic evaluation
Ecological
complete LCA planned, however, just consideration of

energy as of now
Technical
Output is mainly considered
Decision
Support
comparison of a calculation of efficiency ratios to

balance
Uncertainty
necessity was addressed but no studies conducted
74
4 State of Research
Implementation
basically as idea certainly transferable, however very
specific solution
while using relatively easy process models, there is
less effort for modelling
Transferability
Effort
Visualisation
Application
Cycle
no information available
not considered
Table 13 Evaluation of approach developed by Johansson et al.
Main
Author(s)
Institutions
References
Software
tools
Bjrn Johansson et al.

Product and Production Development, Chalmers
University of Technology, Gothenburg, Sweden
(Johansson et al., 2009b; Johansson et al., 2009a)
(juice)
AutoMod, MS Excel
Completeness
Dynamics
TBS
all external energy sources considered, no internal

flows
state based energy consumption of machines, no
cumulative load profiles
TBS not considered at all
Fields of action
Technological
mainly manufacturing system design is focused
Organisational
order sizes and sequencing as important means in case

study
Optimisation
optimisation study was conducted
75
Evaluation
Economic
not considered
Ecological
LCA based on LCI data planned; practically mainly

energy related CO2 emission considered
Technical
throughput times as well as output is considered
Decision
Support
comparison of target variables, efficiency ratios used
Uncertainty
addressed in case study, but not explained in detail
Implementation
Transferability
Effort
Visualisation
Application
Cycle
different branches and questions can be considered,

unclear whether appropriate for flexible multi-product
production, rather expert solution
if modules in simulator available, relatively fast
modelling possible, expensive simulators and
expertise necessary
material flow visualised in 3D, due to model approach
no energy related results during runtime, manual
post-processing of data necessary, no support
regarding meaningful visualisation of results
not provided
studies (food/automotive) and underlined the flexibility of the approach. In

comparison to the original work, the authors introduced PPC measures and
optimisation within one study and also slightly addressed the issue of uncertainty
with the possible range of results.
Dietmair and Verl consider the issue from a machine control oriented perspective
and present a Petri-Net based approach to predict the energy (electricity)
consumption of machine tools. In several publications they successfully proved that
based on detailed state models, individual machine energy consumption (with
supply units like coolant pumps) can be described with high accuracy. This enables
to optimise process respectively machine control, identify critical components or
realistically evaluate different machine alternatives. In a further step they argue that
the same principle can basically also be applied to model manufacturing system with
several connected machines (Figure 42) in order to support tactical and strategic
manufacturing system planning.
76
4 State of Research
Fig. 42 High accuracy modelling of aggregate systems referring to (Dietmair and Verl, 2009)
Table 14 Evaluation of approach developed by Dietmair and Verl
Main
Author(s)
Institutions
References
Software
tools
Anton Dietmair, Alexander Verl

Institute for Control Engineering of Machine Tools and
Manufacturing Units (ISW), Universitt Stuttgart,
Stuttgart, Germany
(Dietmair and Verl, 2009; Dietmair and Verl, 2010)
?
Completeness
Dynamics
TBS
as of now just electricity consumption is considered

very accurate state based modelling of electricity
consumption
no consideration of TBS
77
Fields of action
Technological
Organisational
Optimisation
strong focus on machine improvement, manufacturing

system improvement addressed but some important
questions remain unsolved as of now (e.g. material
flow handling, production planning and control
implementation)
organisational aspects like scheduling shortly
addressed
optimisation possibility on machine level addressed
Evaluation
Economic
Ecological
Technical
Issues like production time slightly addressed
Decision
Support
no discussion of different target objectives
Uncertainty
not addressed
Implementation
Transferability
Effort
Visualisation
Application
Cycle
basically flexible and scalable approach which can be

used to tackle diverse questions, not clear whether
applicable to different manufacturing systems
Detailed modelling demands time and expertise,
relatively fast simulation time assumed due to reduced
approach
No material flow visualisation in runtime, very
detailed diagrams for analysis can be derived
not provided
78
4 State of Research
Material flow networks are a special form of Petri-Nets and can be used to
model material and energy flows in multi-stage manufacturing systems, including
its environmental aspects (Wohlgemuth et al., 2006) and are an important base
for environmental management information systems (EMIS). Thereby transitions
(material or energy transformations calculated based on equations/sub-models,
shown as circle in diagram notation), places (inventories, squares) and flows
(arrows) are being distinguished (see Figure 43). Material flow networks serve as
accounting systems for material and energy flows (Wohlgemuth et al., 2006) and
involve a set of rules like the principle of double-entry-bookkeeping of flows.
However, cumulative values for defined intervals are typically used in transitions.
Material flow networks do not focus on describing the time dependent behaviour
and interactions of production processes and process chains in detail. Against this
background Wohlgemuth et al. aim at combining material flow networks with
discrete event simulation. Two different concepts were pursued so far: on the one
hand the transformation processes within one transition can be calculated based on
an inventory simulation which is being executed when the transition occurs. The
results of the simulation are directly transferred to the material flow network
software (e.g. Umberto by IFU Hamburg GmbH) and serves as output of the
considered transition (see Figure 43). While being able to depict dynamics within
one transition this concept does not allow the inclusion of interactions between
different transitions of the system. As next evolutionary step Wohlgemuth et al.
developed an approach, which combines discrete event material flow simulators
like em-plant or Milan with material flow network software like Umberto. The
production process chain is completely modelled within the simulation
environment and experiments can be conducted. Per command the whole
simulation model structure, as well as the results of the conducted experiment, are
automatically transferred into Umberto. Based on the specific sub-models of the
simulation all transitions, places, flows and result values can directly be used in
order to e.g. conduct detailed economic or environmental evaluation within the
material flow network software environment. A separate modelling of simulation
model and material flow network is not necessary.
In 2007, the German based company Siemens AG overtook Tecnomatics
Technologies Ltd./UGS Corporation as the developers of eM-Plant (formerly
Simple++), which is a well-known and quite popular material flow simulator. In
2005 it was renamed to Tecnomatix Plant Simulation already. Plant Simulation
can be clearly classified to discrete event simulation and is a typical material flow
simulation software which supports the design and control manufacturing systems
while e.g. revealing bottlenecks through simulation. While strongly focusing on
the material flow itself supporting processes are typically not considered in detail.
Continuous flows (like energy flows in interaction with technical building
services) can hardly be depicted. Energy consumption of technical equipment is
not an issue as of now. Being one of the standard tools for material flow
simulation, some of the presented approaches use Plant Simulation as base for
more detailed considerations of energy consumption. However, Siemens
themselves strive to integrate energy consumption in more detail directly into their
tool. Unfortunately very scarce information is available publicly as of now. On the
79
Fig. 43 Linking a Discrete Event Inventory Simulation to a Material Network (Wohlgemuth

et al., 2006)
Table 15 Evaluation of approach developed by Wohlgemuth et al.
Main
Author(s)
Institutions
References
Software tools
Volker Wohlgemuth
FHTW Berlin, Department of Technics II, Industrial
Environmental Informatics, Berlin, Germany
(Wohlgemuth, 2005; Wohlgemuth et al., 2006)
MILAN/emPlant, Umberto
Completeness
Dynamics
basically very comprehensive view on energy and

resource flows possible through energy and material
flow network logic
due to the inner logic all consumption pattern are
modelled per piece and not state-based
80
4 State of Research
TBS
not clearly shown but consumption of supporting

processes can probably be considered, time-based
interactions not
Fields of action
Technological
Organisational
Optimisation
while rather approach for accounting and evaluation

actual improvements not in focus of consideration
while rather approach for accounting and evaluation
actual improvements not in focus of consideration
not considered
Evaluation
Economic
Ecological
Technical
Decision
Support
Uncertainty
material flow network software allows detailed flowbased cost analysis, realistic energy cost models not
included
material flow network software allows detailed flowbased analysis of environmental impact
not in focus of approach, certainly by-product of
simulation studies
not explicitly offered
not considered, in material flow software just one
sample can be considered
Implementation
Transferability
Effort
Visualisation
Application
Cycle
very specific simulation sub-models and expert

knowledge are necessary for considered case
high efforts and expertise for specific modelling, two
tools necessary
material flow shown in simulator, meaningful
visualisation of results at the end of simulation
not provided

Table 16 Evaluation of approach developed by Siemens
Main
Author(s)
Institutions
References
Software
tools
Siemens PLM Software/UGS

Corporation/Tecnomatics
Technologies
Siemens AG
(Siemens AG, 2010)
Plant Simulation
Completeness
Dynamics
TBS
just electricity is considered

state based consideration of electricity consumption
(idle, load, start-up)
not considered
Fields of action
Technological
improvement of manufacturing system considered
Organisational
not considered
Optimisation
certainly possible but not considered as of now
Evaluation
Economic
Ecological
Technical
throughput as main criterion
Decision
Support
short discussion of throuphput in relation to energy

consumption
Uncertainty
not considered
81
82
4 State of Research
Implementation
Transferability
Effort
Visualisation
Application
Cycle
basically transferable to all Plant Simulation

application cases with certain limitations
(e.g. flexible manufacturing systems)
no second software and model necessary,
one-stop-solution
material flow in runtime, automatic processing of
results
not provided
Siemens PLM webpage only a very simple case of a conveyor is shown where
production performance and connected energy consumption are considered
simultaneously. However it illustrates the technical feasibility of the direct
integration as well as the possible direction of development.
4.3 Discussion and Comparison

In the following chapter identified research approaches will be compared and
research focuses as well as possible gaps are being discussed. This allows the later
derivation of necessary research work in the field.
As base for further discussion Table 17 shows the summarised evaluation of all
approaches in terms of the criteria. Although all approaches aim at a similar
general objective the simulation of the energy consumption of manufacturing
system their individual specifications are evidently very different.
In the next step the evaluation was transformed into quantitative values (each
quarter counts for a value of 0.25 summing up to a maximum of 1.0), which
allows a more detailed comparison of the fulfilment of criteria. Based on this
transformation Figure 44 shows the average value (overall 12 considered
approaches) for each criterion as well as overall criteria (dashed vertical line).
Additionally the range is depicted as average of the six highest and lowest values
in each category. Table 17 also contains the average value of all columns
(approach based analysis). Based on this analysis several findings can be stated:
Having in mind that 4.0 is the maximum and ideal value of each criterion it
becomes clear that there is generally significant room for improvement in
all areas towards the vision of a comprehensive integration of energy and
resource flows into simulation based planning procedures.
83
Some approaches fulfil certain criteria very well however they involve
significant drawbacks in other areas. There is no approach with balanced
and high fulfilment of all criteria.
Criteria completeness (of energy and resource flows) and dynamics are
fulfilled higher than the average (still at relatively low level though).
Technical variables are usually considered for evaluation, often in
combination with other economic and/or ecological variables. In addition the
criterion visualisation is fulfilled above average.
On average some specifically critical issues can be identified: the lacking
consideration of uncertainty and optimisation studies, the lack of an application
cycle which supports implementation as well as the significant effort that is
necessary to setup an energy oriented simulation of manufacturing systems.
Johannsson
Dietmair
Wohlgemuth
Plant Simulation
Lfgren
TEEM
EnoPA
Junge
Weinert et al.
Solding et al.
Rahimifard
Heilala et al.
Table 17 Comparison of evaluation results
0.35
0.20
0.28
0.22
Energy and
Resource Flows
Completeness
Dynamics
TBS
Fields of action
Technological
Organisational
Optimisation
Evaluation
Economic
Ecological
Technical
Decision Support
Uncertainty
Implementation
Transferability
Effort
Visualisation
Application Cycle
Average
0.25
0.18
0.27
0.18
0.52
0.47
0.38
0.32
84
4 State of Research
Fig. 44 State of research - degree of fulfilment regarding identified criteria towards energy
oriented simulation
Based on the detailed analysis of the research approaches three different

general paradigms for simulating energy flows in manufacturing systems can be
identified (Figure 45):
In paradigm A the manufacturing system is simulated through (typically

commercial) a discrete event based material flow simulator. While these tools
originally neglect necessary possibilities to consider energy flows, relevant
information (e.g. operation states of machines) is transferred to a static
evaluation application through a defined interface after the simulation run is
finished. Through combination with energy related information (e.g.
consumption in different states) an evaluation with respect to energy
consumption can be conducted.
Paradigm B depicts the coupling of different simulation approaches (plus
optional evaluation) in order to respect the dynamic nature of energy flows as
well as the complex interactions taking place in different sub-systems of a
factory. Examples are the combination of material flow simulation with
building (heat flow), simulation (Junge) or the EnoPA approach which
additionally integrates technical building services simulation and a separate
evaluation module.
85
Paradigm C describes the direct extension/integration of material flow

simulation with energy related flow information and evaluation schemes
within one simulation environment (e.g. Siemens Plant Simulation).
Paradigms for simulating energy flows in manufacturing systems

A
coupling of DES and external

evaluation layer
(dynamic) coupling of DES and

further simulation approaches plus
internal or external evaluation layer
DES
DES
Evaluation
tool
additional
simulation
(e.g. TBS)
DES and evaluation within one

application
DES
plus evaluation
Evaluation
tool
Heilala et al. 2007, Weinert et al.
2009, Johannsson 2009, Wohlgemuth
2005
EnoPA 2008, Junge 2007
Rahimifard 2009, Solding et al. 2005,

Fraunhofer 2009, Lfgren 2009,
Dietmair 2010, Siemens 2010
Fig. 45 Identified paradigms for simulating energy flows in manufacturing systems based
on discrete event simulation (DES)
In the next step the analysis of all research approaches allows to derive
characteristics of each of those paradigms. Figure 46 shows the results of this
consideration. Due to its specific inherent logic each paradigm shows characteristic
profiles in terms of the considered criteria set.
On average, simulation approaches classified as paradigm A (interface between
material flow simulation and ex-post evaluation) offer relatively good coverage of
manifold energy and resource flows, the possibility of extensive and distinctive
evaluation systems as well as the low additional effort for modelling/simulation
and good transferability. The main reason is the strict separation of simulation and
evaluation which allows an uncomplicated and extensive design of the evaluation
environment, which is also independent from specific processes and simulation
models. As a disadvantage, through this isolated consideration certain energy
oriented dynamics and interdependencies cannot or hardly be depicted (e.g. often
no cumulative load profiles on system layer, missing energy flow induced events
and interactions). This also reduces the degree of freedom in terms of considering
fields of action towards better energy and resource efficiency. Additionally,
simulation induced issues like the consideration of uncertainty or optimisation are
more difficult to be addressed.
The characteristic profile of paradigm B (connection of different sub-system
simulation models) as shown in Figure 46, reflects the dilemma of those research
approaches. Energy and resource flows can be considered on a very detailed level
including dynamics and interactions between different sub-systems. This allows
86
4 State of Research
Application Cycle
Completeness
1
Dynamics
0,20
0,40
0,8
Visualisation
0,6
0,80
1,00
Fields of action
Evaluation
0,4
Effort
0,60
TBS
Technological
Implementation
0,2
0
Transferability
Organisational
Uncertainty
Optimization
Optimisation
Decision Support
Technical
Economic
Ecological
Fig. 46 Criteria fulfilment of energy flow simulation paradigms
covering all fields of actions on a realistic level and complex evaluation schemes
including the possibility to consider uncertainty and optimisation. On the other
hand the complexity is also the challenge in those cases: different specific
simulation models in different expert tools are needed and have to be connected
with appropriate dynamic interfaces. This reduces extremely the transferability
and increases the effort for both modelling and simulation (e.g. simulation run
times, certain computer performance necessary). While being advantageous in
theory, a practical application on a broader base does not seem possible yet.
In contrast to paradigm A, paradigm C (direct integration into discrete event
material flow simulation) can better depict the dynamics of energy consumption
on a system layer. This also enables a quantitatively and qualitatively better
consideration of improvement measures at least on the process/machine and
manufacturing system level. Additionally this one-stop-solution does not need
any other tools and allows seamless transfer to other simulation models. However,
evidently the user is depending on available functions and restricted by possible
limitations of the utilised simulation tool or the possibility to add/change features
is necessary. The integration of energy oriented aspects was not in focus of
material simulation developers so far. Additionally, the necessary logic to
integrate dynamic energy consumption, connected processes like those taking
place in technical building services and also appropriate evaluation methods are
also not easy to combine with the strong discrete event and material flow oriented
perspective of available tools. Each addition to the simulation means certain effort
for software implementation. As a result, the considered approaches with this
background offer less comprehensive energy flow considerations and evaluation
systems as well as no, or less detailed modelling of technical building services.
4.4 Derivation of Research Demand

The research demand can be directly derived from the discussions above. The
vision and ultimate objective is to be able to realistically consider all relevant
energy and resource flows in the planning and operation of manufacturing systems
simultaneous with time, quality and cost related variables. As mentioned, this
necessarily involves simulation as well as diverse additional criteria that have to
4.4 Derivation of Research Demand
87
be fulfilled. The analysis of the current state of research clearly reveals that
additional work is needed on many aspects:
The diversity of relevant energy flows in manufacturing systems is not
reflected so far. For obvious reasons (e.g. relevance in industry, easy to measure,
direct cost impact) external energy carriers and specifically electricity is mainly
focused up to now. However, other important energy carriers need to be integrated
in future research. This specifically involves internal energy carriers like
compressed air or process heat.
While the state based energy consumption behaviour of single production
machines is often addressed sufficiently already, their interactions and
cumulative effects on a system level need to be considered in more detail as
this is the main base for e.g. realistic energy cost calculation as well as planning
and control of technical building services.
Technical building services are rarely and qualitatively not sufficiently
considered in current approaches. One main reason is that this involves a totally
different discipline compared to production engineering. As a result, in production
planning there is a lacking awareness of their relevance as well as lacking
knowledge about relevant connected energy flows. Additionally, productionplanning tools like material flow simulation, currently does not focus, or are even
not able to depict those important energy consumers.
Regarding fields of action to improve energy efficiency, most approaches
focus on selected measures and do not provide an integrative view of the
whole factory system although there may be conflicts of goals or rebound effects
as well as further potentials for improvement. The possibility of actual
optimisation of critical parameters in terms of energy efficiency is not considered
at all in most cases.
Besides technical issues, existing energy oriented simulation approaches mostly
offer either an economic or ecological evaluation but rarely both dimensions
simultaneously. With regard to the demand of increasing sustainability in
manufacturing this should be considered in more detail in future research. This
also includes more extensive discussions and integrative evaluation systems to
balance possible conflicts of goals between different dimensions. Additionally,
evaluation methods themselves should be more detailed. Specifically the
economic evaluation mostly lacks detailed energy cost models which are based
on energy supply contracts (including e.g. peak surcharges). The issue of
uncertainty is very important in simulation studies due to stochastic effects that
might significantly influence results. This has to be addressed in future research.
At the moment energy oriented simulation is still rather a research topic and
less used in industrial practice. Developed solutions are very specific and just
transferable with significant effort in terms of time, software costs and necessary
expertise. Additionally, the general challenge for material flow simulation is the
applicability for flexible manufacturing systems, which is also true here.
Altogether there is a strong need for flexible and easy to use solutions in order
to enable broad applicability especially in SME. Another important weakness
until now is the total lack of an application cycle that guides the user through the
whole process of setting up an appropriate simulation model (e.g. selection of
88
4 State of Research
system boundaries, validation) and its parameterisation with measured

(consumption) data. Finally, active decision support is necessary, which needs a
meaningful visualisation of energy and resource flows (above standard material
flow visualisation) and connected key figures (e.g. distribution of consumption to
different machines and operating states).
All these aspects underline the main challenge which has to be met in context
of an energy oriented simulation approach: complex tasks, simple tools meaning
the necessary consideration of dynamics and complexity of the problem while
simultaneously facilitating broad applicability in industry through appropriate
design of the simulation environment. In this context the discussion of different
implementation paradigms shows possible solutions with their specific strengths
and weaknesses. All these considerations will be of main interest when deriving
an own energy oriented simulation approach in the following chapters.
Chapter 5
Concept Development
The previous chapters pointed out the necessity and potentials of energy flow
oriented manufacturing simulation and showed that there is significant research
demand. Against this background the proposed solution approach will be
presented in the following chapter. This includes the conceptual development and
the detailed description of involved elements.
5.1 Synthesis of Requirements into Concept Specifications

For the development of the proposed solution approach a deductive procedure was
chosen. This means that comprehensive general thoughts are brought together in
order to form a solution, which is able to address specific problems towards
energy efficiency in manufacturing companies. Figure 47 uses the factory life
cycle (according to Schenk, 2004) to depict intended fields of application for the
concept as a whole. On the one hand, it aims at supporting the development of
new facilities (greenfield) through testing of design and control alternatives of the
factory system. On the other hand, the concept can also contribute during
operation of a factory (brownfield). It allows deriving and evaluating control
strategies and also technical and organisational measures, which adapt running
production towards more energy efficiency.
Planning
Installation
Ramp Up
Operation
Dismantling/
re-utilisation
Factory life cycle
Simulation
Simulation based application cycle towards energy efficiency
Fig. 47 Classification of proposed concept in factory life cycle according to (Schenk, 2004)

springerlink.com
90
5 Concept Development
Based on the necessary scientific-technological theoretical background and

industry driven obstacles, diverse requirements for the solution approach were
identified. Herewith, energy oriented manufacturing system simulation was
identified as a promising approach for improvement. In order to assess the state of
research in this field, evaluation criteria were derived through further specification
of the requirements (also incorporating additional simulation specific
requirements). While further research demand was identified, all these criteria also
frame the necessary background and characteristics for the proposed solution
approach. On this base the objective is to develop an energy flow oriented
manufacturing system simulation approach which
is not related or restricted to a specific case but generic in nature and

applicable to manifold production situations in the sense of a generic
simulation environment.
explicitly pursues a holistic perspective including all relevant energy and
resource flows as well as their interdependencies.
is also applicable for small and medium sized enterprises typically facing
obstacles towards energy efficiency measures and usage of simulation.
is embedded in a guided methodology for goal-oriented identification and
realistic as well as multi-dimensional evaluation of improvement measures
in all relevant fields of actions.
With these objectives in mind, Figure 48 shows the mapping of the criteria with
actual means as to how they shall be addressed. It can be distinguished between
rather general specifications and more specific features envisioned to be
integrated. However, all aspects are important and altogether determine the
embodiment of the concept as a whole. The specific characteristics are of major
importance for understanding the idea of the approach and will be briefly
explained in the following. More details to specific aspects can be found in later
detailed description of functional principle and modules.
Establishment of flexible simulation environment: It is crucial to

understand that in contrast to some approaches presented in the state of
research a generically applicable concept is proposed. It is not a single
simulation study of a specific company or addressing specific industry sectors
or types of machines. The concept aims at providing a simulation
environment, which is highly flexible and able to consider manifold
production situations. As an extension to available approaches this also
includes flexible production structures with no rigid machine linkage and
changeable flow of products (flexible process chains with free planning of
capacity allocation) throughout the factory. While being reality in many
companies (often in SME) these structures are usually hard to model and
simulate with conventional tools.
Evaluation
Fields of action
E&R Flows
Criteria
91
Characteristics of proposed solution
Completeness
Development of generic, flexible simulation environment which can be used for

manifold production situations including flexible manufacturing systems.
Dynamics
A one-stop solution which is able to consider all sub-systems of a factory with

sufficient accuracy in one approach while still being easy to use.
Interaction with TBS
All forms of energy can be considered. This includes external acquired energy
(e.g. electricity, oil, gas) and internal energy flows (e.g. steam, compr. air).
A hybrid simulation approach is used which combines event driven logic of e.g.
machines with continuous energy flows.
Technological
Organisational
Optimisation
Hierarchical and modular structure which delivers generic, verified modules to be

combined and parameterised for the specific case.
Ensuring verification and validation as crucial tasks through generally verified
modules and validation for specific cases.
Scale and scope (Scalability) of specific simulation study is highly flexible and can
range from e.g. few machines to factories or even whole supply chains.
Economic
Additional models can be easily added to the existing structure through modular
structure and clearly defined interfaces (Extendibility).
Ecological
For all machines specific state based consumption and emission patterns can
be considered (fixed values per state or equations).
Technical
For easy access to simulation capabilities modeling is basically also possible

through interface with MS Excel with no necessary simulation expertise.
Decision Support
Various other software solutions for e.g. energy flow visualisation or detailed
process simulation can easily be connected (Connectivity).
Uncertainty
Realistic energy cost models based on actual contract data can be used which
includes varying prices over the day, peak surcharges and fees.
Implementation
Energy related ecological assessment is provided through conversion of energy

consumption (electricity, gas, oil) to related GHG emissions.
Transferability
Appropriate key figures are provided which support decisions while helping to
solve conflicts of goals and the treatment of uncertainty in evaluation.
Effort
Goal oriented visualisation towards systematic improvement is continuously

given (e.g. breakdown of consumers, cost composition, value creating energy)
Visualisation
Besides testing diverse scenarios also actual optimisation experiments are possible
(with OptQuest algorithm)
Application Cycle
A comprehensive application cycle provides a guided holistic methodology and

ensures goal oriented modeling and simulation towards improvement.
Fig. 48 Mapping of criteria and specific characteristics of the proposed solution
One-Stop approach: The previous chapter revealed different paradigms, which

were already applied for energy oriented manufacturing simulation. As
described, a combination of detailed simulations of different relevant subsystems in a factory is a promising approach (paradigm B). This paradigm is
able to consider energy flows and interactions realistically and offers a broad
range of improvement opportunities. However, resulting complexity and effort
is a major disadvantage which impedes implementation. The proposed solution
aims at using the advantages while overcoming this drawback. It will be able to
simulate not only the manufacturing system itself but also other relevant subsystems in the factory with sufficient accuracy. However, this will all take place
seamlessly within one simulation environment, which significantly facilitates
usability, increases transparency of energy flows in the factory system and
opens new perspectives towards improvement. Thus, it can be described as a
synergetic integration of paradigms B and C (Figure 45).
92
Consideration of all relevant energy inputs: The proposed solution

approach explicitly does not exclude per se any energy flows. It can basically
consider all relevant external (acquired by the company from suppliers) and
internal energy flows (conversions from external forms of energy) and their
interactions (see Chapter 3.2). For reasons of meaningful limitation, the
proposed approach will focus on the most important forms of energy which
are electricity, gas and oil (external), compressed air and steam as well as (to
a certain extend) waste heat (internal).
Hybrid simulation: To be able to cope with the interactions between
production machines and TBS equipment through material and specifically
energy flows, a combined discrete-continuous or hybrid simulation approach
was applied (see chapter 3.4). It combines state based and event driven logic
of machines and material flow with continuous variables connected with
energy flows. The advantage lies in the combination: neither discrete event
nor continuous simulation alone would serve the purpose here.
Hierarchical and modular structure: The whole concept is based on the
combination of different pre-structured, verified modules which stand for
relevant sub-systems/functions and are structured according to the holistic and
hierarchical understanding of the factory system (see Chapter 2.1 and 3.2). These
modules can be combined and parameterised in order to model the structure and
behaviour of the specific case with sufficient accuracy. Thus, complexity can be
handled: specific models can be built relatively fast and maintained with
justifiable effort. Each single module itself as well as the interactions through
defined interfaces will be described in the following chapters.
Ensuring verification and validation: Verification and validation are crucial
tasks in order to guarantee reliable simulation results. In the modular
approach of the proposed solution, verification will be realised in general for
each single module (submodel testing, Rabe et al., 2008) and so verified
modules will be part of the concept. Validation can just be done in context of
the specific case. Thus, it will be part of the suggested application cycle.
Scalability: Scale and scope of the specific simulation study can be adapted
according to the actual purpose. Models can be built on a relatively detailed
level with consideration of each single manual or automatic process in
production. However, if the requirements concerning validity and study
objectives are met, it is also possible to leave out irrelevant steps or
summarise them into one process. Additionally, there is flexibility regarding
system boundaries: studies can either consider a selected process chain of few
processes (e.g. for a specific process), a factory as a whole or even complete
supply chains (e.g. each company is just depicted as one black-box with
inputs and outputs in the simulation).
Extendibility: The modular structure and clearly defined interfaces allows
the extension with any additional models. As mentioned, for reasons of
meaningful limitation some sub-systems will not be considered in detail here
(e.g. air conditioning). However, they can be easily embedded as additional
modules at any time. Similarly, other relevant energy or material flows could
also be integrated.
93
State based consumption and emission patterns: As described energy

related consumption and emission is not static but depending on the operation
mode of the machine (see Chapter 2.3.3). This is addressed through the
development of a generic machine state model, which can be parameterised
with certain consumption (and other) values or equations for each individual
state. As sum of single machine profiles, cumulative load profiles for the
whole considered system are also generated during runtime.
Easy access to simulation capabilities: Chapter 3.4 pointed out that the
application of simulation techniques in companies may demand for expertise
and significant effort in terms of time and connected costs. This often
impedes the usage and therewith the utilisation of opportunities towards
energy efficiency, specifically in SME. Thus, the proposed solution aims at
enabling broad utilisation of simulation capabilities without facing these
entry barriers. Therefore two modelling alternatives can be pursued by the
user: on one hand the manufacturing system can be modelled directly in the
simulation environment as usual. Through graphical modelling and predefined modules and interfaces this does also not require extensive
simulation experience. On the other hand, through appropriate interfaces the
configuration of structure, parameters and production management issues
can even be done with familiar software like MS Excel. In this case, apart
from one click for starting the simulation run, users do not have to get in
touch at all with the simulation software itself while still be able to tap the
advantages of the approach.
Connectivity: At different stages of the proposed solution the usage of
additional software tools might be promising for extended applications. This
is possible through the open architecture and interfaces. Examples are the
aforementioned embedment of office software for model setup, extended
evaluation and visualisation opportunities (e.g. energy flow visualisation) or
detail simulations of processes or machines (e.g. FEM process simulation).
Realistic energy cost models: Chapter 2.4.1 pointed out that specifically
electricity costs are based on individual supply contracts and contain diverse
cost components. In order to enable a realistic calculation of costs for the
specific case, detailed contract models can be explicitly configured.
Simplified energy related ecological assessment: The proposed approach
focuses on energy inputs and less on a comprehensive gathering of all
(auxiliary) material flows even though that would be possible. However, this
would be necessary for detailed environmental evaluation like LCA (Life
Cycle Assessment). In this book ecological evaluation will concentrate on
energy related assessment, meaning realistic conversion into the related green
house gas (GHG) emissions.
Appropriate key figures: In order to be able to give necessary decision
support and solve conflicts of goals, appropriate key figures will be provided.
This also includes appropriate statistical treatment in case of probabilistic
models (uncertainty). Detailed information will be given later in this chapter.
Visualisation towards systematic improvement: The proposed solution will
not only provide cumulative results for general evaluation. It shall also deliver
94
detailed breakdowns of consumption, environmental impact or cost

composition, which help to consciously focus on main drivers.
Integration of optimisation possibilities: Optimisation studies are not in the
main focus of this book. However, some possibilities will be considered in the
solution approach.
Provision of application cycle: The concept as a whole consists of two main
elements. Besides the simulation approach itself, this is a comprehensive
application cycle that shall enable implementation in companies. It provides a
guided methodology and ensures a systematic modelling and simulation
towards system improvement.
5.2 Abstraction of Conceptual Framework

