GaBi Database & Modelling Principles 2012

Version 6.0 November 2012

Authors:
Dr. Martin Baitz, Dr. Cecilia Makishi Colodel, Dr. Thilo Kupfer, Julia
Pflieger, Dr. Oliver Schuller, Fabian Hassel, Morton Kokborg, Dr. An-
nette Köhler, Alexander Stoffregen
with contributions of the PE expert teams:
Energy, Transport, Metals, Chemicals and Polymers, Construction, Re-
newables, Electronic, End-of-Life, Water

PE INTERNATIONAL AG

Hauptstraße 111 – 113

70771 Leinfelden – Echterdingen
Germany
Phone +49 711 341817 – 0
Fax +49 711 341817 – 25
E-Mail info@[Link]
Internet [Link]
 List of Contents

List of Contents
List of Figures ....................................................................................................................................5
List of Tables ......................................................................................................................................7
Abbreviations .....................................................................................................................................8
1 Introduction and aim of document...............................................................................9
2 GaBi LCA database framework ..................................................................................10
2.1 GaBi Database concept and management................................................................................. 10
2.2 GaBi Database development, maintenance and update ............................................................ 12
2.3 Structure of the Master Database contents ................................................................................ 13
2.4 Standardisation, compliance and application issues of LCI databases ...................................... 20
2.5 Databases in reference networks, standards and principles ...................................................... 21
2.6 GaBi LCI Team........................................................................................................................... 22
3 Methodological framework .........................................................................................24
3.1 Definition of tasks in database work ........................................................................................... 24
3.2 Goal ............................................................................................................................................ 24
3.3 Scope ......................................................................................................................................... 25
3.3.1 Function and Functional Unit ...................................................................................................... 26
3.3.2 Definition of terms within system boundaries ............................................................................. 26
3.3.3 System boundaries for the creation of standard LCI cradle to gate datasets ............................. 27
3.3.4 Cut-offs ....................................................................................................................................... 34
3.3.5 Gap closing ................................................................................................................................ 35
3.3.6 Infrastructure .............................................................................................................................. 36
3.3.7 Transportation ............................................................................................................................ 38
3.3.8 Water .......................................................................................................................................... 38
3.3.9 Wastes and recovered material or energy.................................................................................. 46
3.3.10 Aspects of biomass modelling .................................................................................................... 47
3.3.11 Aspects of primary energy of fossil and renewable energy sources........................................... 48
3.3.12 Land Use Change ...................................................................................................................... 49
3.4 Sources and types of data .......................................................................................................... 52
3.4.1 Primary and secondary sources of data ..................................................................................... 52
3.4.2 Unit process and aggregated data ............................................................................................. 53
3.4.3 Units ........................................................................................................................................... 53
3.4.4 LCI data and supported LCIA methods ...................................................................................... 53
3.4.5 Production and consumption mix ............................................................................................... 54
3.5 Data quality issues ..................................................................................................................... 56
3.5.1 Technology and region coverage ............................................................................................... 56
3.5.2 Data aggregation ........................................................................................................................ 57
3.5.3 Precision, completeness, representativeness ............................................................................ 59
3.5.4 Consistency ................................................................................................................................ 60

3
 List of Contents

3.5.5 Uncertainty ................................................................................................................................. 60

4 System Modelling Features ........................................................................................66
4.1 Data collection ............................................................................................................................ 66
4.1.1 Quality check and validation of collected data............................................................................ 67
4.1.2 Treatment of missing data .......................................................................................................... 67
4.1.3 Transfer of data and nomenclature ............................................................................................ 67
4.2 Geographical aspects of modelling ............................................................................................ 69
4.3 Parameter................................................................................................................................... 70
4.4 Multifunctionality and allocation principle ................................................................................... 70
4.5 Generic Modules as background building block ......................................................................... 71
4.6 Special modelling features for specific areas ............................................................................. 72
4.6.1 Energy ........................................................................................................................................ 72
4.6.2 Transport .................................................................................................................................... 77
4.6.3 Mining, metals and metallurgy .................................................................................................... 82
4.6.4 Chemistry and plastics ............................................................................................................... 84
4.6.5 Construction ............................................................................................................................... 88
4.6.6 Renewables................................................................................................................................ 90
4.6.7 Electronics .................................................................................................................................. 94
4.6.8 Recycling or End-of-Life measures ............................................................................................ 97
5 Review, documentation and validation ...................................................................104
5.1 Review procedures and check routines.................................................................................... 104
5.1.1 Technical information and documentation routines in GaBi ..................................................... 104
5.1.2 Important material and energy balances .................................................................................. 105
5.1.3 Plausibility of emission profiles and avoiding errors ................................................................. 105
5.2 Documentation ......................................................................................................................... 106
5.2.1 Nomenclature ........................................................................................................................... 106
5.2.2 Documentation of Flows ........................................................................................................... 107
5.2.3 Documentation of LCI process data ......................................................................................... 107
5.3 Validation.................................................................................................................................. 109
6 Literature ....................................................................................................................110
Supplement A Description of result and impact categories....................................................................... 115
Supplement A 1 Primary energy consumption ............................................................................................................ 116
Supplement A 2 Waste categories .............................................................................................................................. 117
Supplement A 3 Global Warming Potential (GWP) ..................................................................................................... 117
Supplement A 4 Acidification Potential (AP) ............................................................................................................... 121
Supplement A 5 Eutrophication Potential (EP) ........................................................................................................... 123
Supplement A 6 Photochemical Ozone Creation Potential (POCP) ........................................................................... 125
Supplement A 7 Ozone Depletion Potential (ODP) ..................................................................................................... 127
Supplement A 8 Human and eco-toxicity, USEtox ...................................................................................................... 129
Supplement A 9 Resource depletion ........................................................................................................................... 132
Supplement A 10 Particulate matter formation (PM) ............................................................................................... 134
Supplement A 11 Normalization .............................................................................................................................. 135
Supplement A 12 Weighting .................................................................................................................................... 135
Supplement B Background information on uncertainty ............................................................................. 137

4
 List of Figures

List of Figures
Figure 2-1: GaBi Database concept embedded in 3 pillar solution approach ................................... 10
Figure 2-2: Database Management at PE INTERNATIONAL .......................................................... 11
Figure 2-3: Hierarchical system in GaBi ........................................................................................... 13
Figure 2-4: Aggregated dataset in GaBi ........................................................................................... 14
Figure 2-5: Polymerisation subsystem in GaBi Master DB ............................................................... 15
Figure 2-6: Tetraflourethylene subsystem in GaBi Master DB. ........................................................ 16
Figure 2-7: R22 subsystem in GaBi Master DB. ............................................................................... 17
Figure 2-8: Chlorine production mix in GaBi Master DB. .................................................................. 17
Figure 2-9: Chlorine membrane technology production in GaBi Master DB. .................................... 17
Figure 2-10: Refinery model in GaBi Master DB. ............................................................................... 18
Figure 2-11: Crude oil import mix and country specific oil extraction in GaBi Master DB. .................. 18
Figure 2-12: Power plant models of the Grid Mix modelling in GaBi Master DB. ............................... 19
Figure 2-13: German Natural Gas Mix in GaBi Master DB. ................................................................ 19
Figure 2-14: German Natural Gas production in GaBi Master DB...................................................... 20
Figure 2-15: GaBi DB in the international context of databases and frameworks. ............................. 22
Figure 2-16: GaBi LCI Expert Teams and the core LCA content team ............................................... 23
Figure 3-1: Graphic representation of different (sub-) system boundaries ....................................... 26
Figure 3-2: Generic example product system of a dataset development ......................................... 27
Figure 3-3: Application water flows in open-loop and closed-loop cooling systems in various
industrial settings ........................................................................................................... 42
Figure 3-4: Application water flows in once-through cooling and cooling towers in
energy/electricity generation .......................................................................................... 44
Figure 3-5: Application water flows in hydropower generation ......................................................... 45
Figure 3-6: Ad hoc example of a simple plan system including different processes and water
flows ............................................................................................................................... 45
Figure 3-7: Difference between "production mix" and "consumption mix" (for power
generation) ..................................................................................................................... 55
Figure 3-8: Principle graphical explanation of the relation of completeness, precision .................... 58
Figure 4-1: Hierarchical structure of the processes and plans ......................................................... 69
Figure 4-2: Conventional natural gas production in Germany .......................................................... 73

5
 List of Figures

Figure 4-3: Natural gas supply for Germany .................................................................................... 74

Figure 4-4: US, East electricity grid mix ........................................................................................... 75
Figure 4-5: Parameterized US Coal gas CHP power plant............................................................... 76
Figure 4-6: Excerpt of the organic network considered in GaBi ....................................................... 85
Figure 4-7: Consumption mix of Epoxy resin in Germany ................................................................ 86
Figure 4-8: Example of PVC resin - compound- part ........................................................................ 88
Figure 4-9: Schematic life cycle of a building ................................................................................... 89
Figure 4-10: Nitrogen in the agrarian system ..................................................................................... 92
Figure 4-11: Creation of a model for an electronic product - modular structure via Generic
Modules ......................................................................................................................... 96
Figure 4-12: Exemplary incineration model with in GaBi (here average European domestic
waste treatment with dry offgas cleaning) ...................................................................... 99
Figure 4-13: Details of incineration and dry offgas cleaning in GaBi incineration model .................. 100
Figure 4-14: Exemplary landfill model in GaBi (here commercial waste composition for certain
geographic example regions) ....................................................................................... 101
Figure 4-15: Exemplary wastewater treatment model in GaBi (here municipal wastewater for
German circumstances) ............................................................................................... 103
Figure 5-1: Example documentation in GaBi (excerpt) [GABI 2012] ............................................... 108

6
 List of Tables

List of Tables

Table A: Background system boundaries ..................................................................................... 28

Table B: Relevancy of infrastructure of a natural gas power plant in GaBi Master DB
(selected representative sample power plant)................................................................ 37
Table C: Publicly available example value for a medium size gas power plant ............................ 37
Table D: Publicly available example values for CO2 for a gas power plant .................................. 37
Table E: General treatment procedure (if no specific information is available) for common
materials/wastes ............................................................................................................ 47
Table F: General procedure for some hazardous waste flows ..................................................... 47
Table G: Overview of the Land Use Change Indicator Units ........................................................ 50
Table H: Overview of qualitative importance of “quality indicators” in GaBi DBs .......................... 56
Table I: Coefficients of variation .................................................................................................. 64
Table J: ILCD set of recommended impacts .............................................................................. 116
Table K: Normalization references ............................................................................................ 135
Table L: PE Weighting 2012 ...................................................................................................... 136
Table M: Chemical substance datasets available for various countries in GaBi ......................... 138
Table N: Chemical substance datasets available for various technology routes in GaBi ........... 140

7
 Abbreviations

Abbreviations

AP Acidification Potential
ADP Abiotic Depletion Potential
B2B Business-to-Business
B2C Business-to-Customer
CHP Combined Heat and Power Plant
DeNOx NOx emission reduction
DeSOx SOx emission reduction
DB Database
ELCD European Reference Life Cycle Data System
EoL End-of-Life
EP Eutrophication Potential
FAETP Freshwater Aquatic Ecotoxicity Potential
GWP Global Warming Potential
HTP Human Toxicity Potential
IEA International Energy Agency
IPCC International Panel on Climate Change
ILCD International Lifecycle Reference System
KEA Cumulated Energy Approach (Kumulierter Energieaufwand)
LCA Life Cycle Assessment
LCI Life Cycle Inventory
LCIA Life Cycle Impact Assessment
MAETP Marine Aquatic Ecotoxicity Potential
MSW Municipal Solid Waste
NDA Non-Disclosure Agreement
ODP Ozone Depletion Potential
POCP Photochemical Ozone Creation Potential
SCR Selective catalytic reduction (DeNOx type)
SNCR Selective non catalytic reduction (DeNOx type)
TETP Terrestric Ecotoxicity Potential
WtE Waste-to-Energy

8
 Introduction and aim of document

1 Introduction and aim of document

Consistency, Relevancy, Quality, Continuity are the main drivers in the GaBi database. The GaBi databases
contain over 300 person-years of direct data collection and compilation. PE staff with an experience of over
1000 person-years contributes constantly to the management and development of the GaBi databases.
The goal of “GaBi Database and Modelling Principles” is to transparently document the environment, back-
ground, important aspects and details of the GaBi databases, as well as the basis of the models.
Furthermore, the database management is described – an important aspect of professional databases in
practise – as well as update and maintenance procedures and strategies.
Finally, important, general, methodological aspects and branch- or expert-specific methodological aspects are
documented.
This document is the basis of all GaBi databases, the professional database and the extension databases, as
well as all data-on-demand datasets.

9
 GaBi LCA database framework

2 GaBi LCA database framework

Successful, continuous and effective database provision needs
 a professional database concept and management,
 consistent and central database development,
 database maintenance as well as frequent and efficient update routines.
To enable a flexible use of the database content in different Life-Cycle- related applications, Life-Cycle Man-
agement tasks and professional decision situations, the data should be suitable and adaptable to different
schemes and standards of industrial and professional practise and should most importantly reflect the real
supply chain and technology situation. Well-educated and broadly experienced teams of different branches
and expert groups with broad experience in their areas of expertise are important.
The methods and methodological choices used have been selected to reflect the supply networks in the most
appropriate way. “Method follows reality.”

2.1 GaBi Database concept and management

The database concept is embedded into the solution concept of PE INTERNATIONAL. This Master DB con-
cept is one core pillar of our three pillar solution approach, which is accompanied by engineering consulting
knowledge and a professional software environment.

Figure 2-1: GaBi Database concept embedded in 3 pillar solution approach

Database development at PE involves experts on LCA methodology, with technical expertise (see Chapter 2.6
for details on the different teams), as well as knowledge of the supply chain. Relevance checks and routine
quality assurance checks are applied. The generation of new data follows a standard procedure with “cascade
quality checks” and is embedded into the GaBi Master DB concept.
Internal entry quality checks: Newly generated data first passes a purely internal quality check, by two LCA
experts with engineering skills at PE INTERNATIONAL, in a dependent internal review, before entering the
database environment.
Internal resulting quality checks: Depending on the type of data and its intended use, field of expertise and
the sources providing the data (internal or external sources and/or organisations), a second check or valida-
tion by our cooperative partners LBP University of Stuttgart and Fraunhofer IBP or other independent organi-
sations are undertaken, as an independent internal review.

10
 GaBi LCA database framework

External resulting non-public quality checks: Data, which is generated together with industry or associa-
tions for distribution with GaBi databases into the professional LCA user community (e.g. Eco-profile-type data
or other representative averaged industry data of different companies or an individual dataset of single com-
panies), undergo an additional quality check by the respective data providers or selected neutral third party
organisations, as an independent external review or third party review.
External resulting public quality checks: The dataset and systems, which are provided with our software
and databases for public use to a broad user community, are constantly used, compared, benchmarked,
screened and reviewed, and the results are published in various external, professional and third party LCA
applications in industry, academia and politics. User feedback via the online GaBi forum or direct via user
information is standard in the maintenance and update process of the databases and leads to consistent
quality, constant control and improvement of data, if knowledge or technology improves, or industrial process
chains develop and change, as independent public or “cloud” reviews.
The Database Management at PE INTERNATIONAL protects private and project-related information of cus-
tomers and clients, while enabling customers and clients to best benefit from the general, usable part of the
internal information and knowledge and expertise pool of PE.
Any information that leaves the PE internal database area needs release permission and is centrally distribut-
ed.