As a first necessary step towards implementation of the concept, an abstraction of
the real system into a conceptual modelling framework has to take place. Against
the background, the aforementioned theoretical background, Figure 49 shows the
control loop of production management extended with the holistic understanding
of the factory system, which builds the theoretical backbone of the pursued
simulation approach. Based on these considerations five main constituting
elements here called modules - can be distinguished (Figure 49):
reference input variable(s)
PPC Module
planning and control

coordination
inf ormation
EV Module
production management
reference
(actuating variable)
actual state
(feedback)
local
climate
TBS Modules
(compressed air, steam)
cooling
heating
Input
Process Modules
Process Modules
Process Modules
Manufacturing System
raw and
auxiliary
material,
energy
need for defined

production conditions
(e.g. temperature, moisture, purity)
technical
building
services
(TBS)
Output
waste heat
exhaust air
allocation of
media
(e.g. compressed air,
steam, cooling water)
(e.g. steam,water)
backflow of media
production
machines
products,
waste/scrap
(waste) energy
Fig. 49 Contribution of Simulation Modules within Control Loop of Production Management
5.2 Abstraction of Conceptual Framework
95
Referring to the definitions given before, Process Modules (PM, module I in

Figure 51) describe the activities which transform inputs into outputs
(DIN 9000) taking place manually or on production machines. In practice,
certain activities [value creating and supporting activities like transformation,
combination, transport, control, measure or storage (Barbian, 2005)] are
executed simultaneously or successively directly with one or more entities
(parts) in order to strive towards final products. While being generic in nature,
through appropriate parameterisation a process module can imitate the
behaviour of any production process and connected machine in terms of
process times, failures or energy consumption. Detailed descriptions can be
found in Chapter 5.3.1.
TBS Modules (technical building services, (module III in Figure 51) supply
the necessary process energy in different forms (Hesselbach et al., 2008b) and
therewith enable the operation of process modules. However, while being
crucial for running the system they are not interacting with product related
entities themselves so their contribution to value creation is rather indirect.
As described before a broad variety of different types of TBS relevant for
energy flows in the factory can be distinguished (e.g. compressed air, steam,
cooling, space heat). All these types are based on different physical
coherences and work very differently. Therefore, each type of TBS subsystem will be depicted by separate module classes, e.g. there will be general
modules for compressed air and steam. However, each of these modules is
again generic in nature, meaning that through parameterisation the supply
system for the specific case can be considered. Detailed descriptions of
implemented TBS module can be found in Chapter 5.3.3 and 5.3.4.
The generation of products usually involves several process steps, which
form process chains (time based logic sequence towards final product) and
manufacturing systems (fixed set of available processes and machines used
for one or several process chains). Also the supply with different forms of
energy usually does not relate to single processes but systems with several
consumers. Thus, there is a need for an additional layer that allows the
combination of process modules and mutual interfaces with TBS modules.
This combinatory layer is realised through the Manufacturing or Main
System Module (MS, module V in Figure 51, detailed description in
Chapter 5.3.7).
Process and TBS modules coordinated through the manufacturing system module
constitute the real, physical factory system. As described before all these
elements strongly interact and altogether determine both production performance
and energy consumption of the system as a whole. Single consumption profiles of
production machines lead to cumulative load curves for the manufacturing system.
TBS-related energy demand of the production equipment (e.g. compressed air)
serves as input for appropriate partial TBS-models (e.g. for generation of
compressed air). Herewith additional energy consumption (e.g. electricity needed
to generate compressed air) of TBS is calculated together with direct energy
96
energy demand of production site

(base for economic and ecological evaluation)
load profiles
production
system
(time based energy

and media demand)
possible energy
and media supply
technical
building
services
Fig. 50 Simulation based interaction of manufacturing system and technical building services
consumption of production equipment, this leads to the total energy demand of the
production site (Figure 50). Additionally TBS models simulate the possible supply
with energy or media. Interacting with the manufacturing system a lack of e.g.
compressed air (air pressure to low) leads to failures of production machines.
Besides the physical embodiment of the factory system, the production
management is considered through different modules. Their allocation within the
control loop of production management is also shown in Figure 49.
The Evaluation and Visualisation (EV) Module (module IV in Figure

51) gathers necessary data in order to assess the state of the considered
system and processes this information in meaningful form as key figures
or diagrams (detailed description in Chapter 5.3.6).
Finally, the Production planning and control (PPC) Module (module
II in Figure 51) allows the determination of technical and organisational
framing conditions for the specific case. This includes the definition of
product process chains in general and also their flow in terms of timing
and quantity of production machine occupation (detailed description in
Chapter 5.3.5).
The diversification of production management into two separate modules was

made because two different logic functions can be distinguished. From a technical
point of view a separation facilitates modelling as well as verification and
validation. In practical application, the connecting element of those modules is
basically the user itself. Based on the evaluated results of simulation runs he will
develop and adopt production management alternatives, which can be embodied
through the PPC module.
As a combination of all thoughts above, Figure 51 shows the conceptual
framework of simulation approach as a whole with its five main modules. All
modules will be described in detail in the following sections. Afterwards, in order
to support the goal-oriented implementation of the simulation approach, a
comprehensive application cycle will be developed.
5.3 Description of Simulation Approach

II
PPC module
production machine
parameters
(Production Planning and Control)
(e.g. MTBF, cycle times)
product specific
process chains (e.g.
sequences)
state-related
consumption/
emission patterns
Process Module
(PM)
parametrisation of generic
process modules
realistic state-based depiction
of machine and energy
consumption/emission
behavior
flexible degree of accuracy
running
97
Manufacturing /Main System Module

(MS) with interlinked Process Modules
P1
P2
P3
P4
P5
P6
P7
P8
Pn
12000
80 00
ramp up
off
failure
10000
standby
main switch
on
specific energy
contract data,
basic environmental
impact data
PPC data: orders

schedule,
quantities
IV
production
performance
indicators
Evaluation and
Visualisation (EV)
integrated technical, economic
and ecological evaluation for
decision support
energy cost calculation
visualisation of energy flows
production
related energy
consumption
60 00
40 00
20 00
0
80%
65%
20%
time
90%
35%
10%
III
TBS parameters
(e.g. compressor data,
puffer tanks,
capacities)
load profiles
(time based
demand)
possible
energy and
media supply
TBS Modules
Compressed
Air
Steam
(Process Heat)
further TBS
modules (e.g.
space heat,
cooling water)
TBS related
energy flows
further relevant
technical TBS variables
TBS interaction
Fig. 51 Conceptual Framework of the proposed simulation approach

5.3.1 Implementation and General Functional Principle
According to the intended specifications and conceptual framework, the
simulation environment was implemented in a working solution, the energy
oriented manufacturing system simulation environment. Besides PM, EV, MS
and PPC modules, with compressed air and steam supply two types of TBS
modules were already developed, verified and applied to different cases. The
proposed simulation environment was embodied using the simulation software
tool AnyLogic 6.X from XJTek which offers all prerequisites for implementing
the characteristics of the solution approach. As one of few tools commercially
available, it supports multi-paradigm simulation, which is a basic requirement for
many energy flow related features. The development of the innovative simulation
solution also demands certain flexibility and freedom which cannot be provided
by mature, large scale simulation tools like Plant Simulation. This is also the
reason why those tools were just applied in context of energy simulation through
using interfaces with additional external programs. However, this would
contradict one of the basic ideas of the proposed solution approach.
98
Having in mind the conceptual respectively logic framework shown in Figure 51,
Figure 52 shows a general overview of the practical embodiment of the energy
oriented manufacturing system simulation environment and connected interactions
with the user. For usage a standard MS Windows based computer is necessary.
AnyLogic needs to be installed and serves as operating platform. However, due to
the nature of the developed approach with its predefined modules and also the
possible control via MS Excel interface, there is no deep AnyLogic expertise
necessary. Additionally, AnyLogic offers the opportunity to comfortably export
models as self-executable Java applets. This allows running the model with
standard web browsers (supporting Java) and therewith broadens the field of
possible users. However, once finished and exported to the applet, changing the
model structure and parameters is limited. An example for an applet based model is
given later in Chapter 6.4.
standard MS Windows
based computer
Interfaces
MS Excel
MS Excel
Input
Output
MS, PM, PPC, TBS modules
EV module
Logic layer
(built up with AnyLogic 6 and Java Code programming)
Energy oriented manufacturing system simulation environment

Fig. 52 Practical implementation and user interactions with developed energy oriented
manufacturing system simulation environment
The following description of the functional principle concentrates on the more

typical case of application with installed AnyLogic software. When starting a new
project basic modules PPC, EV and TBS are already given and their interfaces are
properly connected. For the modelling of the manufacturing system structure
generic process modules (PM) are used, which can depict any operations with
sufficient accuracy depending on their specific parameterisation (a detailed
overview regarding available parameters follows in the next chapter). Two
alternative ways can be followed at this point:
For the case of logically and/or technically linked processes respectively

process chains, the manufacturing system structure (MS module) can be
modelled directly within the simulation tool through drawing the process
modules in the main area and linking them. The parameterisation is as well
conducted within the tool itself.
99
A generic set of process modules can be used and configured through the PPC
module, which is based on MS Excel. Thus, modelling within the
simulation tool itself is not necessary.
Also specific production programs (sequences, starting times, lot sizes of orders)
can be defined within the PPC module. However, simple runs without complex
production programs (e.g. only continuous production of certain quantities) can be
configured and started independently from this module. When running simulation
all relevant variables are transferred to the evaluation and visualisation module.
Therewith, relevant key figures and meaningful diagrams are generated and
displayed continuously during runtime (and can be stored in MS Excel based
databases). Additionally, energy costs and environmental impact are computed
based on parameterised calculation schemes (e.g. electricity contract model).
Additionally, energy flow data is continuously transferred to E!Sankey which
allows a live visualisation in order to depict main energy flows.
In the following sections all five modules will be described. In order to support
the understanding of module functionality and interfaces with involved inputs and
outputs a standardised illustration will be used (Figure 53). The description of
each module will also be conducted in standardised manner. For the core PM and
TBS modules the necessary abstraction of the real world system to the functional
logic of the module will be explained first. Afterwards, the practical embodiment
with involved parameters and state variables will be described and finally a
verification study will be conducted for each of those generic modules. MS, PPC
and EV modules serve more superior purposes and are responsible for ensuring
the general functionalities of the simulation approach. Their necessity and relation
to the real system was already pointed out in Chapter 5.2. However, parameters
and state variables will also be explained for those modules. While being strongly
connected, verification can hardly be conducted independently. Thus, it is realised
for MS, PPC and EV module at once in Chapter 5.3.7.
Parameters
Input
Variables keyed in by
user in order to
parametrise module.
(interfaces to other modules)
Output
Automatic data/
information flow from
other modules.
module name
Automatic data/
information flow to other
modules.
Variables that describe

the current state of the
module.
State variables
(selection)
from/to Process Module
from/to MS Module
from/to PPC Module
from/to TBS Module
from/to EV Module
Symbol/colour code to depict interactions with other modules.
Fig. 53 Description of standardised illustration for modules
100
5.3.2 Process Module

Abstraction from Real World to Model
As described before manufacturing processes are a transformation from inputs to
wanted and unwanted outputs. For automated or semi-automated process
execution typically production machines are involved. For those machines
different distinctive operation states connected through different transition
conditions can be distinguished (Dietmair and Verl, 2008; Dahmus and Gutowski,
2004; Devoldere et al., 2007). This well-established perspective on operation
states and transitions builds the background for the embodiment of the process
module. Figure 54 shows a simplified illustration of the underlying logic of the
process module as state chart (in the actual model there are a couple of additional
states to ensure functionality etc, which will not be considered here as they are not
necessary for the general understanding).
process time
processing
10
10000
entity occurs
8
8000
standby/idle
defined end
ramp up time
failure
MTTF
process time
12000
12
power
[kW]
post-production
MTTF
post p. time
6
6000
ramp up
setup time
MTTR
setup
4000
4
2000
2
defined start
off
00
time [sec.]
Fig. 54 Underlying state chart logic of process module and connected modelling of
(e.g. energy) consumption of machines
Setup is the state when preparing for operation e.g. exchange of tools/dies.
Ramp up denotes the state after the machine is turned on thereby all components
are enabled and eventual heating phases etc might be necessary. In standby/idle
mode the machine is basically ready for production but certain requirements for
processing are not fulfilled (e.g. missing part or lacking supply with energy).
Operating/processing stands for the actual treatment of a part. While being an
abstracted consideration of manufacturing processes here, it is less relevant what
physically happens during the process. In the end a manufacturing process
somehow transforms one or more entities at once, which consumes time and
additional inputs like e.g. energy. This necessary abstraction allows a very broad
applicability to any manufacturing processes. The state post-production was
introduced because machines may include certain automatic procedures after
101
processing (e.g. cleaning). As displayed in Figure 54, the transition between

different states can be triggered by either time or occurrence of events:
time: pre-defined duration of states depending on machine and productspecific process characteristics.
events: stochastic failures, control signals from e.g. PPC module (e.g. shift
start), occurrence of parts in order to be processed.
As formerly mentioned, the failure state is of specific importance in context of the

time based behaviour of the process module. This state is determined by two
characteristic variables, which can be derived from company data: Mean Time To
Failure (MTTF) is an expression for the average lifetime and can serve as
parameter for describing the probability of a failure. Mean Time To Repair
(MTTR) determines the duration of a failure (Bertsche, 2004). For mathematically
describing the failure behaviour, distribution functions (e.g. normal, exponential)
are used. For practical purposes and its broad applicability the usage of the
Weibull function is suggested (Abernethy, 2006; Bertsche, 2004; Birolini, 2010).

(19)
The necessary input parameters b (shape parameter, describes the mode of failure)
and a (scale parameter, characteristic life time at which 63.2% of considered units
fail) can be derived as a result of statistical analysis of e.g. process times or failure
occurrence and transformation through mathematical/statistical methods like the
Median Rank Regression (MRR) or Maximum Likelihood Method (MLE). The
Weibull function is very useful due to its flexibility: all kinds of different
distributions can be described with this equation through varying the shape and
scale parameters (Figure 55) (Bertsche, 2004; Birolini, 2010). When using the
Weibull distribution the determined MTTF is transferred to the scale parameter for
a while using the gamma function :

(20)
with

(21)
or (for positive integer values)

(22)
As the equations show MTTF will be equal to a if b=1 ((2)=1), which depicts a
random failure behaviour. It is crucial to note that the usage of distribution
function automatically leads to a probabilistic nature of the simulation model and
any results are only one possible outcome of the simulation. Hence, sufficient
simulation runs and appropriate statistical treatment are necessary. Further
explanations in this context can be found later in Chapter 5.3.6.
density function f(t)
102
Iife time t
Fig. 55 Weibull function with different shape parameters b (Bertsche, 2004)
As shown in Figure 54 each state is connected with the consumption of energy

or/and auxiliary materials. Any kind of consumption can be allocated to different
states with either single value for each state (sufficient for many applications) or
based on equations. An equation based formulation might make sense if further
coherences shall be included. For instance, bearing in mind the thoughts from
Chapter 2.3.3 increasing load or machine speed usually lead to higher demand on
energy.
Practical Embodiment
Against the background of the conceptual thoughts of the process module, Figure
56 shows an illustration of its main elements and interfaces. In accordance with
the definitions in Chapter 2.1 a process involves inputs and outputs which are
being determined by the specific parameters. Certain variables can be used to
assess the state of the process.
As already mentioned in context of the general conceptual structure, process
modules are embedded within the main (MS) module. Thus, technically they do
not have direct interfaces with other modules. In terms of input, process modules
are triggered by the arrival of one or several entities (parts), which have to be
processed. This is determined through production planning and control as well as
the processes in the upstream process chain. Additionally the operation of process
modules depends on energy supply, e.g. a certain compressed air pressure or the
availability of necessary steam. Although provided by TBS modules, technically
this information is as well transferred to each process module via MS module.
103
Parameters
Input
general type of process

specifications in terms of
time, failure behaviour,
consumption, quality rate
startup/shutdown regimes
material flow from upstream

processes
energy supply
Output
Process Module
material flow to downstream

processes
single energy consumption
profiles
utilisation
produced volume, quality rate
current energy demand and
consumption
share on system electricity consumption
State variables
(selection)
from/to MS Module
from/to PPC Module
from/to TBS Module
from/to EV Module
Fig. 56 Constituting factors of Process Module
Parameters
A summary of relevant parameters is also shown in Figure 56. Due to their
importance Table 18 gives a detailed overview of all parameters including their
actual identifier within the simulation, a short description as well as the unit and
the range of values. In general three different types of parameters can be
distinguished for the process module:
As a very basic option the type of process needs to be defined. As mentioned

before, the simulation does not aim at a detailed representation of any
processes. However, two basic alternatives are distinguished, which can
describe the generic logic of most processes relevant in manufacturing. If the
transportation parameter is enabled, the part is constantly moving with a
certain speed and for a specific distance (together this determines the
duration) and leaves the process module afterwards. A typical application
would be e.g. a conveyor belt. Alternatively (transportation parameter off),
each single part (respectively a lot of parts) is delayed for a certain period of
time without movement where normally the value creating processing takes
place. In contrast to transportation activities, other parts cannot be processed
while the process is occupied. This is the case for many manufacturing
processes like e.g. metal machining (machine tools).
In the next step, individual technical specifications need to be inserted in
order to enable realistic behaviour. Thereby the described state based
perspective on processes and machines is relevant to keep in mind. The
duration and related energy consumption for each state can be keyed in with
fixed values or also based on equations/distribution functions. Failure
behaviour can be configured through applying the Weibull function.
Practically, the standard syntax (provided by AnyLogic) weibull(shape
104
parameter, scale parameter, minimum value) needs to be used here. For

instance, a normal distribution (around the value of e.g. 500 seconds) can be
realised with the term weibull(3.4, 500, 0). Besides state based characteristics
further general or more product oriented technical specifications can be
included. Examples are the necessary compressed air pressure for operation,
the quality rate or the material efficiency as ratio from weight of finished part
to input material per part.
Startup and shutdown procedures play an important role for machine
operation and energy related outcome. Thus, they can be defined for each
machine individually and different strategies can be integrated.
o The individual machine can start which means turning it on, not
necessarily processing - at a definable point of time during the
workday, usually at the beginning of the shift (e.g. 6am). The
alternative is to turn it on when the first order appears.
o The machine can stop operation at a specific point of time, e.g. at the
end of the shift.
o Even more, an automatic shutdown can be enabled. Therewith, the
machine shuts down automatically if idle for a definable period of time.
Table 18 Parameter list of process module
Parameter
Name
process_time
dynamic_processtime
set_up_time
ramp_up_time
post_production_time
start_machine_regime
start_machine_time
end_machine_time
lot_size
Description
Dimension
Labelling of process
module
Necessary time per product
if dynamic process time is
off.
Process time is not fixed
but individual per product.
Time for setup of machine.
Time for rampup of
machine.
Time necessary for post
production processes.
String variable,
free naming
Process can either start at

fixed hour of the day or
when first order occurs.
Denotes starting hour, e.g.
shift start (if
start_machine_regime=2)
Denotes machine stop
hour, e.g. shift end (if
start_machine_regime=2)
Number of parts being
processed at once.
in [sec.]
in [sec.]
in [sec.]
in [sec.]
in [sec.]
1: Machine start if
first parts occur
(starting with
setup)
2: Machine start at
specific hour
[hour of the day],
from 0-24
[hour of the day],
from 0-24
[number of parts]
105
Parameter
Description
Dimension
quality_rate
Average percentage of
good to total parts.
MTTF
Mean Time To Failure.
MTTR
Mean Time To Repair.
[from 0..1 meaning

0-..100%]
[in sec.] plus
distribution
function, free
range
[in sec.] plus
distribution
function, free
range
transportation
conveyor_length
conveyor_speed
accumulating
material_efficiency
nec_compair_pressure
power_consumption_state
(for setup, ramp_up, standby,
producing, post_production)
compair_consumption_state
steam_consumption_state
gas_consumption_state
(for standby, producing)
aux_consumption_state
auto_shutdown
auto_off_time
Denotes whether it is a
[1: yes, 0: no]
transportation process.
Length of conveyor (if
[in meter]
transportation on)
Speed of conveyor (if
[in meter/second]
transportation on)
Describes whether several
entities can use conveyor
[1: yes, 0: no]
simultaneously.
Ratio of product weight and [from 0..1 meaning
raw material input.
0-..100%]
Minimal compressed air
pressure necessary for
[in bar]
operation.
State based allocation of
electrical power demand.
[in kW]

compressed air demand.
[in m/h]

steam demand.
[in kg/h]

gas demand.
selected auxiliary material
demand.
Activation of auto-off
function for machine.
Time until machine turns
automatically off (if
enabled).
[in kW]
[any continuous
unit]
[1: yes, 0: no]
in [sec.]
106
State Variables
The state of each process module can be assessed continuously during simulation
run time through different variables. An overview of considered state variables
and the graphical depiction of a process module during simulation gives Figure
57. The current state is depicted with a colour scheme and with one glance the
operation mode of each machine can be recognised. Two variables are of specific
interest to evaluate the performance of each process. On the one hand, the
utilisation rate is used which gives a clear indication of the process contribution
towards value creation and the employment of provided capacity.

(23)
On the other hand, the current energy demand of each process is displayed. Even
more, instead of just displaying the energy consumption continuously which
would not deliver meaningful information when shown alone the share of the
single process consumption related to the total energy consumption of the system
is displayed. Because of its importance electricity consumption is specifically
addressed.

(24)
With this key figure hotspots regarding electricity consumption can be identified
at a glance. This supports the systematic improvement process towards energy
efficiency in manufacturing.
Name
Delivered good
parts
Number of Failures
Waste parts
Operation state and

utilisation rate
Current power
demand
Process time
Current compressed
air demand
Share on total electr.

consumption
Current steam
demand
processing
idle
failure
post production
off
Fig. 57 Screenshot of graphical depiction of process module in simulation
Verification
For verification of the process module a simple case was built up, which can be
modelled and simulated but also statically calculated. This is possible as the case
is deterministic in nature no probabilistic/stochastic elements and just one
process with defined product arrival are considered. This enables the comparison
of two separate models - static calculation and simulation - to describe machine
107
behaviour and calculate energy consumption. If both models come to the same
results independently from each other the process module can be considered as
verified since its behaviour is basically consistent with other models (Chapter
3.4 based on Rabe et al., 2008). Obviously typical simulation applications will be
more complex and not solvable with static calculation (which reasons the usage of
simulation). However, basic functionalities of the process module itself remain the
same and so verification is possible with the simplified case.
The parameters are shown in Figure 58. For this example, two parts of any kind
are processed and their appearance at this generic machine is determined at 60
seconds (part 1) and 400 seconds (part 2). Figure 58 shows the simulation results
in terms of state variables at the end of the simulation run (1000 seconds).
Additionally, the load curve for electricity and compressed air consumption is
shown. The first analysis underlines that the simulation model works as expected:
the machine starts with the arrival of the first part at 60 seconds and is processing
the part after a 30 second ramp-up phase. After some time of idleness at 400
seconds the second part appears and is started to be processed. As intended there
is a failure within the operation which leads to stoppage of 120 seconds (MTTR).
The machine is restarted after that and is in idle mode until the enabled AutoOff
function shuts down the machine automatically after 300 seconds. For this
deterministic case, this machine behaviour could be easily predicted with the data
given. Therewith, the basic functionality of the process module is verified.
3
5 6
power [kW]
Compressed air demand [m/h]
1 2
Process Module Parameters

Setup time =
0 sec.
Ramp-up time =
30 sec.
Process time =
120 sec.
MTTF=
200 sec. (static)
MTTR=
200 sec. (static)
AutoOff enabled, shutdown after 300 sec. Idleness
Electrical Power consumption
Ramp-Up: 10kW; Idle: 3 kW; Producing: 7kW
Compressed air consumption
Ramp-Up: 0 m/h; Idle: 1 m/h; Producing: 5 m/h
time [sec]
Simulation results (Verification Run)
Event description
1: Ramp-up / 2: Processing first part / 3: Idleness, waiting
for next part / 4: Processing second part / 5: induced failure
during processing / 6: Restart / 7: idleness
Selected state variables
Simulation time =
Electricity consumption =
Compressed air consumption =
Utilisation =
time [sec]
Fig. 58 Results of verification run for process module
1000 sec.
0.896 kWh
0.414 m
20%
108
Additionally, the means for calculation shall be tested as well. The

consumption W can be calculated mathematically through the following equation
which combines the multiplied power (P) and time (t) for different states of the
machine (ramp-up, idle, process).