Figure 2-2: Database Management at PE INTERNATIONAL

Any confidential project or customer-related information is protected by a “Non-Disclosure Agreement (NDA)”

and securely separated from any publicly available database.

11
 GaBi LCA database framework

2.2 GaBi Database development, maintenance and update

After observation of the LCA developments over the last 20 years it becomes obvious that the motivation to
start using LCA approaches two decades ago is industry-driven. Naturally all data should ideally be “industry
borne,” meaning the data are validated or sourced from or within industry to ensure the proper representation
of real circumstances. So “data” is a typical topic for LCA work with and in industry.
The need for sound methodological approaches within database and LCI modelling has attracted the academ-
ic world and later, standardisation bodies.
LCA databases began appearing in the early 1990s - some with software systems. GaBi was the only pioneer
having both database and software system at hand from the beginning, opening up synergies and unique
possibilities.
LCA Databases are growing in relevance. GaBi databases evolved and established LCA in daily use early
within research and industry. Only professionally managed, maintained and updated databases continue to be
relevant in practise.
Maintaining and updating databases is an important, although time- and management-intensive task. Accura-
cy of data, new (practical, proven) methods and user needs are just three examples that need constant atten-
tion. And constant attention requires a consistent group of people taking care of specific topics and branches.
 New scientific findings, new data and technologies and new methods need constant database
development.
 Decisions for development of new products based on LCA and optimisation or investment de-
pend on reliable results, applicability and continuity in daily practise.
GaBi database development combines the important aspects of these requirements. The GaBi database
employs proven “best practice” data and approaches. New scientific methods and data are applied only after
feasibility checks to reduce risks of wrong (product or process) decisions. “Best practice” is based on the
“latest science.”
The GaBi database work is considered to be done “for practice with information from practise” and therefore
considers the “critical success factors” in professional LCA applications. GaBi data is therefore not just any
randomly available data, but rather best practise data.
We have access to raw data sources developed by PE and in-house engineering expertise, which enables PE
to develop and deliver within scope, on time, with high quality and provide guidance towards suitable data and
data selection. A standard format for all LCI datasets is mandatory for all PE-owned data.
PE considers its data to be “Industry-borne” as we aim for stakeholder involvement and validation on data
developments with suitable contacts, sources and published industry data. PE always welcomes feedback,
constructive criticism and suggestions for improvement.
We model real supply chains for inter-sectoral use for all B2B and B2C relationships. The data reflect specific
and up-to-date technology and technology routes for individual branches. Region-specific background sys-
tems are combined, wherever suitable and possible, with local/regional process technology information. Indi-
vidual, user-specific modification, adaptation and extension on local situations with customer-owned data or
parameterised data are possible. Individual data on demand can be constructed according to the same rou-
tine for best consistency.

12
 GaBi LCA database framework

The ultimate goal is to attain flexibility in the application of our data to address different topics and allowing
flexible assimilation between political and industrial decision contexts. In other words the same database can
be used in making a quick decision or in an intensive investment decision.
Regarding development, maintenance and update environment, a suitable group structure (see Chapter 2.6
for details) with different responsibilities at PE is in place. There is a direct relationship between software and
database development, which supports practical and relevant solution pathways, seeing as many issues deal
with both fields.
Maintenance and support routines are installed and updates are regularly conducted with the least possible
user effort required including smart database/software updates with automated addition of new standard LCI
or LCIA data.

2.3 Structure of the Master Database contents

The Master Database is the core data knowledge memory and contains about 10,000 generic plan systems,
each with one or more unit processes and several sub-systems.
In some cases single cradle-to-gate systems involve several thousand individual plan systems and tens of
thousands of individual processes tracing back to the resources.

Figure 2-3: Hierarchical system in GaBi

Each PE-owned, aggregated process provided in the public available databases has a corresponding plan
system, unit processes and sub-systems with sub unit-processes in the Master Database.
Huge systems result, which are hardly manageable without suitable LCA software support. In principle it
would be possible to display all sub-systems of all processes and plans of the complete Master DB. The re-
sulting document would probably have about a quarter of a million pages1. This is one main reason why GaBi
and its corresponding Master database were developed: To be able to transparently and simply manage and
use large process chain systems of real supply chains.

1
Rough estimate assuming two screenshots per page.

13
 GaBi LCA database framework

The graphical display for this document is therefore limited to an example. It aims to transparently document
the structural background of the Master Database. Further publicly available process chain and technology
information on all datasets and systems is covered in the documentation.
We offer to share more details and process chain knowledge through bilateral business relationships.
The publicly-available databases contain plan systems, unit processes, partially aggregated processes and
aggregated processes.

Figure 2-4: Aggregated dataset in GaBi

Aggregated processes are often the only way to provide relevant, suitable and up-to-date information of indus-
trial sources to the LCA user community. Many users consider aggregated processes the best way to reliably
and representatively model existing background systems.
PE has added value from unit process data collection and compilation, including verification of technical real-
istic boundaries, to country-specific supply chain modelling.
Opening the first level of the related polytetraflourethylene production in the Master database shows the
polymerization step with the respective unit process in the centre. Upstream sub-systems are shown on the
right (in the unit process only technical flows are visualised; elementary flows such as resources or emissions
are not visualised, but definitely physically and mathematically present in the individual unit processes).

14
 GaBi LCA database framework

Figure 2-5: Polymerisation subsystem in GaBi Master DB

We follow one single upstream pathway from Tetraflourethylene (indicated by the red circle; details are shown
in the next figure)…

15
 GaBi LCA database framework

Figure 2-6: Tetraflourethylene subsystem in GaBi Master DB.

…to R22 details and chlorine mix details (marked in red)…

16
 GaBi LCA database framework

Figure 2-7: R22 subsystem in GaBi Master DB.

Figure 2-8: Chlorine production mix in GaBi Master DB.

… then to chlorine membrane technology details (marked in red) and back to rock salt mining.

Figure 2-9: Chlorine membrane technology production in GaBi Master DB.

The previous example showed the journey from polymer back to rock salt. The following example gives insight
to the fossil fuel and organic process chain. Starting with the various refinery products diesel, gasoline, naph-
tha and gases on the right side….

17
 GaBi LCA database framework

Figure 2-10: Refinery model in GaBi Master DB.

… the refinery products progress through the different refinery stages to the crude oil input on the left….

Figure 2-11: Crude oil import mix and country specific oil extraction in GaBi Master DB.

…and from the right side of crude oil import mix to country-specific oil extraction and the bore hole at the
source.

The last example shows the electricity modelling in GaBi Master Database.

18
 GaBi LCA database framework

Figure 2-12: Power plant models of the Grid Mix modelling in GaBi Master DB.

The output which results on the right side of above screenshot is 1 kWh of electricity. On the right next to the
hydro, wind, waste and nuclear power plants, the necessary fuels (hard coal, lignite, oil and natural gas)….

Figure 2-13: German Natural Gas Mix in GaBi Master DB.

…which are provided by the German consumption and import mix of natural gas…

19
 GaBi LCA database framework

Figure 2-14: German Natural Gas production in GaBi Master DB.

…can be traced all the way back to the natural gas production at the source.
The above screenshots represent only a very small amount of the total process chain network involved in the
chosen PTFE example.
In summary we can conclude that a pre-calculated dataset integrates a large amount of valuable information,
which would otherwise be barely manageable.
Thousands of pre-modelled, real world subsystems and engineering information are included. Data collection
time, industry research and compilation and consistency checks create real B2B supply chains. Knowledge of
technical aspects of supply chains has been documented, along with the approximately 300 person-years
work on the database and content.

2.4 Standardisation, compliance and application issues of LCI databases

The customer or case specific foreground model must be compliant to the desired approach in first instance.
GaBi supports in various ways due to its flexible modelling features.
GaBi Databases are developed for use within different situations and applications as upstream, downstream
and background data and seek to be in line with relevant existing standards, reference documents and best
practise documents.
In this context we primarily consider:
 LCA / LCI / LCIA: [ISO 14040 : 2006, ISO 14044 : 2006]
 Ecolabels [ISO 14020 : 2000 ] , Typ II [ISO 14021:1999], Typ III [ISO 14025:2006],
Environmental product declarations (EPD) [ISO 21930:2007], [ISO 15804], Institute

20
 GaBi LCA database framework

Construction and Environment [IBU 2011], Fiches de Déclaration Environnementales et

Sanitaires (FDES) [NF P 01 010 : 2004]
 Greenhouse Gases / Carbon Footprint: [ISO 14064-1:2006], [ISO 14067], GHG Protocol Corpo-
rate Value Chain (Scope 3) [GHGPC 2011] and Product Life Cycle [GHGPP 2011], [PAS
2050:2011]
 Carbon Disclosure Project (CDP)
 Environmental Management ISO 14001, EMAS II, EMAS III
 Database reference systems and guidelines: ILCD reference system [ILCD 2010],
SETAC/UNEP Global Guidance on databases [UNEP/SETAC 2011], Eco-profiles and Environ-
mental Declarations, PlasticsEurope [PLASTICSEU 2011]
 CDP Water Disclosure and Water Footprint Network Manual
Because LCA is a multi-function/multi-application method, the GaBi data is generally developed to be used
consistently within the aforementioned framework (please visit also [Link]
[Link]/international/solutions/ for further details). It may be possible that some frameworks defined in
certain specific applications result in contrary requirements, that one single background dataset cannot match
both by default. Therefore the GaBi system supports and allows for specific addition/modification/adaptation of
the dataset.

2.5 Databases in reference networks, standards and principles

GaBi databases are known to be of practical relevance and therefore are often used to support different initia-
tives, industry or national databases and schemes. And vice versa, initiatives, industry or national databases
and schemes influence GaBi databases. This coexisting symbiosis enables practicability, applicability, com-
patibility and distribution of data within the relevant professional frameworks. The following graph illustrates
the dependencies within this coexisting symbiosis.

21
 GaBi LCA database framework

Figure 2-15: GaBi DB in the international context of databases and frameworks.

Potential data and metadata flows are visualized among the different professional frameworks.

2.6 GaBi LCI Team

GaBi databases result from the teamwork of 10 expert teams and one core content team. Each expert team is
responsible for modelling its specific system, as well as documenting the generated LCI. Each team requires
experts which have a solid background in the following fields:
 Technical knowledge specific to the given industry branch
 Performing LCAs and specifically having experience in analysing technical production routes
 Good understanding of the analysed production technologies applied for material production
and/or power generation
 Sensitivity to the industry’s current state and having an appropriate understanding of the role of
LCA within industry
 Self-directed work in effective cooperation with industry
The coordination of all expert teams is the task of the core content team.
The content team provides the technical platform and methodological guidelines to all expert teams to ensure
a consistent and synchronized database in the end. It also serves as an interface to clients, to the market and
to the scientific community to receive feedback on the existing database, to make sure the GaBi databases
are in line with the development of methodologies and the demands of the market, as well as to constantly
improve the internally-used work flow and guidelines. In this way consistency throughout all GaBi databases
can be assured.

22
 GaBi LCA database framework

Energy Chemistry Plastics Metals Renewables End-of-life

Expert teams provide

data input product LCIs or unit processes as well
Coating &
Transport as documentation
Textile

Electronics GaBi Master Database Construction

Content team

Provision of technological platform and

methodological guidance

Feedback from clients Creation of final database

and market content (after validation)

Figure 2-16: GaBi LCI Expert Teams and the core LCA content team

23
 Methodological framework

3 Methodological framework
This chapter summarises important methodological principles, which are applied in GaBi database modelling
and are utilised if new datasets are developed or existing datasets are updated for Life Cycle Inventory (LCI)
purposes.

3.1 Definition of tasks in database work

Database work can be separated into the following categories:
 Data and database development and set-up
 Data and database maintenance
In Data and Database development new LCI data and databases are produced using suitable raw data
sources and appropriate methodological approaches to set-up new data the first time in line and consistent to
existing data.
Data and Database maintenance keeps existing LCI data and databases constantly up-to-date in terms of
(relevant and practically proven) evolutionary aspects of data formats, flow formats, flow hierarchies and
methodological findings and to correct possible errors. Data and Database maintenance further involves fre-
quent upgrades on new technological background information of unit processes, upstream technology infor-
mation and technology routes, consumption and production mix figures for commodities, new impact factors,
as well as all new combined software-database functions that enable the use of generic data in a broader,
more flexible and extended way.
For any of the above mentioned tasks in database work we use the phrase ”modelling”.
These modelling processes follow principally the known steps of an LCA and contain the following main steps:
 Goal, Scope and System boundaries
 Data collection/validation/check and system modelling
 Data quality requirements and checks
 Documentation and publication
The “GaBi Database and Modelling Principles” are the basis for consistent database work. These guidelines
address the important points but are not exhaustive. Transferring theory into practise requires interpretation
and experience and, as a result, a degree of responsibility is held by the practitioner.

3.2 Goal
The results of an LCA study, as a rule, are related to a specific question. Therefore, the goal definition of an
LCA study is of vital importance.
In the development of generic and representative (single) datasets, deciding on the goal of the dataset is of
vital importance.
The main goal of all datasets in GaBi is to reflect the reality of our industrial and business networks and to be
as flexible as possible to address all different aspects.

24
 Methodological framework

GaBi datasets therefore incorporate best available practise and information from internal or external sources.
Consistency is important in that all sources used fit with each other and verify the final resulting data with
existing data and our engineering knowledge.
Concerning the ISO standards [ISO 14044 : 2006], the goal of GaBi data can be understood as follows:
 Intended application: All practical life cycle-related applications that aim to maintain links to-
wards or are based upon the ISO 14040/44 series.
 Reasons: Not applicable in the generic data context. Reasons to be specified within context of
the system.
 Intended audience: All LCA practitioners in industry, research, consulting, academia and politics
that aim to base their individual work on relevant data based in reality.
 Comparative assertions: No comparative claims are intended or supported on solely an invento-
ry level from the database level. The databases are a consistent compilation of different da-
tasets per functional unit, but direct comparison on the database level is not appropriate be-
cause proper (user case-specific) modelling is needed. The user is, however, able to take data
and set up comparative assertions disclosed to public, which is its own responsibility.

3.3 Scope
The scope of the dataset and data systems depend on the type of dataset requested (see Gate to Gate, Cra-
dle to Gate and Cradle to Grave2).
In most cases the complexity of the answer or result interpretation is strongly dependent on the degree of
desired general validity of the answer or result interpretation.
Models of specific circumstances tend to be described with less complex systems, fewer possible varying
circumstances or sensitivities that must be addressed. However, specific circumstances often call for more
specific data.
Models of general circumstances tend to be described with more complex systems, because more possible
varying circumstances or sensitivities must be addressed. Circumstances that are more general enable the
use of more generic data.
In other words: For specific results or a specific company product, specific foreground primary data from the
related company is needed. For general results concerning an average product, generic background data can
be suitable and for unspecific results, such as sector-related results, even more general data (such as I/O
table-type LCA data) can be used.
To avoid misinterpretation due to the use of data and datasets, the type of data and its boundaries, the specif-
ic product systems and its upstream technology routes must be documented and understood. The GaBi da-
taset and the related documentation of the GaBi dataset provide the necessary information to avoid misinter-
pretation.

2
To avoid confusion by using any “vogue terms” of non-standardised concepts and visions the well-known and established
term “Cradle to Grave” is used. The broadly used “Cradle to Grave” approach is able to include all kind of End-of-Life op-
tions and recycling options. So the “Cradle to Grave” approach is used to model all kind of cycles and recycling issues and
is not used in contrast to any other method, as all aspects of technical and natural cycles e.g. like carbon, water and nutri-
tions can be covered.

25
 Methodological framework

3.3.1 Function and Functional Unit

The functional unit is a “quantified performance of a product system for use as a reference unit” in a life cycle
assessment study [ISO 14044 : 2006]. It should be representative of the goal of the dataset/data system and
should allow the comparison of similar systems, processes or products, if needed.
In GaBi datasets the goal of functional unit is always defined as the related output product flow. Depending on
the product, the functional units used in the GaBi databases [GABI 201] are essentially physical metric [SI]-
units related to the amount of product, e.g. 1 kg, 1 MJ, 1000 kg, 1 m3. The functional unit of each process is
defined within the process. The choice of the SI-unit does not influence the results of a comparison, seeing as
all compared systems can be described in the chosen SI-unit.

3.3.2 Definition of terms within system boundaries

Within this sub-chapter the different bases for the data collection and system modelling (building up the LCI
dataset) is described. The system boundary defines what is included in the dataset and depends on the kind
of dataset: a ‘gate to gate’ unit process, a ‘cradle to gate’ aggregated or a ‘cradle to grave’ aggregated da-
taset.

Figure 3-1: Graphic representation of different (sub-) system boundaries

Figure 3-1: Graphic representation of different (sub-) system boundaries

is a pictoral representation of the system boundary definitions.
 Gate to Gate: All company or site-related activities from material acquisition or procurement,
beginning at entrance gate through all the production steps on site, until final commissioning
steps before leaving the site gates again.
 Cradle to Gate: All activities from resource mining through all energy and precursor production
steps and on site production, until final commissioning steps before leaving the site gates.

26
 Methodological framework

 Cradle to Grave: Cradle-to-Gate extended through the use, maintenance and the end of life
(disposal, recycling, reuse) of a product.
During development of a dataset the system boundaries can be subjected to step-by-step adjustments due to
the iterative nature of data system set up and validation procedures.

Figure 3-2: Generic example product system of a dataset development

Figure 3-2 gives an example of an example product system. Elementary flows enter and leave the system
environment, as do product flows to and from other systems. Included within the system environment are
different transports, energy supply, raw material acquisition, production, use, recycling/reuse, and waste
treatment, depending on system boundaries. The respective system boundaries are defined by the type of
dataset.

3.3.3 System boundaries for the creation of standard LCI cradle to gate datasets
Within this section the system boundaries for the generation of standard life cycle inventories are described.
System boundaries are defined by the included and excluded processes of the foreground and background
systems.
The foreground system boundaries are described in the documentation of the GaBi dataset ([Link]
[Link]/international/databases/).
The background system boundaries of the GaBi datasets are described in the following tables.
The models are set-up over hundreds of engineering parameters in the software, which would be difficult to
list, thus is one reason why GaBi relies on the combined software-database approach to couple functionality

27
 Methodological framework

with precision. PE offers the opportunity to share more details and process chain knowledge under bilateral
business relationships. In the following tables the system boundaries of the main operations in the back-
ground system of GaBi dataset are documented.

Table A: Background system boundaries

within system boundary3 outside system

Crude oils and primary, secondary and tertiary production per country offshore supply ves-
natural gases sels, onshore drilling
onshore processes of exploration and drilling per country
transports and some
offshore processes of exploration and drilling per country minor drilling chemi-
resource extraction cals

venting and flaring emissions

drilling meter length
generators (diesel/gasoline) and electricity
thermal and mechanical energy
water use and wastewater treatment
waste and hazardous waste treatment
share of spilled crude oil from well testing
share of vented natural gas from well testing
bentonite and barium sulphate use
infrastructure
see also [Link]

3
If relevant in the context of the country- or technology specific data system

28
 Methodological framework

Table Background System Boundaries (continued)

within system boundary3 outside system

Coals and open pit operations per country production of convey-
Lignites ers and mining vehi-
under ground operations per country
cles
soil removal and digging
overburden
mining trucks and excavators
conveyors
water pumping
water use and wastewater treatment
air conditioning
explosives
dust and explosion emissions
specific pit methane, CO2, chloride
fuels and electricity
Power plants all relevant combustion and off gas cleaning steps (see screen- construction processes
(electrici- shot in Chapter 2.3) per country of power plant
ty/heat)
power plant park per country
fuel characteristics per country
imports of other countries
all relevant emission country and technology specific
DeNOx and DeSOx units
electricity/heat shares
distribution losses
off ags treatment chemicals
infrastructure
see also [Link]

29
 Methodological framework

Table background system boundaries (continued)

within system boundary3 outside system

Refinery all relevant refining steps, approx. 30 different (see screenshot Construction and infra-
operations in Chapter 2.3) per country structure
crude oil characteristics per country
H2 production in reformer and use
external H2
process water
all relevant refining emissions per country
desulphurisation and treatment
internal energy management
methanol, bio-methanol
product spectrum of 21 products per country
see also [Link]
Mining ores ores concentrations and combined ore shares per country production of convey-
and minerals ers and mining vehi-
open pit operations
cles
under ground operations
soil removal and digging
landfill overburden
mining trucks and excavators
conveyors
water pumping
water use and treatment
air conditioning
explosives
dust and explosion emissions
thermal energy propane
fuels and electricity

30
 Methodological framework

Table background system boundaries (continued)

within system boundary3 outside system

Ore benefication process chemicals infrastructure and
machinery
fuels and electricity
thermal energy
process water
wastewater treatment
ammonium sulphate use
waste and tailings treatment
end of pipe measures and emissions

Metal smelter, electricity specific per electrolysis infrastructure and

electrolysis and materials of facilities
silica use, oxygen use
raffination
compressed air
coke and related reduction media
waste and slag treatment
hazardous waste treatment
auxiliary chemicals, caustics, chlorine, HCl, formic acid, soda,
ammonia
thermal energy LPG, naphtha use
water use and wastewater treatment
see also [Link]
[Link]/international/databases/
Chemical Syn- all relevant educts or monomers some catalysts of con-
thesis, Formula- fidential or patented
electricity specific per reaction type
tions and composition and mate-
Polymerisations thermal energy use or production rials of reactors and
waste treatment facilities

hazardous waste treatment

auxiliary chemicals
water use and wastewater treatment
purge purification of recycling (if any)
see also [Link]

31
 Methodological framework

[Link]/international/databases/
Mineral pro- all relevant mineral inputs and fuels infrastructure and
cessing and kiln materials of machinery
electricity specific per kiln and operation type
processes
thermal energy
waste and hazardous waste treatment
end-of-pipe operations
auxiliary chemicals
water use and wastewater treatment
particle and combustion emissions
see also [Link]
[Link]/international/databases/
Agrarian prod- CO2 uptake, sun light and nitrogen balance farm infrastructure and
ucts and renew- materials of machinery
rain water, irrigation water, water pumping
ables
individual pesticides per crop
individual fertilizers per crop
land use and reference systems
fertilizing effects of by-products
tillage and all related soil preparation
tractor and all related machinery
transports to field border / farm
electricity and fuels for cultivation
electricity and fuels for harvesting
see also [Link]
[Link]/international/databases/

32
 Methodological framework

Table background system boundaries (continued)

within system boundary3 outside system

Electronic prod- NF-metal and precious metal materials infrastructure and
ucts and com- materials of machinery
polymer and resin components
ponents
Solders
housing and frames
fire retardant
printed wiring boards
processing and assembly
Etching and processing chemicals
see also [Link]
[Link]/international/databases/
Water supply water withdrawal and pumping infrastructure and
materials of machinery
mechanical and chemical (pre-) treatment
chemicals for processing (ClO2, O3,...)
electricity and thermal energy technology specific
reverse-osmosis and membrane technology
see also [Link]
[Link]/international/databases/
EoL water mechanical and chemical (pre-) treatment infrastructure and
treatment materials of machinery
chemicals for processing (ClO2, O3,...)
sludge and slag treatment (fertilizer or incineration)
see also [Link]
[Link]/international/databases/
EoL incineration waste input specific (composition, calorific value) infrastructure and
materials of machinery
fuels, co-firing, combustion, boiler, SNCR/SCR
active filter, end-of-pipe, DeSOx
chemicals, water
Efficiency and energy recovery (electricity/heat)
Combustion calculation incl. all relevant emissions
see also [Link]

33
 Methodological framework

[Link]/international/databases/

All datasets of commodities and products are modelled within the foreground system boundaries described in
the documentation and within the background system boundaries described above. For any of the PE-owned
datasets, the underlying plan systems are accessible in the Master database and PE can grant access rights
(e.g. for review purposes) under bilateral contracts. PE Master database content is valuable, privately-
financed information, developed, collected and compiled with a tremendous amount of recourses and costs
with no public funding. It is therefore not possible to grant free public access to the Master DB in its totality.

3.3.4 Cut-offs
Cut-off rules are defined to provide practical guidelines to be able to omit specific less relevant process chain
details, while creating a specific product system. The ISO 14044 : 2006 mentions three criteria used to decide
which inputs are to be included: a) mass, b) energy and c) environmental significance.
There are three different types of cut-offs:
1. A known input or substance is not connected to an upstream process chain due to lack of information
2. A known inconsistency in a mass or energy balance with a known reason
3. An unknown or known inconsistency in a mass or energy balance with an unknown reason
The GaBi database has very few cut-offs of type 1). The only two reasons for cut-offs of type 1) are starting
conditions (at the very beginning of the supply chain) or confidentiality reasons of competitive formula-
tions/substances (see table in Chapter 3.3.3). Due to the magnitude of the database content, most information
is available or can be developed. If a substance for which no LCA data exists is needed and is not available
as a dataset, the GaBi Master database uses information for a chemically/physically-related substance and
creates a worst case scenario for the substance causing the gap. If the influence of the worst case scenario
on the overall result is smaller than 5%, the worst case scenario can stay (gap-closing insignificantly overes-
timates to the actual value, precautionary principle). If the influence on the result is higher, more information is
gathered or the sensitivity quantified.
The GaBi database has acceptable cut-offs of the type 2), if the environmental relevance on the overall result
can be justified as small. An example of a justifiably small environmental relevance is a known inconsistency
in a mass or energy balance with known reason, such as missing or imprecise quantified mass information in
the input. These can be minor variations in moisture content or minor amounts of diffuse water input, reaction
or combustion air, which is directly taken from the atmosphere and normally not quantified in a “bill of materi-
al” or process flow chart. Known inconsistencies in a mass or energy balance with known reason on the out-
put side can be undocumented “emissions” or energy flows such as evaporated water, used air, “clean” off-
gas streams or off-heat. These cut-offs are acceptable, if their quantification would raise the effort drastically
and marginally improve the overall results.
All GaBi unit processes seek to reflect actual physical and thermodynamic laws and the mass balance of the
key substances and fuels in the input must best match the product, waste and emission output. As a general
rule in GaBi unit process modelling, the mass and energy balances are closed and cut-offs are avoided. Pro-
jects and data collections with industry and associations showed that on the unit process level mass balance
inconsistencies of less than 1% are achievable with practically feasible effort.

34
 Methodological framework

On the unit process level of GaBi datasets, a best practise value of < 1% cut-offs (or unknown omissions,
sources or sinks) is applied for flows that are less environmentally-relevant.
Diffuse emissions (which are normally calculated or estimated according to local regulations) are considered,
if there is any indication that they are relevant in the respective process. Many processes limit or virtually
prevent diffuse emissions by using specific sealing technologies or by operating with pressures below atmos-
pheric condition (which can prevent unwanted substances to leave the system).
Unintentional cut-offs (mistakes) or forced cut-offs (non-closable gaps) of type 3) (unknown or known incon-
sistency in a mass or energy balance for unknown reasons) are due to missing information or due to a mis-
take. If cut-offs must be applied in the foreground system, they are mentioned in the dataset documentation in
GaBi [Link] and limited as much as possible or practicably
feasible. If reviews, validations or usages of the Master database reveal unintentional cut-offs, these are doc-
umented in the “GaBi database bug forum” and corrected with the next appropriate maintenance activity with-
in the GaBi database maintenance and service schemes.
Straightforward application of mass-% cut-off rules can lead to significant inaccuracies, if no possibilities exist
to properly quantify the environmental relevance properly (e.g. on the basis of comparable existing systems).
Therefore, the definition and use of cut-off rules should essentially be done or validated by experienced LCA
professionals who
 know the respective process chain technically, and
 know the field of potential environmental effects caused by the related material and energy
flows that are intended to be cut-off.
Only this combined knowledge ensures proper application of cut-off rules. Therefore, cut-off rules are indeed
essential elements when preparing, collecting and validating data. These rules are especially important for
processes with a large amount of different substance flows (such as pesticides in agriculture) or systems that
employ large material flows of less environmental relevance and few minor mass flows of substances with
potentially high impact (such as heavy metals in a mineral mass production process or precious metals in
catalyst production). In such cases even small amounts (<1% mass) can sum up to relevant cut-offs due to
their environmental relevancy in comparison to the main mass flows.
It can be concluded that the best rule for cut-offs is: “Only cut off what can be quantified.” The definition of
useful cut-off criteria is therefore quite complex for those stakeholders and users who have limited access to
the relevant technical background data.

3.3.5 Gap closing

Suitable application of cut-off rules on the input side defines the amount of relevant and included upstream
processes and process-chains. The possibilities to avoid cut-offs were discussed in Chapter 3.3.4.
This chapter documents gap-closing possibilities on the output side, primarily for “data on demand” requests.
“Data on demand” are datasets, which are additionally ordered and developed on request and enhance the
standard database content.
On the output side the cut-off rules mainly influence the degree of detail in terms of by-products, emissions
and wastes.

35
 Methodological framework

On the output side, the procedure is as follows:

 All known by-products are recorded (primary data is the first choice, if applicable).
 All known emissions are recorded (primary data is the first choice, if applicable).
 In case no data is available, emissions from similar processes or suitable literature data are
used.
 Emission data can alternatively be calculated overreaction equations, mass-energy balances,
known efficiencies and yield figures with adequate engineering expertise.
 Optionally, gaps in the data are identified and provided with a worst-case scenario (such as le-
gal limit, which is in most cases higher than the actual value).
 The ecological relevance of the individual emissions of concern (and their sensitivities) is quan-
tified with software. Sensitivity analyses are supported by GaBi software solutions and can
therefore easily be done during data collection and validation process.
 If the contribution is less relevant, the worst-case scenario may remain. If the contribution is rel-
evant, the emissions of concern must be investigated in detail (maybe an iterative step of prima-
ry data acquisition needed).
The seven steps above are used in any customer specific “data on demand requests,” as well as for any new
internal or external datasets, whose goal is to be consistent with the rest of the GaBi data and where the first
choice, primary data, cannot be used.