(25)
As mentioned, in this simple case with no probabilistic behaviour and only a

single process, all events can be determined in advance. Thus, it is possible to
calculate statically:

With the same equation also the compressed air consumption can be determined
with:

The utilisation of a production machine can be calculated through:

These results are calculated statically and completely independent from the
simulation. However, the simulation results in very similar numbers with a
consistency of over 99% for each variable. Altogether, with the successful
assessment of the functionality and the proof that calculations are working
correctly the process module can be seen as verified.
5.3.3 TBS Module Compressed Air

Chapter 2.4.2 already described the supply with compressed air with one or
several compressors being the main component(s) for transforming electrical into
mechanical energy. Determining variables of a specific compressor in this context
are the possible compressed air supply (volume) rate, the electrical power demand
and the control scheme (Bierbaum and Htter, 2004; Ruppelt, 2003). Four
different control approaches are available in industrial practice, which are shown
in Figure 59. As illustrated in the figure, these control regimes directly determine
the compressed air output and the energy consumption of the compressor. The
actual compressed air pressure in the system is the reference variable for
compressor control while keeping up a defined pressure to ensure operation of
production machines. Thus, main influencing variables are the different possible
operation states (running, idle, off) as well as values for startup and shutdown
pressure of the individual compressor.
109
p [bar]
p [bar]
p [bar]
p [bar]
pmin
pmin
pmin
pmin
max
max
system air pressure
power
[kW]
[t]
power
[kW]
100%
100%
max
system air pressure
[t]
power
[kW]
electrical power
[t]
system air pressure
[t]
electrical power
[t]
0%
electrical power
[t]
[t]
100%
0%
electrical power
system air pressure
power
[kW]
100%
0%
0%
max
[t]
Fig. 59 Integrated control schemes for compressors (Bierbaum and Htter, 2004)
Orienting on the nominal pressure, the operating compressed air pressure is the
result of supply and demand in the system. Based on the equations given earlier
this actual compressed air pressure can be calculated with

(26)
This inherent logic for determining compressed air system pressure and
controlling a compressor builds the background for the compressed air module. As
shown in Figure 60, the compressed air supply rate and electrical power demand is
modelled based on the distinctive states of a compressor.
Reference variable
Compressor State chart model
power
[kW]
operating
p [bar]
pmax
<pmin
>=p max
pmin
standby/idle
shutdown
delay time
system air pressure
[t]
<pmin
off
Resulting air supply and power

demand
100%
0%
electrical power
[t]
comp. air supply rate
[t]
air supply
[m/h]
100%
0%
Fig. 60 State based control of compressor in compressed air module
A compressed air system as a whole can contain several compressors in order

to provide the necessary capacity but also flexibility for air supply. For control of
multi-compressor systems different control schemes are available. For many
years, a cascading approach was mainly applied. Thereby each compressor is
responsible for a certain pressure range and all work together with cascading
overlapping. Nowadays all compressors usually work in one shared pressure range
and are controlled electronically (Ruppelt, 2003). Both control schemes can be
realised in the module through appropriate parameterisation.
110
Figure 61 shows the main elements and interfaces of this module. The main input
is the total compressed air demand of the system. Based on this demand and the
calculated supply the resulting operating pressure is determined and handed over
to the MS module from where it serves as input for process modules. Additionally,
the compressor induced electricity consumption flows into the EV module in order
to calculate the total electrical power demand.1 Parameters and relevant state
variables are described in the following.
Parameters
Input
basic compressor data

nominal system volume and
pressure
energy consumption and air
generation rate patterns
compressor control schemes
Output
actual operating pressure

system compressed air
consumption rate
Compressed Air Module

compressor electrical power
demand
actual operating pressure
actual energy consumption and air
supply rates
on/off-cycles
available exhaust heat
State variables
(selection)
from/to MS Module
from/to PPC Module
from/to TBS Module
from/to EV Module
Fig. 61 Inputs, Outputs and Parameters of the Compressed Air Module
Parameter
The parameterisation can be done in either the simulation environment itself or via
MS Excel user interface which is provided as well. All parameters are listed in
Table 19. As explained above (Chapter 2.4.2) the compressed air systems consists
of one or several compressors and auxiliary components (e.g. filter), which all
need electricity for operation. In the simulation, a compressed park with up to
seven compressors can be modelled. For each compressor relevant technical data
1
The compressed air module can also be used as a separate application (without connection to
other modules). In this case, a measured compressed air load profile of the factory is
imported as MS Excel document. Based on this data the compressed air module calculates
the state variables of the system (e.g. electrical power demand, air pressure) with its specific
characteristics. Without the need of building up a comprehensive model of the whole factory
with all production processes, this allows the analysis of different compressed air system
configurations against the background of a given demand of compressed air over time. The
consideration of dynamic interactions between production and compressed air supply is
naturally limited in this application case.
111
Table 19 Parameter list of compressed air module (n: number of compressor)
Parameter
identifier_compressorn
leakage_rate
plan_operating_pressure
plan_total_volume_compressed_air
running_power_rate_compressorn
idle_power_rate_compressorn
auxiliary_energy_consumption_rate
startup_pressure_compressorn
shutdown_pressure_compressorn
shutdown_delay_compressorn
control_scheme_compressorn
air_generation_rate_compressorn
compressor_efficiency
heat_exchanger_efficiency
Description
Dimension
Labelling of
compressor 1..n.
Air losses in
system due to
leakage.
Nominal air
pressure of
system.
Nominal air volume
of system.
Power demand for
running
compressor.
Power demand for
idling compressor.
Power demand for
auxiliary
components.
Trigger pressure
for compressor
startup.
Trigger pressure
for compressor
shutdown
Delay time before
shutdown from idle.
String variable,
free naming
[m/h]
[bar]
[m]
[kW]
[kW]
[kW]
[bar]
[bar]
[seconds]
(1) interrupting
control, (2) idle
Control regime for
run control, (3)
single compressor.
delayed
interruption, (4)
continuous control
Possible air supply
[m/h]
of compressor.
Efficiency
compressor energy
[%]
conversion.
Efficiency of heat
[%]
exchanger.
like the electricity consumption rate for different operation state, the possible air
generation rate and configuring variables for the underlying control scheme (e.g.
type, startup/shutdown pressure) can be keyed in. Auxiliary system electrical
power demand can also be added. From a system perspective the nominal air
112
pressure and system volume (as a result of tanks and pipes) are crucial variables to
calculate the actual air pressure in the system.
Additionally efficiency factors for the compressor itself and a possible heat
exchanger can be defined which are typically given in technical documentation.
These values allow further considerations in combination with other modules.
State Variables
Different variables are used to track the state of the compressed air system.
Figure 62 shows a screenshot of the graphical user interface of the compressed air
module which gives an overview of relevant state variables.
Fig. 62 Overview of relevant compressor state variables (screenshot from GUI of

compressed air module)
Of main relevance for the whole factory is the current operating compressed air
pressure as explained above it is calculated based on the compressed air system
characteristics and the total air supply and demand of the system. Additionally, the
electrical power demand is an important state variable and can have significant
influence on total power demand of the factory. It is the result of the electrical power
demand of each compressor (depending on specific operating state) and the auxiliary
equipment (e.g. filter). Based on the electricity consumption of the compressor
system and the supply with compressed air, the specific compressor power demand
(kWh/m/min) related to the classification shown in Figure 23 is continuously
calculated. From the technical perspective on compressors with the number
switching operations another variable is very relevant: constantly turning on and off
the compressor leads to deterioration and reduces lifetime. Therefore, there are
certain limits for switching operations, which should be respected (Figure 63).
113
compressor power [kW]

from
to
max. switching
cycles allowed [1/h]
7,5
30
11
22
25
30
55
20
65
90
15
110
160
10
200
250
Fig. 63 Allowed switching operations for compressors (Mller et al., 2009)
As mentioned above compressed air involves significant energy losses in form

of exhaust heat. Therefore, occurring heat is also an important state variable since
it could be used for energy recovery (e.g. for steam generation).
Verification
For verification, data from a literature case study from (Bierbaum and Htter,
2004, p. 146) is used. Thereby, constant compressed air consumption is assumed
which depicts leakage losses in the network. The respective key figures are shown
in Figure 64. If the simulation model would be able to come to the same results its
functionality can be considered as verified. Energy consumption values were not
available but are assumed to be 30kW (operating) respectively 10kW (idle) in
order to verify consumption calculation as well. Additionally, a control scheme
with delayed (60 seconds) compressor shutdown was selected. Therewith, based
on the calculated values of the given case study the energy consumption can be
calculated:

Figure 64 shows the simulation results. As the diagrams show, the model of the
compressed air system acts as intended and expected. Also the simulated state
variables are very similar to the calculated results provided by the external case
study. The accuracy through all variables is at least 98%. This is also true for the
electricity consumption, which reflects the statically calculated value very well.
Similar to the verification of the process module this is certainly a relatively
simple case with constant compressed air consumption. However, only this
simplicity enables a correct verification; the general logic stays the same anyway,
even for more complex cases with fluctuating consumption patterns.
Altogether, with the successful assessment of the functionality and the proof
that calculations are working correctly, also the compressed air module can be
seen as successfully verified.
air pressure [bar]
114
Verification Study Parameters
power [kW]
operation
Idle/off
Basic data
System volume =
nominal pressure =
start pressure =
CA consumption =
CA generation =
Power consumption =
time [sec]
3000 l = 3 m
10 bar
8 bar
2.035 m/min = 122 m/h
3.350 m/min = 201 m/h
30kW (operating) / 10 kW (idle)
Calculated results
Compressor run time
=
Compressor off/idle time =
Switching operations/hour=
4.56 min
2.95 min
8 1/h
Simulation results (Verification Run)
Selected state variables

Simulation time =
Compressor run time
=
Compressor off/idle time =
Switching operations =
Electricity consumption =
3600 sec.
4.54 min
2.90 min
8 1/h
19.385 kWh
time [sec]
Fig. 64 Verification study for compressed air module
5.3.4 TBS Module Steam Generation

The descriptions in Chapter 2.4.3 underlined that, from a combustion perspective,
the steam generation process is determined by boiler capacity and efficiency as
well as the quantity and type of fuel. With respect to the medium water its specific
enthalpy for evaporation (depending on pressure and temperature, a steam table
can be found in the appendix) and the share and temperature of both fresh and
condensed water are relevant. The underlying understanding of the steam supply
module with the considered parameters is depicted in Figure 65.
steam flow in [kg/h] with certain pressure
[bar] and temperature [K]
energy input [kW]
boiler
efficiency
capacity
water supply as combination of

condensate/fresh water
consumer
tank
condensate water with certain

temperature [K]
fresh water with certain

temperature [K]
Fig. 65 Abstraction of steam supply system as underlying model logic
115
The main purpose is to calculate the necessary energy input in kW (and finally
the necessary quantity of fuel) based on the steam supply ( , constant flow or
dynamic but limited by capacity of boiler). With combining the variables
described above (share of fresh water/condensate nFW/C, the related temperature
differences to the steam T, water specific variables cm and hw) the necessary
energy input ES can be calculated with

(27)
Through additionally applying the equations given in Chapter 2.4.3 the time based
energy and finally fuel demand can be calculated. Steam generation is a
continuous process that is well describable through thermodynamic equations.
Thus, these equations build the inherent logic of the steam module as continuous
simulation model. The configuration of the specific system can be modelled with
sufficient flexibility and accuracy.
Figure 66 shows the main elements and interfaces of the steam module. The
module can use the steam demand of the production equipment as dynamic
requirement for steam generation. Alternatively, a constant generation volume can
be determined. Based on the specific boiler data, technical specifications of the
evaporation process and water supply, the necessary energy input and actual steam
generation rate is calculated. This data is handed over to the MS and EV module.
As further option, data on available and usable exhaust heat from the compressor
system can serve as input of the steam module. With this heat recovery option the
demand on oil or gas for steam generation can be reduced.
Parameters
basic boiler data
basic physical data
water supply
Input
Output
steam demand of production
current steam supply
Steam Module
exhaust heat from
compressed air generation
boiler energy demand

necessary energy input
steam generation rate
State variables
(selection)
from/to MS Module
from/to PPC Module
Fig. 66 Inputs, Outputs and Parameters of the Steam Module
from/to TBS Module
from/to EV Module
116
Parameters
Table 20 shows a list of all parameters of the steam module. It can be
distinguished between
basic technical specifications of the boiler and control strategy applied in

the specific system (efficiency, maximum steam generation),
information about the evaporation process itself like the targeted
temperature or specifications of water supply,
general physical variables like the specific heat capacity or the
evaporation heat, which is depending on the pressure and temperature of
the steam and can be derived from steam tables.
These specifications are widely used for dimensioning steam generation system
and can be found in technical documentation of the boiler or basic literature (e.g.
steam tables).
Table 20 Parameter list of steam module
Parameter
boiler_efficiency
max_steam_generation_rate
dynamic_supply_control
constant_supply_rate
target_temperature
share_condensedwater
share_freshwater
temperature_condensedwater
temperature_freshwater
heat_capacity
steaming_energy
Description
Efficiency of boiler
for steam
generation.
Maximum supply
with steam per hour.
Steam supply
depending on
demand but limited
to maximum rate.
Constant supply rate
with steam per hour
(if dynamic control is
off).
Designated
temperature of
steam.
Share of condensed
water on water
supply.
Share of fresh water
on water supply.
Temperature of
condensed water
backflow.
Temperature of
fresh water.
Specific heat
capacity of used
medium
Specific evaporation
heat.
Dimension
[%]
[kg/h]
1: on / 0: off
[kg/h]
[K]
[%]
[%]
[K]
[K]
[kJ/kg K], for water
4.187 kJ/kg K
[kJ/kg]
117
State Variables
Two variables are of main interest for the steam module, the steam generation rate
and the necessary energy input. The steam generation rate is being determined by
the control strategy applied. Steam generation can either be a constant flow based
on a fixed generation rate or connected with production demand (with certain
minimum generation rate and limited by maximum capacity of the boiler). Based
on this steam generation and the relevant technical/physical variables of the
system the necessary energy input is being calculated.
Verification
For verification, a case study with specific numbers is applied. In contrast to the
previous modules the steam modules is mainly based on equations. Thus, only the
correct calculation of values needs to be verified. Therefore, a static calculation is
compared with the simulation outcome (Figure 67).
steam flow [kg/h]
Verification Study Parameters
Basic data
steam demand =
share fresh water =
share cond. water =
spec. heat capacity =
evaporation heat =
boiler efficiency =
1000 kg/h
40%
Temperature = 10C
60%
Temperature = 90C
4.187 kJ/kg K (water)
2202 kJ/kg (for 120C/10bar steam)
65%
Static calculation:
Simulation result:
Energy demand = 1046.15 kW

Energy demand = 1051.83 kW
time [sec]
Fig. 67 Verification results for steam module
For many applications a constant supply with heat is necessary; therefore this
case is being considered with a constant demand of 1000 kg steam per hour. The
additional relevant background data is shown in Figure 67. These values were also
used in the simulation model. As Figure 67 shows, the simulation behaves
correctly and comes to the same results. Thus, calculation is working properly and
the steam module can be considered as verified.
5.3.5 PPC Module

Figure 68 gives an overview over the production planning and control (PPC)
module. It shows the specific role of this module as superior layer which is
responsible for coordinating all production activities in the manufacturing
system.
118
Parameters
process module parameters
production program
general product information
Input
Output
Production Planning
and Control
process time, machine energy

consumption
sequences, lot sizes, starting
time
product weight and value
State variables
(selection)
from/to MS Module
from/to PPC Module
from/to TBS Module
from/to EV Module
Fig. 68 Inputs, Outputs and Parameters of PPC Module
The PPC module has no input from other modules or any state variables. At first
this seems to contradict the understanding of a control system with a necessary
feedback variable regarding the state of the system. The reason for that is the logic
separation of production management functions into MS, EV and PPC module. As
described in Chapter 5.2, the feedback is realised through the EV module. The PPC
module is not an automated control unit; the user himself is the connecting element
and changes parameters in PPC while considering the system state provided by the
EV module. Via these different parameters the manufacturing system as a whole and
complex production process chains can be configured. The module is built up in MS
Excel. Thus, it is easily accessible for users who are not familiar with simulation
applications. However, basic production configurations for simpler cases can also be
done directly in the simulation (which is sufficient for many cases and might be a
reasonable option for advanced users).
Parameters
Within the PPC module three different kinds of parameters are considered. Firstly,
general product information can be set, which play a role in later evaluations for
example. These are the (start) weight and a certain value associated with the product.
These are for instance the profit margin or the total added value per product.
Secondly, process module parameters can be defined through the PPC
module (e.g. process time, machine starting time). As mentioned before this is also
possible via direct process module configuration whereas for two reasons it might
make sense to do it via the PPC MS Excel interface. Besides the easier
accessibility for non-experienced simulation users this additionally allows
configuring product specific process module parameters, e.g. dynamic process

order size x process time
Processe Module 1
Processe Module 2
119
starting points
product specific process chains
1
2
Processe Module 3
Processe Module n
06:00
08:00
10:00
12:00
14:00
16:00
18:00
20:00
time
Fig. 69 Input parameters of PPC module
times depending on the actual product. Finally, the production program with
product specific process chains can be configured as shown in Figure 69.
Besides the aforementioned individual process times, the production program is
determined by order sizes and starting points for specific products as depicted in
Figure 69. As background for that individual, product specific production process
chains can be defined.
Verification
As depicted above, the PPC module works very differently compared to process
and TBS modules. It has a superior function and is strongly interacting with both
process and MS modules. A separate verification is not possible because this
automatically involves all other modules. Therefore verification will be done for
all modules together in context of the MS module.
5.3.6 Evaluation/Visualisation (EV) Module

Figure 70 shows the main elements and interfaces of the evaluation and
visualisation module. It plays a crucial role for the whole approach as this module
brings together data from different other modules.
As inputs, the EV module gets the energy demand of TBS modules and
manufacturing system (MS module) as well as further information e.g. regarding
production performance. Based on diverse parameters, relevant key figures and
diagrams are generated, which are depicting the state of the simulation model as a
whole. The EV module is also responsible for transferring data to external tools
for the visualisation of energy flows and extended (manual) evaluation.
120
Parameters
energy contract data
environmental base data
general background data
Input
Output
(interfaces to external software)
boiler energy demand
Evaluation and
Visualisation Module
CA energy demand/supply
production data in context of
energy and performance
data interface for visualisation

via E!Sankey
data interface MS Excel
based evaluation
energy consumption variables

energy costs
energy related environmental impact
technical evaluation
State variables
(selection)
from/to MS Module
from/to PPC Module
from/to TBS Module
from/to EV Module
Fig. 70 Inputs, Outputs and Parameters of the EV Module
Parameter
All parameters of the EV module are shown in Table 21. These parameters are
mainly background data for the calculation of certain key figures. Therefore no
further explanation will be given at this point but later, in context of the
calculation of state variables.
State Variables
As mentioned before, the calculation and visualisation of several state variables is
the core of the EV module since they depict the state of the whole system. Against
the objective as a flexible simulation approach for manufacturing systems the user
shall not be limited beforehand. A diversity of variables for evaluation is being
offered, which allows the selection of the most appropriate ones for the specific
case. Different perspectives are being pursued in this context. Besides technical
evaluation and pure energetic view, this includes the calculation of energy costs
and related environmental impact. Figure 71 shows a screenshot of the graphical
user interface of the simulation environment, which is mainly implemented
through the EV module. The relevant key figures will be explained in the
following.
Technical Analysis
The evaluation of the general technical capability of a manufacturing system is the
usual application for material flow simulation studies. Since the proposed
approach aims at an integrated evaluation these issues need to be considered as
well. The most important state variables in this context are the product output of
the whole system (production output - number of finished parts) and the relation
121
Table 21 Parameter lists of EV module
Parameter
CO2_in_kg_per_kWh_electricity
CO2_in_kg_per_kWh_gas
CO2_in_kg_per_kWh_oil
base_electricity_price_per_kWh
peak_electricity_price_per_kWh
contract_load
dynamic_fees
peak_surcharge
static_fees
daily_worktime
monthly_working_days
oil_price_per_kg
gas_price_per_kWh
heating_value_gas
heating_value_oil
compressed_air_efficiency
gas_usage
oil_usage
electricity_base_load
Description
CO2 conversion
factor for electricity
usage.
CO2 conversion
factor for gas
combustion.
CO2 conversion
factor for oil
combustion.
Electricity price for
base time.
Electricity price for
peak time.
Electrical power as
defined in contract.
Various fees
depending on
consumption.
Surcharge for peaks
over contract load.
Fixed monthly fees.
Daily working hours
as base for
forecast.
Working days as
base for forecast
Price per kg for oil.
Price per kWh for
gas.
Heating value for
oil.
Heating value for
gas.
Efficiency
compressor.
Usage of gas for
steam generation.
Usage of oil for
steam generation.
Base load e.g. sum
of small consumers.
Dimension
[kg CO2, eq/kWh]
[kg CO2, eq/kWh]
[kg CO2, eq/kWh]

[/kWh]
[/kWh]
[kW]
[/kWh]
[/kW]
[]
[h]
[days]
[/kg]
[/kWh]
3
[kWh/m ]
[kWh/kg]
[%]
1: yes / 0: no
1: yes / 0: no
[kW]
122
Cumulative load profiles for

electricity, compressed air
and steam
Process modules structured

according to manufacturing
system logic
Production output, composition

of absolute energy consumption
(kWh), energy consumption costs
() and related CO2 emissions and
relative efficiency indicators
(different base units)
Detailed analyses of electricity

consumption
estimation of monthly electricity
costs (based on contract)
composition of consumption
Fig. 71 Screenshot of simulation environment with sample model and diagrams/key figures
for evaluation
to time as time per part. Additionally, the specific weight and value per product
can be defined and therewith production output can also be measured as
cumulated value or weight of finished products. These are numbers for the
system as a whole; on individual process base, further key figures are of
importance. Besides the number of failures and waste parts the utilisation rate
value plays a key role and was already explained above in context of the process
module.
Energy Consumption Analysis
In the context of energy consumption different means for analysis are provided.
The current demand of the whole system on electricity, compressed air and
steam are visualised through time based diagrams (load profiles) and allow the
analysis of consumption patterns over time. The sampling rate is usually one
second. For electricity in addition a 15 minutes average is tracked since this
reflects the sampling rate relevant for billing purposes.
The calculation of the total energy consumption gives an overview from an
energetic perspective. Thereby, the focus lies on all forms of energy externally
acquired by the company and so it is distinguished between electricity, gas and oil.
For visualisation a pie chart is provided, which shows the contribution of all
considered forms of energy on total consumption.
Electricity usually plays a major role in context of energy consumption.
Therefore further means for analysis are provided. One chart is provided, which
shows the shares of production versus compressor electricity consumption.
Another pie chart introduces the perspective on valuable and waste electricity
consumption. Electricity consumption is valuable if used for the value creating
123
process. All consumption connected to other operation modes (e.g. idle) is

considered as waste. This differentiation helps to identify potentials for
improvement. As already explained for each individual process module its current
energy demand and the specific contribution to total electricity consumption is
also displayed which are strong means for decision support.
Energy Cost Analysis
As Chapter 2.3.1 revealed the pure energetic view is only one necessary but not
sufficient perspective. Even more important for a company is the conversion to
related energy costs and environmental impact. Energy costs are calculated with
two schemes.
Firstly, the costs of energy consumption of electricity, gas and oil are calculated
through multiplying consumption with certain price rates which can be determined
as parameters. The composition of total energy consumption costs is illustrated
as a pie chart.
However, Chapter 2.4.1 clearly revealed that at least for electricity a pure
conversion of consumption to costs is not sufficient. Electricity costs consist of
diverse components depending on different variables. Therefore, electricity supply
contract models can be applied in the proposed approach.
Different consumption price rates for base (6am-8pm) and peak (8pm-6am)
times can be inserted in order to reflect daily price changes.
Typically a negotiated power load limit is part of the contract and can be
keyed in. Exceeding this limit results in peak surcharges with a specific rate
per kW. Thereby the highest peak per month is relevant.
Fees can be distinguished in fixed monthly fees and dynamic fees which
are depending on consumption. Both types of fees can be considered
simultaneously.
Based on this contract data and the current electricity load profile (aggregated to
15 minute interval as usual for billing) the monthly electricity costs are being
estimated and displayed in a pie chart. While simulation runs are typically
covering rather hours or a few days, a similar consumption pattern is assumed for
the whole month and forecasted with given parameters in terms of working days
and daily work time. The monthly perspective is relevant since this is the usual
billing period and certain cost aspects are calculated on this base as well.
Altogether, monthly electricity costs describe what companies actually have to
pay for electricity and therefore this view is very relevant for consideration.
Energy Related Environmental Impact
For assessing the energy related environmental impact, energy consumption and
usage is converted into related green house gas emissions. Typically, CO2 or CO2
equivalents (which subsumes different GHG) are applied as unit in this context.
Necessary conversion factors need to be inserted as parameters. For electricity this
factor depends on the specific primary energy mix for electricity generation (see
Chapter 2.2). Gas or oil is burned directly at the company in order to e.g. generate
124
heat; in these cases a certain factor for GHG through combustion needs to be
keyed in.
Efficiency Key Figures
Besides absolute values the calculation of relative values is a crucial option for
multi-criteria evaluation and solving of conflicts of goals. In this context, the
usage of appropriate efficiency figures is proposed. However, Chapter 2.5 already
pointed out that there are difficulties of defining the right efficiency indicators
for the specific case. The proposed approach does not want to restrict any
possibilities for the user, therefore different options were integrated. Either the
number of produced parts, the weight of all parts or the economic value of the
parts can be used as base. These values are set in relation to pure energy
consumption but also energy consumption costs and GHG emissions. Altogether,
in accordance with the efficiency definitions given in Chapter 2.5 the following
options can be considered:

(28)
(29)
(30)
All these efficiency key figures are calculated continuously during runtime. The
displayed reference unit (number of parts, weight or economic value) can be
switched simply through the graphical user interface.
Evaluation under Uncertainty
Within the proposed simulation approach the consideration of both deterministic
and probabilistic models (see Chapter 3.4) is possible. Stochastic effects become
relevant if e.g. failure behaviour of machines or certain process times are modelled
through distribution functions. As a result, each single simulation run leads to only
one possible result in probabilistic models; another run may lead to totally
different outcomes. This uncertainty is not acceptable for providing decision
support in manufacturing companies. This is specifically true when it comes to
comparing different scenarios. It has to be ensured that different outcomes of
scenarios are statistically significant and not just a result of stochastic effects. To
overcome this obstacle and gain acceptable statistical power, a sufficient amount
of simulation runs has to be conducted. Variables to estimate the necessary
number of runs are (Soper, 2011)
the statistical power which should be typically >0.80,

the alpha level or error rate/p-value, whereas a value <0.05 expresses
statistical significance
125
the effect size d (also named Cohens d) which expresses the strength of the
relationship of two data sets. It is calculated with the mean of the data sets
x1,2, their variance s1,2 and the sample size n1,2

(31)
(32)
For the effect size d three different categories can be distinguished. A value of
d=0.2 refers to a small, d=0.5 to a medium and d>=0.8 to a large effect
(Cohen, 2009).
1500
1400
1300
1200
1100
1000
900
800
700
600
500
400
300
200
100
0
small effect size (d=0,2)
error rate = 0.05
medium effect size (d=0,5)
sample size / simulation runs
sample size / simulation runs
Figure 72 shows the necessary number of runs to gain a specific statistical power
with respect to the effect size d and an error rate of 0.05 and 0.01 respectively.
large effect size (d=0,8)
0,8
0,82
0,84
0,86
0,88
0,9
0,92
statistical power
0,94
0,96
0,98
1500
1400
1300
1200
1100
1000
900
800
700
600
500
400
300
200
100
0
small effect size (d=0,2)
error rate = 0.01
medium effect size (d=0,5)

large effect size (d=0,8)
0,8
0,82
0,84
0,86
0,88
0,9
0,92
0,94
0,96
0,98
statistical power
Fig. 72 Necessary sample size depending on effect size, statistical power and error rate
(calculated according to Soper, 2011)
The diagrams underline the strong coherences of intended statistical power with
sample size (referring to number of simulation runs), effect size and accepted
error. Specifically the effect size has a strong influence: this practically means the
lower the effect, the more runs have to be conducted to statistically prove the
difference between two data sets. The diagrams help to estimate the necessary
sample size. For medium and large effects (which are of main interest in analyses)
and an acceptable statistical power of 0.8-0.9 with an error rate of 0.05, approx.
20-90 simulation runs are often sufficient to gain reliable results. However, for
detailed considerations 500 to >1000 runs might be necessary, which lead to
significant effort in terms of computing time. However, any of those analyses are
possible with the proposed approach support. The applied software tool AnyLogic
offers proper functionalities to efficiently run hundreds of runs (which may take a
couple of hours). To ease application, an automatic continuous export of relevant
result variables into a MS Excel database was integrated, which allows
automated simulation and data recording. With this data statistical treatment can
be done. Besides the calculation of the statistical confidence, typical key figures
for evaluation are the average and standard deviation of the data set. Also the
statistical probability to stay below certain limits might be relevant to consider
(e.g. monthly energy costs will be <10000 with a certainty of 95%, see also
Figure 73) (Black, 2008; Anderson, 2002).
126
Fig. 73 Selected statistical key figures for a normal distribution (e.g. Black, 2008;
Anderson, 2002)
Visualisation of Energy Flows

Besides key figures and charts for supporting evaluation and decision support,
general energy flows within the factory can be visualised with Sankey diagrams.
Sankey diagrams were developed by H.R. Sankey and are an established method
specifically for those kinds of tasks. They can strongly increase the transparency
of energy and material flows for complex systems. Therewith, it supports the
systematic derivation of most promising improvement measures.
An example for a Sankey diagram was already shown with Figure 14. Another
example gives Figure 74 with an old Sankey illustration of a steam plant (which is
also the first Sankey diagram) by H.R. Sankey himself (Sankey, 1898).
To understand Sankey diagrams some basic assumptions are important to know
(Schmidt, 2006; Schmidt, 2008b):
variables are related to a specific period of time (in this case simulation time)
considered values are extensive (different values can be added)
the arrow width is proportional to its value
no stocks, what goes in a process has to go out as well
To realise automated Sankey visualisation of simulation results a dynamic

interface with the software tool E!Sankey (developed by IFU Hamburg GmbH,
www.ifu.com) was established. E!Sankey is a specific tool for easy
visualisation of Sankey diagrams, which are often difficult to draw manually.
Every 10 seconds the simulation tool writes energy flow related values in a
database, which is being read out by E!Sankey and then automatically updates
the pre-defined energy flow model.
127
Fig. 74 Example for Sankey diagram for the case of a steam plant (Sankey, 1898 also
shown in Schmidt, 2008a)
Verification
The EV module is highly relevant for the system as a whole and strongly
interacting with both PPC and MS module. It has a supporting function and its
calculations depend on input values of other modules; a separate verification with
blending out other modules is not reasonable and hardly possible. Therefore
verification will be done for all modules together. If this can prove that
calculations work correctly, this module can be considered as verified.
5.3.7 Main Level MS Module

As depicted in Figure 75 and also earlier in Figure 51, the MS module is the
backbone of the whole simulation approach. On this layer all generic modules are
logically put together (if not given already when using generic approach configured
by PPC module) and parameterised in order to establish an individual model for the
specific case. For simulation, the MS module combines process module and PPC
data and executes the experiment runs. Thereby, relevant data about energy demand
and supply as well as the technical performance of the manufacturing system are
transferred to EV, TBS and also back to process modules.
While being highly relevant from a technical perspective as hub for all other
modules, the MS module is only a necessary platform and has limited meaning
from a content oriented perspective. Several variables are tracked as raw data but
directly transferred to the modules where further processing takes place. This
simplicity of the MS module is established on purpose: if not accessed via the
Excel interface anyway, this layer is the main interface for the user when
modelling. To faciliate usage but also for avoiding accidental changes of critical
model code of this module, most functionalities were inserted in separate modules
and the MS module itself was kept simple and transparent.
128
Parameters
structure (if not
determined by PPC
module)
Input
Output
(interfaces to external software)
process data in context of

energy and performance
production energy demand
production program / process

chain
Main System
MS Module
energy/performance input data

feedback from TBS
State variables
(selection)
from/to MS Module
from/to PPC Module
from/to TBS Module
from/to EV Module
Fig. 75 Inputs, Outputs and Parameters of the MS Module
Verification
As indicated before, due to their close relation and interactions the three modules
EV, PPC and MS shall be verified all at once at this point. A small case is built up,
which allows checking functionalities and calculation schemes connected with those
modules. This case will consist of three process modules and is deterministic in
nature, which allows transparent assessment of model behaviour and static
calculation of state variable for comparison with simulation results. The
manufacturing system is completely configured with the PPC module over MS
Excel while using the generic simulation approach. There is no rigid linkage
between the machines but a free flow of products through the system (just
determined by PPC). An order of 10 parts shall be produced on three different
machines in defined sequences. Certain process times, a simple energy consumption
pattern (idle and processing), and starting points were determined for each machine.
For reasons of simplicity, only electricity consumption was considered. However,
calculation patterns etc for other forms of energy are basically the same. Figure 76
shows the configuration of the verification experiment and the results from static
calculation as well as from simulation.
The simulated electricity load profile reflects the activities which were planned
in the PPC module. All processes start and end as intended (delays through e.g.
failures were not considered). Basic machine data like state based power demand
or process time was transferred correctly. Therewith the PPC module can be
considered as verified. The comparison of static calculation and simulation
outcome reveals that both results in similar values for energy consumption and
production output. These are the main variables as all other variables are based on
the correct calculation of those figures. With these positive results, the EV module
can also be considered as verified. The MS module is the hub and basically the
5.4 Application Cycle
129
Verification Study Configuration and static Calculation
Process Module 1
300s
Process Module 2
off
3 kW
idle
600s
Process Module 3
500s
process 1, 3.08kWh, 27%
900s
1000s
1500s
3 kW 10 kW
idle
2000s
2500s
5 kW
3 kW 15 kW
idle
off
0s
Power demand
idle
proc.
3000s
3500s
time
E = (300s*5kW+3200s*3kW)+(600s*15kW+2500s*3kW)+(900s*10kW+1500s*3kW) = 11,417 kWh

idle: 2.08kW
idle: 1.25kW
idle: 6.00kW
power [kW]
idle: 2.67kW
time [sec]
Simulation Results - Verification Run
Fig. 76 Verification results for MS, EV and PPC module
enabler of all those functionalities. Hence, due to the positive verification of all
other modules, the MS module is also verified. Additionally, the results proved
again the correctness of the process modules and their interaction with the MS
module.