3.3.6 Infrastructure
The integration and omission of infrastructure in LCA systems are closely related to its respective relevance
within the system, which can significantly differ.
Infrastructure is relevant for processes which show comparatively fewer direct emissions during operation but
involve material-intensive infrastructure per product output. This is the case for some renewable resource-
based operations like hydropower plants (mainly reservoir), wind converters (blades, tower, gear) and geo-
thermal power plants (turbines halls, well equipment). For wind converters the majority of all established im-
pacts (> 90%) are from infrastructure because virtually no relevant emissions appear in the use phase. For
hydro and geothermal power plants the impact of infrastructure can be up to 80%, in our experience. The
impacts of storage hydropower plants especially depend upon the latitude of the site of the reservoir. The
degree of relevancy of degrading organic matter in the reservoir of warm climates can reduce the infrastruc-
ture’s relevance, such as in the case of hydro, as far down as 20%. For geothermal power plants the kind of
geological underground situation (rocks, soil) may influence the share of impacts concerning infrastructure
and maintenance.
The relevancy of infrastructure of mainly fossil operated power plants is significantly lower; according to our
records much less 1% across some main impacts. We will document the relevancy of fossil operations in two
ways: Based on non-public LCA data of the GaBi database and based on an internet public domain calcula-
tion.
GaBi Master DB:

36
 Methodological framework

Table B: Relevancy of infrastructure of a natural gas power plant in GaBi Master DB (selected rep-
resentative sample power plant)

emissions + chemical mainly concrete +

natural gas supply steel EoL, recycling
fuel supply operation infrastructure others
Acidification [kg SO2-
Equiv.] 79.7% 20.3% 0.06% 0.02%
Eutrophication [kg
Phosphate-Equiv.] 60.1% 39.8% 0.05% 0.02%
Global Warming [kg
CO2-Equiv.] 21.7% 78.2% 0.02% 0.004%
Photoch. Ozone Creat.
[kg C2H4-Eq.] 83.6% 16.3% 0.05% 0.02%
Fossil Primary energy
[MJ] 99.9% 0.1% 0.02% 0.003%

Larger plants with large throughput and longer life times tend to have lower impact shares in infrastruc-
ture/operation than smaller plants with shorter life times.
Public internet sources:
The above given evaluation can be cross-checked (e.g. by interested parties without access to LCA data)
taking publicly available power plant information from many internet sources. We consider the following fig-
ures of a medium power plant as a public domain example:

Table C: Publicly available example value for a medium size gas power plant

Cross check Example value (considered as public domain)

Operation time 30-50 years
Installed capacity (electrical) 400-500 MW
Emissions Operation 400-450 kg CO2 emissions / MWh electricity output
Total emissions Operation 40-90 Mio. t CO2 over the life time of the power plant

Furthermore, we considered the following main material intensity of a power plant for the cross check of a
public domain example (see various public and easily accessible internet sources).

Table D: Publicly available example values for CO2 for a gas power plant

Cross check Example value (considered as public domain)

Steel infrastructure 2000 t to 4000 t steel per 1 Mio kWh electricity output
Concrete infrastructure 16,000 – 20,000 t concrete per 1 Mio kWh electricity output
Asphalt infrastructure 1000 t to 2000 t asphalt per 1 Mio kWh electricity output

37
 Methodological framework

Considering additional publicly available CO2 intensity factors of the ELCD database
([Link] for the aforementioned materials the infrastruc-
ture is responsible for about 60000 to 80000 t CO2, which amounts to about 0.09%-0.15% of the CO2 emis-
sions of the operation (neglecting the supply of gas and recycling possibilities of the power plant materials). If
the gas supply and recycling are also included, the CO2 intensity of infrastructure would be further reduced
and a distribution similar to GaBi model above could be expected.
Summary:
As a consequence the degree of relevance of infrastructure is strongly case-specific. Even if one considers
the side effects of construction of vehicles and machinery as several factors more impact-intensive than the
material supply for infrastructure, infrastructure and construction would still have very low relevance for fossil-
fuelled power plants.
Huge conversion processes show the most likely comparable characteristics (of high throughput and long life
times), so we consider the infrastructure for those operations as very low in relevance for a background data-
base4.
Regardless of the degree of relevancy, all energy datasets in GaBi databases (fossil and renewable) include
the power plant infrastructure for consistency.

3.3.7 Transportation
As a general rule all known transportation processes have been included to remain consistent. Pipeline,
ocean vessels, river boats, trucks, railroad and cargo jets are used as parameterised processes, meaning
they are scaled and parameterised according to technology, distance, utilisation, fuel type, road type, river or
sea conditions and cargo specifications.
Transportation processes, including fuel production and utilisation, is especially relevant if the process in the
considered system is known to be relevant due to:
 Weight of material/product to be transported or
 Distance of transportation.
If an LCI database is structured into many sub-systems of producing and consuming systems, the transporta-
tion system should be modelled in the respective and consuming system. This ensures the generic use of the
same producing system in other applications while reflecting specific transportation situations in the consum-
ing plan system.

3.3.8 Water
Water use is understood as an umbrella term for all types of anthropogenic water utilisation. Water use is
generally differentiated in consumptive water use (i.e. water consumption) and degradative water use.
Freshwater consumption describes all freshwater losses on a watershed level which are caused by evapora-
tion, evapotranspiration (from plants), freshwater integration into products and release of freshwater into sea
(such as from wastewater treatment plants located at the coast line). Freshwater consumption is therefore

4
Be aware: This documentation relates to a background database. For a specific goal and scope of a specific study it can of
course be important to consider infrastructure (maybe even in the foreground system).

38
 Methodological framework

defined in a hydrological context and should not be interpreted from an economic perspective. It does not
equal the total water withdrawal, but rather the associated losses during water use. Note that only the con-
sumptive use of freshwater (not sea water) is relevant from an impact assessment perspective because
freshwater is a limited natural resource. Sea water is plentifully available and therefore not further assessed in
the life cycle impact assessment.
Degradative water use, in contrast, denotes the use of water with associated quality alterations, in most cases
quality degradation (e.g. if tap water is transformed to wastewater during use). Quality alterations are not
considered (fresh) water consumption. Also noteworthy is that the watershed level is regarded as the appro-
priate geographical resolution to define freshwater consumption (hydrological perspective). If groundwater is
withdrawn for drinking water supply and the treated wastewater is released back to a surface water body (river
or lake), then this is not considered freshwater consumption if the release takes place within the same water-
shed; it is degradative water use.
In a GaBi balance the above terms can be understood as:
Fresh water use = total fresh water withdrawal = water (river water) + water (lake water) + water
(ground water) + water (rain water) + water (fossil groundwater)
Fresh water consumption = total freshwater use – total fresh water release from Technosphere =
water vapour + water (incorporated in product inputs) – water (incorporated in product outputs) +
water (fresh water released to sea)
Furthermore, new and different water flows are being introduced for hydropower (e.g. “water (river water from
technosphere, turbined)”) and a new approach to consider cooling water is implemented, which takes into
account the latest developments of assessing thermal emissions to the aquatic environment.

Additionally applied water flows in GaBi database to enable consistent modelling of water:
 “Water (fresh water)”: This is a composite flow. Individual water elementary flows shall be doc-
umented (river/lake/ground water) and given priority. Use this flow only in very exceptional cas-
es if the former is not possible.
 “Water (fossil ground water)5”: The consideration of fossil groundwater is important because the
use of fossil water directly contributes to resource depletion which is specifically addressed by
some LCIA methods.
 “Water (surface run-off)”: Note that surface run-off is differentiated in surface run-off from soil
(considered as resource elementary flow) and surface run-off from technosphere (considered
as technosphere flow). Surface run-off from a landfill surface can be represented by the flow
“water (wastewater, untreated)” due to the pollutant load included in the flow.
 “Water (tap water)”: We used the term “tap water” as general term encompassing tapped water
with different qualities. It includes non-drinking-water quality water and high-quality drinking wa-
ter produced from groundwater and/or surface or seawater by desalination.

5
Fossil water or paleowater is groundwater that has remained sealed in an aquifer for a long period of time. Water can rest
underground in "fossil aquifers" for thousands or even millions of years. When changes in the surrounding geology seal
the aquifer off from further replenishing from precipitation, the water becomes trapped within, and is known as fossil water.

39
 Methodological framework

 “Water (wastewater, untreated)”: This flow is generally treated in a Wastewater Treatment

Plant, connected to a wastewater treatment plant module.
 Water vapour: Note that only water vapour stemming from evaporation (not steam) is used as a
term here. Steam is an output from a process and therefore a technosphere flow.
 Resource flows from technosphere: Water resource flows from the technosphere are introduced
in order to facilitate complete water mass balances on the level of plan systems including fore-
ground processes and aggregated background data (supply chains).
 Water (evapotranspiration)6: Evapotranspiration can be an output from either rain water or/and
irrigation water stemming from e.g. rivers or lakes.
 Water (brackish water): Brackish water has more salinity than fresh water, but not as much as
seawater. It may result from mixing of seawater with fresh water, as in estuaries, or it may occur
in brackish fossil aquifers.
Examples of how water was addressed in GaBi databases:
 Process using process water as input:
o Input flow: Apply “water (process water)” and connect flow to a water treatment/supply
module (see Figure 3-6)
o Output flow: Apply “water (waste water, untreated)” and connect flow to a wastewater treat-
ment plant module (see Figure 3-6)
 Process using tap water as input:
o Input flow: Apply the appropriate GaBi dataset for tap water production (see Figure 3-6)
o Output flow: Apply “water (waste water, untreated)” and connect flow to a wastewater treat-
ment plant module (see Figure 3-6)
 Process using cooling water as input:
Note that for cooling water we distinguish between use in 1) general production processes and 2) ener-
gy/electricity generation. Waste heat released to the water environment will also be properly recorded (see
Figure 3-3) as both the information on the volume of released cooling water and the incorporated waste heat
are necessary to perform the subsequent LCIA. Different technologies for cooling are differentiated as out-
lined below.
1) General production process (in different industrial settings):
Open-loop and closed-loop cooling are differentiated (see Figure 3-3).
 Input flow: Identify whether the cooling water input is...
o directly withdrawn from the environment (e.g. from a river or lake)  then apply the
appropriate water resource flow (e.g. “water (river water)”).

6
Evapotranspiration (ET) is a term used to describe the sum of evaporation and plant transpiration from the Earth's land
surface to atmosphere. Evaporation accounts for the movement of water to the air from sources such as the soil, canopy
interception, and waterbodies. Transpiration accounts for the movement of water within a plant and the subsequent loss of
water as vapour through stomata in its leaves.

40
 Methodological framework

o taken from a connected upstream water treatment process (e.g. water deionisation) 
then apply the appropriate water technosphere flow/operating material (e.g. “water (de-
ionised”).
 Output flow: Identify whether the cooling water output is...
o directly released to the environment (e.g. back to the river the cooling water was with-
drawn from)  then apply the appropriate resource flow from technosphere (e.g. “wa-
ter (river water from technosphere, cooling water”)). Consider also water vapour and
waste heat, if applicable.
o released as wastewater to the sewer system  then apply the flow “water (waste wa-
ter, untreated)” and connect flow to a wastewater treatment plant module. Consider al-
so water vapour and waste heat, if applicable.

41
 Methodological framework

Water vapour
[Inorganic Water (river water
emissions to air] from technosphere,
Water (river water) cooling water)
Open-loop cooling
Waste heat
[Other emissions to
fresh water]
Water vapour: if no information is available, estimate 5 % losses as water vapour
due to evaporation/leakage.

Water vapour
[Inorganic
emissions to air]
Water (river water)
Closed-loop cooling

Water vapour: if no information is available, estimate 5 % losses as water vapour

due to evaporation/leakage.
Note that the amount of water vapour lost equals the amount of the resource
input “water (river water)” due to the closed-loop set-up.

Water vapour
[Inorganic Water (river water
emissions to air] from technosphere,
cooling water)
Water Water
Water deionisation Open-loop cooling
(river water ) (deionised)
process Waste heat
[Other emissions to
Water vapour: if no information is available, estimate 5 % losses as fresh water]
water vapour due to evaporation/leakage.

Water vapour
[Inorganic
emissions to air]

Water Water
Water deionisation Closed-loop cooling
(river water ) (deionised)
process

Water vapour: if no information is available, estimate 5 % losses as water vapour

due to evaporation/leakage.

Figure 3-3: Application water flows in open-loop and closed-loop cooling systems in various
industrial settings

42
 Methodological framework

2) Energy/electricity generation:
Once-through cooling and cooling towers (also denoted open-loop cooling in electricity production)
are distinguished (see Figure 3-4).
 Input flow: Identify which water source is used for cooling (e.g. river water, lake water)  then
apply the appropriate water resource flow (e.g. “water (river water)”).
 In the case of cooling plants located at the coastline and using sea water for cooling purposes,
consider a desalination process as an additional water treatment process and apply the appro-
priate water technosphere flow/operating material (e.g. “water (desalinated, deionised)”).
 Output flow: Apply the appropriate resource flow from the technosphere according to the water
source used for cooling (e.g. “water (river water from technosphere, cooling water)”). Consider
also water vapour and waste heat, if applicable.

43
 Methodological framework

Water vapour
[Inorganic Water (river water
emissions to air] from technosphere,
Water (river water) cooling water)
Once-through
cooling Waste heat
[Other emissions to
fresh water]
Water vapour: if no information is available, estimate 1 % losses as water vapour
due to evaporation of heated cooling water from the river after release (Goldstein
R., Smith W. 2002).
Waste heat embodied in the cooling water release according to heat balance.

Water vapour
[Inorganic
emissions to air]
Water (river water)
Cooling tower
(open-loop cooling)

Water vapour: Amount of evaporated water equals amount of the resource input “water
(river water)”.

Water vapour
[Inorganic Water (sea water
emissions to air] from technosphere,
Sea water cooling water)
Water Water Once-through
desalination
(sea water) (desalinated, cooling Waste heat
process
deionised) [Other emissions to
sea water]
Water vapour: if no information is available, estimate 1 % losses as water vapour due to
evaporation of heated cooling water from the sea after release (Goldstein R., Smith W. 2002).
Waste heat embodied in the cooling water release according to heat balance.
Output flow “water (sea water from technosphere, cooling water) denotes the origin of the water
applied for cooling, namely the sea, and at the same time indicates that the cooling water is
released back to the marine environment (assumption!).

Figure 3-4: Application water flows in once-through cooling and cooling towers in ener-
gy/electricity generation

 Use of water in hydropower generation:

For hydropower generation the following four generation technologies are considered: run-of-river
power station, pump-storage and storage power stations, and tidal/wave power plants. See the fol-
lowing graphs (Figure 3-5) for instructions for inventorying the appropriate water flows.

44
 Methodological framework

Water (river water

Run-of-river power from technosphere,
Water (river water) station turbined)

Water (xxx water

Water (xxx water) (Pump-) storage from technosphere,
power station turbined)

Both on the input and output side, the water type needs to be defined: either
river water or lake water.
Example: The input of an alpine dam (e.g. storage power station) is “water (lake
water)” and the output is generally a river (i.e. “water(river water) from
technosphere, turbined”)!

Water (sea water

Water (sea water) Tidal/wave from technosphere,
power plant turbined

Figure 3-5: Application water flows in hydropower generation

Ecosphere Water vapour

(resource [Inorganic
flows) GaBi plan system
emissions to air] Water (river water
from technosphere,
cooling water)
Power plant
Water with Waste heat
(river one-through [Other emissions to
water) cooling fresh water]
located at a
river

Electricity
Water Process Water (river
(river Water
water Waste water from
water) (process Manufacturing technosphere,
treatment water
water) process of treatment
Tap water product X waste water)
Water Water plant
production
(river (tap
& supply
water) water)

Figure 3-6: Ad hoc example of a simple plan system including different processes and water
flows

45
 Methodological framework

For the GaBi background database water that has been treated generally (chemically or physically deion-
ised/decalcified) is used for process and cooling water purposes, which reflect the standard case. Untreated
water (tap or even surface water) is only used where it is explicitly known that it was used.

3.3.9 Wastes and recovered material or energy

Waste volumes or masses are known and commonly used to describe the environmental relevance of outputs
of processes. Waste volumes or masses are not an environmental intervention. The environmentally relevant
intervention occurs in the incineration, treatment or landfill after waste is turned into emissions like landfill gas
or water leaching.
According to ILCD [ILCD 2010] all product and waste inputs and outputs should be completely modelled until
the final inventories exclusively show elementary flows (resources in the input and emissions in the output).
Therefore waste treatment is integrated throughout the whole system during modelling wherever possible and
known to occur7. For all known treatment pathways (e.g. for regulated waste with calorific value) the incinera-
tion and landfilling processes of the residues are integrated.
Different waste treatment options are provided in the GaBi databases (inert matter landfill, domestic waste
landfill, hazardous waste landfill underground / above ground, waste incineration of domestic waste, waste
incineration of hazardous waste). The waste fractions of the processes are identified by the composition and
their appropriate treatment and the respective GaBi process applied.
“Waste” going to any kind of reuse or recycling is modelled in loops or allocated/substituted, if a considerable
positive market value (a product) exists.
There are many products which are legislatively considered a waste, but which must be treated as products in
life cycle analysis. It should be noted that the same market value is applied at the point where the waste (or
waste products) accumulates and at the point where the waste is recycled. For suitable modelling feedback
from both sides (producer of waste product and user or processor of waste product) is necessary. Waste to be
recycled without a market value will stay (virtually) as waste in the producer process and is documented as
such.
Standard procedure (general waste treatment)
In the case that specific information is not available for the respective situation, a standard procedure is
adopted according to secondary material markets.
 Any secondary material that already has a recycling market is treated as recycled according to
the market share (see examples in following table).
 All waste generated within the EU that has a calorific value and can be disposed with municipal
solid waste (MSW), is treated in an incineration plant (see selected examples).
 If case-specific treatment is specified and known, and the waste cannot be mixed with MSW,
specific treatment is modelled.
 All other waste (mainly inert waste) goes to landfill.

7
Due to the integration of treatment pathways for known waste or residue streams it might be possible that (intermediate)
waste flows are deleted from existing plan systems (because those are now modeled further).

46
 Methodological framework

Table E: General treatment procedure (if no specific information is available) for common materi-
als/wastes

Material/waste Treatment Process

Mixture of plastics Incineration, waste to energy
Polyofefin and PVC Incineration, waste to energy
Wood Incineration, waste to energy
Aluminium, non ferrous metals Recycling
Steel Recycling
Coating and sealing Incineration, waste to energy
Glass, concrete, stones Inert landfill

Standard procedure (Hazardous waste treatment)

Hazardous waste streams are often hard to define as default in a background database, because, depending
on various options to mix different waste streams, several disposal options exist. Hazardous waste streams in
the upstream chains are modelled according to their specific fate, if it is known (e.g. in tailing ponds). Hazard-
ous slags/sludges are treated via vitrification, incapsulation and landfill. If unspecific hazardous waste streams
appear, a worst-case scenario (precaution principle rule) is used. The worst-case scenario models incinera-
tion, vitrification, macroencapsulation and the inert landfill of the remains. Carbon-rich and carbon-free haz-
ardous waste is differentiated, as are other emissions which occur in incineration.

Table F: General procedure for some hazardous waste flows

Kind of waste treatment treatment treatment final treatment

Slag/Sludge Vitrification Macroencapsulation Inert Landfill

Non-specific source Incineration Vitrification Macroencapsulation Inert Landfill

If hazardous waste treatments become relevant, a check must be performed to determine if specific data for
the treatment pathway is available.

3.3.10 Aspects of biomass modelling

The carbon cycle in LCA can be defined as: CO2 in atmosphere  CO2 intake/H2O/sunlight/surface  plant
growth  harvested biomass  biomass use as fuel or matter  CO2 combustion/decomposition  CO2
intake in atmosphere  …
Depending on the situation one can understand “biomass” as a certain status at different points in the cycle:
As a plant, as harvested biomass and as a renewable product.
The definition of “biomass resource” is therefore somewhat arbitrary and can be chosen according to the
given goal and scope.
Biomass in GaBi is further modelled towards carbon dioxide, water, solar primary energy and the land use
[GABI 201]. This modelling assures mass balance consistency especially of the carbon-balance, for example,

47
 Methodological framework

biomass storage in the product and fuel and the incineration or decomposition releases of CO2, which had
been released previously.
The solar primary energy embedded or stored in the biomass is exactly the amount of solar energy which has
been converted by the biomass (the calorific value). The efficiency of conversion does not play a role, as the
source (sun) can be understood as infinite in human timeframes.

3.3.11 Aspects of primary energy of fossil and renewable energy sources

Energy evaluation in the GaBi database is based on the principle of “cumulated energy approach (KEA)“ or
often also referred to as embodied energy. The primary energy needed to supply certain materials or energies
often serves as indicator of the energy efficiency. The indicator can be misleading, if renewable and non-
renewable energy sources are compared or summed and not separately interpreted. Renewable and non-
renewable energy sources must be interpreted separately, as implemented in the GaBi database. The inter-
pretation is usually done in LCA reporting practise. A combined evaluation of the primary energy (renewable
and non-renewable figures) may be required.
It is relatively common to compare non-renewable energy production procedures with a uniform parameter
like the calorific value of the primary energy needed to provide a certain usable energy. However such a uni-
form parameter does not intuitively exist for renewable energy sources like hydro and wind or for nuclear
energy. Different approaches exist (technical efficiency8, physical energy content method with virtual 100%
efficiency for renewables9, substitution approach to avoid renewable efficiencies with virtual thermal fossil
efficiencies for renewables 10) to define or compare the „primary energy demand“ of a related usable energy
form.
The IEA states11: “Since these types of energy balances differ significantly in the treatment of electricity from
solar, hydro, wind, etc., the share of renewables in total energy supply will appear to be very difficult depend-
ing on the method used. As a result, when looking at the percentages of various energy sources in total sup-
ply, it is important to understand the underlying conventions that were used to calculate the primary energy
demand”.
In principle the method of the technical efficiency differentiates between renewable and non-renewable prima-
ry energy needs, while others do not.
ISO 14040 frameworks do not call for an explicit method for the aggregation/separate representation of the
primary energy.
The ILCD framework [ILCD 2010] does not call for an explicit method either, but a recommendation is given
for a differentiation between non-renewable energy resources and renewable energy resources.
In GaBi consequently the method of the technical efficiency with differentiation between non-renewable ener-
gy resources and renewable energy resources is applied as it illustrate the situation adequately, comprehen-
sively and transparently. This is especially important in countries with significant portions of renewables in the

8
See Richtlinie, VDI 4600, 1997: VDI 4600 Kumulierter Energieaufwand - Begriffe, Definitionen, Berechnungsmethoden.
9
See International Energy Agency (IEA) (Hg.): Methodology of International Energy Balances. Unter Mitarbeit von Karen
Treanton. Paris, France, 2001. Online verfügbar unter [Link] zu-letzt geprüft
am 2010-09-07.
10
See Murtishaw, S.; et al.: Development of Energy Balances for the State of Cali-fornia. Lawrence Berkeley National Labo-
ratory. Berkeley, USA, 2005. Online verfügbar unter [Link] zuletzt geprüft am 2010-09-
07.
11
International Energy Agency (IEA) (Hg.): Energy Balances of Non-OECD Countries 2010. Paris, France, 2010.

48
 Methodological framework

grid (e.g. Norway, Austria and Denmark). The international trade of energy is accounted for individually to
avoid a virtual efficiency of 100% for imported electricity, which is relevant for countries with a high share of
imported energy
The value and burden of the use of 1 MJ of renewable primary energy is not directly comparable with 1 MJ of
fossil primary energy, because the availability of the fossil resources is limited and depletion occurs. The topic
cannot be discussed in detail here, but the guidelines will help to prevent “double counting” as well as “per-
petual motion.”
1 MJ of electricity from wind power is produced using (virtually) approx. 2.5 MJ of primary wind energy (an
efficiency of approx. 40%, due to usable kinetic energy of wind).
For 1 MJ of electricity from hydropower (virtually) 1.15 - 1.25 MJ of primary hydro energy is used (an efficien-
cy of 80 - 85%, due to usable kinetic energy of water).
For 1 MJ of electricity from geothermal power (virtually) 5 – 6.5 MJ of primary geothermal energy is used (an
efficiency of approx. 15 - 20%, due to energy content of usable temperature gradient).
For 1 MJ of electricity from nuclear power approx. 2.5 - 3.3 MJ of primary nuclear energy is used (an efficien-
cy of approx. 30 - 40%, due to energy content of used fissile material).
For 1 MJ of electricity from photovoltaic approx. 10 MJ of primary solar energy is used (an efficiency of ap-
prox. 10%, due to the usable part of the solar radiation).
For 1 MJ of electricity imports the specific efficiency of the import country is applied.

3.3.12 Land Use Change

Apart from the classical impact categories like GWP, AP, EP and POCP, land use as an environmental issue
is widely considered to be important and constantly gains attention in the Life Cycle Assessment (LCA) com-
munity, especially for agrarian products, forestry and sealed areas.
In the software and database system GaBi 6 land use parameters are integrated. The methodology behind
integrated land use parameters is based on the dissertation of Martin Baitz [BAITZ 2002] and subsequent work
that was carried out at the University of Stuttgart, Chair of Building Physics (LBP), Dept. Life Cycle Engineer-
ing (GaBi; former Institute for Polymer Testing and Polymer Science) and PE INTERNATIONAL AG [Beck,
Bos, Wittstock et al. 2010]. According to Baitz 2002, a set of indicators has been defined to model land use
aspects in LCA and incorporate them into the software:
 Erosion Resistance
 Mechanical Filtration
 Physicochemical Filtration
 Groundwater Replenishment
 Biotic Production
These land use indicators are calculated for several land-intensive processes with the support of the LANCA
tool (Land Use Indicator Calculation Tool) based on country-specific input data and the respective land use
types. A detailed description of the underlying methods can be found in Beck, Bos, Wittstock et al. 2010.

49
 Methodological framework

The values calculated according to these methods are then integrated into the GaBi 6 database and software,
and aggregated over the process chain to form environmental indicators that are representative for the entire
life cycle of many of the aggregated processes in the GaBi database system. Land use can be considered an
additional aspect in LCA to extend its environmental impact evaluation.
Country-specific input data for the tool has been derived from the ISRIC database [ISRIC WISE 2002] for the
input parameters of humus content, skeletal content, declination and effective cation exchange capacity and
from the [Mitchell 2003], TYN CY 1.1 dataset for precipitation, summer precipitation and evapotranspiration.
For the input parameter of “distance to groundwater,” a default value (0.8-10m) is used.
For the calculation of indicator values, indicator qualities Q must be calculated for the state and land use types
of the land before transformation (t1), during use (t2 and t3) and after regeneration of the land (t4) [Beck, Bos,
Wittstock et al. 2010]. At this stage it is assumed that the occupation phase is a static situation. Consequently,
for all processes calculated t2 and t3 have the same land use type. For each indicator, occupation and trans-
formation are calculated according to the following equations:
Occupation indicator value = (Q(t4, ref)-Q(t2,3))*area used*time of occupation
Transformation indictor value = (Q(t4)-Q(t1))*area used
The resulting units of qualities, transformation and occupation indicator values as used in GaBi are shown in
Table G.

Table G: Overview of the Land Use Change Indicator Units

50
 Methodological framework

Positive occupation indicator values can be interpreted according to the following:

 Erosion Resistance (expressed by kg of erosion): kg of soil eroded in addition to naturally-
occurring soil erosion due to the effects caused by the production of one functional unit
 Mechanical Filtration: amount of water that could not be filtered due to the effects caused by the
production of one functional unit
 Physicochemical Filtration: amount of cations that could not be fixed to the soil due to the ef-
fects caused by the production of one functional unit
 Groundwater Replenishment: amount of groundwater, which could not be replenished due to
the effects caused by the production of one functional unit
Biotic Production: amount of biomass not produced due to the effects caused by the production of one func-
tional unit. Positive transformation indicator values (permanent impacts) can be interpreted according to
the following:
 Erosion Resistance: kg of soil eroded in addition to naturally-occurring soil erosion per year in
the time following the considered land use due to the permanent transformation impacts caused
by the production of one functional unit
 Mechanical Filtration: amount of water that cannot be filtered in the time following the consid-
ered land use per day due to the permanent transformation impacts caused by the production of
one functional unit
 Physicochemical Filtration: amount of cations that cannot be fixed to the soil in the time follow-
ing the considered land use, due to the permanent transformation impacts caused by the pro-
duction of one functional unit
 Groundwater Replenishment: amount of groundwater, which cannot be replenished in the time
following the considered land use per year, due to the permanent transformation impacts
caused by the production of one functional unit
 Biotic Production: amount of biomass that is not produced in the time following the considered
land use per year due to the permanent transformation impacts caused by the production of one
functional unit
Negative indicator values show the respective positive impacts.
According to BAITZ 2002, the reference situation “tref” is assumed to be the same as the situation “t4.”
With the integration of consistent land use information in the GaBi 6 database, it is possible to examine and to
quantify the effects of a product or a process on the land. The resulting information refers to land quality pa-
rameters and is summable and scalable over the whole process chain.
Seeing as this is the first inclusion of such land use information into a database, there are some limitations to
be mentioned:
 Land use information is only included in land-consuming unit processes.

51
 Methodological framework

 Entry data is often derived from databases where only country-specific values can be gathered.
More site-specific values would enhance the accuracy of the results. These values are often dif-
ficult to obtain.
 Allocation of transformation effects, especially for agricultural processes, has not been done
and is subject to further research.
The background data provides valuable information which can be used to compare and contrast different
products on a country scale. For the comparison of foreground processes and land uses in their exact loca-
tions, indicator values can be calculated using LANCA and can be included into unit processes.
The land use database in GaBi 6 gives the user a good starting point in order to determine the main contribu-
tors to land use effects. Due to further development of the method and tools, and increasing data availability,
similarly to other LCA data, land use data will be periodically updated to ensure currency of results.

3.4 Sources and types of data

Many sources and types of data exist. Whether the source or type of data is suitable is a matter of the goal
and scope of the exercise, and the capability of the data modeller to turn raw data and process information
into LCI data. The raw data and resulting LCI data used in the generic GaBi background databases seek to
reflect the reality of a certain point in time as representatively as possible.

3.4.1 Primary and secondary sources of data

(Primary) data and information from industry sources is the preferred choice of GaBi raw data and background
data, wherever possible and approved.
Primary data can be collected via the classical approach of collecting data from several companies producing
the same product and averaging the resulting inventories. Primary data is obtained from specific facilities as a
primary source of information. This data is measured, calculated or acquired from the bookkeeping of a par-
ticular facility.
Secondary data is obtained from published sources and used to support the set-up of the LCI. Examples of
secondary data sources include published literature, environmental reports of companies or LCI and LCA
studies, emissions permits and general government statistics (e.g. mineral industry surveys, Bureau of Labor
statistics and Energy Information Administration data).
This secondary data of industrial operations is used to develop, calculate and set-up LCI data by experienced
PE engineers with background in the technology and capability in the field, with the support of technical refer-
ence literature or branch encyclopaedias.
PE engineers are in constant contact with industrial companies and associations to update their knowledge
about representative process-chain details and new technologies.
PE’s developed capabilities and critical-constructive feedback from industry confirms PE’s approach to model
real process chain circumstances. Due to this process of continuously learning about industrial operations, we
consider PE data the best available “industry-borne” data.
PE’s strategy is proactive cooperation with industry. In the event of an unavailability of data, confidentiality or
missing access to (company or process) specific data, PE can bridge the gap with developed capabilities and
possibilities to generate generic data of comparable quality.