As an integral part of the concept as a whole an application cycle will be
developed. It shall
support users to independently (without external help) and systematically

identify potentials to improve the energy efficiency in their specific
manufacturing company.
guide users through the goal-oriented and correct application of the
simulation approach for their specific case
enable users to derive and evaluate different measures with the proposed
simulation approach
The development of the application cycle and the description of its individual
phases will be presented in the following.
130
5.4.1 Application Cycle Synthesis

Against the background of the proposed objectives of the concept as a whole and
the application cycle in particular three other fields of interest need to be
considered:
In literature, a couple of approaches are available which present systematic

procedures towards energy efficiency assessment and measures in
manufacturing. Examples can be found in (Gopalakrishnan, 2005; Wohinz
and Moor, 1989; Gesellschaft Energietechnik, 1998). They all differ in
specific aspects (e.g. transferability, level of detail); however, a general
structure can be identified, which is shown in Figure 77. Typically top-downapproaches are pursued, which start with a rough analysis of existing energy
data and general energy consumption of the company. Afterwards more
detailed steps are to follow and based on estimated energy (cost) savings
potential measures are prioritised and implemented. Due to their similar
objective these procedures can certainly give a good orientation for the
proposed application cycle. However, none of them incorporates energy
oriented simulation approaches, which underlines the necessity for a further
development.
In recent years the development and implementation of energy management
systems has become more and more the focus of companies. An Energy
Management System (EnMS) systematically records the energy flux and []
helps a company to continuously and systematically improve its energy
performance while taking into consideration other relevant and legal
requirements. An EnMS includes the organisational and informational
structures required for implementing energy management, including
resources (Federal Ministry for the Environment, 2010). Based on previous
experiences from different countries (e.g. VDI, 2006) with DIN EN 16001 a
European standard was introduced in 2009 (DIN EN 16001). On this base, an
international standard is planned for 2011 with ISO 50001 (Pinero, 2009).
DIN EN 16001 reflects the Plan-Do-Check-Act (PDCA) cycle for continuous
improvement. Its main phases are displayed in Figure 77. In contrast to the
systematic improvement procedure presented before it is broader in nature
and explicitly includes extended aspects like energy policy development,
organisational and infrastructure setup, auditing/reporting or communication.
The proposed application cycle does not aim to include these issues; it should
rather be compatible with EnMS initiatives and support their implementation
and operation.
In the proposed application cycle the necessary steps for conducting
simulation studies need to be respected (Banks, 2010; Law, 2007; Wenzel et
al., 2008; VDI, 2007). A procedure was already introduced in Chapter 3.4 and is
display again in Figure 77. Main steps are the definition of the system/problem,
data acquisition and modelling, verification and validation as well as the
iterative conduction of experiments towards the implementation of promising
measures.
131
Systematic procedures towards

energy efficiency
Implementation
2
3
4
Planning
Implementation and
operation
7
9
10
Inspections and
corrective measures
Continuous Improvement
Corrective and
preventive action
Simulation Study procedure

Problem formulation
Objectives/ Project
plan
1
2
Model
conceptualization
Data collection
3 5
3 4
Model translation
Verification
Monitoring and
measurements
Estimating Savings/Priorisation
Energy policy
Management
review
Detailed Analysis
Internal audit
Rough Analysis
Energy Management System
Validation
Experimental design
7
9
10
Production runs and

analysis
7
9
Documentation/
Reporting
Implementation
10
10
Simulation oriented Application Cycle towards Energy and Resource Efficiency
Input
Output
Processing
1
System Definition
2
Identification of
Energy Consumers
4
improvement measures
(contract perspecitve)
check
machine list with

norm values and run
times
Energy Cost/Contract
Analysis
Comprehensiveness
energy bills, contract

details, load profiles
Data Metering and

Processing
5
Modeling
Modelling
6
Validation
Improvement
measures / Scenarios
Legend
8
data
Simulation
process steps
9
identified measures
plant and PPC perspective
evaluated measures
10
Evaluation
evaluated measures
Numbering of application
cycle phase and allocation
to related phases of
relevant fields of interest
optimisation
optimization
representative
production program
(machine perspective)
improvement
production
structure/layout,
TBS specifications
Implementation
Fig. 77 Synthesis of proposed application cycle
132
Based on these prerequisites and the requirements of the developed simulation

approach the proposed application cycle is shown in Figure 77. The figure also
shows the allocation and contribution of certain phases to the mentioned fields of
interest. As depicted, the application cycle brings together all necessary actions
towards a simulation based improvement of energy efficiency.
In general, it is also a top-down procedure which starts with a general analysis
of energy costs and flows. In the next step energy consumption as a whole is
broken down to single consumers whose consumption patterns need to be
measured in detail. This data serves as input for modelling the considered system
with the developed simulation approach. Since the simulation environment itself is
already verified, only the validation for the specific case is necessary in the next
step. With the validated model simulation runs can be conducted, which allow
deriving and evaluating diverse measures for energy efficiency improvement.
As mentioned above, the application cycle aims at enabling the user in a
manufacturing company to independently conduct studies for systematic
derivation and evaluation of energy efficiency measures for his specific case.
Thereby the usage of the developed energy oriented simulation approach is
motivated. The user is guided through the whole application cycle. No specific
requirements are connected to the first implementation. In the following each
phase and connected tools of the application cycle will be presented. Thereby it is
assumed that the user has no experiences in energy efficiency studies. Besides
some standard documents (e.g. energy bills) no further information about energy
consumption is needed at the starting point. For reasons of simplicity and general
relevance the explanations will have a certain focus on electricity. However, as
also indicated before, all forms of energy can be considered with the presented
application cycle and its inherent tools.
5.4.2 Step 1: Objective and System Definition

The very first step is the definition of the designated task, and the considered
system as well as connected system boundaries. The usual objective is to minimise
energy consumption while still being able to fulfil at least the necessary
production tasks. This leads to the objective of increasing energy efficiency with
the aforementioned definition as ratio of production output to energy input.
Companies are certainly less interested in improving these issues from a pure
energetic point of view typically the minimisation of energy costs and/or
environmental impact with given output are the main objectives. This may be
connected with defined quantitative targets for a specific period of time, e.g. an
energy efficiency increase of 5% in one year.
In context of system definition, the following alternatives might be possible
whereas this is neither necessarily a complete list nor are these strict categories:
The analysis of a company as a whole is certainly an ideal case because total

energy costs and all energy carriers are considered. Therewith the danger of
problem shifting and focusing on minor relevant issues is at least minimised,
potential cost reduction effect is not restricted and might be optimal and there is
a maximum leverage/synergy effect over the whole company (e.g. measures can
133
be applied in different areas of the company). However, the complexity of

analysis, connected efforts in terms of costs and time as well as the uncertainty
of considerations might increase strongly specifically for large companies
with several complex production structures.
Therefore, concentration on a certain area of production (e.g. specific
factory/manufacturing system) can be a promising approach. This might
compensate some of the challenges mentioned above which might restrict
some opportunities.
Considering a process chain for a specific product can provide an
alternating perspective for the analysis. This might involve several processes
in one or even over diverse companies. Typical questions would be to
minimise the environmental impact (e.g. carbon footprint) or the associated
costs of product. Whereas the calculation of direct consumption is not a
problem, the allocation of indirect consumption has to be addressed carefully
while it may also be related to different products.
Diverse variables may influence the system definition. The main drivers are the
energy related objectives of the company (e.g. energy cost decrease, improving the
environmental impact of a specific product) but also data availability or technical
as well as organisational structures (e.g. size of company, specific structure of cost
allocation, different production sites of a company) can influence this decision.
5.4.3 Step 2: Total Energy Consumption and Contract Analysis

In the next step the energy costs and the energy related environmental impact of
the selected system need to be assessed. This includes all kinds of energy acquired
from external sources for most manufacturing companies, electricity plus gas
and/or oil. Base for these analyses are the related bills from energy suppliers
(which might be internal suppliers when being different cost/profit centres) as well
as their underlying contracts. Whereas the costs for gas and oil are quite clearly
defined depending on consumption, the electricity cost calculation can be more
complex as already shown in Chapter 2.4.1. Hence, the electricity supply contract
should be analysed carefully and the composition of electricity costs has to be
determined. For the determination of energy related environmental impact
conversion factors for electricity usage and combustion of gas or oil can be used.
Altogether the following questions can serve as guideline for this step:
Which forms of energy are relevant for the specific case?

What is the contribution to total energy consumption (in kWh), costs (in e.g. )
and environmental impact (in e.g. kg CO2,eq) for each energy form?
Which cost portions contribute (and to what extend) to total electricity costs?
This information gives clear indications for further priorities; depending on the
outcomes next steps shall concentrate on the most important forms of energy.
Besides the necessary prioritisation for the next steps, from an economic point of
view first fields of action can be identified through contract analysis. Figure 78
gives an overview of means to influence electricity costs for the case of Germany.
134
electricity cost driver
(different cost portions)
Taxes
Licence fee
Renewable Energy Sources Act
Combined heat and power generation Act
Measuring price
System usage fee/network access
energy acquisition
influence through
pricing/contract
no influence
influence through
usage pattern
x
x
x
x
x
x
x
x
x
<
<
Price/costs for electrical work

Price/costs for electrical power
Base/peak price; mixed price
Reactive energy price
Minimum/contract power
Maximum load utilisation period
Load factor
energy usage
x
x
x
x
x
no influence
x
x
<
x
x
x
x
<
x
x
x
x
x
Fig. 78 Matrix for means to influence electricity costs
As mentioned before, electricity costs consist of manifold components, which

may be influenced through modifying contract conditions through negotiating
with the supplier or influencing consumption behaviour in the company (which is
addressed in later steps of the application cycle). Since there are almost no
standard contracts in industry, specifically for large consumers there are
opportunities to consciously influence contract contents. Typical examples are the
determination of a mixed price rate (instead of having different rates for peak and
base times), contract load or the ratio of consumption related price rate to peak
demand surcharges. Due to the complexity of the problem the evaluation of
alternatives is difficult for the user. However, basic possible influencing variables
have to be identified at this stage while alternatives can be evaluated with the
simulation approach at a later stage.
Additionally, time based analyses of energy consumption have to be conducted.
Therefore monthly consumption values (from energy bills) can be used to gain an
overview of e.g. seasonal influences. Even more, for electricity it is possible to get
the 15 minute load profiles for specific periods of time. This gives a good hint
regarding the electricity consumption pattern of the system and may reveal first
saving potentials (e.g. electricity consumption at night or weekends).
As an example, Figure 79 shows a sample electricity load profile for two
months of a sample manufacturing company (SME). In combination with the
yearly consumption and the details of the contract model some interesting aspects
can already be observed. There are very few distinctive energy peaks which
exceed 200 kW evidently in the morning hours when machines are turned on
simultaneously. A reduction of these few peaks (which might be easily avoidable
if being aware of the impact) would lead to cost saving of approx. 600 per year.
Additionally, it can be observed that the base load is at least approx. 5-6 kW.
Assuming that this load is demanded 24 hours a day and 365 days a year it gets
clear that this alone accounts for approx. 15% of total electricity consumption
without any value creation.
power [kW]

250,000
135
Reducing peak to 200 kW
potential saving of approx. 60
per year,
Load profile November/December 2008
200,000
with 190 kW further saving of

appox. 450
150,000
100,000
50,000
0,000
power [kW]
time [min]
250,000
annual peak 235 kW
8.12.-12.12.2008
200,000
150,000
100,000
50,000
basic load approx. 5-6 kW

estimated with 365 days / 24hours
approx. 50.000 kWh = 15% of total electricity
consumption
0,000
time [min]
Fig. 79 Example load profile of manufacturing company
All these considerations are most useful when considering the company as a
whole since all consumers together and the specific contract model determine the
energy costs. However, those analyses might also give interesting hints if only
specific areas of production are considered. On the one hand, realistic energy costs
calculation schemes (e.g. price rates) can be assumed. On the other hand, later on,
the contribution of the considered area to total energy consumption can be
estimated.
5.4.4 Step 3: Identification of Energy Consumers

While now knowing the important energy flows of the system in general, they
have to be broken down to single consumers in the next step. As it is assumed that
no detailed consumption information is available, an alternative way for
identification and first prioritisation is needed. For this purpose, a list of all
technical equipment of the considered system needs to be generated. This involves
production machines but explicitly also includes supporting machines like
compressors. Besides general data (like type, name, year of construction etc), two
values are critical to gather in this context:
The nominal value for energy consumption (nominal energy demand) of a

machine can be found in the technical documentation and/or on type plates
directly on the machine. As mentioned before, specifically in the case of
electricity the nominal value tends to be too high and rarely reflects the real
power demand in operation. However, it can serve as base for first estimation.
The operation time per year is a usual machine related key figure and
available in ERP systems or manual calculations of e.g. utilisation rates.
136
With those values an estimated energy consumption value can be calculated for
each machine of the system.

(33)
Therewith the machines of the system can be sorted according to this value which
is giving a clear indication of priorities for the next steps. It is crucial to ensure
that all potential consumers of the system were considered. An example for the
case of electricity is shown in Figure 80 (based on a real company case).
Evidently, electricity consumption is dominated by few consumers: in this case
just six machines sum up for half of the estimated consumption, approx. 15
machines account for 80%. With not even 30 machines determining consumption,
the majority of consumers are most likely not worth being considered in detail
(diagram is even cut off; altogether 78 machines are in the list).
40000
estimated consumption
[kWh/a]
250000
30000
200000
25000
20000
150000
approx. 80%
approx. 50%
15000
100000
10000
50000
5000
0
Machine 1
Machine 2
Machine 3
Machine 4
Machine 5
Machine 6
Machine 7
Machine 8
Machine 9
Machine 10
Machine 11
Machine 12
Machine 13
Machine 14
Machine 15
Machine 16
Machine 17
Machine 18
Machine 19
Machine 20
Machine 21
Machine 22
Machine 23
Machine 24
Machine 25
Machine 26
Machine 27
Machine 28
Machine 29
Machine 30
Machine 31
Machine 32
Machine 33
Machine 34
Machine 35
Machine 36
Machine 37
Machine 38
Machine 39
Machine 40
cum. estimated consumption

[kWh/a]
300000
35000
operation time
nominal value
estimated
consumption
h/a
kW
kWh/a
736
48
35328
2070
12
24840
CY
920
27
24840
Machine 4
XZ
800
31
24800
Machine 5
XY
8200
2,2
18040
Machine 6
368
45
16560
Machine n
Typ
manufacturer
Machine 1
XY
Machine 2
ABC
Machine 3
Fig. 80 Example for estimation of electricity consumption with pareto analysis
For further considerations and distinctive decision support this data is

transferred into an energy portfolio as shown in Figure 81 in general and also as
an example. Therewith the composition of consumption in its determining parts
power demand and operating time is presented separately. This allows the
137
classification of energy consumers into four distinctive categories, leading to

different strategies for further actions:
High power demand, low operating time (category I): these consumers may
be critical if peak surcharges make up a significant amount on energy costs.
Detailed measurements should be made to understand if and in which
operation mode demand peaks occur. Means for improvement can be
technically or organisational in nature, examples are an energy sensitive
production planning and control or load management.
High power demand, high operating time (category II): these consumers are
most critical as they are responsible for most of the energy consumption and
should be analysed in detail through certain measurements. Means for
improvement will typically be more technical driven, e.g. improvement of
machine components to reduce energy demand.
Low power demand, high operating time (category III): these consumers can
contribute significantly to total energy consumption due to their relatively
long operating time over the year while demanding little power (a pump or
electric driver on conveyor). Detailed measurements are often not even
necessary. These kinds of consumers might offer some feasible approaches
for improvement: with relatively low necessary investment and high leverage
effect through operating time, appropriate amortisation times are very likely.
Low power demand, low operating time (category IV): these consumers are
less relevant in context of the energy consumption of the system. No further
efforts should be spent for detailed consideration.
high power demand,

low operating time
Leverage
power demand
60
average
II
high power demand,
high operating time
critical machine
avg.
IV
III
low power demand,

low operating time
low power demand,

high operating time
noncritical
Leverage
operating time
50
Machine 1
nominal value [in kW/kVA]
nominal value [kW]
average
Machine 6
40
Machine 4
30
Machine 3
20
Machine 9
10
Machine 2
average
0
0
operating time per year [in h]
Machine 7
Machine 11 Machine 8
500
1000
1500
2000
2500
operation time per year [in hours]
Fig. 81 Energy portfolio as a tool for classifying energy consumers
5.4.5 Step 4: Data Metering and Processing

Previous classifications already gave a clear indication about priorities for detailed
energy consumption measurements. Typically, consumers of categories I and II
should be considered in further detail. In the next step the necessary accuracy for
this measurement needs to be determined. Against this background, Figure 82
shows the influence of different sampling rates in context of the accuracy of
resulting energy consumption data. It reveals that too rough measurements might
138
1 main switch on
1,05
2 start up
10000
power [W]
8000
1s
3 ready for production
1min
4 value creation
5min
0,95
0,9
correlation to 1 sec. sampling rate
12000
2
4
6000
4000
3
2000
0,85
0,8
0,75
0,7
0,65
0,6
0,55
0,5
0,45
0,4
0,35
0,3
0,25
0,2
0,15
15 sec.
0,1
0
0
60
120
180
240
time [sec]
300
360
420
0,05
0
0
50
100
150
200
250
300
350
sampling rate [sec]
Fig. 82 Influence of different sampling rates on accuracy of energy consumption patterns
lead to a significant loss of information. A later allocation of consumption to

different operation states is hardly possible. For the same measurement, the figure
also shows a correlation analysis compared to data measured with a sampling rate
of one second. It reveals that differences to the measured curve are fast becoming
strongly significant. To achieve a correlation value of 0.9, the necessary sampling
rate should not be larger than 15 seconds.
A lot of further variables influence the determination of an appropriate measuring
strategy. As one example, smaller sampling rates lead to a larger amount of data that
has to be handled. If process times are long, extremely detailed measurement will
most likely not provide much more valuable information. Additionally, it has to be
mentioned again, that the sampling rate relevant for billing purposes is only 15
minutes so things happening irregularly and with duration of just a few seconds
will have limited impact on a 15 minute power value at all.
Since a user with no previous energy related experiences is assumed, all
measurements will normally be conducted with mobile devices. States of the art of
those portable devices are sampling rates around one second. The data volume of
the single measurements is relatively low and standard office software can easily
be used for data processing. Therefore in this context a sampling rate of at least 10
seconds (better using 1 second) is suggested (for continuous measurement other
criteria might lead to different suggestions).
With measured energy data for all machines, the priority list with estimated
yearly energy consumption as well as the energy portfolio can be updated. With
this now more realistic data the contribution of considered machines on real total
power demand and energy consumption (derived from step 2) can be calculated.
This ensures the completeness of the analysis and also serves as criterion for
stopping measurements (e.g. if 80% of energy consumption is caused by machines
already considered further measurement activities are not necessary).
Besides laying the necessary data background for further steps, the analysis of
energy consumption behaviour directly helps to derive machine based measures for
improvement. Obvious saving opportunities can often be identified and employed at
139
once. Again it shall be mentioned that this is also true and relevant for other forms of
energy: whereas certainly being different in terms of measuring equipment and
strategy those analyses are relevant and applicable for e.g. compressed air, steam or
gas consumption as well.
5.4.6 Step 5: Modelling

In the next step the considered manufacturing system will be transferred into the
proposed simulation environment. Practical and technical aspects of modelling
were already presented in detail in Chapter 5.2. At this point, not the question
how to model? but rather what to model? is addressed. This directly leads to
the determination of the necessary level of detail.
Besides the data being gathered so far, further information on production
machines and structure as well as technical building services is necessary.
For production machines at least process times for the considered products are
necessary. Information on failure behaviour (MTTR, MTTF) will be helpful
for more realistic considerations. Additionally, product specific process chain
sequences are crucial for modelling it has to be clarified which processes
determine the material flow of certain products within the system. Value
stream maps (VSM) are an ideal base for modelling because they illustrate
both, process times and sequences.
For the relevant equipment of technical building service further system
specifications and operational data are necessary. In the case of compressed
air these are e.g. the system volume, nominal air pressure, amount and
technical data of compressors and their modes of operation. More details on
necessary parameters (also for steam generation) can be found in Chapter 5.2.
As already mentioned in Chapter 5.2, the proposed simulation approach is very

flexible in terms of level of detail. While extensive models increase the
complexity of analysis and necessary computing time, it is beneficial to
concentrate on most relevant machines. Figure 83 shows a decision tree that
supports the selection of an appropriate level of detail for modelling each
consumer. While keeping in mind the energy related relevance of each consumer
(classification in categories), different general alternatives can be distinguished.
An individual process module is necessary if the process is important for the
material flow as a whole and integration with other consumers is not possible or
feasible (e.g. consumer has to be treated separately to analyse behaviour in detail).
In many cases it is favourable to combine consumers to integrated process
modules. This is possible if e.g. processes are directly connected in the process
chain (e.g. neighbouring processing) or there is a supplying process whose energy
consumption is related with the operations of another process (e.g. machine
related cooling device). Practically, energy consumption is just added to the
consumption of the main process in this case. A separate TBS module should be
modelled if the consumer does not have a constant consumption and is responsible
for supplying the system as a whole (e.g. compressed air system). As described
before it is possible to subsume less relevant energy consumers as constant base
consumption of the system.
140
starting point
relevant for
material flow
no
yes
no
constant
consumption
similar
upstream/
downstream
processes
no
directly
related to
other main
machine
yes
yes
individual process
module
yes
integrated process
module
no
separate supply/
TBS module
base consumption
Fig. 83 Decision tree for level of detail while modelling
5.4.7 Step 6: Validation

Verification and validation are essential issues when conducting simulation studies.
Verification is ensured already since certain studies were conducted for each single
module in the development phase. However, the validity of the simulation model
has to be assessed for the specific case. Since validation is the assessment whether
the model behaves with satisfactory accuracy consistent with the study objectives
(Banks, 2010). Several techniques are available (see Figure 35); typically it is
realised through comparison of a simulation run with the comparable situation of
the real system.
In this context the ideal validation background would be the detailed
consumption data of the system (load profile and total energy consumption) for all
relevant forms of energy plus the real production data (production program and
individual machine behaviour) for the same period of time. This would allow
conducting simulation runs with the given production program and comparing
both energy as well as performance related target variables for validity check.
However, in practical cases, specifically energy related data will be rarely
available in the designated form and even less in combination with related
production performance data. Against this background also an isolated
consideration can be considered as sufficient:
Validation with given production data: with the given production program
validation runs can be conducted. Results in terms of selected key figures like
output or utilisation are compared to real production data. If system behaviour
is clearly determinable (e.g. automated production line with rigid linkage) it is
also possible to exclude probabilistic effects to test model behaviour (Fixed
Value Test - Rabe et al., 2008, e.g. theoretical output rate can be calculated
141
statically and compared to simulation results). If the system behaviour is

correct (e.g. similar production output), the model can be considered as
validated. Since energy consumption patterns rely on measurements and the
simulation environmental itself is verified also energy consumption behaviour
of the system can be assumed to be correct.
Validation with given energy data: vice versa - it might be possible that
only some energy related data is available, e.g. the energy consumption or a
load profile for a certain period of time. However, there might be no related
production data. In this case the simulated energy demand and consumption
should be compared with given data. Additionally, through careful checks of
assumption and process times as well as critical review with experts the
production performance related validity should be ensured.
5.4.8 Step 7: Scenario Building

The validated simulation model can now be used to derive and evaluate different
improvement measures towards energy efficiency, which leads to different
scenarios being considered (the validation run can serve as base scenario for
comparison). Thereby, single production machine and TBS measures derived
earlier can be tested in context of their impact on the system as a whole. It can be
assured that no negative consequences are occurring on system level (or, vice
versa, positive effects occur) and the contribution towards improvement from a
holistic perspective can be determined. This helps to prioritise those measures.
Furthermore, there are specific measures on factory/manufacturing system
level (e.g. load management, energy aware production planning and control).
Their functionality and impact on diverse target variables can be tested as well. As
described earlier, one of the target criteria is the energy cost estimation based on
realistic contract models. Vice versa, this aspect also allows the testing of
alternative energy contract models against the background of the behaviour of
the manufacturing system. Finally, also optimisation studies are possible while
using the solver OptQuest which is integrated in AnyLogic. Therefore it is
necessary to define a target variable (e.g. minimising energy costs), certain frame
conditions (e.g. ensuring defined production volume) and parameters (e.g. speed
of production) that can be changed.
5.4.9 Step 8: Simulation Runs

In an iterative procedure, diverse simulation runs are conducted, which reflect
different scenarios as formerly described. The results of each run should be
written down and transferred into e.g. MS Excel to be able to conduct extensive
evaluation afterwards. If the model contains probabilistic elements a sufficient
number of simulation runs need to be conducted. Against the background of the
considerations from Chapter 5.3.6 the conduction of 50 simulation runs is a
reasonable way to obtain a good estimation of value distribution. Through using
the functionalities of AnyLogic it is possible to conduct these runs
automatically through the function parameter variation/replications. This may
142
take several hours - thereby, relevant key figures are automatically written in MS
Excel as raw data to enable later statistical treatment and evaluation. For
instance, with the equations given it is possible to calculate the effect size d for
two scenarios. Herewith and the intended statistical power (typically at least 0.8)
as well as error rate (by convention maximal 0.05), the necessary amount of
simulation runs can be determined according to Figure 72. This allows identifying
whether the differences of two runs are proved to be statistically significant or
further runs are necessary.
5.4.10 Step 9: Evaluation

As presented before diverse key figures and diagrams support transparency and
the derivation of promising fields of action. For testing improvement measures,
the evaluation and comparison of different scenarios is necessary. Once having
identified a promising measure the economic profitability has to be ensured
through appropriate calculations.
Evaluation and Comparison of Scenarios
Several alternative target variables are provided which may serve as base for
evaluating scenarios. The set of relevant variables is individual for the specific
case and subject of e.g. study objectives (step 1). While several targets are
addressed simultaneously, multi-criteria evaluation including balancing of
conflicting goals might be possible. Several sophisticated methods for MultiCriteria Decision Making (MCDM) are available in literature (e.g. PROMETHEE
by Brans et al., 1986). For reasons of easier applicability in industrial practice a
simpler approach is proposed here. Sample results of scenario studies are shown in
Figure 84. In this case with energy consumption, energy costs and production
performance (in terms of necessary time for 100 parts) three target variables were
considered (all of these variable shall be as low as possible). Each variable is
associated with a certain weighting factor and normalised to relative values
(compared to base scenario).
Therewith, the weighted sum can be calculated and serve as a base for decisions.
( 34)
Description
Energy
consumption
(w. factor =1)
relative
values
Base
Standard scenario
10000 kWh
1.00
Improved machine
9000 kWh
0.90
Alternative PPC
7500 kWh
Scenario
Energy
costs
(w. factor =1)
relative
values
1000
200 seconds
0.90
250 seconds
0.75
1200
1.00
900
1.20
150 seconds
0.75
Fig. 84 Sample evaluation of simulation results
Prod. time
(100 parts)
(w. factor =1)
relative
values
weighted
sum
Energy efficiency
consumption/parts
1.00
1.000
100 kWh/part
1.25
1.013
90 kWh/part
0.675
75 kWh/part
143
These results can also be depicted via a spider diagram as shown in Figure 84.
This supports decisions while showing all targets and scenarios graphically with
one view. While all target variables are supposed to be as low as possible in this
example, the goal is to minimise the area of the triangle in this case.
energy
consumption
1,5
1
0,5
0
relative values
base
production
time
1
2
energy
costs
Fig. 85 Graphical representation of simulation results
As the sample results show, conflicts of goals between different target variables
might occur. Besides the weighted sum of all target variables, efficiency indicators
(see Chapter 5.3.6) can be used to bring together values and balancing targets in
order to assess the favourability of scenarios. These key figures are often more
intuitive and easier to use but may not be sufficient if manifold target variables
have to be considered simultaneously. A hierarchical combination of relative and
absolute values might be a promising way. For example, the absolute achievement
of most critical goals, like production output can be assured in the first step.
Scenarios that pass the defined threshold are compared afterwards with respect to
e.g. efficiency indicators in order to identify the best solution.
The consideration shows that a general recommendation is hard to give as it is
always a matter of the specific study. However, the presented means for
evaluation are generic in nature, which allows a broad applicability for all cases.
Investment Decisions
Measures towards energy efficiency are usually connected with necessary
investment. After having identified promising alternatives from a technical and
144
energetic perspective, their profitability has to be calculated before implementation.