52
 Methodological framework

Publicly available information such as internet sources, environmental reports, scientific or application reports
with industry participation, other industry publication or other LCI relevant literature is constantly screened and
used for benchmark purposes. The quality of technical data of many publications varies considerably. The
sole fact that the information is officially published or publicly available ensures neither the consistency nor
quality of the content. The professional user of publicly available data should either know and trust the source,
or be able to judge and ensure the quality.
All generic GaBi data seeks to directly involve feedback of users, companies and associations by validation or
benchmarks with various industry or process information. PE offers and maintains a constant connection with
suitable users and diverse information sources from industry.

3.4.2 Unit process and aggregated data

GaBi databases deliver unit processes, aggregated and partly-aggregated data and complete life-cycle (sub-)
systems (plans), which include varying combinations of the aforementioned data. Any delivered dataset and
system is based on suitable raw data and process chain data.
As stated in the “Global Guidance Document for LCA databases” UNEP/SETAC 2011 – to which PE contrib-
uted considerably with its expertise to reflect professional issues through the provision of a global software
and multi-branch database - there exist many good reasons to provide and use any of the aforementioned
datasets.
The main goal of GaBi data is to enable the utilisation of best available information from reliable and suitable
technical sources. GaBi does not follow certain paradigms or patterns concerning data or data types. All data
types are welcome, used and supported, if they are determined to be suitable.
The reliability and representativeness of the data source are important aspects to ensure the data’s appropri-
ateness and quality. The possible level of (public) disclosure of data is subject to individual circumstances, the
source and the proprietary nature of the information provider. In LCA and business practise many different
circumstances related to ownership, rights, patents and property exist.
In practise anti-trust and competition regulations exist, aside from those dealing in the proprietary, which are
properly maintained by GaBi database. It works to ensure compliance with related laws and regulations.
Regarding reliability and representativeness, unit process data must ensure that it technically fits within each
other if used in one system. Random connection without a suitable check of technical consistency may lead to
wrong results, even if unit processes are disclosed. The fact that a unit process for a certain operation exists,
does not necessarily mean that it is technically suitable, up-to-date or appropriate. Background knowledge
concerning the real B2B supply chains is essential.
Transparency is an important aspect. In aggregated processes GaBi databases ensure transparency through
suitable documentation that covers all important technical facts. Parts of the Master Database are used to
share more details and process chain knowledge under bilateral business relationships.

3.4.3 Units
All data should be presented in metric (SI) units. When conversions are required from imperial or non-SI units,
the conversion factor must be clearly stated and documented.

3.4.4 LCI data and supported LCIA methods

It is important to clearly define the kind of data which will be covered by creating an LCI dataset for a system.

53
 Methodological framework

The GaBi LCI datasets are generally full-range LCI datasets. These datasets seek to cover all LCI data infor-
mation, which are of environmental relevance in relation to LCA best practise.
The sum of input and output (like resources and emissions) are a compendium of 20 years of LCA work in
industrial practise and the harmonised sum of all LCI interventions which could be measured, calculated or
documented in LCA practise.
Important impact methodologies have influenced the flow list – and hence the data collection – seeing as
GaBi considers the relevant impact categories and evaluation methods.
Basing the work on a harmonised and constantly growing flow list provides consistency among different da-
tasets provided by different groups or branches. A list of the supported impact categories including a brief
description is given as a supplement.
The GaBi database delivers full-range LCIs, which enables the use of any (existing and future) impact meth-
ods for which corresponding characterisation factors exist. For the following impact assessment methods
GaBi delivers already implemented default values.
Input-dependent quantities
 Abiotic Depletion (elements and fossil)
 Primary energy non-renewable (entered as an additional quantity)
 Primary energy renewable (entered as an additional quantity)
 Demands on natural space (surface)
Output-dependent quantities
 CML 2001 (versions 2001 – 2010) categories (GWP 100, ODP, AP, EP, POCP, HTP, TETP,
FAETP, MAETP) [CML 2001]
 TRACI categories [TRACI 1996]
 TRACI 2.1 categories [TRACI 2012]
 EDIP 2003 [HAUSCHILD 2003]
 USETox 2010 [USETOX 2010]
Further assessment methods are:
 ReCiPe Endpoints and Midpoints [RECIPE 2012]
 Ecoindicator 99 [ECO-INDICATOR 99 : 2000] or 95 [ECO-INDICATOR 95 : 2000]
 Impact 2002+[IMPACT 2002]
 UBPs [UBP 2006]

3.4.5 Production and consumption mix

In LCA practise process chain networks working toward one common product contain different levels of repre-
sentative situations:

54
 Methodological framework

a. “production mix:” This approach focuses on the domestic production routes and technologies ap-
plied in the specific country/region and individually scaled according to the actual production volume
of the respective production route. This mix is generally less dynamic.
b. “consumption mix:” This approach focuses on the domestic production and the imports taking
place. These mixes can be dynamic for certain commodities (e.g. electricity) in the specific coun-
try/region.
Figure 3-7 shows the differences between the two principle approaches. Electricity generation has been se-
lected as an example to explain the two approaches. The electrical power available within Country C is gen-
erated by operating different types of power plants. The fuels necessary for the operation of the power plant
will be supplied by domestic resources, as well as by imports from different countries. In addition to the do-
mestic power generation, electric power might also be imported.

Power grid mix Power Imports

Country C

Figure 3-7: Difference between "production mix" and "consumption mix" (for power genera-
tion)

The part of the Figure 3-7 which is coloured in grey represents the domestic part of the production and repre-
sents the “production mix” approach.
All parts of the supply chain of the power generation process coloured in orange (dark grey if b/w print) repre-
sent the imports of supplies for the power generation (imports on fuels). Imports on end energy level (imported
power) are indicated by a (yellow, bright grey) criss-cross. The “consumption mix” includes the “production
mix” as well as all imports.
The GaBi database supplies both the electricity consumption and electricity production mixes. The inclusion of
the imports in the LCI data requires country-specific information about supply generation and whether final

55
 Methodological framework

products are available or will be gathered during data collection. Not included in this example is the export as
the reverse of import.
It is apparent that for every commodity contained in the database, a screening of domestic production and
imports must be done, since this combination can be different for each commodity.
The GABI 2012 database aims to provide consumption mixes wherever possible.

3.5 Data quality issues

Data quality is probably one of the most discussed issues of databases and most widely interpreted. For the
development of the current GaBi databases the following importance of “quality indicators” can be stated.

Table H: Overview of qualitative importance of “quality indicators” in GaBi DBs

indication of importance
Indicator less more
credibility and source of data
access to industry information
relation of data to technology issues
consitency
representativeness of data
age / validity of data
transparency of documentation
country/region specificness
completness of data
precision of data
transparency of final data set
reduction/management of data uncertainty
uncertainty of data
public access of raw and unit process data

Several methods and approaches have already been proposed, but no single approach so far can be estab-
lished as the “best practice.” The GaBi data quality approach follows a golden rule: Be as precise and specific
as needed, and as simple and applicable to all circumstances as possible. The following paragraphs address
the important quality issues that arise in GaBi modelling.

3.5.1 Technology and region coverage

GaBi datasets aim to be technology specific. Various technologies may produce basically comparable prod-
ucts. GaBi datasets aim to provide
 the most likely “representative” case
 if suitable, a range of different technologies for the same product
 if suitable, the local consumption (or market) mix based on capacities

56
 Methodological framework

Where distinctly different technology pathways are used to produce the same materi-
als/products/commodities, they are kept separate and the local consumption (or market) mix is additionally
provided. Below are some examples of important technology differences:
 Electricity from different power plants (CHP, coal or gas, hydro, or wind)
 Steel making: electric arc, basic oxygen furnace, HiSmelt technology
 Blast furnace or electro-refined metals
 Wet or dry process cement clinker production
Plain average values for the above-mentioned processes (regardless of unit process level or aggregated
level) would not be representative of any of the technologies.
There is also a rationale for regional production models for commodities which are predominantly traded with-
in a certain region.
 Electricity, gas and petroleum products
 Wood panels and timber products
 Cement , aggregates and sand
 Waste management services
For some low impact materials, transport is the dominant impact on their production and transport distances
and modes may crucially affect the LCI results with sometimes counter-intuitive outcomes. For example:
 Aggregates shipped long distances by sea from coastal quarries may have lower net impacts
than more local sources travelling by road.
Therefore, the GaBi databases focus on the most relevant aspects first, after screening and identifying the
most important issues of a specific life-cycle model.

3.5.2 Data aggregation

The aggregation of datasets is often necessary and requested by users and providers of data in order to se-
cure the privacy of confidential information. This enables the use of accurate and up-to-date information; fur-
ther aggregation speeds up LCAs (and therefore lowers costs) as the handling of datasets and complete
process chains becomes feasible for both experts and users.
Almost any LCI dataset is aggregated: Either on the unit process level (either several production steps are
aggregated towards a unit process or different unit processes producing a comparable product are aggregat-
ed into an average unit process) or on the process chain level (different subsequent processes are aggregat-
ed). For a good description of the various types of aggregation see the UNEP/SETAC 2011 database guid-
ance.
Some systems are characteristically complex and therefore only understandable for LCA experts and experts
of the related technology. In order to make the handling for non-experts possible, some complex and often
used datasets must be aggregated in a representative and applicable way to make them suitable for use in
practise.
A prominent example is the aggregation of electricity mix data for a specific country; a complex background
model, consisting of a large amount of processes and parameters (see Chapter 2.3 for details). The user has

57
 Methodological framework

access to information transparency concerning the underlying model and data in the documentation. Most
users have an interest in accurate data and are less interested in power plant details, so an aggregation of
datasets is suitable and meaningful for a wide range of users12.
Two types of aggregation exist:
 horizontal
 vertical
The following figure describes the difference.

Company A Company B Company N

Mining
(Mz) M1 M2 Mn
+

Benefication

(Bz) B1 B2 Bn

LCI of vertical environmental profiles M, B, etc. for each company

Figure 3-8: Principle graphical explanation of the relation of completeness, precision
The horizontal aggregation (M1+M2+M3+...) and (B1+B2+B3+...) is applied in the creation of a process for an
average production step of a specific product by taking (different) technologies into account. The upstream or
downstream processes are not integrated into this step of aggregation. The horizontal aggregation must be
sure to lead to understandable and interpretable datasets, as technical information and upstream substances
of different processes is aggregated and provided side by side (whilst never appearing in reality as one pro-
cess). Not all unit processes of the same kind are automatically suitable for horizontal aggregation or are
subject to easy misinterpretations.
The vertical aggregation (M1+B1+...) and (M2+B2+...) is carried out by considering a specific technological
route and aggregating process chain parts that exist in reality. In this case the upstream and/ or downstream
processes are included in the aggregated dataset.
Depending on the case, in GaBi databases vertical and horizontal aggregation are applied to the datasets.

12
A power plant operator or energy provider may have another view on this and wants to deal with the effects of the power
plant parameters within the electricity mix. However, users that are interested in their own foreground system behaviour
should rather model on basis of their specific foreground situation and should take generic background data to set up their
respective background system or use it as reference or validation. Specific results on foreground systems request specific
foreground data.

58
 Methodological framework

3.5.3 Precision, completeness, representativeness

Each item’s precision, completeness and representativeness provide specific details of the composition of the
background data and are documented in provided datasets. The users can use these indicators to describe
their own data in GaBi.
Precision
Precision determines the variance of data, whether it has been measured, calculated or estimated. In the case
of the GABI 201 database, the following procedure is adopted:
 Measured: Values measured directly by the LCA practitioner, producer or project partner. Val-
ues from reports, which were measured and allowed to be published, can be also considered as
measured.
 Literature: Values obtained from literature which are not explicitly stated, whether the value
was measured or estimated.
 Calculated: The values were calculated, e.g. stoichiometrically.
 Estimated: Expert judgement, e.g. referring to comparable products/processes or legislations.
Completeness
Completeness provides information regarding the percentage of flows that are measured, estimated or rec-
orded, as well as unreported emissions. In the GaBi databases the following procedure is adopted:
 "all flows recorded": The entire process is covered due to complete access to process data or
the process was modelled in a very detailed form. Processes in which the cut-off rules were ap-
plied and checked can also be considered complete.
 "all relevant flows recorded": The relevant flows of the process are covered. When all flows
cannot be recorded, this is the next option, which still enables reasonably good quality of results
in terms of evaluation.
 "particular flows recorded": Only particular flows are recorded. It must be clear that in this
case some important flows can have been left out, so only medium quality of data can be
achieved. If possible, further research should be performed.
 "some relevant flows not recorded": If good quality is desired, this case should not occur. In
the case that no data is available, reasons for using this kind of data should be documented.
The technical, geographical and time related-representativeness of the background process is also stated in
the documentation and the process name. Aside from the description of the underlying background data the
proper application of the data by the user (goal and scope dependent) and its respective documentation is
also important. GaBi offers several possibilities to document the proper application of the background data in
the user-specific case. This can be done on the plan-system level in GaBi, by indicating the technical, geo-
graphical and time-related representativeness.

59
 Methodological framework

Technical Representativeness
Information about data representativeness is assessed qualitatively and reflects the extent to which the da-
taset represents the reality of a certain process or process chain, e.g. completely, partly or not representative.
GaBi data aims for best technological representativeness from the commission gate back to the resource
extraction. Technology matters.
Geographical representativeness
The GaBi databases have a 4 level regionalisation approach.
 Transferring existing technology information into other countries by adapting the energy supply
 Adapting the important upstream processes with regional supply data
 Collecting technology mix information of used in the region to adapt the existing information
 Collecting and validating primary data in the regional industry networks
Inventory data that shows the necessary geographical representativeness for the foreground data, site or
producer/provider specific data for the foreground system, supplier-specific data is used for the products that
connect the foreground with the background system. Generic data of geographical mixes can be used also in
parts of the foreground system if it is justified for the given case to be more accurate, precise and complete
than available specific data (e.g. for processes operated at suppliers). For the background system average
market consumption mix data can be used.
Time-related representativeness
The time-related representativeness indicates a reasonable reference value for the validity of the dataset.
That means for unit processes the dataset is most representative for the indicated year. This year is neither
the year of the most recent source that is used nor the year of the oldest one. The time at which the data
collection occurred should be used as a reference.
In GaBi the ‘most representative’ year indicates the current year of the modelling or validity checking of the
data, if our engineers did not have any evidence that something changed or developed in process technology
concerning this production step.

3.5.4 Consistency
Consistency refers to the uniformity of the data, methodology and procedure used in the data set-up and
database maintenance and additions. The GaBi database is consistent since all datasets follow the same
methodology and principles as described in this document. The PE database content uses consistent data
sources and background systems (e.g. transport, energy processes).

3.5.5 Uncertainty
Uncertainty in the LCA is often discussed from two different viewpoints: There is a scientific discussion on one
side, as to which approach is the best to calculate something rather uncountable13.
And there is a discussion about practise, dealing with how to limit uncertainty of results and how to judge its
importance regarding stability of results and proper decision support.
In GaBi database work PE INTERNATIONAL chooses the following approach to minimise uncertainty:

13
Not everything that can be counted counts and not everything that counts can be counted. Albert Einstein

60
 Methodological framework

1. Completing correct data collection (and close mass and energy balances).
2. Choosing a representative LCA data for the upstream and background data, which represent the ac-
tual technology
3. Understanding the technical processes and defining technical parameters that are uncertain
4. Completeness of the system (no unjustified cut-offs)
5. Consistent background data
Consistent data collection and background data are the basis to reducing uncertainty. In addition useful sce-
narios, sensitivity calculations and technical understanding of the LCA modeller (as well as the reviewer)
ensure minimum uncertainty.
If LCA modeller and the reviewer have no indication how the identified technical parameters may perform or
how the parameters are independent from each other, the Monte Carlo Analysis is an alternative. It allows the
examination of consequences of random uncertainties for some selected technical parameters. The quality of
the resulting “uncertainty statements” strongly depend on the selection of these technical parameters, which
should be as representative (in terms of uncertainty) as possible.
In principle the Monte Carlo analysis should consider each parameter in the model which is uncertain (all
inputs, outputs, parameters, impact values). This analysis is not yet implemented in GaBi. Challenges in this
context are: broad methodological acceptance, availability of (useful) uncertainty information for all model
parameters, implementation effort and probably the calculation performance.
Based on the above discussion, a practical approach to quantify the uncertainty issue was developed for the
GaBi background database.

Quantifying uncertainty in GaBi

Uncertainty in LCA can be split into two parts:
 data uncertainty (the uncertainty of the modelled, measured, calculated, estimated) and data
within each unit process
 model uncertainty (uncertainty introduced in the results of a life cycle inventory analysis due to
the cumulative effects of model imprecision, input uncertainty and data variability)
Uncertainty in LCA is usually related to measurement error-determination of the relevant data, e.g. consump-
tion or emission figures. Since the ‘true’ values (especially for back-ground data) are often unknown, it is vir-
tually impossible to avoid uncertain data in LCA. These uncertainties then propagate through the model and
appear in the final result. Small uncertainties in input data may have a large effect on the overall results, while
others will be diminished along the way. This article addresses PE INTERNATIONAL’s recommendations for
addressing the quantification of uncertainty in an LCA study, and how it can be done practically and with rea-
sonable accuracy.
Quantifying the uncertainty of primary data points on company-specific processes can be relatively
straightforward and easy for a company to calculate using the mean value and its standard deviation
over a certain number of data points.
But quantifying the uncertainty in the background systems (hundreds of upstream pro-cesses including mining
and extraction) and then performing error propagation calculation is typically neither practical nor feasible due

61
 Methodological framework

to the cost and time constraints in an industrial setting. In addition to put the issue in a general perspective,
one should be wary of data with an extremely precise uncertainty value to each inventory flow, as these can-
not be calculated with the accuracy that the value implies.
A common rule estimates that the best achievable uncertainty in LCA to be around 10%. This was supported
by [KUPFER 2005] on the forecast of environmental impacts in the design of chemical equipment. The actual
degree of uncertainty can vary significantly from study to study.
The overarching question that really must be answered is:
How robust is my overall result when taking into account the combined uncertainties?
The effort to come up with a reasonable estimate can be significantly reduced by following a two-step ap-
proach:
1. Understand the model structure and its dependencies
Keep it simple at first and start by setting up your model with values you have. Then try to develop an un-
derstanding of the most relevant aspects of your LCA model, i.e. those life cycle phases, contributors, or
data points that have the largest impact on your result. This is usually done by a contribution or ‘hot spot’
analysis and a subsequent sensitivity analysis. Both of these functions are available to GaBi users in the
LCA balance sheet through the Weak Point Analysis and the GaBi Analyst Uncertainty in the LCA is often
discussed from two different viewpoints: There is a scientific discussion on one side, as to which approach is
the best to calculate something rather uncountable14.
And there is a discussion about practise, dealing with how to limit uncertainty of results and how to judge its
importance regarding stability of results and proper decision support.
In GaBi database work PE INTERNATIONAL chooses the following approach to minimise uncertainty:
1. Completing correct data collection (and close mass and energy balances)
2. Choosing a representative LCA data for the upstream and background data, which represent the ac-
tual technology
3. Understanding the technical processes and defining technical parameters that are uncertain
4. Completeness of the system (no unjustified cut-offs)
5. Consistent background data
Consistent data collection and background data are the basis to reducing uncertainty. In addition useful sce-
narios, sensitivity calculations and technical understanding of the LCA modeller (as well as the reviewer)
ensure minimum uncertainty.
If LCA modeller and the reviewer have no indication how the identified technical parameters may perform or
how the parameters are independent from each other, the Monte Carlo Analysis is an alternative. It allows the
examination of consequences of random uncertainties for some selected technical parameters. The quality of
the resulting “uncertainty statements” strongly depend on the selection of these technical parameters, which
should be as representative (in terms of uncertainty) as possible.

14
Not everything that can be counted counts and not everything that counts can be counted. Albert Einstein

62
 Methodological framework

In principle the Monte Carlo analysis should consider each parameter in the model which is uncertain (all
inputs, outputs, parameters, impact values). This analysis is not yet implemented in GaBi. Challenges in this
context are: broad methodological acceptance, availability of (useful) uncertainty information for all model
parameters, implementation effort and probably the calculation performance.
Based on the above discussion, a practical approach to quantify the uncertainty issue was developed for the
GaBi background database.

Quantifying uncertainty in GaBi

Uncertainty in LCA can be split into two parts:
 Data uncertainty (the uncertainty of the modelled, measured, calculated, estimated) and data
within each unit process
 Model uncertainty (uncertainty introduced in the results of a life cycle inventory analysis due to
the cumulative effects of model imprecision, input uncertainty and data variability)
Uncertainty in LCA is usually related to measurement error-determination of the relevant data, e.g. consump-
tion or emission figures. Since the ‘true’ values (especially for back-ground data) are often unknown, it is vir-
tually impossible to avoid uncertain data in LCA. These uncertainties then propagate through the model and
appear in the final result. Small uncertainties in input data may have a large effect on the overall results, while
others will be diminished along the way. This article addresses PE INTERNATIONAL’s recommendations for
addressing the quantification of uncertainty in an LCA study, and how it can be done practically and with rea-
sonable accuracy.
Quantifying the uncertainty of primary data points in company-specific processes can be relatively
straightforward and easy for a company to calculate using the mean value and its standard deviation
over a certain number of data points.
But quantifying the uncertainty in the background systems (hundreds of upstream pro-cesses including mining
and extraction) and then performing error propagation calculation is typically neither practical nor feasible due
to the cost and time constraints in an industrial setting. In addition to put the issue in a general perspective,
one should be wary of data with an extremely precise uncertainty value to each inventory flow, as these can-
not be calculated with the accuracy that the value implies.
A common rule estimates that the best achievable uncertainty in LCA to be around 10%. This was supported
by [KUPFER 2005] on the forecast of environmental impacts in the design of chemical equipment. The actual
degree of uncertainty can vary significantly from study to study.
The overarching question that really must be answered is:
How robust is my overall result when taking into account the combined uncertainties?
The effort to come up with a reasonable estimate can be significantly reduced by following a two-step ap-
proach:
1. Understand the model structure and its dependencies
Keep it simple at first and start by setting up the model with values you have. Then try to develop an under-
standing of the most relevant aspects of the LCA model, i.e. those life cycle phases, contributors, or data
points that have the largest impact on your result. This is usually done by a contribution or ‘hot spot’ analy-

63
 Methodological framework

sis and a subsequent sensitivity analysis. Both of these functions are available to GaBi users in the LCA
balance sheet through the Weak Point Analysis and the GaBi Analyst.
Here is an example: The contribution or ‘hot spot’ analysis of an energy-using product may show that the use
phase is dominating the life cycle greenhouse gas emissions, closely followed by the production of a printed
circuit board and logistics. Sensitivity analyses may then show that the parameters that influence these con-
tributors the most are the split between online and stand-by mode during use, the amount of precious metals
in the circuit board and the distance from the Asian production facility to the local distribution centre.
2. Test the robustness of the model’s results
The next step is to focus efforts on estimating the level of uncertainty of each of the identified key parameters.
Do some more research to establish upper and lower bounds for the relevant parameters. The higher the
uncertainty, the larger these intervals will be. It may even be possible to find data that allows for the calcula-
tion of a standard deviation in literature.
The combined effect of these uncertainties can then be assessed using the Monte-Carlo simulation availa-
ble in the GaBi Analyst. By defining uncertainty intervals around the key parameters, the Monte Carlo simula-
tion is able to produce a statistical estimate (mean value) of the end result (e.g. X kg of CO2 equiva-
lents) as well as its standard deviation across all simulation runs. To do this it simply draws random num-
bers from the defined intervals and calculates a single result using that set of numbers. By repeating this
procedure a multitude of times (10,000 runs is usually a good choice), it will produce a probability distribution
of 10,000 individual results. The lower the standard deviation associated with it, the more robust or
‘certain’ your result is. The resulting mean value is also closer to the ‘real’ value than the value obtained
when doing a simple balance calculation based on the basic parameter settings.
To make the assessment even more robust towards any additional, unknown uncertainties, it is possible to
increase the ascertained intervals around the key parameters by a specific ‘safety factor.’ This will provide
a sound estimate of the robustness of the model.
For more quantified results on uncertainty issues in LCA see Supplement B.
Coefficients of variation
As seen in the above discussion and from quantified results in Supplement B, the percentage maximum error
can easily reach several orders of magnitude for the ‘choose max’ cases. These numbers can be misleading,
though, since they heavily depend on the magnitude of the respective denominator, i.e. the minimum values.
A more unbiased way to look at the variability across the evaluated datasets is to calculate the coefficients of
variation across the absolute indicator results, which is defined as the standard deviation divided by the
modulus of the mean value. When the modulus is used, the coefficient is always a positive value.
The following table displays the maximum coefficients of variation across datasets for each impact category
separately. Again, knowing the country of origin but not knowing the specific technology route can be
worse than the inverse case. The coefficients of variation are significantly higher for the latter case.

Table I: Coefficients of variation

Impact known technology / unknown country of unknown technology / known country of

origin origin

64
 Methodological framework

PED 32% 88%

AP 92% 98%
EP 63% 123%
GWP 47% 89%
POCP 86% 132%

This chapter answered two questions: First, how do I assess the uncertainty of my LCA model in GaBi, and
second, how large are the uncertainties across different datasets assuming that either the country of origin or
the technology route is not known?
While it is known from experience, as well as a recent PhD thesis, that the model uncertainty can scarcely be
kept below 10% once the most appropriate datasets have been chosen, the uncertainty around this choice
can be significantly higher. For most considered datasets, the relative error is between -75% and +250%,
while the coefficient of variation is roughly between 90% and 130%.
Based on these results, the following conclusions can be made:
1. The appropriate choice for dataset is a higher concern for the uncertainty on the elementary flow lev-
el. The selection of the most representative technology route has a large influence on the resulting
environmental profile. The most ‘certain’ dataset can introduce a massive error to your model if it is
not representative to the process / product at hand.
2. When the most representative datasets have been identified and deployed, the next concern is about
the accuracy of your model structure and parameter settings. Here the described functionalities of
the GaBi Analyst can help you understand the dependencies and assess the overall effect on your
results.

In Chapter 3 the methodological framework of goal, scope, types, sources and quality issues of data was
addressed. In Chapter 4 the technical framework of system modelling is built upon this framework.

65
 System Modelling Features

4 System Modelling Features

The GaBi software system was developed to support the complete work flow of LCA work: Starting at data
collection, over life-cycle system modelling, data storage and handling, as well as interpretation.
Appropriate results call for appropriate system modelling and appropriate data. In the following chapter the
technical framework of system modelling is described.

4.1 Data collection

Data collection is the basis for all following modelling steps: Analysing the gathered data, the use of this data
for the set-up of the process models and as the basis for the inventory calculation. The quality of the dataset
will finally depend upon the type, sources, consistency and appropriateness of data collection. A standardised
procedure is therefore defined and applied for GaBi data collections:
 Understanding the core production technique.
 Identifying the generic situation of the manufacturing of the product system to be analysed (e.g.
how many competitive producers exist, what are the applied technologies).
 Identifying the essential single process steps which are dominating the manufacturing phase of
a certain product system. Ideally this process is done in cooperation with industry, validated or
accompanied by experts of the related branch.
 Creating a customised data collection sheet. Golden rule: data collection should be as detailed
as necessary and as efficient as possible; staying on a realistic level, which can be supported
by the data source but also fulfils LCI quality issues. A flow chart of the process helps to have a
good overview and to keep track in technical discourse.
 Inspection of the returned data applying general rules which focus on consistency and quality of
the gathered data, which includes:
o mass and energy balance
o emission and substance balances
o plausibility check focusing the general process characteristics (energy efficiency, yield,
purge streams, residues, by-products, loop substances, recovered matter)
 Provision of feedback to the data supplier or validator.
For the process of data collection different techniques can be used which differ in type, technique and effort.
The following types of data collection can be used:
1. Manual informal (generally not used in GaBi data collection procedures)
2. Manual predefined formats (Word® or Excel® documents)
3. GaBi process recording tool
4. Web based applications (e.g. GaBi 6 web questionnaire)
Collection types 3 and 4 comfortably support the user to integrate data consistently and while saving time into
GaBi.

66
 System Modelling Features

4.1.1 Quality check and validation of collected data

During the process of data collection, our experts prepare a check-list of general points that ensure the data
quality requirements are fulfilled. As previously mentioned these methods include: mass and energy balance,
emission balances, plausibility check, in addition to whether all relevant processes steps and inputs and out-
puts are included.
If anomalies occur, problems are iteratively checked with the data provider or the data-providing expert team
within PE. The goal would be to clarify whether it is a data or methodological problem and whether it is a spe-
cial case or a common issue.
Apart from this technical check, aspects covered by the data quality issues (Chapter 3.5), data sources
(Chapter 3.4) or principles such as goal (Chapter 3.2) or scope like functional unit and system boundaries
(Chapter 3.3) must be checked in order to assure consistency over all data collected. All data aims to repre-
sent the reality, but the kind and detail of needed data sources can differ.
After this check the data considered as “validated” and can be used for modelling in the GaBi framework.

4.1.2 Treatment of missing data

Missing data is a common problem of LCA practitioners (see also Chapter 3.3.5 for gap closing strategies).
This can happen due to unavailability of data or missing access to data. In this case it is up to the expert team
to decide which procedure to adopt.
The goal is to find the missing data and close the gap as efficiently as possible, without unacceptable simplifi-
cations.
There is no standard rule for this problem as each case should be analysed separately, but the following
measures can be taken:
 Literature: reports, papers, books can be checked (standard way, but often no LCA suitable in-
formation available)
 For chemical reactions, often an estimation can be provided by the stoichiometry and estimation
of the reaction’s yield
 Estimation based on similar processes/ technologies
 Expert judgement of a skilled person (supported by one or more above aspects).
The chosen procedure for the treatment of missing data shall be documented according to the ISO 14044
[ISO 14044 : 2006].

4.1.3 Transfer of data and nomenclature

The system modelling starts with the transfer of gathered data into the GaBi software system. GaBi 6 is or-
ganised into modules. Plans, processes and flows, as well as their functions, are formed into modular units.
The fundamental basis of modelling using GaBi 6 is the object type flow. A GaBi 6 flow is a representative of
an actual product, intermediate, material, energy, resources or emission flow.
Elementary flows are resources and emissions that are released from unit processes directly into the envi-
ronment without further treatment, causing a specific environmental impact.