Diverse calculation schemes and criteria are available and shall not be addressed in
detail here. Examples are the calculation of the payback period or the net present
value, which includes financial aspects like interest rates (Brigham and Houston,
2009; Khatib, 2003). The simulation results can directly be used in these
calculations while e.g. very specific and realistic energy cost (savings) data can be
provided.
5.4.11 Step 10: Implementation

The technical and/or organisational implementation of promising measures is
certainly the ultimate goal of the proposed concept but rather a practical issue and
therefore not explicitly focused. After implementation it has to be ensured that the
simulation model is also updated as this serves as the new standard for further
simulation studies.
Chapter 6
Application of Concept
In this chapter the developed simulation based concept for improving energy
efficiency will be applied to several case studies in order to underline applicability
and potentials of the approach. Thereby different cases are being considered to
show the opportunities. Each case will follow the steps of the presented
application cycle. While the developed concept is generic in nature, the examples
will underline case specific implementation and practical implications of each
step. The case studies are based on real data from several companies; for reasons
of confidentiality many values will just be given in a relative manner.
6.1 Aluminium Die Casting

The first case study focuses on a company which produces die casted parts for the
automotive industry (Herrmann et al., 2011a). Products are manufactured in mass
production with continuous material flow in transfer lines with rigid linkage
(highly automated). Production is running in a three shift system (24 hours a day).
Step 1: Objective and System Definition
In this case study one specific transfer line for the production of a specific product
shall be considered in detail. The objective is to minimise direct electricity costs
while still obtaining the same output on products.
Step 2: Total Energy Consumption and Contract Analysis
Production machines in the designated area directly consume electricity as well as
compressed air. However, for the purpose of this study just direct electricity
consumption is considered. The actual electricity consumption of this area is not
known due to missing measuring points. The considered transfer line is part of a
large production site with diverse facilities. A separate bill is not available. In
terms of costs, only consumption induced cost portions need to be considered (no
peak surcharges or fees).
Step 3: Identification of Energy Consumers
The production area that is analysed is clearly defined. No further prioritisation is
necessary the objective is to analyse a specific production line as a whole, which
springerlink.com
6 Application of Concept
146
does not contain too many different consumers. Therefore all consumers will be
considered. An overview of the system structure and all machines involved is
shown in Figure 86. Three parallel die casting cells (which contain robots for
spraying and handling as well as heating/cooling units) convert liquid aluminium
into the raw part which is further processed through a saw. Afterwards, parallel
CNC milling takes place on two parallel machines before abrasive blasting and
palletising finalise the process chain.
Casting cell
Saw, Robot
Casting cell
Saw, Robot
CNC-milling
Abrasive
blasting
Palletizing
Palletising
CNC-milling
Casting cell
Saw, Robot
90000
power [kW]
80000
Die casting machine
70000
peripheries
60000
50000
heating/
cooling
40000
30000
spraying robot
handling robot
Casting cell
20000
10000
0
t
time [sec]
Fig. 86 Structure of considered manufacturing system
Step 4: Data Metering and Processing

In the following step relevant data was obtained for all machines. As mentioned
above, it is a highly automated transfer line aiming for mass production of one
specific product. Thus, all process times are well defined and can be considered as
constant. Based on production respectively specifically maintenance data MTTF
and MTTR were calculated which describe the failure behaviour of the machines.
Finally, the electrical power demand of all consumers was measured with a
portable measuring device and a sampling rate of one second. A sample result of
these measurements can be seen in Figure 86. For all consumers the load profiles
were analysed and consumption values were allocated to the different operation
states.
Step 5: Modelling
The whole manufacturing system was modelled with the developed energy
oriented manufacturing system simulation approach. Thereby, the procedures
explained in Chapters 5.2 and 5.4.6 were used. As mentioned above, a casting cell
contains diverse separate consumers besides the die casting machines (DCM)
these are cooling and heating units as well as spraying and handling robots. The
147
Energy oriented manufacturing

system simulation environment
power kW]
electricity demand was measured for all of those machines. However, referring to
the modelling guideline explained in Chapter 5.4.6 these consumers were
integrated within one process module while their behaviour is directly connected
and their integration would not limit any possibility nor lead to different results.
Therewith, process time and energy consumption of the casting cell reflects the
combination of these machines. Similar to this, conveyor belts that partly connect
different processes automatically were not considered separately but are integrated
within the related module. All other consumers were considered as separate
process module whereas this most realistically reflects real circumstances and
offers most flexibility for potential analysis. A screenshot of the model of the
manufacturing system shows Figure 87.
time [sec]
Fig. 87 Simulation model for Aluminium die casting case (results based on scenario A)
Based on the electricity measurements distinctive machine operation states

could be identified that were also reflected in the model. For most machines just a
distinction between off, idle and processing is reasonable. However, the die
casting machine involves a necessary spraying process after value created
processing, which was considered separately.
Step 6: Validation
As mentioned before, for the considered production area no detailed energy
related data was available. Therefore, validation is ensured through comparing
production performance data with simulation results. For validation the maximum
(nominal) production performance of all machines was collected from the
companys production planning and therefore the maximum output of the system
148
could be calculated statically. When excluding all machine failures and a quality
rate of 100% in a deterministic model, similar values should be achieved through
simulation (Fixed Value Test). For the purpose of validation one day of
production was simulated and compared with the nominal output of the real
system. The results are shown in Figure 88 (for reasons of confidentiality no
absolute value can be given as this point).
production output
120,0%
100,0%
80,0%
60,0%
100,0%
97,9%
actual system
simulation
(validation run)
40,0%
20,0%
0,0%
Fig. 88 Results of simulation run
The results reveal that the simulation is very capable to reflect real system
behaviour. The small difference is explainable the simulation run started with an
empty system with a certain ramp-up phase until all machines were in full
production. This results in less output compared to the calculated maximum
production of the real system. For testing the energy consumption related
behaviour a logic check based on the simulated load profile was conducted. It
revealed that the simulated power demand reflected the expected behaviour and
the model worked correctly. Altogether, based on these results, the simulation
model of this aluminium die casting process chain can be considered as validated.
Step 7: Scenario Development / Step 8: Simulation Runs / Step 9: Evaluation
As described in the application cycle steps 7 to 9 are closely linked and part of an
iterative procedure. For practical reasons these steps will be considered together in
this case.
The first and obvious scenario which has to be considered is a standard scenario
which depicts the real system and serves as base for later comparisons. For this
purpose the values from the successful validation run were used (scenario A). Similar
to the validation machine failures will be excluded at this point to avoid uncertainty
and facilitate first analyses. Based on an analysis of the results improvement
potentials can be identified and measures will be part of further scenarios.
149
The base run (see screenshot in Figure 87) reveals the high level of electrical
power demand with an average value of approx. 700 kW. Due to the continuity of
production (continuous flow in transfer line, no shutdown of machines when
idling, no failures) demand is not strongly volatile and always stays below 800
kW. As the key figures show, main drivers of electricity consumption are the three
die casting cells (which sum up for 87% of consumption), the blasting (9%) and
the milling (4%) processes. From an electricity consumption perspective, the other
consumers can be neglected in further considerations. The bottleneck is evidently
the milling process (conducted by two machines) as the utilisation rates underline.
The die casting cell has an utilisation rate of 46% which is not a result of
significant idle time but rather of the necessary (but not value creating) spraying
process just after each part. The blasting process is utilised with 61% which is a
result of waiting for part supply from upstream processes.
The objective is to reduce energy costs (in this case directly related to
electricity consumption) while still achieving comparable production output.
Therefore different fields of action could be identified. Due to the nature of the
system, organisational approaches are less promising since the influence of both
PPC (continuous production of one product) and employees (highly automated
system) is very limited. Based on the first analysis the following improvement
measures will be further evaluated, which are more technical in nature and strive
to decrease the level of consumption while avoiding non-value creating losses in
idling times:
The blasting process is not fully utilised but demands significant power even
in idle mode. Therefore, a batch production with a lot size of 10 (instead of
the current one piece flow) and an automatic shutdown after each batch is
proposed (scenario B).
Since the die casting machines are not the bottleneck of production it shall be
analysed whether a production with just two die casting machines is
feasible. This would dramatically decrease the power demand of the
manufacturing system (scenario C).
Process optimisations allow a decrease of approx. 8% of the process time in
die casting. It shall be elaborated whether this actually has an effect from
system perspective. By applying this measure the power demand of the die
casting machine itself will not be reduced (scenario D).
To reduce non-valuable idle electricity consumption, all machines with
certain idle time (blasting, packaging, saw) are equipped with systems for
automated shutdown (scenario E).
Since it is technically possible, scenario F is a combination of scenarios B, C
and E.
Table 22 shows the results of the simulation runs relatively to the validated base
scenario. Thereby energy (electricity) consumption and production output are the
target variables. However, energy efficiency is used as a key figure for evaluation.
Additionally, based on 252 working days in three shift system and an assumed
electricity price of 0.10 /kWh the potential savings on electricity costs per year
are depicted.
150
Table 22 Results of simulation runs for aluminium die casting case

Scenario
A
C
D
Description
Base Scenario
Batch process for
blasting (lot size 10) and
automatic shutdown in
between
Just two die casting
machines are used.
Process time of die
casting machines
reduced by 8%.
Automatic shutdown for
Blasting/ Packaging/
Saw when idling.
Energy
consumption
Production
output
Energy
Efficiency
yearly electricity
cost savings
100%
100%
100%
92%
95%
103%
34.396,64
70%
89%
126%
127.294,12
100%
100%
100%
829,12
98%
100%
103%
10.642,33
Combination B+C+E.
62%
81%
130%
162.864,83
B*
Optimised batch
production for blasting.
93%
101%
109%
29.841,76
Based on these results the potential impact of certain measures is proved based
on the energy oriented simulation. For example, the reduction of the process time
in the die casting machine has limited impact which is logical since these
machines are not the bottleneck of production. Scenario E shows an improvement
potential of approx. 2% regarding energy consumption while keeping the same
production performance. Over the year this can total cost savings of over 10000.
As the implementation of these measures do not involve much effort in terms of
time and investment (adaption of machine control or automatic shutdown system)
this results in very attractive payback periods. The most interesting results are
certainly those from scenarios B and C.
The batch production can reduce electricity consumption by approx. 8% and
generate savings of almost 35000 per year. As a drawback the output of
production is also decreased by approx. 5% which might not be acceptable.
However, the lot size of 10 for this scenario was rather randomly selected to get an
impression of general potentials. With regard to the underlying potential, further
analyses were conducted. Figure 89 shows the results of parameter variation
simulation experiments in order to determine an optimal batch size for the blasting
process (range from 1 to 20 as technically maximal capacity).
These results show that a batch size of 10 is too large as it restricts production
flow. As the progression of energy consumption depicts, similar saving effects can
already be generated with a smaller batch size. An optimum is found at a batch
size of 4 where the production outcome is not restricted and a significant reduction
of energy consumption can be realised. This is further emphasised when
considering the energy efficiency value. As already shown in Table 22, applying
this scenario (labelled B*) would still lead to saving of almost 30000 per year
while keeping production output constant.
110%
151
optimum, scenario B*
105%
Energy consumption
Product output
Energy Efficiency
100%
95%
90%
base scenario A
scenario B
85%
80%
1 2 3 4 5 6 7 8 9 10 11 12 13 141516 17 18 19 20
lot size blasting process
Fig. 89 Results of parameter variation experiment for batch size of blasting process
Scenario C results in a very drastic reduction of energy consumption of about

30% compared to the base run. Also the energy efficiency ratio shows very
significant improvement. However, this scenario is a good example while just
relying on the efficiency ratio may lead to unsatisfying results. Further criteria
need to be considered and serve as necessary constrain in this case the reduction
of the production output of 11% is hardly acceptable in a mass production
environment (although this decrease is relatively lower compared to energy saving
potential). Nevertheless the scenario is very relevant to consider since it gives
clear indications about the immense leverage of this measure, which might be an
alternative if e.g. product demand is lower.
Altogether some promising means for reducing energy costs could be derived.
However, all these analyses are based on a deterministic model with e.g. machine
failures being disabled. While this shows the general functionality and potential
there is some uncertainty regarding the outcome when assuming more realistic
circumstances. Therefore extensive simulation studies were conducted which
consider realistic failure behaviour of machines and shall assure applicability and
potentials of the most promising scenarios B* and C. Therefore derived values for
MTTR and MTTF are used and assumed as to be normally distributed. The
relevance of stochastic effects transfers the former deterministic model into a
152
energy consumption
T-Test results
A-B*: p<0.0001
A-C: p<0.0001
B*-C: p<0.0001
54%
56%
58%
60%
62%
64%
66%
68%
70%
72%
74%
76%
78%
80%
82%
84%
86%
88%
90%
92%
94%
96%
98%
100%
102%
104%
106%
70%
60%
50%
40%
30%
20%
10%
0%
scenario A - avg. 100%, STD 1.7%

scenario B* - avg. 91.9%, STD 1.7%
scenario C - avg. 70.7%, STD 2.3%
scenario A - avg. 100%, STD. 1.7%

scenario B* - avg. 99.1%, STD. 2.0%
scenario C - avg. 77.4%, STD. 3.1%
production output
T-Test results
A-B*: p=0.02
A-C: p<0.0001
B*-C: p<0.0001
54%
56%
58%
60%
62%
64%
66%
68%
70%
72%
74%
76%
78%
80%
82%
84%
86%
88%
90%
92%
94%
96%
98%
100%
102%
104%
106%
frequency
70%
60%
50%
40%
30%
20%
10%
0%
frequency
probabilistic simulation model which needs a certain number of runs to be valid.

There is not a single value as a result but rather a distribution of values, which has
to be evaluated as a whole. Figure 90 shows the results of these simulation studies.
All values are put in relation to the average outcome of scenario A. The analysis
of energy consumption confirms the results of the deterministic runs and shows
that significant electricity costs savings are possible with scenario B* and C.
Statistical tests (e.g. t-Test) confirmed the expected effects of derived measures
while differences in value distribution proved to be statistically significant (all pvalues<0.05). Altogether, the probabilistic simulation runs are important to prove
that measures do have the intended effect while taking into account the
uncertainties in production. This can be clearly stated for scenario B* - the batch
wise processing is a promising measure to reduce energy costs while keeping
constant output. Compared to the existing situation, the output would be similar
and an energy consumption saving potential of 8% could be confirmed. Producing
with only two die casting machines (scenario C) definitely has an even higher
potential from an energetic perspective, which could also be confirmed. However,
probabilistic analyses even emphasise the major drawback of output reduction.
This measure should only be applied if a 20-30% reduction of production output is
acceptable in order to save energy costs. The range of possible values is also
broader compared to other scenarios which underlines the strong dependence from
failure behaviour and the strong uncertainty/risk connected with applying this
measure.
energy consumption/production output in % f rom scenario A (base run) average
Fig. 90 Results of probabilistic simulation runs
6.2 Weaving Mill
153
6.2 Weaving Mill

The next case study considers a small and medium (SME) sized company running
a very energy intensive and large scale weaving mill to produce technical textiles
that are being used for e.g. industrial purposes (e.g. supporting material for
abrasive papers, printing industry) (Herrmann et al., 2011a). The raw yarn is
running through pre-treatment phases like warping and sizing before the actual
weaving process is conducted on specialised weaving machines. As a last step an
extensive (partly automated) inspection is conducted to ensure high quality
products.
The company as a whole shall be considered; the objective is to minimise energy
costs while still keeping up the necessary production volume.
The analysis of the companys energy inputs reveals that electricity and oil are
necessary to run production. As Figure 91 shows, electricity is the main energy
cost driver. Due to continuous three shift production in main production areas, the
electrical power demand is relatively constant on a high level (dropdowns of
demand depict production pause on Sundays). Electricity costs are mainly
determined by consumption related costs and standard components (fees, tax);
peak surcharges play a minor role. Additionally, oil is needed for generating steam
in order to run the sizing process as pre-treatment before weaving. Also warm
water and space heating (of e.g. administration building) supply is ensured
through oil but this is significantly less relevant compared to the demand of
production.
Energy costs
electricity cost breakdown
electricity load profile for sample month
power [kW]
oil
10%
other
20%
peak
9%
consumption
71%
electricity
90%
time [min]
Fig. 91 Energy consumption analysis for weaving mill case

According to the application cycle, main energy consumers shall be identified
through applying the proposed energy portfolio. Since the consumers of oil are
154
known, the portfolio was only used for the case of electricity. Therefore the
nominal values for electrical power demand of different production areas and their
annual operating time were derived from technical documentation and production
data. Figure 92 shows the energy portfolio.
(relative) nominal value [in kW/kVA]
avg.
50%
compressors
45%
40%
35%
30%
weaving machines
25%
20%
air conditioning
15%
10%
avg.
5%
0%
0%
5%
10%
15%
20%
operating time per year [in hours]

Fig. 92 Prioritisation of electricity consumers for weaving mill case
The analysis reveals that compressors and weaving machines are very likely to
be by far the most relevant electricity consumers. They count up for approx. 75%
of total nominal power demand and approx. 80% of estimated electricity
consumption. Interestingly, both consumers are closely connected since the
electricity demand of compressors is majorly induced by the compressed air
demand of the weaving machines. The third largest electricity consumer is the air
conditioning system which is also mainly needed for the weaving area of the
company. However, on a yearly basis the consideration of the nominal power
demand and the operating time may be misleading in this case. In contrast to the
production itself, air conditioning is directly influenced by external climate and
therewith seasonal changes. Hence, compared to the value in the portfolio its
contribution to electricity consumption will be significantly lower in most times of
the year. Altogether, for detailed analyses of electricity consumption, compressors
and weaving machines will be considered due to their importance. All other
consumers are significantly less relevant and offer minor leverage to improve
energy costs.
In the next step the energy consumption of the weaving machines was measured.
As a result of the previous analysis it is necessary that both electricity and
6.2 Weaving Mill
155
compressed air consumption need to be considered. The weaving mill consists of

41 machines of four different types. Consequently those four different types of
machines were analysed separately with a sampling rate of one second.
Additionally, the influence of different speeds was determined through varying
process parameters. Sample results of those measurements are shown in Figure 93.
electrical power
8.000,00
compressed air demand

200,00
20000
150,00
15000
100,00
10000
50,00
5000
electrical power [W]
25000
compressed air [m/h]
effective electrical power [W]
7.000,00
100,00
y = 0,0335x + 57,566
R = 0,9848
6.000,00
90,00
80,00
compressed
air
electricity
5.000,00
70,00
60,00
4.000,00
50,00
y = 7,1272x - 105,69
R = 0,9894
3.000,00
2.000,00
time [sec]
0,00
30,00
20,00
1.000,00
10,00
0,00
0
40,00
200
400
600
800
1000
compressed air demand [m/h]
250,00
30000
0,00
1200
speed of weaving machine [1/min]
Fig. 93 Energy measurements and modelling of weaving machines
The necessary data for the compressed air system could be gathered from an
online monitoring system provided by the manufacturer of the system. Based on
measurements with a sampling rate of ten seconds it offers distinctive information
on e.g. electrical power demand and compressed air supply as well as switching
cycles of each single compressor of the system.
Step 5: Modelling
All weaving machines, the compressed air system and also the steam generation
were modelled with the developed energy oriented manufacturing system
simulation. The machine park of the weaving mill consists of a total of 41
weaving machines based on four different machine types operating independently
(no linkage, every machine with own production program) and almost
continuously in a three shift system. Based on an analysis of production data, each
one of the weaving machines was modelled as one process module, with its
specific characteristics on process time, failure behaviour and energy
consumption. The base unit which defines one product needs to be kept in mind
for modelling and later evaluation is one m of textile. As shown for one example
in Figure 93 for each type of weaving machine the demand on electricity and
compressed air could be transferred into equations. These consumption functions
reveal very good consistency with real values and allow the consideration of
energy demand with respect to the speed of the machine. Additionally, both
compressed air system with its seven large scale compressors and steam supply
system were modelled as separate TBS modules. The necessary information
regarding system structure (e.g. puffer tank size, number of compressors) and
control (e.g. start-up air pressures) was derived from technical documentation and
existing monitoring data.
156
Step 6: Validation
For validation purposes different criteria are used in order to compare results from
simulation runs with real data. The validation results are shown in Figure 94 (real
data=100%).
The production output rate was derived from production data. Machine
failures and setup times were excluded to ease comparability. Simulation
runs with similar conditions result in very similar results with a
consistency of >99%.
Regarding electrical power demand and electricity consumption no data
broken down for the considered system (weaving machines plus
compressors) was available. However, based on the aforementioned
estimated shares of these consumers on a companys total electricity
consumption (75% of power demand, 80% of consumption) and real total
values (from electricity bills) at least a rough check could be made. It
turns out that simulation results fit pretty well and show the right order of
magnitude.
For the compressed air system detailed information was gathered from the
online monitoring system. As base for validation, average values of selected
state variables for one month of production were taken. For all values a
sufficient consistency of measured and simulated results can be observed.
To validate the calculation of the steam generation module simulated
results were compared to real data regarding monthly oil demand (average
value over several years). Again, a very good fit of the simulation results
can be stated.
120,0%
consistency with actual data
100,0%
system performance and

energy consumption
99,5%
95,8%
99,0%
compressed air system state

variables
101,3%
103,8%
102,3%
switching
cycles
[1/h]
average
system
air pressure
[bar]
oil demand for

steam generation
102,3%
103,9%
80,0%
60,0%
40,0%
20,0%
0,0%
production
output
[m/h]
monthly
power
demand
electricity
consumption cons. system
[kW]
[kWh]
power
demand
compressors
[kW]
Fig. 94 Validation results for weaving mill case
Efficiency
Oil
compressed consumption
air
per month [kg]
[kW/m/min]
6.2 Weaving Mill
157
Altogether the simulation achieves very satisfying results for all the considered
criteria, which cover all necessary sub-systems of the company. Therewith, the
simulation model can be considered as validated.
elec. power [kW]
Figure 95 shows the results of the base simulation run as Sankey diagram, which
are automatically generated through the E!Sankey interface. The energy flows
depict the real situation in the company.
Heat
comp. air [m/h]
time [sec]
Heat
Electricity
Compressed air
steam [kg/h]
time [sec]
Steam
Oil
Electricity
time [sec]
Fig. 95 Simulated load curves and automatically generated Sankey diagram of simulated
energy flows (base run, in kW)
Significant amounts of electricity and oil serve as input for compressed air and
steam generation. As mentioned in Chapter 2.4.2, specifically the compressors
have a low efficiency, which leads to high heat losses. Only a small share of
inserted electricity ends up as energy in form of compressed air. The weaving
machines are the main consumer of compressed air; therewith they are directly
and indirectly majorly responsible for electricity consumption. Steam is consumed
in a pre-treatment process (sizing) just before weaving. However, the consumption
does not depend on the behaviour of the weaving machines. Based on the results
of the base simulation run (scenario A), the following fields of action shall be
considered towards improvement (Table 23).
As the measurements underline the demand on electricity and compressed air

is strongly depending on the speed of the weaving machines. Therefore, the
leverage of weaving machine speed regarding the influence on energy costs
shall be analysed for the system as a whole (scenario B and C).
158
A main driver of losses within compressed air systems are leakages. For the
considered case the impact of leakages shall be determined to investigate the
feasibility of actions towards leakage reduction. Pre-analysis show that a
reduction by 50% could be achievable with certain effort (scenario D).
The volume of the compressed air system is one of the determinants of system
behaviour. Therefore, the impact of increasing volume by adding puffer
tanks will be analysed (scenario E).
Static calculations show that a stable compressed air supply of production
could theoretically be achieved with only four compressors instead of seven.
This shall be analysed through applying the simulation approach (scenarios F,
G, H and I).
As mentioned above, a significant amount of heat is occurring as a result of
extensive compressed air demand. Since heat is needed to generate steam the
general potential of applying a heat recovery system will be analysed
(scenario J).
Table 23 shows the results of the simulation runs in relation to the base run. As
relevant target variables of energy consumption (in kWh, electricity plus oil), the
production output (in m) and the energy efficiency (m/kWh) are considered.
Additionally the potential annual energy cost savings are displayed (based on
production situation and market prices, which do not reflect actual price schemes
of the company). The simulation runs explicitly consider failures of the machines
which are - from a single machine perspective - an important influencing factor on
energy demand and production output. For the base scenario A extensive analyses
were conducted to evaluate the uncertainty of results against the background of
stochastic occurrence of failures. Those runs reveal that the resulting values for
the considered system are very robust against stochastic effects. The energy
consumption has a standard deviation of only 0.5% around the average value; the
production output ends up with a standard deviation of 1.3%. This can be
explained through the large number of machines (minor relevance of single
consumer), the continuous production situation (high share of value creating time)
with the independent behaviour of the weaving machines (no linkage, so failure
does not lead to failure of other machines). Thus, fluctuating failure behaviour of
single machines does have limited statistical influence on key target variables
from a system perspective. In further considerations, for all scenarios the same
basic data set and frame conditions compared to the base run are assumed (e.g.
failures will occur in a similar manner). Hence, comparisons are possible and also
statistically legitimate due to a high level of statistical certainty.
Scenarios B and C assume a general (all weaving machines) reduction
respectively increase of machine speed by 10%. Not surprisingly energy
consumption and production output are directly influenced. However, a 10% change
of speed led to just 2% change of energy consumption (direct and compressed air
induced) whereas the effect on production output is 9-11%. Motivated by the
outcomes of these scenarios more detailed runs were conducted. The results are
shown in Figure 96 and consolidate the results. The leverage on energy consumption
is relatively low while production output is heavily influenced. Thus, from an energy
6.2 Weaving Mill
159
Table 23 Results of simulation runs for weaving mill case

Scenario
A
B
C
D
E
F
G
H
I
Description
Base Scenario
Reducing production
speed by 10%.
Increasing production
speed by 10%.
Leakage is reduced by
50%
Increasing puffer tank
size. (doubled)
Just compressor 1-4
used.
Just compressor 1-4
used+doubled puffer
Just compressor 1-4
used+reduced leakage
Just compressor 1-5
used
Energy
consumption
Production
output
Energy
Efficiency
yearly energy
cost savings
100%
100,0%
100%
98%
91,0%
93%
15.247,87
102%
111,3%
109%
-18.081,79
98%
99,5%
102%
18.724,61
100%
99,6%
100%
2.965,25
86%
52,0%
60%
112.444,42
92%
74,7%
81%
64.903,68
96%
99,7%
104%
32.908,03
100%
99,4%
100%
2.592,00
Heat recovery
76%
100,0%
132%
73.345,54
Combination of
scenarios H and J
74%
99,7%
135%
99.479,81
energy consumption/produciton output/energy efficiency
220%
Energy consumption
200%
Production output
180%
Energy Efficiency
160%
approx. 80000 less
energy costs per
year
140%
120%
100%
80%
60%
40%
approx. 80000
higher energy costs
per year
20%
0%
0,5 0,6 0,7 0,8 0,9
1,1 1,2 1,3 1,4 1,5
relative machine speed (1=standard speed)
Fig. 96 Impact of changing speed of weaving machines
160
efficiency perspective operating as fast as possible is the most promising strategy.

However, it has to be mentioned that coherences of failures probabilities and
machine speed are not considered here due to lacking data. Significantly faster (or
also slower) production is likely to result in more failures. Still the impact on energy
costs per year is very considerable with an average of approx. 16000 per 10%
speed change. While certainly being no strategy in times of economic upturn (due to
less output), reducing machine speed could be a useful strategy for low utilisation
phases and is a way to bring down operating costs.
With the next scenarios a closer look on the highly relevant compressed air
system is taken. Scenario D reveals that reducing leaks bears a significant cost
reduction potential, which sets the limit for the necessary efforts in terms of time
and costs/investment to achieve those reductions. Increasing the puffer tank size
has limited effect on energy consumption and production output (scenario E).
However, analysing further technical variables shows that it reduces the number of
cycling switches of compressors by 50% which may extend their lifetime. Based
on those results further investment calculations regarding the feasibility of
integrating additional tanks are conducted.
Having in mind the measured and simulated compressed air demand of
weaving machines it seems possible to significantly reduce the necessary number
of compressors. This would bear an enormous saving potential in terms of energy
and other operating costs and investment. While this would theoretically cover
weaving machines demand, in the first step, the operation of only the largest four
of the existing compressors was assumed (scenario F). This option turns out to be
not feasible for practical application with the given circumstances. The simulation
reveals that the necessary air pressure of the system would drop down too much,
causing failures of the weaving machines and enormous output losses. With the
simulated time depending consumption behaviour and assumed leaks in the
system, four compressors are not sufficient for a stable production. To ease the
problem of dynamically fluctuating demand with some peaks, the puffer was
doubled in the next step (scenario G). This can in fact improve the situation
however, the result in terms of production output is still not acceptable. While
leakage of the system is evidently a significant factor, similar to scenario D a
reduction of 50% was assumed in scenario H. In fact, this reduction in constant
compressed air demand enables a stable production with just four compressors.
This gives further incentives to focus on reducing leaks in the compressed air
system. Without reducing leaks or increasing puffer size a stable production is
possible in any case with five compressors. However, from an energy saving
perspective, just minor effects can be expected, which underlines that the
additional two compressors of the existing system (and base run) are rarely
operating anyway. Still this information is very important for further investment
decisions. Finally, the base run reveals the immense heat potential, which is
available due to compressor operation while the neighbouring boiler needs heat to
generate steam. Therefore, the potential of applying heat recovery was analysed
(scenario J). With a given efficiency (oriented at state of the art heat exchangers)
the exhaust heat of the compressors can serve as energy input for steam
generation. As the results show a significant reduction of additional energy
6.3 PCB Assembly
161
demand can be achieved. In contrast to the other scenarios the energy cost
reduction is a result of less oil needed. While this is just a rough estimation and
detailed technical analyses are necessary, cost savings of >70000 are possible,
which justifies the investment in necessary heat recovery technology. Finally,
scenario K combines the promising scenarios H (only four compressors used with
50% reduction of leakage losses) and J (heat recovery) to evaluate the overall
impact. This is a very good example that underlines the necessity of a holistic and
simulation based evaluation of measures to be able to analyse the impact of their
interactions. In this case both measures counteract to some extent. Still the saving
potential is very promising even though it is not the sum of the possible savings of
the isolated consideration of those scenarios: the reduction of compressor power
demand directly leads to significant energy savings. However, also less waste heat
as source for heat recovery is available. Thus, less oil can be saved compared to
the original heat recovery scenario J.
6.3 PCB Assembly

The third case study considers a SME company which assembles printed circuit
boards in a very flexible production environment (no coupling of machines, free
flow of orders). Main processes are the mounting of parts on PCBs with SMD
(surface-mounted device) or THT (through-hole technology) techniques, which
involve reflow soldering machines/ovens (SMD) and solder waves (THT).
The company as a whole shall be considered; the objective is to minimise energy
costs while still keeping up the necessary production volume.
In this case the only externally acquired form of energy is electricity. Figure 97
shows an analysis of the daily electrical power demand as measured by the
electricity supplier with the usual sampling rate of 15 min.
It is a statistical analysis from 60 working days meaning that the value of each
depicted 15min period is a result of 60 single values (e.g. average power demand
at 10 a.m. is the average of 60 different days at this point of time). It is important
to note in this context that this analysis can just provide general tendencies
regarding consumption behaviour the company consists of a highly flexible
production environment so basically each day can look quite unique. The
consumption pattern of a typical working day is not as even as the average curve.
Also minimum and maximum values are not representative for a specific day but
provide a corridor to describe the most likely power demand pattern. This corridor
can be narrowed down, by considering the average plus respectively, minus
standard deviation per definition this range contains 68.2% of all values. In
general it can be stated that coming from an average base load of around 4 kW,
production typically starts at around 6 a.m. with switching on diverse machines,
162
60
AVG
AVG +
50
AVG -
MAX
power
[kW]
electrical
load
in kW
40
MIN
30
20
10
23:15
22:30
21:45
21:00
20:15
19:30
18:45
18:00
17:15
16:30
15:45
15:00
14:15
13:30
12:45
12:00
11:15
10:30
09:45
09:00
08:15
07:30
06:45
06:00
05:15
04:30
03:45
03:00
02:15
01:30
00:45
00:00
time [min]
Fig. 97 Electrical power demand of PCB assembling company
which leads to a peak demand on electrical power. After 2-3 hours, consumption
is relatively constant on a certain level. After 4 p.m. a significant dropdown of
power demand can be observed which is related to the typical daily shift ending.
Sometimes selected production machines are operating longer to cover necessary
production volumes.
In the specific case of this company electricity, costs are directly related to
consumption with one constant price rate for the whole day. There are no peak
surcharges.
Figure 98 shows the energy portfolio of the considered company with respect to
nominal electrical load and estimated operating time per year. It can be observed
that electricity consumption is heavily determined by four main consumers which
sum up to over 87% of the estimated yearly consumption: two solder waves, one
reflow oven and the compressor for compressed air generation. From an energy
point of view all other consumers are of minor relevance.
Based on the outcomes of the previous step, measurements of electricity
consumption were conducted for diverse consumers. Figure 99 shows an example
measurement for the reflow oven (sampling rate of one second). The main
consumers involve extensive time and related electricity consumption for heating
up, which is necessary prior to any value creating activity at each working day.
Based on measured data, a comprehensive check was conducted in order to
6.3 PCB Assembly
163
confirm results of the aforementioned portfolio and ensure that all main
consumers were actually considered. The results of this check are also displayed
in Figure 99. To ensure the comparability the maximum load within a 15 minute
interval as base for electricity consumption measurements done by the supplier was calculated for each consumer (based on measurements with sampling rate of
just one second). The comparison of these values with the power demand of the
company as a whole (Figure 97) confirms the previous considerations. The typical
demand (average and standard deviation range) can be well explained through
these consumers; the higher absolute maximum is a one-time value and is most
likely the result of incidentally switching on the second reflow oven.
avg.
35
(relative) nominal value [kW/kVA]
reflow oven
30
solder wave 2
alternative
reflow oven
(not in use)
25
solder wave 1
20
15
avg.
10
machines for PCB mounting

and related processes
compressor
(4.000-5.000h)
0
0
500
1.000
1.500
2.000
2.500
operating time per year [hours]
Fig. 98 Energy portfolio of PCB assembling company
Step 5: Modelling
In contrast to the other case studies, the generic modelling approach was used for
this case. As presented in Chapter 5.3.1 and even more specifically in Chapter 5.3.5
it is possible to setup the whole model via parameterisation of generic process (and
also TBS) modules outside the actual simulation environment in a standard MS
Excel datasheet (PPC module). This addresses entry obstacles and eases
utilisation of simulation capabilities without having the necessary detailed expertise.
The considered company is a typical SME which does not have the necessary
experts and financial possibilities to run complex simulation tools (whereas
commercial energy oriented simulation approaches are not available anyway as
shown above).
All necessary data was keyed in the PPC module. Thereby, all relevant
consumers were modelled according to the proposed procedure depicted in Figure
83. Thereby it is necessary to mention that - in accordance to the proposed
procedure - certain machines were modelled as separate process module although
164
they are per se not relevant from energy perspective. However, these machines are
crucial to consider as they determine the material flow within the company which
again influences the operation of main energy consumers. In total, seven different
production machines were modelled (reflow oven, two solder waves, two devices
for PCB mounting, automated optical quality inspection, machine for automatic
application of soldering paste). Based on the data analysis an electricity base load
was modelled subsuming minor consumers, which do not need to be considered
separately (according to Figure 83). As mentioned before production machines are
not technically linked with each other and operate independently. The material
flow is determined by production planning of orders that run through the company
in flexible process chains utilising the given production equipment e.g. in diverse
sequences and lot sizes. Based on realistic circumstances, certain general process
chains were modelled via the PPC module for three different types of orders
(products) and a typical production program for a working day (involved
machines and sequence for each order, lot sizes, starting time) was developed.
Supplementing the production equipment and management, the compressor was
modelled as TBS module based on its technical specifications. Compressed air is
mainly used by PCB mounting devices. Their compressed air demand was
estimated based on nominal values.
30000
Power [W]
Setup/Heating up
Producing (continuous)
25000
20000
15000
10000
5000
09:03:49
09:07:35
09:11:21
09:15:07
09:18:53
09:22:39
09:26:25
09:30:11
09:33:57
09:37:43
09:41:29
09:45:15
09:49:01
09:52:47
09:56:33
10:00:19
10:04:05
10:07:51
10:11:37
10:15:23
10:19:09
10:22:55
10:26:41
10:30:27
10:34:13
time [sec]
45
40
35
30
25
20
15
10
5
0
kW
PCB mounting
compressor
solder waves
reflow ofen
electr. power demand

[15min interval]
Fig. 99 Example measurement result of reflow oven and cumulated maximum power
demand in 15 minute interval for main consumers
Step 6: Validation
Due to the aforementioned flexible production environment in combination with
lacking production data the validation is conducted through assessing the energy
related behaviour of the system. If this turns out to be realistic the production
performance as a whole can also be considered as sufficiently realistic since
product individual setup and process times were actually measured. Figure 100
shows the simulated electrical load profile (calculated with 15 minute interval) for
the validation run, which considers a typical working day with production start at
6a.m. and consequent shift end at 4p.m. A comparison with the given real
electrical load profile intuitively shows the very similar consumption behaviour
over the day. This can also be expressed quantitatively: with the given production
scenario the validation run results in an energy consumption of 363 kWh over the
6.3 PCB Assembly
165
considered period of time. For the same period the analysis of real data reveals an
average consumption of 312 kWh with a standard deviation of 51 kWh (261kW363kWh). Therewith the simulated consumption reflects the real consumption
behaviour while even being in the range of the most likely values. An alternative
scenario (B, later explained in detail) with a slightly modified machine start in the
morning results in a less distinctive power peak while leading to a consumption of
357 kWh. Altogether, it could be proved that the simulation model is capable to
describe the consumption pattern of the PCB company, with sufficient accuracy
and can be considered as validated.
50
45
scenario A
scenario B
solder waves
(2 machines)
21%
40
power [kW]
35
30
44%
25
PCB mounting
(3 machines)
reflow oven
20
15
25%
10
compressor
5
0
time [min]
10%
Fig. 100 Simulated electrical load profile for PCB case (second based values converted to
15min interval) and consumption composition for scenario A (base scenario)

The simulation run for validation purposes helps to understand the energy
consumption dynamics of the company. As a result of this base run, Figure 100
shows the composition of energy consumers. The distinctive power peak in the
morning is a result of turning on and heating up soldering waves and the reflow
oven. Similar to the real load profile there is also a dropdown observable after the
reflow oven reaches the necessary temperature for operation. There is again a
slight increase when the soldering waves start operation. In the base scenario all
machines are on for the whole shift and are turned off at shift end (4p.m.), which
seems to be a realistic assumption when comparing to real values. However, they
are not necessarily completely utilised for value creation due to involved heating
periods and waiting for upstream supply of parts (utilisation rate of 10-20%).
Against the background of these observations, different scenarios are derived in
order to improve the energy cost situation of the company. Besides technical
measures organisational aspects are of specific interest here due to the flexible
production environment and possible influence of machine operators.
Additionally, a closer look shall be taken on the energy contract situation of the
company. An overview of the considered scenarios and respective results is given
in Table 24. All scenarios were simulated with a deterministic model while
uncertainty is a minor issue in this case (very high availability of machines, clearly
determinable process times).
166
Scenarios B-E focuses on alternative strategies for production planning and

control (PPC) the activities of the manufacturing system. As described above
within the base run, production machines are turned on and off at shift start
and end independently from their utilisation. Scenarios B-E aim at a more
efficient operation of machines with reducing waste energy in idling times
through purpose oriented machine start and e.g. more batched production. In
scenario B each machine is firstly turned on (including heating) when the first
part from the upstream process is arriving. As further extension scenarios C-E
introduce an energy oriented production program (capacity) planning with
distinctive starting times for the critical machines while aiming at still
providing the necessary output.
The generation of compressed air also has a relevance regarding the energy
costs of the company. Since the existing compressor is dimensioned relatively
large for the necessary compressed air, in scenario F it was analysed whether
an alternative compressor with lower air supply, but also lower power
demand could keep up production and save energy costs.
As described before, the structure of the current electricity contract of the
company is rather simple and only includes a single consumption related price
rate independent from daytime. In scenario G it shall be investigated whether
an alternative contract setup could lead to energy cost savings. Therefore
energy costs of the base scenario were calculated with the contract from a
comparable SME company. This contract consists of e.g. peak power
surcharges, which allow a lower price rate for consumption.
Table 24 Simulation results overview for PCB company case

Scenario
Description
Base Scenario
Machines are
turned on if first
upstream part
available (PPC1).
Energy oriented
production
planning (PPC2)
Energy oriented
production
planning (PPC3)
Energy oriented
production
planning (PPC4)
alternative
compressor
alternative
electricity supply
contract
Energy
consumption
Energy
costs
Production
output
Efficiency
100%
100%
100%
100%
98%
98%
100%
102%
60%
60%
77%
128%
68%
68%
100%
148%
66%
66%
100%
151%
98%
98%
100%
102%
100%
119%
100%
100%
6.3 PCB Assembly
167
power [kW]
power [kW]
The simulation results reveal the significant potential of energy sensitive

production planning and control. For the considered case some PPC strategies
offer saving potentials of 30-40% compared to the base scenario. As an example,
Figure 101 shows the load profile of the base scenario A and scenario E. However,
there is evidently also the possibility of negative impact on production output
(scenario C), which is usually not an option in industrial practice. Additionally,
these measures require the necessary flexibility with a certain degree of freedom
in daily operation. This is not given in high volume production with rigid linkage
like e.g. in case study one (die casting). However, in the considered case there is
this degree of freedom. Production orders do not require the full daily capacity of
available machines and operators are flexible to be appointed (can be involved in
other activities within the company, flexible time contracts). Against this
background scenario E is a promising way to gain energy efficiency. Instead of
continuous running but often idling energy intensive production machines over the
whole day, certain production volumes are batched and processed in the afternoon.
The simulation confirms that the same production output can be achieved.
Operators are free to be appointed to other processes in the morning or can start
their shift later.
time [sec]
time [sec]
Fig. 101 Selected simulated electrical load profiles
Certainly, the scenarios only cover a sample day with selected production
orders and involved process chains; the effect of those measures needs to be
verified for differing production scenarios and in industrial practice. However, the
general potential is promising and the achievable order of magnitude encourages
the application of energy sensitive planning and control in the company. Since just
selected scenarios were considered, through a more systematic design of measures
even more saving potential seems to be possible for certain cases. The ultimate
goal should be to include those considerations in daily planning procedures, e.g.
within ERP systems.
As a more technical measure, an alternative compressor was used for
compressed air generation (scenario F). The simulation proved that the smaller
sized compressor is still able to sufficiently supply the manufacturing system
while leading to some savings in energy consumption.
168
Finally, the alternative contract model is not appropriate when being applied on
base scenario A. While reducing the consumption costs, the influence of the power
induced cost component is highly significant and weighs more. Still it could be
proven that alternative contract models can be included in the simulation approach
and may have significant impact on occurring energy costs.
6.4 Application in Education of Production Engineers

Besides the application in industry towards an improvement of energy efficiency,
the proposed simulation approach was already used in the education of engineers.
Since autumn 2009 the Product- and Life-Cycle-Management Research Group
within the Institute of Machine Tools and Production Technology (IWF) of TU
Braunschweig offers an innovative lecture named "Sustainability in Production
Engineering" (Herrmann et al., 2010b). Therewith students of mechanical and
industrial engineering shall be sensitised and enabled to handle issues like energy
efficiency in manufacturing. From a longer term perspective this is a very
promising approach since these students will be the relevant experts to plan,
evaluate and decide on those issues in companies later on. Thus, there is a strong
leverage by motivating them and providing necessary methods and tools. The
lecture pursues a practical approach: besides the theoretical background provided
in the lecture itself, laboratory classes and further tutorials are offered. In groups
of approx. five people, the students independently solve a comprehensive case
study, which aims at increasing the energy efficiency of a defined machining
process chain. They practically need to apply knowledge and experience related
challenges.
The task of the case study is to produce a shaft with defined geometry as energy
efficient as possible. In the first step the students take a closer look at necessary
processes. For achieving the necessary final diameter and surface quality (fine
machining) as the main production task, with turning or grinding two process
alternatives are available. Within the laboratory environment of the institute, the
students conduct themselves measurements on machines to assess process times
and electrical power demand. With these results process recommendations for the
most energy efficient way of manufacturing can be given. In the next step, the
process chain as a whole is considered which includes several working steps
rough turning, hardening, fine machining and cleaning. The task is to produce 100
shafts as energy efficient as possible. For assessing the technical performance and
energy consumption of the process chain, the proposed energy oriented simulation
approach is used. For facilitating access in terms of training time and necessary
software installations, the possibility of AnyLogic of generating an executable
Java applet was used. Therewith it is possible to provide a generic model of the
process chain which can be run in any web browser. The model with its graphical
user interface is depicted in Figure 102.
6.4 Application in Education of Production Engineers

start simulation stop/restart
share of parts running via Machine 1

(0 = Machine is off!)
169
input field for measuring results
share of parts running via Machine 1..n

(0 = Machine is off!)
lot size of
hardening process
lot size of
cleaning process
evaluation
Fig. 102 Screenshot of Java-applet for energy oriented manufacturing system simulation for
educational purposes
It provides the general framework of the considered process chain while

leaving sufficient freedom to test alternative means to structure and control the
system.
Number and type (turning, grinding) of processes and the allocation of

production volume shares to single machines can be changed.
Measured data for processes can be integrated; therewith the effect of process
alternatives can be analysed from process chain perspective.
Lot sizes for hardening and cleaning can be changed.
As formerly mentioned the objective is to produce 100 shafts as energy efficient

as possible (in accordance with the measurements, only electricity is considered
here). The presented influencing options lead to significantly different results in
terms of energy consumption and production output, which need to be balanced.
Therewith the simulation model can strongly support the comprehension of
interactions in the process chain and possible conflicts of goals. It was
successfully applied in the education of engineering students and the feedback was
very positive. In future the simulation based case study will be developed further
and continuously used for teaching.
Chapter 7
Summary and Outlook
In the final chapter, the contents in general and the research outcomes in particular
will be briefly summarised. Additionally, an evaluation of the developed concept will
be conducted and an outlook on further possible work will be given.
7.1 Summary
This book aims at contributing towards improving energy efficiency in manufacturing
through developing a generic energy flow oriented manufacturing simulation
environment. The environmental and economic motivation is outlined in Chapter 1,
which also stresses the necessary consideration of all dimensions of sustainability as
new paradigm in manufacturing.
In Chapter 2 the relevant theoretical background based on the known state of
the art is described. First of all, necessary definitions and basics in the field of
production and energy (supply) are given. Afterwards both aspects are brought
together when describing the background of energy consumption and energy
efficiency in manufacturing. It is pointed out that the topic in general is certainly
not new and good basic knowledge exists in literature. However, the reasoning
also underlines that implementation in industry is still unsatisfying and worthwhile
potentials can still be tapped. The description of the theoretical background also
stresses the diversity of questions, challenges and disciplines which are connected
with the topic of energy efficiency in manufacturing.
Taking this into account, Chapter 3 considers the derivation of general
requirements from both scientific/technical as well as industrial/business
perspective, which should be addressed to successfully foster energy efficiency in
manufacturing. Thereby the discussion reveals that considering the interacting
dynamic consumption profiles of all relevant energy flows for the factory system
as a whole is crucial for the evaluation of improvement measures on both single
machine as well as factory level. To assess those consumption profiles with static
calculations, fuzzy logic, artificial neural networks and manufacturing system
simulation, different methodological approaches are available. The comparison
with the requirements clearly shows that manufacturing system simulation is the
most appropriate methodological approach for the considered research task.
springerlink.com
172
7 Summary and Outlook
Whereas not available in industrial practice, in research, first approaches

towards energy oriented manufacturing system simulation were developed. The
requirements were further concretised into specific criteria, which serve as a basis
for the evaluation of the state of research in Chapter 4. It turns out that there is
significant room for improvement in all areas and there is no approach with a
balanced and high fulfilment of all criteria.
This evaluation shows the necessary research demand, which sets the frame
conditions and necessary specifications of the concept development being
conducted in Chapter 5. The core element is an innovative energy oriented
manufacturing simulation approach. It is a one-stop solution with modular and
flexible structure that aims at facilitating access to simulation capabilities for
broad fields of application and provides decision support towards a systematic
improvement of energy efficiency. Thereby it is explicitly able to consider all
relevant forms of energy needed by production equipment or technical building
services as well as their interactions. The developed simulation environment is
embedded within an application cycle, which supports the goal-oriented
implementation for the derivation and evaluation of improvement measures. Even
without usage of the simulation it is a helpful methodology to identify and
prioritise main energy consumers in manufacturing companies.
Finally, in Chapter 6 (concept application) the developed simulation based
concept as a whole is applied to a couple of very different case studies in order to
demonstrate the broad applicability and potentials. The case studies show that the
concept is working for diverse practical applications (large and SME companies of
different branches, engineering education) and underlines the diversity of energy
efficiency measures which can be evaluated on a realistic base.
7.2 Concept Evaluation

Finally, the developed simulation based concept for improving the energy
efficiency in manufacturing shall be evaluated. To support a structured and
objective assessment the same criteria as for evaluating the state of research are
used (Chapter 4.1). The results are shown in Table 25.
The evaluation shows that all criteria could be addressed and improvement
towards the state of research could be made in all categories. The average
fulfilment of criteria is 0.88 (88%) which is by far the highest value for all
comparable approaches. This is certainly not too surprising since these criteria
served as requirements for the development of the proposed energy oriented
simulation approach. However, it shows that these identified critical aspects
could be successfully implemented. Figure 103 shows a more detailed
comparison with the state of research and underlines that significant advances are
made in all areas.
7.2 Concept Evaluation
173
Table 25 Evaluation of proposed simulation approach
Main
Author(s)
Institutions
References
Software
tools
Sebastian Thiede
TU Braunschweig, Institute of Machine Tools
and Production Technology, Product- and
Life-Cycle-Management Research
Group, Braunschweig, Germany.
(Herrmann and Thiede, 2009a, Thiede and Herrmann,
2010, Herrmann and Thiede, 2009b, Herrmann and
Thiede, 2008; Herrmann et al., 2011a)
AnyLogic, MS Excel (parameter input, extended
evaluation), E!Sankey (Energy flow visualisation)
Completeness
Dynamics
TBS
all external energy sources and internal flows can be

considered
state based consumption of all energy flows,
interaction with two TBS systems already
implemented, extension possible
Fields of action
Technological
Organisational
Optimisation
basically all technical fields of action can be addressed,

also from integrated perspective
diverse PPS and employee behaviour related aspects
can be addressed
Optimisation studies are possible through OptQuestsolver; however, not main focus and just exemplarily
conducted
Evaluation
Economic
costs of several energy carriers with realistic contract

based cost model for electricity, no extended cost
model with other cost portions
Ecological
conversion of different energy flows to GHG emissions
Technical
output/production rate and further key figures (e.g.

utilisation) available
174
diverse key figures and methods for integrated

evaluation are introduced, from scientific perspective
extension towards MCDM possible
is addressed and appropriate methods for coping with
uncertainty are presented
Decision
Support
Uncertainty
Implementation
highly transferable to diverse production cases, eased
access also for non-expert
little expertise necessary, fast modelling possible,
possible export function as java applet, simulation time
depending on actual computer - complex models might
involve significant computing time
material flow/machine states and all results
continuously displayed during runtime
comprehensive application cycle provide which guides
user through usage towards systematic improvement
Transferability
Effort
Visualisation
Application
Cycle
Fields of
action
Energy and
Resource
Flows
0,00
0,10
0,20
0,30
0,40
0,50
0,60
0,70
0,80
0,90
Completeness
Dynamics
TBS
Technological
Organisational
Optimisation
Optimization
Evaluation
Economic
Ecological
Technical
Decision Support
Implementation
Uncertainty
Transferability
Effort
Visualisation
Application Cycle
average
Fig. 103 Comparison of proposed simulation based concept with state of research
1,00
7.3 Outlook
175
The comparison reveals the success factors of the proposed simulation

approach compared to the state of research which are:
The consideration of the manufacturing system as a whole with all relevant

energy flows and their time dependent dynamics.
The applicability for a broad range of production cases (including flexible
manufacturing systems) of different branches with relatively low effort and
necessary expertise.
The integrative derivation and evaluation of diverse improvement measures
from different fields of action.
The provision of clear decision support through meaningful key figures and
diagrams while also addressing the issue of uncertainty.
The integration within a comprehensive application cycle, which guides also
non-experienced users through a systematic improvement process while using
the simulation approach.
The simulation based concept was developed and practically implemented;

applicability and potentials were successfully proved through application to
several very different case studies.
7.3 Outlook
Although significant advances in comparison to the state of research could be
made there are diverse opportunities for future research, which would extend
functionalities of the developed approach and support embedment in industrial
business processes.
Extension and addition of TBS modules: Within this book the development
of compressed air and steam generation as TBS modules is presented, because
they are not only of main relevance in general but also particularly for the
case studies. These modules were verified and can be easily parameterised in
order to depict any supply systems in manufacturing companies. However,
further generic TBS modules are worth to be developed. As one example air
conditioning systems are an interesting field of action while being relevant for
many companies and often a major consumer of energy. Also the integration
of electricity supply systems is a promising approach. Being neglected so far,
the local generation of electricity certainly bears potential to foster
sustainability in manufacturing specifically when it comes to renewable
energy sources (e.g. solar or wind power). Through simulation of electricity
demand of production the developed approach can strongly support
dimensioning and control of those systems. Again, this is specifically of
interest when different forms of energy act together, e.g. in combined heat
and power generation (CHP). Additionally, further TBS modules are certainly
worth to be developed for decentralized cooling respectively heating devices
for production machines (e.g. cooling for laser cells, process heat for casting).
These devices are widely distributed in manufacturing companies and so the
provision of generic TBS module would be very helpful.
176
Consideration of further resource flows: This book focuses on the

development of an energy flow oriented simulation of manufacturing systems.
The consumption and emission of other resources (e.g. auxiliary materials,
waste streams) is only partly considered (as long it is relevant for energy
related considerations). Technically an implementation is not too difficult due
to the flexible structure of the provided simulation environment. Some of
those variables are already integrated in e.g. process modules but for
reasons of comprehensibility and the actual definition of the designated
research task were not put in the main focus of the considerations here.
However, in further research, consumption and emission of additional
resources should be integrated while this provides a further perspective on
sustainability in manufacturing.
Extended cost models: The embedded cost evaluation concentrates on the
realistic evaluation of energy costs. Further cost portions are not explicitly in
the simulation itself and need to be calculated separately. In fact it would
make sense to integrate those issues in future. Through the extension with
other resource flows (see above) and worker operations, a comprehensive
calculation of total operating costs is possible. Even more, through
integrating more basic data (on e.g. necessary investment, financial conditions
or maintenance strategies) even a calculation of life cycle costs is basically
thinkable. Altogether extended cost models would support the decisions
towards energy efficiency measures but also broaden the perspective to
evaluate any kind of changes in the manufacturing system.
Extension of environmental evaluation towards Life Cycle Assessment
(LCA): While up to now an energy oriented conversion into GHG emissions
is provided for environmental evaluation, an extension with other relevant
resource consumption and emission patterns would enable a more
comprehensive assessment of environmental impact. Through interfaces with
life cycle inventory databases (e.g. EcoInvent) even an automatic life cycle
assessment (LCA) as ultimate goal is possible.
Integration of complex operations research (OR) and statistical methods:
Methods for optimisation, multi-criteria decision making and evaluation
under uncertainty are an integral part of operations research as business
related research discipline. Very complex methods are available in this
context and can be (manually) applied in context of the data which is
generated by the developed simulation environment. However, in this book
mostly simplified approaches are used in order to facilitate practical
application. To improve the quality of decisions, the integration of proper
methods (e.g. optimisation algorithms for optimal production capacity or lot
size planning, guided PROMETHEE evaluation, automated statistical
treatment of stochastic simulation data) as an inherent part into the simulation
is a promising approach. This would enable the usage of those methods also
for users who are not experts in operations research and statistics.
Integration in industrial data environment: Up to now the simulation
based concept is based on data gathered manually from other sources (e.g.
production data acquisition systems, maintenance data) respectively with one
7.3 Outlook
177
time measurements on technical equipment (e.g. energy consumption).

Without a doubt a continuous data exchange through embedment of the
simulation approach into the company`s data environment would bear
potential for improvement. A coupling with e.g. applications for energy
monitoring and/or production data acquisition could provide the simulation
with the latest data and improve the consistency with the real manufacturing
system. With an interface with ERP (enterprise resource planning) systems
e.g. alternative scenarios for the planning of production capacities could be
automatically evaluated with the simulation in order to provide decision
support. Thinking even further, coupling with machine control might be
beneficial as from system perspective suggestions for ideal modes of
operation can be given or even automatically triggered.
Coupling with process simulation: For most manufacturing processes specific
process simulation applications are available which allow the detailed analysis
on a physical level (e.g. FEM simulation in machining or die casting). This is
not the focus of the proposed simulation approach but a coupled consideration
with those applications is technically possible and might be advantageous (e.g.
Herrmann et. al. 2011b). Whereas the manufacturing system simulation
considers the process as black box with inputs and outputs, with a detailed
physical model the actual impact of certain variables determined on system
level can be realistically assessed and useful information can be generated. As
an example, fluctuating process times or volatile supply with heat might
influence the quality of the product.
References
Abernethy, R. B. (2006): The new Weibull handbook - Reliability & statistical analysis for
predicting life, safety risk, support costs failures and forecasting warranty claims. 5. ed.,
North Palm Beach, Fla.: R.B. Abernethy.
Anderson, D. R. (2002): Statistics for business and economics. 8. ed., Cincinnati, Ohio:
South-Western.
Andreassi, L.; Ciminelli, M. V.; Feola, M.; Ubertini, S. (2009): Innovative method for
energy management: Modelling and optimal operation of energy systems. Energy and
Buildings, 41/4, pp. 436444.
Arnold, D. (2002): Handbuch Logistik - mit 77 Tabellen. VDI-Buch, Berlin: Springer.
Babcock and Wilcox Company (2010): Steam Its Generation and Use. Kessinger
Publishing.
Banks, J. (2010): Discrete-event system simulation. 5. ed., internat. ed., Upper Saddle
River, NJ: Pearson.
Barbian, P. (2005): Produktionsstrategie im Produktlebenszyklus - Konzept zur
systematischen Umsetzung durch Produktionsprojekte. Dissertation, Techn. Univ.
Kaiserslautern.
Bayer, J. (Ed.) (2003): Simulation in der Automobilproduktion. Berlin: Springer.
Bertsche, B. ; Lechner, G. (2004): Zuverlssigkeit im Fahrzeug- und Maschinenbau Ermittlung von Bauteil- und System-Zuverlssigkeiten. 3. ed., Berlin, Heidelberg:
Springer.
Beyene, A. (2005): Energy Efficiency and Industrial Classification. Energy Engineering
Journal, 102/2, pp. 5980.
Bickford, D.; Leong, K.; Ward, P. (1996): Configurations of Manufacturing Strategy,
Business Strategy, Environment and Structure. Journal of Management, 22/4, pp. 597
626.
Bierbaum, U.; Htter, J. (2004): Druckluft-Kompendium. 6. ed., Darmstadt: Hoppenstedt
Bonnier Zeitschriften.
Binding, H. J. (1988): Grundlagen zur systematischen Reduzierung des Energie- und
Materialeinsatzes (Bases for the systematic reduction of the use of energy and material).
Dissertation, RWTH Aachen.
Birolini, A. (2010): Reliability Engineering - Theory and Practice. Berlin, Heidelberg:
Springer.
Black, K. (2008): Business statistics - For contemporary decision making. 5. ed., Hoboken,
NJ: Wiley.
BMWi (2011): German Federal Ministry of Economics and Technology (BMWi) - Energy
Statistics, Available online: www.bmwi.de.
180
References
Bode, H.-O. (2007): Einfluss einer energieeffizienten Produktion auf Planungs- und
Produktprmissen am Beispiel der Motorenfertigung. In: Fraunhofer IPK (Eds.): XII.
Internationales Produktionstechnisches Kolloquium, Berlin, pp. 299305.
Bge, A. (Ed.) (2009): Handbuch Maschinenbau - Grundlagen und Anwendungen der
Maschinenbau-Technik. Wiesbaden: Vieweg+Teubner, available online
http://dx.doi.org/10.1007/978-3-8348-9249-2.
Bonneschky, A. (2002): Energiekennzahlen in PPS-Systemen, Diss. u.d.T.: Bonneschky,
A.: Integration energiewirtschaftlicher Aspekte in Systeme der Produktionsplanung und
steuerung. Dissertation, Technische Universitt Cottbus.
Boos, W.; Kuhlmann, K. (2010): Bewertung der Ressourceneffizienz von
Werkzeugmaschinen. wt Werkstattstechnik, 100/5, pp. 350353.
Borshchev, A.; Filippov, A. (2004): From system dynamics and discrete event to practical
agent based modeling: Reasons, techniques, tools. In: Proceedings of the 22nd
International Conference of the System Dynamics Society, Oxford, UK.
Brans, J.; Mareschal, B.; Vincke, P. (1986): How to select and how to rank projects: The
PROMETHEE method for MCDM. International Journal of Operations Research, 24,
pp. 228238.
Brettar, T. (1988): Konzeption einer betrieblichen Energiewirtschaft. Dissertation,
Universitt Saarbrcken, Betriebswirtschaftliche Beitrge zu Energie, Rohstoff und
Umweltfragen Nr. 5, Frankfurt am Main: Lang.
Brigham, E. F.; Houston, J. F. (2009): Fundamentals of financial management. 12. ed.,
Mason, Ohio: South-Western Cengage Learning.
Brggemann, A. (2005): KFW-Befragung zu den Hemmnissen und Erfolgsfaktoren von
Energieeffizienz in Unternehmen. Publikation der Volkswirtschaftlichen Abteilung,
Frankfurt, Main: KfW.
Brundtland Commission (1987): Our common future. Oxford paperbacks, Oxford: Oxford
University Press.
C. I. R. P. (2004a): Wrterbuch der Fertigungstechnik / Dictionary of Production
Engineering /Dictionnaire des Techniques de Production Mechanique Vol. III Produktionssysteme/Manufacturing Systems/Systemes de Production. Berlin: Springer.
C. I. R. P. (2004b): Wrterbuch der Fertigungstechnik/Dictionary of Production
Engineering/Dictionnaire des Techniques de Production Mechanique Vol. II Trennende Verfahren/Material Removal Processes/Procedes denlevement de matiere.
Berlin: Springer.
C. I. R. P. (2008): CIRP Unified Terminology of Manufacturing Systems. Berlin: Springer
Cannata, A.; Karnouskos, S.; Taisch, M. (2010): Energy efficiency driven process analysis
and optimization in discrete manufacturing. Industrial Electronics, IECON'09 - 35th
Annual Conference of IEEE.
Cassandras, C. G.; Lafortune, S. (2008): Introduction to Discrete Event Systems. New
York: Springer, available online http://dx.doi.org/10.1007/978-0-387-68612-7.
Chadderton, D. (2004): Building Services Engineering. Taylor & Francis.
Chiotellis, C.; Seliger, G.; Weinert, N. (2009): Energy-aware Production Planning and
Control. In: Proceedings of 16th CIRP International Conference on Life Cycle
Engineering (LCE 2009), Cairo, Egypt, pp. 310315.
Chow, W. K. (1996): Application of Computational Fluid Dynamics in building services
engineering. Building and Environment, 31/5, pp. 425-436.
Chung, C. A. (2004): Simulation modeling handbook - A practical approach. Industrial and
manufacturing engineering series, Boca Raton, Fla.: CRC Press.
References
181
Clarke, C.; Bras, B.; Guldberg, T.; Nader, G. (2008): Future Manufacturing Energy
Networks. In: Proceedings of 15th CIRP International Conference on Life Cycle
Engineering (LCE 2008), Sydney, Australia, pp. 420425.
Cohen, J. (2009): Statistical power analysis for the behavioral sciences. 2. ed., New York:
Psychology Press.
Dahmus, J.; Gutowski, T. (2004): An environmental analysis of machining. ASME
International Mechanical Engineering Congress and RD&D Exposition, Anaheim,
California, USA.
Dalzell, J. M. (2000): Food industry and the environment in the European Union - Practical
issues and cost implications. 2. ed., Gaithersburg, Md.: Aspen Publ.
Dehli, M. (1998): Energieeinsparung in Industrie und Gewerbe - Praktische Mglichkeiten
des rationellen Energieeinsatzes in Betrieben. Kontakt & Studium Nr. 535, RenningenMalmsheim: expert-Verl.
DIN 8580 (2003): Fertigungsverfahren. Berlin: Beuth.
DIN EN 16001 (2009): Energiemanagementsysteme. Berlin: Beuth.
DIN 9000 (2009): Qualittsmanagement. Berlin: Beuth.
Devoldere, T.; Dewulf, W.; Deprez, W.; Duflou, J. R. (2008): Energy Related Life Cycle
Impact and Cost Reduction Opportunities in Machine Design: The Laser Cutting Case.
In: Proceedings of 15th CIRP International Conference on Life Cycle Engineering (LCE
2008), Sydney, Australia, pp. 412419.
Devoldere, T.; Dewulf, W.; Deprez, W.; Willems, B.; Duflou, J. R. (2007): Improvement
Potential for Energy Consumption in Discrete Part Production Machines. In: Takata, S.;
Umeda, Y. (Eds.): Advances in Life Cycle Engineering for Sustainable Manufacturing
Businesses - Proceedings of the 14th CIRP Conference on Life Cycle Engineering,
Waseda University, Tokyo, Japan, London: Springer-Verlag, pp. 311316.
Dietmair, A.; Verl, A. (2008): Zustandsbasierte Energieverbrauchsprofile. wt
Werkstattstechnik, 98-7/8, pp. 640645.
Dietmair, A.; Verl, A. (2009): A generic energy consumption model for decision making
and energy efficiency optimisation in manufacturing. International Journal of
Sustainable Engineering, 2/2, pp. 123133.
Dietmair, A.; Verl, A. (2010): Energy Consumption Assessment and Optimisation in the
Design and Use Phase of Machine Tools. In: Proceedings of the 17th CIRP International
Conference on Life Cycle Engineering (LCE 2010), Hefei, China, pp. 116122.
Dyckhoff, H. (1994): Betriebliche Produktion - Theoretische Grundlagen einer
umweltorientierten Produktionswirtschaft. Berlin: Springer.
Dyckhoff, H.; Spengler, T. S. (2010): Produktionswirtschaft - Eine Einfhrung. Berlin,
Heidelberg: Springer.
Dyckhoff, H.; Souren, R. (2008): Nachhaltige Unternehmensfhrung - Grundzge
industriellen Umweltmanagements. Berlin: Springer.
Eckebrecht, J. (2000): Umweltvertrgliche Gestaltung von spanenden Fertigungsprozessen,
Aachen, Germany: Shaker Verlag.
Effenberger, H. (2000): Dampferzeugung, VDI-Buch, Berlin: Springer.
EIA (2009): U.S. Energy Information Administration. Available online:
http://www.eia.doe.gov/.
Eichhammer, W.; Jochem, E.; Patel, M.; Tnsing, E. (1996): Overview of Energy RD&D
Options for a Sustainable Future. European Commission, Luxembourg.
Einstein, D.; Worrell E.; Khrushch M. (2001): Steam systems in industry: Energy use and
energy efficiency improvement potentials. In: Proceedings of the 2001 ACEEE Summer
study on Energy Efficiency in Industry, pp. 535548.
182
References
Energy Star (2010): Website. Available online: http://www.energystar.gov.

Engelhardt-Nowitzki, C.; Krenn, B.; Nowitzki, O. (Eds.) (2008): Praktische Anwendung der
Simulation im Materialflussmanagement - Erfolgsfaktoren und Implementierungsszenarien.
1. Ed., Gabler Edition WissenschaftLeobener Logistik Cases, Wiesbaden: Gabler Verlag /
GWV Fachverlage GmbH Wiesbaden.
Engelmann, J. (2009): Methoden und Werkzeuge zur Planung und Gestaltung
energieeffizienter Fabriken.
European Comission (2006): Action Plan for Energy Efficiency: Realising the Potential.
Brussels.
European Comission (2008): Energy Efficiency in Manufacturing - The Role of ICT. ICT
and Energy Efficiency - Consultation Group on Smart Manufacturing. Available online:
doi 10.2759/43414.
European Comission (2009): Reference Document on best available techniques for energy
efficiency. Available online: http://eippcb.jrc.es/reference/.
EuP Directive 2005/32/EC (2005): Establishing a framework for the setting of ecodesign
requirements for energy-using products (EuP Directive). European Parliament.
Federal Ministry for the Environment (2010): DIN EN 16001: Energy Management
Systems in Practice - A Guide for Companies and Organisations, Umweltbundesamt,
Berlin.
Ferretti, I.; Zanoni, S.; Zavanella, L. (2008): Energy efficiency in a steel plant using
optimization-simulation. In: Proceedings of 20th European Modeling & Simulation
Symposium.
Fichter, K. (2005): Interpreneurship - Nachhaltigkeitsinnovationen in interaktiven
Perspektiven eines vernetzenden Unternehmertums, Habilitation, Universitt Oldenburg,
Theorie der Unternehmung Nr. 33, Marburg: Metropolis-Verl.
Fiedler, T.; Metz, D.; Ott, S. (2007): Knstliche Neuronale Netze zur Lastprognose im Stromund Gasbereich. Querschnitt - Magazin der Hochschule Darmstadt, 2, pp. 135138.
Fleschutz, T.; Rahman, A. A. A.; Harms, R.; Seliger, G. (2010): Assessment of Life Cycle
Impacts and Integrated Evaluation Concept for Equipment Investment. In: Proceedings
of 17th CIRP International Conference on Life Cycle Engineering (LCE 2010), Hefei,
China, pp. 8387.
Forrester, J. W. (1969): Urban Dynamics. Taylor & Francis Inc.
Fraunhofer IWU (2008): Energieeffizienz in der Produktion - Untersuchung zum
Handlungs- und Forschungsbedarf. Mnchen.
Freeman, S. L.; Niefer, M. J.; Roop, J. M. (1996): Measuring Industrial Energy Eficiency:
Physical Volume Versus Economic Value. Report to the U.S. Department of Energy.
Gbel, K.; Tillman, A.-M. (2005): Simulating operational alternatives for future cement
production. Journal of Cleaner Production, 13-13/14, pp. 12461257.
Gauchel, W. (2006): Energy-saving pneumatic systems. O + P lhydraulik und
Pneumatik, 50/1, pp. 1-22.
Innovationen bei der rationellen Energieanwendung (1998) Neue Chancen fr die
Wirtschaft, VDI-Berichte Nr. 1385, Dsseldorf: VDI-Verl.
Gloor, R. (2000): Energieeinsparungen bei Druckluftanlagen in der Schweiz. BfEBundesamt fr Energie, Bern.
Gopalakrishnan, B. (2005): A Systems Approach to Plant-wide Energy Assessment. Energy
engineering, 102/5, pp. 49-79.
Gutenberg, E. (1983): Die Produktion. 24. Ed., Enzyklopdie der Rechts- und
Staatswissenschaft, Abteilung Staatswissenschaft, Berlin: Springer.
References
183
Gutowski, T. G.; Dahmus, J. B.; Thiriez, A. (2006): Electrical Energy Requirements for
Manufacturing Processes. In: Duflou, J. R. (Ed.): Proceedings of the 13th CIRP
Conference on Life Cycle Engineering (LCE 2006), Leuven, Belgium, pp. 623627.
Hall, F.; Greeno, R. (2009): Building Services Handbook: Incorporating Current Building
& Construction. Butterworth-Heinemann.
Heilala, J.; Vatanen, S.; Tonteri, H.; Montonen, J.; Lind, S.; Johansson, B.; Stahre, J.
(2008): Simulation-based sustainable manufacturing system design. In: Mason, S. J.
(Ed.): 2008 Winter Simuation Conference (WSC 2008); Miami, Florida, USA,
Piscataway, NJ: IEEE, pp. 19221930.
Herrmann, C. (2009): Ganzheitliches Life Cycle Management: Nachhaltigkeit und
Lebenszyklusorientierung in Unternehmen. Berlin: Springer.
Herrmann, C.; Thiede, S. (2009a): Process chain simulation to foster energy efficiency in
manufacturing. CIRP Journal of Manufacturing Science and Technology, Vol. 1, Issue
4, pp. 221-229.
Herrmann, C.; Thiede, S. (2008): Increasing Energy Efficiency in Manufacturing
Companies through Process Chain Simulation. In: Proceedings of the 6th Global
Conference on Sustainable Product Development and Life Cycle Engineering, Busan,
Korea, pp. 5257.
Herrmann, C.; Thiede, S. (2009b): Towards Energy and Resource Efficient Process Chains.
In: Proceedings of the 16th CIRP International Conference on Life Cycle Engineering
(LCE 2009), Cairo, Egypt, pp. 303309.
Herrmann, C.; Thiede, S.; Kara, S.; Hesselbach, J. (2011a): Energy oriented simulation of
manufacturing systems - concept and application. CIRP Annals - Manufacturing
Technology, Vol. 60/1, pp. 45-48.
Herrmann, C.; Heinemann, T.; Thiede, S. (2011b): Synergies from Process and Energy
Oriented Process Chain Simulation A Case Study from the Aluminium Die Casting
Industry, In: Proceedings of the 18th CIRP International Conference on Life Cycle
Engineering, Braunschweig, Germany, Springer, pp. 317-322.
Herrmann, C.; Thiede, S.; Heinemann, T. (2010a): Ganzheitliche Anstze zur Erhhung der
Energie- und Ressourceneffizienz in der Produktion, In: Karlsruher Arbeitsgesprche
Produktionsforschung 2010, Karlsruhe, Germany.
Herrmann, C.; Bogdanski, G.; Winter, M.; Heinemann, T.; Thiede, S.; Zein, A. (2010b):
Sustainability in Production Engineering - Holistic Thinking in Education. In: Advances
in Sustainable Manufacturing - Proceedings of 8th Global Conference on Sustainable
Manufacturing, Abu Dhabi, VAE, pp. 25-30.
Herrmann, C.; Bergmann, L.; Thiede, S.; Halubek, P. (2007a): Total Life Cycle
Management - An Integrated Approach Towards Sustainability. In: Proceedings of 3rd
International Conference on Life Cycle Management, Zurich, Switzerland.
Herrmann, C.; Thiede, S.; Stehr, J.; Bergmann, L. (2008a): An environmental perspective
on Lean Production. In: Proceedings of the 41th CIRP International Seminar on
Manufacturing Systems, Tokyo, Japan, pp. 8388.
Herrmann, C.; Zein, A.; Thiede, S.; Bergmann, L.; Bock, R. (2008b): Bringing sustainable
manufacturing into practice: The machine tool case. In: Proceedings of the 6th Global
Korea, pp. 272277.
184
References
Herrmann, C.; Bergmann, L.; Thiede, S.; Zein, A. (2007b): Framework for Integrated
Analysis of Production Systems. In: Takata, S.; Umeda, Y. (Eds.): Advances in Life
Cycle Engineering for Sustainable Manufacturing Businesses - Proceedings of the 14th
CIRP Conference on Life Cycle Engineering (LCE 2007), Tokyo, Japan, London:
Springer-Verlag, pp. 195200.
Hesselbach, J.; Martin, L.; Herrmann, C.; Thiede, S.; Ldemann, B.; Detzer, R. (2008a):
Energieeffizienz durch optimierte Abstimmung zwischen Produktion und technischer
Gebudeausrstung. In: Rabe, M. (Ed.): 13. ASIM - Fachtagung: Simulation in
Produktion und Logistik /// Advances in simulation for production and logistics
applications, ASIM-Mitteilung Nr. 118, Stuttgart: Fraunhofer IRB Verlag, pp. 177185.
Hesselbach, J.; Junge, M. (2005): Reduzierung von Energiespitzen durch Fabriksimulation
(Reducing Energy Peaks by Factory Simulation), Industrie Management, 21/2, pp. S.
35-37.
Hesselbach, J.; Herrmann, C.; Detzer, R.; Martin, L.; Thiede, S.; Ldemann, B. (2008b):
Energy Efficiency through optimized coordination of production and technical building
services. In: Proceedings of the 15th CIRP International Conference on Life Cycle
Engineering (LCE 2008), Sydney, Australia, pp. 624629.
Hornberger, M. (o. J.): Total Energy Efficiency Management (TEEM).
Hornberger, M. (2009): Total Energy Efficiency Management - Vom
Energiemanagementsystem zur Simulation der Energieverbrauchswerte auf
Prozessebene. In: Elektronik ecodesign congress, Mnchen, Germany.
Hufendiek, K.; Kaltschmitt, M. (1998): Einsatz knstlicher neuronaler Netze bei der
kurzfristigen Lastprognose. In: VGB PowerTech e.V. (Eds.): VGB-Konferenz, pp. 16.
ies Industrie Engineering Service GmbH (2009): Corporate Website. Available online:
http://strahlen-mit-trockeneis.com/ServiceProdukte.
IHK (2009): Energiepreise und Unternehmensentwicklung in Baden-Wrttemberg,
Auswertung einer Umfrage der Industrie- und Handelskammern Heilbronn-Franken,
Hochrhein-Bodensee, Karlsruhe and Ostwrttemberg, autumn 2008.
ISO EN15603 (2008): Energy performance of buildings. Overall energy use and definition
of energy ratings.
Jaffe, A. B.; Stavins, R. N. (1994): The energy efficiency gap: what does it mean? Energy
Policy, 22/10, pp. 6071.
Jahangirian, M.; Eldabi, T.; Naseer, A.; Stergioulas, L. K.; Young, T. (2010): Simulation in
manufacturing and business: A review. European Journal of Operational Research, 203,
pp. 113.
Johansson, B.; Mani, M.; Skoogh, A.; Leong, S. (2009a): Discrete Event Simulation to
generate Requirements Specification for Sustainable Manufacturing Systems Design. In:
Proceedings of PerMIS '09 Proceedings of the 9th Workshop on Performance Metrics
for Intelligent Systems.
Johansson, B.; Kacker, R.; Kessel, R.; McLean, C.; Sriram, R. (2009b): Utilizing
Combinatorial Testing on Discrete Event Simulation Models for Sustainable
Manufacturing. In: Proceedings of 2009 ASME Design for Manufacturing and the Life
Cycle Conference, San Diego, California, USA, pp. 1095-1101.
Junge, M. (2007): Simulationsgesttzte Entwicklung und Optimierung einer
energieeffizienten Produktionssteuerung, Dissertation, Universitt Kassel, Produktion &
Energie, Vol. 1, Kassel.
Kaiser, S.; Starzer, O. (1999): Handbuch fr betriebliches Energiemanagement.
Energieverwertungsagentur.
References
185
Kelton, W. D.; Sadowski, R. P.; Swets, N. B. (2010): Simulation with Arena. 5. ed.,
Boston, Mass.: McGraw-Hill.
Khatib, H. (2003): Economic Evaluation of Projects in the Electricity Supply Industry.
Institution of Electrical Engineers, Vol. 44.
Kircher, K.; Shi, X.; Patil, S.; Zhang, K. (2010): Cleanroom energy efficiency strategies:
Modeling and simulation. Energy and Buildings, 42/3, pp. 282289.
Klocke, F.; Schlosser, R.; Tnissen, S. (2010): Prozesseffizienz durch Parameterwahl Evaluierung des Frsprozesses. wt Werkstattstechnik, 100/5, pp. 346349.
Krenn, B. (2007): Management komplexer Materialflsse mittels Simulation: State-of-theArt und innovative Konzepte: Duv.
Kuhrke, B.; Schrems, S.; Eisele, C.; Abele, E. (2010): Methodology to assess the energy
consumption of cutting machine tools. In: Proceedings of the 17th CIRP International
Conference on Life Cycle Engineering (LCE 2010), Hefei, China, pp. 7682.
Lang, B.; Hesselbach, J. (2009): Energiebedarfsvorhersage produzierender Unternehmen
mithilfe Neuronaler Netze. In: Gnauck, A.; Luther, B. (Eds.): ASIM 2009 - 20.
Symposium Simulationstechnik, pp. 132136.
Lanz, M. S.; Mani, M.; Leong, S. K.; Lyons, K. W.; Ranta, A.; Ikkala, K.; Bengtsson, N.
(2010): Impact of Energy Measurements in Machining Operations. In: Proceedings of
the ASME 2010 International Design Engineering Technical Conferences & Computers
and Information in Engineering Conference, Montreal, Quebec, Canada, pp. 1-7.
Larsson, M.; Dahl, J. (2003): Reduction of the Specific Energy Use in an Integrated Steel
Plant The Effect of an Optimisation Model. ISIJ International, 10, pp. 16641673.
Lau, H. C. W.; Cheng, E. N. M.; Lee, C. K. M.; Ho, G. T. S. (2007): A fuzzy logic
approach to forecast energy consumption change in a manufacturing system. Expert
Systems with Applications, 34/3, pp. 1813-1824.
Law, A. M. (2007): Simulation modeling and analysis. 4. ed., Boston: McGraw-Hill.
Lind, S.; Johansson, B.; Stahre, J.; Berlin, C.; Fasth, .; Heilala, J.; Helin, K.; Kiviranta, S.;
Krassi, B.; Montonen, J. (2009): SIMTER-A Joint Simulation Tool for Production
Development.
Lfgren, B. (2009): Capturing the life cycle environmental performance of a companys
manufacturing system, diploma thesis, Chalmers University of Technology.
Mark, K.; Rojk, J.; Stluka, P. (2009): The Role of Predictive Models in Energy Efficiency
Optimization of Complex Industrial Plants, Chemical Engineering, 18.
Massachusetts Institute of Technology (MIT) (2010): What is Operations Management?
Available online: http://mitsloan.mit.edu/omg/om-definition.php.
McKinney, M. L.; Schoch, R. M.; Yonavjak, L. (2007): Environmental science - Systems
and solutions. 4. ed., Sudbury, Mass: Jones and Bartlett.
Mller, E.; Engelmann, J.; Lffler, T.; Strauch, J. (2009): Energieeffiziente Fabriken planen
und betreiben, Berlin, Heidelberg: Springer.
Mller, L. (2001): Handbuch der Elektrizittswirtschaft - Technische, wirtschaftliche und
rechtliche Grundlagen. 2. Ed., Elektrische Energietechnik, Berlin: Springer.
Neumann, K. (1985): Betriebliche Umweltschutzplanung mit Hilfe der Simulation - Ein
integrierter Planungsansatz mit Anwendung auf einen landwirtschaftlichen Betrieb.
Frankfurt am Main, New York: P. Lang.
Offner, K. (2001): Betriebliches Energiemanagement - Qualittsmerkmale Lieferantenfestlegung Qualittstechniken. Dissertation, Technische Universitt Graz,
1. Ed., Techno-konomische Forschung und Praxis, Wiesbaden: DUV Dt. Univ.-Verl.
186
References
stergren, K.; Janestad, H.; Berlin, J.; Sonesson, U. (2008): Integrated Simulation
Technology for safe sustainable Food Processes. In: European Technology Institute EUROSIS: FOODSIM 2008, pp. 181185.
Oxford University Press (2011): Oxford Dictionaries Online. Available online:
http://www.oxforddictionaries.com.
Patterson, M. G. (1996): What is energy efficiency? - Concepts, indicators and
methodological issues. Energy Policy, 24/5, pp. 377390.
Pegden, C. D.; etc.; Sadowski, R. P.; Shannon, R. E. (1995): Introduction to Simulation
Using Siman. McGraw-Hill Education - Europe.
Pinero, E. (2009): ISO 50001: Setting the Standard for Industrial Energy Management.
Green Manufacturing News.
Planck, M.; Psler, M. (1964): Vorlesungen ber Thermodynamik. 11. Ed., Berlin: de
Gruyter.
Pritsker, A. A. B. (1995): Introduction to simulation and SLAM II. 4. ed., New York:
Wiley.
Quadriguasi, J.; Walther, G.; Bloemhof, J.; van Nunen, J. A. E. E.; Spengler, T. (2009): A
methodology for assessing eco-efficiency in logistics networks. European Journal of
Operational Research, 193/3, pp. 670-682.
Rabe, M.; Spiekermann, S.; Wenzel, S. (2008): Verifikation und Validierung fr die
Simulation in Produktion und Logistik - Vorgehensmodelle und Techniken. VDI-Buch,
Berlin, Heidelberg: Springer.
Radgen, P.; Blaustein, E. (Eds.) (2001): Compressed Air Systems in the European Union:
Energy, Emissions, Savings Potential and Policy Actions. Final report, Stuttgart:
LOG_X Verlag.
Rager, M. (2008): Energieorientierte Produktionsplanung - Analyse, Konzeption und
Umsetzung. Dissertation, Universitt Augsburg, Gabler Edition Wissenschaft,
Wiesbaden: Gabler.
Rahimifard, S.; Seow, Y.; Childs, T. (2010): Minimising Embodied Product Energy to
support energy efficient manufacturing. CIRP Annals - Manufacturing Technology,
59/1, pp. 2528.
Rahimifard, S. (2009): Minimisation of Energy Consumption during the Manufacturing
Phase of a Product Life Cycle. In: Manufuture 2009, Gteborg.
Rajan, G. G. (2008): Optimizing Energy Efficiencies in Industry: McGraw-Hill Education Europe.
Rebhan, E. (Ed.) (2002): Energiehandbuch - Gewinnung, Wandlung und Nutzung von
Energie. Engineering online library, Berlin: Springer.
Recknagel, H.; Sprenger, E.; Schramek, E.-R. (Eds.) (2010): Taschenbuch fr Heizung +
Klimatechnik 11/12 - Komplettversion: Oldenbourg Industrieverlag.
Ridder, N. (2003): ffentliche Energieversorgungsunternehmen im Wandel Wettbewerbsstrategien im liberalisierten deutschen Strommarkt. Marburg: Tectum-Verl.
Ruppelt, E. (Ed.) (2003): Druckluft-Handbuch. 4. Ed., Essen: Vulkan-Verl.
Saacke (2009): Faustformelsammlung. 6. Ed., Bremen.
Saidur, R.; Rahim, N. A.; Hasanuzzaman, M. (2010): A review on compressed-air energy
use and energy savings. Renewable and Sustainable Energy Reviews, 14/4, pp. 1135
1153.
Sankey, H. R. (1898): Introductory note on the thermal efficiency of steam-engines.
Minutes of Proceedings of The Institution of Civil Engineers., pp. 278283.
Schenk, M. (2004): Fabrikplanung und Fabrikbetrieb - Methoden fr die wandlungsfhige
und vernetzte Fabrik. Berlin, Heidelberg: Springer.
References
187
Schieferdecker, B. (2006): Energiemanagement-Tools - Anwendung im Industrieunternehmen.

Berlin, Heidelberg: Springer.
Schmid, C. (2008): Energieeffizienz in Unternehmen - Eine wissensbasierte Analyse von
Einflussfaktoren und Instrumenten. ETH Zrich: vdf Hochschulverlag AG.
Schmid, C.; Layer, G. (2003): Mglichkeiten, Potenziale, Hemmnisse und Instrumente zur
Senkung des Energieverbrauchs branchenbergreifender Techniken in den Bereichen
Industrie und Kleinverbrauch. Umweltforschungsplan des Bundesministeriums fr
Umwelt, Naturschutz und Reaktorsicherheit, Rationelle Energieerzeugung und -nutzung,
Karlsruhe: ISI.
Schmidt, J. W. (1970): Simulation and analysis of industrial systems, Homewood, Ill.:
Irwin.
Schmidt, M. (2008a): The Sankey Diagram in Energy and Material Flow Management Part I History. Journal of industrial ecology, 12/1, pp. 82-94.
Schmidt, M. (2008b): The Sankey Diagram in Energy and Material Flow Management Part II Methodology and Current Applications. Journal of industrial ecology, 12/2,
pp.173-185.
Schmidt, M. (2006): Der Einsatz von Sankey-Diagrammen im Stoffstrommanagement,
Beitrge der Hochschule Pforzheim Nr. 124, Pforzheim: Hochsch.
Schmidt, M. (2007): Rohstoffe und Ressourceneffizienz - bereit fr den Wettbewerb um
Nachhaltigkeit?,
In:
"Statussemeinar
2007",
Berlin.
Available
online:
http://www.netzwerklebenszyklusdaten.de/cms/webdav/site/lca/shared/Veranstaltungen/Statusseminar/Status
seminar_2007/Presentations/04_Schmidt_NetLZD.pdf.
Schrter, M.; Weifloch, U.; Buschak, D. (2009): Energieeffizienz in der Produktion
Wunsch oder Wirklichkeit? Energieeinsparpotenziale und Verbreitungsgrad
energieeffizienter Techniken, Fraunhofer-Institut fr System- und Innovationsforschung
ISI, Karlsruhe.
Schufft, W. (2007): Taschenbuch der elektronischen Energietechnik. Hanser
Fachbuchverlag.
Schuh, G. (2006): Produktionsplanung und -steuerung - Grundlagen, Gestaltung und
Konzepte. VDI-Buch, Available online: http://dx.doi.org/10.1007/3-540-33855-1 /
http://www.gbv.de/dms/hebis-darmstadt/toc/177220961.pdf.
Schultz, A. (2002): Methode zur integrierten kologischen und konomischen Bewertung
von Produktionsprozessen und -technologien, Dissertation, Universitt Magdeburg.
Scottish government (2006): Scottish Energy Study, http://scotland.gov.uk.
Seefeldt, F.; Wnsch, M. (2007): Potenziale fr Energieeinsparung und Energieeffizienz im
Lichte aktueller Preisentwicklungen. Prognos AG, Final report 18/06.
Siemens AG (2010): Simulation of the Energy Consumption of Conveyor Lines with
Tecnomatix Plant Simulation, Siemens AG. Available online: slideshare.net/
SiemensPLM, last update 2010.
Solding, P.; Petku, D.; Mardan, N. (2009): Using simulation for more sustainable
production systems methodologies and case studies. International Journal of
Sustainable Engineering, 2/2, pp. 111122.
Solding, P.; Thollander, P. (2006): Increased energy efficiency in a Swedish iron foundry
through use of discrete event simulation. In: Proceedings of the 2006 Winter Simulation
Conference (WSC 06), Monterey, California, USA, Piscataway, NJ: IEEE Operations
Center, pp. 19711976.
188
References
Solding, P.; Petku, D. (2005): Applying Energy Aspects on Simulation of Energy-intensive

Production Systems. In: Proceedings of the 2005 Winter Simulation Conference,
Orlando, Florida, U.S.A, pp. 14281432.
Soper, D. S. (2011): A-priori Sample Size Calculator for Student's t-Test. Department of
Information Systems and Decision Sciences, California State University. available
online: http://www.danielsoper.com/statcalc/calc47.aspx.
Sorrell, S.; Schleich, J.; Scott, S.; O'Malley, E.; Trace, F.; Boede, U.; Ostertag, K.; Radgen,
P. (2000): Barriers to Energy Efficiency in Public and Private Organisations. Final
report to the European Commission.
Specht, H. (2005): Stromliefervertrge im liberalisierten Energiemarkt - Gestaltung von
Sondervertrgen und Ausschreibung von Stromlieferungen. Berlin: VDE-Verl.
Spirax Sarco (2006): Grundlagen der Dampf- und Kondensattechnologie. Konstanz.
Spirax-Sarco Limited (2005): The Steam and Condensate Loop. Spirax-Sarco Limited.
Sterman, J. (2000): Business Dynamics: Systems Thinking and Modeling for a Complex
World. McGraw-Hill Education - Europe.
Stern, P. C. (1984): Energy use - The human dimension. New York: Freeman.
Thamling, N.; Seefeldt, F.; Glckner, U. (2010): Rolle und Bedeutung von Energieeffizienz
und Energiedienstleistungen in KMU. Prognos AG, Basel.
Thiede, S.; Herrmann, C. (2010): Simulation-based Energy Flow Evaluation for Sustainable
Manufacturing Systems. In: Proceedings of the 17th CIRP International Conference on
Life Cycle Engineering (LCE 2010), Hefei, China, pp. 99104.
Thollander, P. (2009): Towards Increased Energy Efficiency in Swedish Industry: Barriers,
Driving Forces & Policies. Dissertation, Linkping University.
Tnsing, E. (1996): Energiekostenreduzierung durch betriebliches Energiemanagement.
Initiative Energie effizient nutzen Schwerpunkt Strom, pp. 19.
U.S. Department of Energy (2010): Building Energy Software Tools Directory. Available
online: http://apps1.eere.energy.gov/buildings/tools_directory.
VDI (2003): VDI 4661:2003-09 Energiekenngren - Definitionen - Begriffe Methodik.
Berlin: Beuth.
VDI (2006): VDI 4602 Blatt 1:2006-04 Energiemanagement-Begriffe, Definitionen. Berlin:
Beuth.
VDI (2007): VDI 3633 - Simulation of systems in materials handling, logistics and
production. Berlin: Beuth.
Verfaillie, H. A.; Bidwell, R. (2000): Measuring eco-efficiency - A guide to reporting
company performance, Dedicated to making a difference - World Business Council for
Sustainable Development.
Viegas, J. (2005): Kinetic and Potential Energy: Understanding Changes Within Physical
Systems. Rosen publishing group.
Wagner, W. (2010): Wasser und Wasserdampf im Anlagenbau. 2. Ed., Wrzburg: Vogel
Business Media.
Wahren, S. (o. J.): Total Energy Efficiency Management - Mit Energiemanagement zur
Kostensenkung in der Produktion.
Waltenberger, G. (2005): Energiemanagement in der Industrie - Die energiewirtschaftlichen
Grundlagen. 1. Ed., Lohmar: Eul.
Warnecke, H. J.; Westkmper, E.; Gottwald, B. (1998): Einfhrung in die
Fertigungstechnik. Wiesbaden: Teubner.
Weinert, N. (2010): Planung energieeffizienter Produktionssysteme. ZWF, 105/5, pp.
503507.
References
189
Weinert, N.; Chiotellis, S.; Seliger, G. (2009): Concept for Energy-Aware Production
Planning based on Energy Blocks. In: Proceedings of the 7th Global Conference on
Sustainable Manufacturing, pp. 7580.
Wenzel, S.; Collisi-Bhmer, S.; Pitsch, H.; Rose, O.; Wei, M. (2008): Qualittskriterien
fr die Simulation in Produktion und Logistik - Planung und Durchfhrung von
Simulationsstudien. VDI-Buch, Berlin, Heidelberg: Springer.
Westerkamp, T. A. (2008): It's not easy being green - energy management appraises the
details. Industrial Engineer, 3, pp. 3741.
Westkmper, E. (2005): Einfhrung in die Organisation der Produktion. Berlin: Springer.
Wischhusen, S.; Ldemann B.; Schmitz G. (2003): Economical Analysis of Complex
Heating and Cooling Systems with the Simulation Tool HKSim. In: Proceedings of the
3rd International Modelica Conference, Linkping, Sweden.
Wohinz, J. W.; Moor, M. (1989): Betriebliches Energiemanagement - Aktuelle Investition
in die Zukunft. Wien: Springer.
Wohlgemuth, V.; Page, B.; Kreutzer, W. (2006): Combining discrete event simulation and
material flow analysis in a component-based approach to industrial environmental
protection. Environmental Modelling & Software, 21/11, pp. 16071617.
Wohlgemuth, V. (2005): Komponentenbasierte Untersttzung von Methoden der
Modellbildung und Simulation im Einsatzkontext des betrieblichen Umweltschutzes Konzeption und prototypische Entwicklung eines Stoffstromsimulators zur Integration
einer stoffstromorientierten Perspektive in die auftragsbezogene Simulationssicht.
Dissertation, Universitt Hamburg, Aachen: Shaker.
Yager, R. R.; Zadeh, L. A. (1992): An Introduction to Fuzzy Logic Applications in
Intelligent Systems: Kluwer Academic Publishers.
Zeigler, B. P.; Praehofer, H.; Kim, T. G. (2000): Theory of modeling and simulation.
Academic press New York, NY.
Own References
2011
Herrmann, C.; Thiede, S.; Kara, S.; Hesselbach, J. (2011): Energy oriented simulation of
manufacturing systems - concept and application, In: CIRP Annals - Manufacturing
Technology: Elsevier, Vol. 60, Issue 1, pp. 45-48.
Herrmann, C.; Kara, S.; Thiede, S. (2011): Dynamic life cycle costing based on lifetime
prediction. International Journal of Sustainable Engineering, Vol. 4, Issue 3, pp. 224235.
Thiede, S.; Herrmann, C.; Kara, S. (2011): State of research and an innovative approach for
simulating energy flows of manufacturing systems, In: Proceedings of the 18th CIRP
International Conference on Life Cycle Engineering, Braunschweig, Germany, Springer,
pp. 335-340.
Herrmann, C.; Heinemann, T.; Thiede, S. (2011): Synergies from Process and Energy
Oriented Process Chain Simulation A Case Study from the Aluminium Die Casting
Industry, In: Proceedings of the 18th CIRP International Conference on Life Cycle
Engineering, Braunschweig, Germany, Springer, pp. 317-322.
2010
Thiede, S.; Herrmann, C. (2010): Simulation-based Energy Flow Evaluation for Sustainable
Manufacturing Systems, In: Proceedings of the 17th CIRP International Conference on
Life Cycle Engineering, Hefei, China, pp. 99-104.
Herrmann, C.; Bogdanski, G.; Winter, M.; Heinemann, T.; Thiede, S.; Zein, A. (2010):
Sustainability in Production Engineering - Holistic Thinking in Education, In: Advances
in Sustainable Manufacturing - Proceedings of 8th Global Conference on Sustainable
Manufacturing, Abu Dhabi, pp. 25-30.
Thiede, S., Herrmann, C. (2010): Energy Flow Simulation for Manufacturing Systems, In:
Advances in Sustainable Manufacturing - Proceedings of 8th Global Conference on
Sustainable Manufacturing, Abu Dhabi, pp. 275-280.
Herrmann, C.; Thiede, S.; Heinemann, T. (2010): A Holistic Framework for Increasing
Energy and Resource Efficiency in Manufacturing, In: Advances in Sustainable
Manufacturing - Proceedings of 8th Global Conference on Sustainable Manufacturing,
Abu Dhabi, pp. 267-273.
Bergmann, L.; Thiede, S.; Herrmann, C. (2010): Mehr Energie- und Ressourceneffizienz in
der Produktion, In: Nachhaltige Produktion, SellersMedia, 2010, Issue 2, pp. 30-32,
ISSN 1868-4181.
Herrmann, C.; Heinemann, T.; Thiede, S. (2010): Identifying levers for Enhancing Energy
and Resource Efficiency in Industrial Process Chains using the example of Aluminium
Die Casting: Proceedings of the 1st International Conference on Automotive Materials
and Manufacturing, Pune, India, pp. 119129.
192
Own References
Herrmann, C.; Kara, S.; Thiede, S.; Luger, T. (2010): Energy Efficiency in Manufacturing
Perspectives from Australia and Europe: Proceedings of the 17th CIRP International
Conference on Life Cycle Engineering, Hefei, China, pp. 2328.
Herrmann, C.; Thiede, S.; Heinemann, T. (2010): Ganzheitliche Anstze zur Erhhung der
Energie - und Ressourceneffizienz in der Produktion, In: Proceedings of Karlsruher
Arbeitsgesprche Produktionsforschung 2010, Karlsruhe.
Herrmann, C.; Thiede, S.; Zein, A.; Heinemann, T. (2010): Holistic Approaches for Increasing
Energy and Resource Efficiency in Manufacturing: Proceedings of the 1st International
Conference on Automotive Materials and Manufacturing, Pune, India, pp. 3948.
2009
Herrmann, C.; Bergmann, L.; Thiede, S. (2009): Methodology for the design of sustainable
production systems, International Journal of Sustainable Manufacturing, Vol. 1, Issue
4, pp. 376395.
Herrmann, C.; Thiede, S. (2009): Process chain simulation to foster energy efficiency in
manufacturing, CIRP Journal of Manufacturing Science and Technology, Vol. 1, Issue
4, pp. 221-229.
Herrmann, C.; Thiede, S. (2009): Towards Energy and Resource Efficient Process Chains,
In: Proceedings of the 16th CIRP International Conference on Life Cycle Engineering,
Cairo, Egypt, pp. 303309.
Herrmann, C.; Thiede, S.; Kuntzky, K.; Bhm, S.; Frauenhofer, M.; Gadhia, D. (2009):
Indian Solar Thermal Technology: Potentials and Challenges, In: Proceedings of the 7th
Global Conference on Sustainable Manufacturing, Chennai, India, pp. 169174.
Herrmann, C.; Thiede, S.; Luger, T.; Zein, A.; Stehr, J.; Halubek, P.; Torney, M. (2009):
Automotive Life Cycle Engineering, In: Proceedings of the 16th CIRP International
Conference on Life Cycle Engineering, Cairo, Egypt, pp. 157164.
Herrmann, C.; Thiede, S.; Zein, A.; Ihlenfeldt, S.; Blau, P. (2009): Energy Efficiency of
Machine Tools: Extending the Perspective, In: Proceedings of the 42nd CIRP
Conference on Manufacturing Systems, Grenoble, Switzerland.
2008
Herrmann, C.; Bergmann, L.; Halubek, P.; Stehr, J.; Thiede, S. (2008): Life Cycle
Engineering - State of the Art and Research Perspectives, In: Proceedings of the 15th
CIRP International Conference on Life Cycle Engineering, Sydney, Australia, pp. 449
454.
Herrmann, C.; Bergmann, L.; Halubek, P.; Thiede, S. (2008): Lean Production System
Design from the Perspective of the Viable System Model, In: Proceedings of the 41th
CIRP International Seminar on Manufacturing Systems, Tokyo, Japan, pp. 309314.
Herrmann, C.; Thiede, S. (2008): Increasing Energy Efficiency in Manufacturing
Companies through Process Chain Simulation, In: Proceedings of the 6th Global
Korea, pp. 5257.
Herrmann, C.; Thiede, S.; Stehr, J.; Bergmann, L. (2008): An environmental perspective on
Lean Production, In: Proceedings of the 41th CIRP International Seminar on
Manufacturing Systems, Tokyo, Japan, pp. 8388.
Herrmann, C.; Zein, A.; Thiede, S.; Bergmann, L.; Bock, R. (2008): Bringing sustainable
manufacturing into practice: the machine tool case, In: Proceedings of the 6th Global
Korea, pp. 272277.
Own References
193
Hesselbach, J.; Martin, L.; Herrmann, C.; Thiede, S.; Ldemann, B.; Detzer, R. (2008):
Energieeffizienz durch optimierte Abstimmung zwischen Produktion und technischer
Gebudeausrstung, In: Proceedings of 13. ASIM - Fachtagung: Simulation in
Produktion und Logistik - Advances in simulation for production and logistics
applications, Stuttgart: Fraunhofer IRB, pp. 177185.
Hesselbach, J.; Herrmann, C.; Detzer, R.; Martin, L.; Thiede, S.; Ldemann, B. (2008):
Energy Efficiency through optimized coordination of production and technical building
services, In: Proceedings of the 15th CIRP International Conference on Life Cycle
Engineering, Sydney, Australia, pp. 624629.
2007
Herrmann, C.; Bergmann, L.; Thiede, S. (2007): An Integrated Approach for the Evaluation
of Maintenance Strategies to Foster Sustainability in Manufacturing, In: Proceedings of
Sustainable Manufacturing V: Global Symposium on Sustainable Product Development
and Life Cycle Engineering", Rochester, NY USA.
Herrmann, C.; Bergmann, L.; Thiede, S. (2007): Developing Life Cycle Oriented
Innovations Within the Turbulent Business Environment, In: Proceedings of ICED 07 16th International Conference of Engineering Design, Paris, France.
Herrmann, C.; Bergmann, L.; Thiede, S. (2007): Gestaltungselemente und Erfolgsfaktoren Ergebnisse einer empirischen Umfrage unter produzierenden Unternehmen, Intelligenter
Produzieren, Issue 3, pp. 20-22.
Herrmann, C.; Bergmann, L.; Thiede, S. (2007): Life Cycle Oriented Design Of Lean
Production Systems, In: Proceedings of the 3rd International Virtual Design and
Automation Conference, Poznan, Poland, pp. 295302.
Herrmann, C.; Bergmann, L.; Thiede, S.; Halubek, P. (2007): Total Life Cycle
Management - An Integrated Approach Towards Sustainability, In: Proceedings of 3rd
International Conference on Life Cycle Management, Zurich, Switzerland.
Herrmann, C.; Bergmann, L.; Thiede, S.; Luger, T. (2007): Total Life Cycle Management Framework and Concepts, In: Proceedings of International Workshop on Sustainability
in Manufacturing - Remanufacturing for a closed-loop economy, Busan, Korea, pp.
165176.
Herrmann, C.; Bergmann, L.; Thiede, S.; Torney, M.; Zein, A. (2007): Framework For The
Dynamic And Life Cycle Oriented Evaluation Of Maintenance Strategies, In:
Proceedings of the 3rd International Virtual Design and Automation Conference,
Poznan, Poland, pp. 277284.
Herrmann, C.; Bergmann, L.; Thiede, S.; Zein, A. (2007): Energy Labels for Production
Machines - An Approach to Facilitate Energy Efficiency in Production Systems, In:
Proceedings of the 40th CIRP International Seminar on Manufacturing Systems,
Liverpool, UK.
Herrmann, C.; Bergmann, L.; Thiede, S.; Zein, A. (2007): Framework for Integrated
Analysis of Production Systems, In: Proceedings of the 14th CIRP Conference on Life
Cycle Engineering, Tokyo, Japan: Springer, pp. 195200.
Herrmann, C.; Bergmann, L.; Thiede, S.; Zein, A. (2007): Life Cycle Innovations in
Extended Supply Chain Networks, In: Proceedings of the 14th CIRP Conference on Life
Cycle Engineering, Tokyo, Japan, Springer, pp. 439444.
Herrmann, C.; Bergmann, L.; Thiede, S.; Zein, A. (2007): Total Life Cycle Management A Systems and Cybernetics Approach to Corporate Sustainability in Manufacturing, In:
Proceedings of Sustainable Manufacturing V: Global Symposium on Sustainable
Product Development and Life Cycle Engineering", Rochester, NY USA.
Appendix
Steam table (taken from Spirax-Sarco Limited)

http://www.buildingdesign.co.uk/mech-technical/spirax-sarco-t2/
Saturated-Steam-Tables.jpg
Cumulative load profiles for

electricity, compressed air
and steam
Process modules structured

according to manufacturing
system logic
Production output, composition

of absolute energy consumption
(kWh), energy consumption costs
() and related CO2 emissions and
relative efficiency indicators
(different base units)
estimation of monthly electricity

costs (based on contract)
composition of consumption
Detailed analyses of electricity

consumption
196
Appendix
Graphical User Interface of Simulation Approach
Index
agent based simulation 48

aluminium die casting 145
application cycle 55, 94, 129
artificial neural networks (ANN) 43
automatic shutdown 104, 149
batch production 149
boiler 28, 115
brownfield 89
building 34, 38, 69
case study 145, 153, 161
climate 39
coal 17, 18
compressed air 2528, 95, 108, 155,
160
compressor 26, 27, 109
concept 89
conceptual framework 94, 97
consistency 3
consumers of energy 19
control system 38
conversion 14, 26
cost composition 24
criteria 52
cumulative load curves 95
cumulative load profiles 40
dimensioning 40
discrete event simulation 48
dynamic systems simulation 47
education 168
efficiency 4, 30, 124, 143
electricity 6, 17, 18, 23
electricity costs 24, 123
electricity supply 23
embodied energy 59
emissions 124
energy 12, 13
energy carriers 14, 19
energy consumers 19, 135, 145, 153, 162
energy consumption 18, 2022, 31, 122
energy consumption peaks 32
energy cost 5, 18, 93, 123
energy demand 22
energy efficiency 4, 5, 30, 31, 35, 36,
130, 131, 132
energy flows 23
energy form 13, 17
energy management system 130, 131
energy mix 16
energy portfolio 137, 154, 162, 163
energy prices 5, 6, 18
energy profiles 22
energy supply chain 15
evaluation and visualisation (EV)
module 96, 119
extended process model 38
factory 10, 32, 38
factory life cycle 89
failure 101, 122
failure behaviour 139
fuzzy logic 43
gas 6, 17, 18
graphical user interface (GUI) 112,
120, 168
green house gas (GHG) 4, 15, 123, 124
greenfield 89
holistic system definition 38
hybrid simulation 92
idle power 22
implementation 97
improvement potential 31
Index
198
Java applet 98, 168
key figures 37, 54, 93, 106, 119, 124
production planning and control

(PPC) 32, 63, 96, 117, 167
production processes 37
life-cycle 41
load curve 32
load management systems 33
load profile 21, 122 128, 134, 135, 153
requirements 3537, 89
research approaches 57
research demand 86
resource efficiency as 4
machine states 21
machines 38
main level MS module 127
manufacturing 9
manufacturing or main system
module 95
manufacturing systems 10, 52, 96
material flow networks 78
measurements 137
multi-criteria decision making 142
nominal pressure 109
nominal value 135
non-renewable resources 4, 15
obstacles 35, 36
oil 6, 17, 18
operation states 100, 108
operations management 9
optimal batch size 150
PCB assembly 161
peak 32
petri-net 75, 78
power 12
PPC module 117
process chains 10, 41
process modules 95, 100, 139
production 9, 10, 38
production engineering 10
production machines 21, 31
production management 11, 12, 41, 94,
96
sample manufacturing company 134

sampling rates 138
sankey diagram 126, 157
scenario 141
simulation 43, 45, 49, 130, 131, 141
SME (small and medium sized
enterprises) 5, 35, 87, 153, 161
specifications 89
state chart 100
steam 28, 114
steam generation 114
stochastic 124
sufficiency 3
sustainability 1, 3, 41
sustainable development 2, 3, 30
sustainable manufacturing 2, 87
switching operations 112
system dynamics simulation 47
TBS module 95, 108, 114, 139
technical building services (TBS) 20,
33, 38, 66, 69, 87
transferability 54
uncertainty 87, 124
utilisation 122
utilisation rate 106
validation 48, 50, 92, 140, 147, 156, 164
verification 48, 50, 92, 106, 113, 117,
119, 127, 128
weaving mill 153
weibull function 101

Energy Efficiency in Manufacturing Systems (2012) PDF

Uploaded by

Copyright:

Available Formats

Energy Efficiency in Manufacturing Systems (2012) PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Energy Efficiency in Manufacturing Systems (2012) PDF

Uploaded by

Copyright:

Available Formats

Sustainable Production, Life Cycle Engineering

Prof. Sami Kara

Joint German-Australian Research Group Sustainable Manufacturing and Life

For further volumes:

Sustainable Production, Life Cycle Engineering

Due to the increased economic and environmental concerns, a systematic

Prof. Christoph Herrmann

Prof. Sami Kara

List of Figures ..................................................................................................... XI

Theoretical Background ................................................................................. 9

Derivation of Requirements and Methodological Approach ................... 35

Concept Development .................................................................................. 89

Application of Concept .............................................................................. 145

Summary and Outlook .............................................................................. 171

References ......................................................................................................... 179

Fig. 1: Drivers for sustainability in manufacturing companies (adapted from

Fig. 41: Conceptual framework of ENOPA coupled simulation approach

Fig. 67: Verification results for steam module. ...................................................117

Fig. 97: Electrical power demand of PCB assembling company. .......................162

Table 1: Energy consumption for German producing industry with respect to

Table 21: Parameter lists of EV module. ............................................................121

List of Symbols and Abbreviations

scale parameter of Weibull function

List of Symbols and Abbreviations

heat input / combustion capacity

boiler output / boiler capacity

material processing rate

List of Symbols and Abbreviations

List of Symbols and Abbreviations

1.1 Sustainability as New Paradigm in Manufacturing

(e.g. research funding,

(e.g. Global Warming

(e.g. self commitment,

S. Thiede: Energy Efficiency in Manufacturing Systems, SPLCEM, pp. 18.

Therefore, besides traditional economical production objectives (e.g. cost, time,

1.1 Sustainability as New Paradigm in Manufacturing

Therefore, all technological and organisational measures within manufacturing

Use of resources / environmental impact

As shown in Figure 3 the sufficiency strategy may involve the conscious

harmful materials are avoided. While having significant improvement potential

Green house gas (GHG) emissions with significant contribution to global

Price index (2000=100%)

1.3 Objectives and Work Structure

1.3 Objectives and Work Structure

Contribution towards energy and

Easy to use, also for SME

all relevant energy flows

highly flexible, generic

specific means and characteristics to

Derivation of requirements and solution

Fig. 6 Hierarchy of objectives and related structure of the book

is not related or restricted to a specific case but generic in nature and

All these considerations are incorporated in an own innovative solution

2.1 Production and Production Management

Figure 7 underlines the understanding of production in context of

raw material (e.g. steel),

Input / Initial state

Output / Final state

Like any other process, a production process is a set of interrelated activities