67
 System Modelling Features

Intermediate flows (material or energy) are technical flows between unit processes or a product flow leaving
the final process for further use in a system.
Intermediate flows are used are the link between processes within a life cycle system.
Plans (or plan systems) are used in GaBi 6 to structure the processes in a product system. Essentially, plans
are the “process maps” which visually depict a stage or sub-stage in the system and help to understand the
technical reality behind the system.
A clearly defined nomenclature of flows is needed. GaBi 6 defines all known and used flows consistently by
avoiding double entries (e.g. synonyms).
A clear and defined nomenclature is needed to ease or enable data transfer with other nomenclatures and
systems (like e.g. ILCD 2010). Different nomenclature system are proposes by academia and practise. No
global standard nomenclature currently exists, because theoretical and practical approaches still call for dif-
ferent aspects.
For each modular unit a clearly defined nomenclature is necessary to specify flows, processes and plans. In
the following the most important nomenclature aspects are listed.

Flows
 Name (most commonly used or according to existing systems)
 CAS code
 Abbreviation (e.g. polypropylene PP)
 Chemical formula (e.g. carbon dioxide CO2)
 Technical aspects like calorific value, element content or impact category
 Reference unit (e.g. kg, MJ, Bq, Nm3)
The GaBi software system has a substantial list of consistently predefined elementary flows, so that ideally
only new intermediate or product flows need to be created (look out for synonyms before creating new ele-
mentary flows).

Processes
 Specification of the country
 Name (mostly the name of the product created which is also the functional unit of the process
analysed)
 Addition to the name (e.g. polyamide 6 granulate (PA 6))
 Production technology (if several technologies exist to produce the material)
 Reference year
 Data quality and completeness

68
 System Modelling Features

Plans
The name of the plan system should enable to understand its related system boundaries, the core technology
route and the core location of the operation.
Goal is a consistent naming of the flow, the related process and the related system plan.
GaBi Databases [GABI 2012] have already integrated elementary and product flows for more than 8000 da-
tasets and the respective used flows are documented directly in the process headline.

Figure 4-1: Hierarchical structure of the processes and plans

Since the efficient and flexible combination of processes and plans in GaBi affect the appropriate result analy-
sis, a certain structure of the desired system should be known beforehand. The processes and plans can be
individually structured (shown in the figure above) to represent any desired degree of detail.

4.2 Geographical aspects of modelling

To set data in the correct regional context is an important aspect of LCI modelling. Users in multinational
companies, as well as national and international programs and requirements, call for realistic geographical
representation. Realistic regionalisation is as dynamic as markets. The core issue of regionalisation is not the
methodological approach, but rather the necessary background information on technology and the market
situation.
Country-specific energy (pre-) chains are called for throughout the database (electricity, thermal energy, re-
sources). The most relevant industry processes, including the technology route, in the respective region must
be country or region-specific. If use phase or utilisation (losses or other performance issues) data are rele-
vant, a country-specific situation is necessary. Recycling rates and waste (water) treatments may be adopted,
as well as the crediting of materials and energies in EOL.
In GaBi database work and “data on demand” business a “4 level regionalization approach” is used, which
depends on the goal and scope of the data and the relevance of the related measure on the overall result.
1. Transferring existing technology information into other country by adapting the energy supply

69
 System Modelling Features

2. Adapting the important upstream processes with regional supply data

3. Collecting regional technology (mix) information to adapt existing information
4. Collecting and/or validating primary data in the regional industry networks
If a GaBi dataset is country-specific, at least level 2 is applied. For individual information please consult the
respective documentation.

4.3 Parameter
Parameters are variables within a dataset, which allow the variation of process input and output flows to de-
tach from a strict relationship between input and output flows (scaling). Parameters can therefore be used to
calculate flow quantities (e.g. due to the characteristics of a used substance) based on technical conditions,
such as efficiency of power plant using energy carrier properties or sulphur dioxide emissions depending on
the sulphur content of the used fuel or other parameters.
A typical application of parameterised models (processes) is the modelling of transportation processes. It is
possible to calculate the CO2 emissions by means of a mathematical relation depending on the travelled dis-
tance, the utilisation ratio and the specific fuel consumption of a truck (see Chapter 0).
Important parameterised (background) processes are:
 crude oil, natural gas and coal extraction
 power plants
 refinery operations
 water supply
 wastewater treatment, recycling and incineration processes
 transports
 agricultural processes
 certain metal beneficiation and refining processes
Suitable parameterisation can reduce the error probability seeing as one individual (quality-checked) process
can be applied in many generic situations.

4.4 Multifunctionality and allocation principle

GaBi modelling principles follow the ISO 14040 series concerning multifunctionality.
Subdivision for black box unit processes to avoid allocation is often possible but not always [ILCD 2010].
Subdivision is therefore always the first choice and applied in GaBi database work. This includes the use of
the by-products in the same system (looping).
System expansion (including substitution) is applied in GaBi database work, wherever suitable. The system
boundaries are the key issue. ISO says: “Expanding the product system to include the additional functions
related to the co-products, taking into account the requirements of appropriate system boundaries [ISO 14044
: 2006].

70
 System Modelling Features

It is to carefully check, if the function of the system would be enlarged inappropriately. If this is the case and
the explicit and unique function of the dataset is not clear anymore, the system expansion should not be ap-
plied.
In practise, system expansion can lead to the need for further system expansion because the additional sys-
tems are often multifunctional. In other cases the alternative processes exist only in theory or are of no quanti-
tative relevance in practise. Another challenge is to identify the superseded processes, which will prove to be
complex [ILCD 2010].
The aspects of a (virtually) enlarged system can cause interpretation and communication problems and needs
special attention. The interpretation of the results can grow weaker and less transparent.
System expansion (including substitution) is applied, if it does not lead to misinterpretation or to an enlarge-
ment of the functional unit, because this would be in a conflict with the aim to provide single datasets with
respective functional unit.
In GaBi database work system expansion is frequently applied to energy by-products of combined or integrat-
ed production, where direct use in the same system is not feasible.
Allocation is the third method to deal with multi-functionality. Allocation has long been discussed and debat-
ed, despite the fact that often only one feasible or useful allocation rule is applicable and the relevance of
different allocation keys is frequently of rather low relevance on the results.
Identification of the most appropriate allocation key is essential and often intuitive. The inputs and outputs of
the system are partitioned between different products or functions in a way that reflects the underlying physi-
cal relationships between them, i.e. they should reflect the way in which the inputs and outputs are changed
by quantitative changes in the products or functions delivered by the system. Wherever possible, physical
relationships are utilised to reflect meaningful shares of the burden.
Whereas physical relationships alone cannot be established or used as the basis for allocation, the inputs are
allocated between the products and functions in proportion to the economic value of the products.
Sensitivity analysis of possible choices is helpful to justify a decision. Allocation always works and the sum of
the allocated emissions is 100% of the actual total amount of emissions. Allocation is applied in GaBi, where
subdivision and system expansion (including substitution) fail on the practical level.
If there is a significant influence on the results due to an allocation, a sensitivity analysis can transparently
show the effects and enable interpretations of the results. Different datasets for the same product with differ-
ent allocation keys may be supplied to document relevant sensitivity and to be able to choose the right one in
a given goal and scope.
Our experiences from research and industry projects have shown over time that allocation - using appropriate
allocation keys - is a suitable tool for distributing environmental burdens to specific products. Scenario calcula-
tion and sensitivity analysis to quantify the influences of changing allocation keys are particularly effective.

4.5 Generic Modules as background building block

Some industrial processes or natural systems are highly complex (see Chapter 2.3). Their complexity is not
only characterised by the amount of required materials and processes, but also by their non-linearity in relat-
ing to each other. Complex systems can be often found in electronic products (many materials, parts and
process steps), agrarian systems (natural processes interfering with technical processes with unclear bounda-

71
 System Modelling Features

ries) and construction systems of complex use and secondary effects. If the required materials and processes
are the same for several different systems, the model can be parameterised once and adapted for each pur-
pose individually – as long as the complex relationship is the same and integrated in the model.
The generic module approach is applied to manage complex product models and provides the opportunity to
produce transparent and summarised results within an acceptable timeframe. Generic modules comprise
flexible models with parameter variations, including already-modelled materials and parts. These parameters
allow the variation of system models based on technical dependencies (technically understandable and inter-
pretable parameters). The parameter variation offers the possibility to adapt the models to specific product
properties or modelling design scenarios without the need to create entirely new models.
Generic modules are used for single processes, system parts or the complete manufacturing of a product.
Varying significant parameters allows each individual module of the product chain to be specified. By imple-
menting the entire manufacturing process into a modelled Life Cycle, all effects to each life cycle phase can
be recognised according to the different variations.

4.6 Special modelling features for specific areas

In the following paragraphs specific modelling issues are addressed for key areas, which are applied in the
GABI 2012 database [GABI 2012]:
 Energy
 Road Transport
 Metals and steels
 Chemistry and Plastics
 Construction
 Renewables
 Electronics
 End-of-Life

4.6.1 Energy
Energy is a core issue because its supply and use influences the performance of most industrial products and
services.
Energy supply systems differ significantly from region to region, due to individual power plant parks and indi-
vidual energy carrier supply routes.
Due to its specific situation in different regions and the related complexity, the modelling of the energy supply
takes place at different levels:
 Supply of different energy carriers (e.g. different energy resources)
 Creation of country-/ region-specific mixes for each single energy carrier (e.g. natural gas mix
Germany, crude oil mix EU-27)
 Supply of final energy from conversion to liquid fuels such as gasoline and diesel fuel
 Supply of the final energy by conversion to electricity, thermal energy and steam

72
 System Modelling Features

For detailed modelling the technical processes necessary for the supply of renewable and non-renewable
sources of energy, as well as the analysis of the power plant technology/refinery used in each case for the
production of electricity/fuel, are required.
Supply of Energy Carriers
The supply of an energy carrier includes exploration and installation of the production site, production and
processing. Figure 4-3 shows the natural gas production in Germany as an example to clarify how the energy
carrier supply is modelled. Among the considerations is the need for auxiliary materials for the drilling during
exploration of the gas fields, the energy demand for exploitation of the energy carriers, as well as further con-
sumption and losses, such as venting and flaring of gas during production.

Figure 4-2: Conventional natural gas production in Germany

For the combined crude oil and natural gas production, allocation by energy content (based on net calorific
value) is applied.
Associated gas and wastewater from crude oil production is allocated only to crude oil production. Vented gas
and wastewater from natural gas production is only allocated to natural gas production.
Energy Carrier Mix
For the countries addressed in the GaBi Database, the energy carrier supply mixes (consumption mixes) have
been analysed and modelled. The consumption mixes of the main energy carriers, natural gas, crude oil and
hard coal, have been analysed and modelled in great detail to ensure the needed specification. The infor-
mation about the different shares and sources are based on statistical information.
.

73
 System Modelling Features

Figure 4-3: Natural gas supply for Germany

Production of electricity, thermal energy and steam

Through the utilisation of different energy carriers like gas, oil and coal in their respective power plants, elec-
tricity, thermal energy and steam is produced. The country-specific power plant technologies (efficiency of
conversion, exhaust-gas treatment technologies and their efficiencies) are considered.
In addition direct and combined heat and power generation are considered separately, depending upon the
country-/region-specific situation.
Generic modelling of the power plants enables consideration of both fuel-dependent (e.g. CO2) and technolo-
gy-dependent (e.g. NOX, polycyclic aromatics) emissions, including the effects of emission reduction
measures (e.g. flue gas desulphurisation).
Mass and energy flows, including auxiliary materials (e.g. lime for desulphurisation), are considered during the
energy conversion. The emissions of the power plant and the material and energetic losses (waste heat) are
also taken into consideration. Figure 4-4 shows the modelling of the US, East power grid mix.

74
 System Modelling Features

Figure 4-4: US, East electricity grid mix

The parameterised unit process models in the centre of the plan system are all comprehensive input-output
relations based on several technology settings and calculation steps to represent the given regional technolo-
gy. The following figure provides insight to the degree of engineering detail of the GaBi power plant models.

75
 System Modelling Features

Figure 4-5: Parameterized US Coal gas CHP power plant

For the combined heat and power production, allocation by exergetic content is applied. For the electricity
generation and by-products, e.g. gypsum, allocation by market value is applied due to no common physical
properties. Within the refinery allocation by net calorific value and mass is used. For the combined crude oil,
natural gas and natural gas liquids production allocation by net calorific value is applied.
Energy consumption by power plants themselves and transmission losses of the electricity from the power
plants to the consumers are included in the analysis.
The difference of thermal energy and process steam
The GaBi database offers country-specific datasets for thermal energy and process steam by energy carrier.
For example, the datasets “US: Thermal energy from natural gas” and “US: Process steam from natural gas
90%” are available for natural gas. In the GaBi databases all process steam and thermal energy datasets
refer to the same functional unit of 1 MJ of final energy delivered (“at heat plant”).
The difference between the two types of datasets is related to the conversion efficiency of the energy carrier
consumed to the final energy (steam, thermal energy) produced by the conversion process (heat plant).

76
 System Modelling Features

While the LCI datasets for process steam are provided with several conversion efficiencies, i.e. 85%, 90% and
95%, the thermal energy datasets are calculated with an efficiency of 100% by definition. The thermal energy
datasets therefore represent emission equivalents of the energy carrier consumed in the conversion process.
For practical LCA modelling:
If the amount of fuel (energy carrier), which is converted to final energy, e.g. litres of heavy fuel oil or kilo-
grams of coal consumed, is known, then use the thermal energy processes. In contrast, if the amount of final
energy, e.g. MJ of process steam, is known, then use the process steam processes. The latter is also to be
used if the process steam in MJ is further translated into kg of process steam.
In addition to calculating conversion efficiencies, both types of LCI datasets also consider the energy self-
consumption by the heat plants. Due to this fact, the “overall process system efficiency” is in reality lower than
the conversion efficiency (mentioned above). The conversion efficiencies of 100%, 95%, 90% and 85% should
be documented accordingly as conversion efficiencies.
Summary of most important aspects applied in GaBi energy modelling
 Country-/region-specific resources extraction technology (primary, secondary, tertiary)
 Country-/region-specific power plant and conversion technology
 Country-/region-specific production and consumption mix of energy
 Country-/region-specific transport chains (pipeline, tanker, LNG tanker)
 Specific efficiencies and specific emission equivalents per fuel use
 Specific resource/fuel characterisation per region
 Qualities and characteristics of fuel properties used in power plant models
 Parameterised models for emission calculations (specific standards adapted)
 Country-/region-specific refinery technology
 Unit process modelling based on engineering figures (no black box unit processes)
 Modular energy data provision (separate upstream data, fuel data, consumption mix data, fuel
specific electricity generation data, country grid mix data)
 Deep regionalisation of energy data on all levels and layers of the life cycle model
 Adaptable electricity grid mix data
These main aspects ensure a reliable background database and enable the GaBi user to use the best prac-
tise energy data.

4.6.2 Transport
Transport is the link between process chain steps at different locations. Road, Rail, Air, Ship and Pipeline
transports are the main modes of transport; however, the GaBi background model contains other modes of
transport such as excavators, mining trucks and conveyors.

77
 System Modelling Features

Road transport
Transportation systems are found in the use phase, which contains the fuel demand and released emissions.
The functional units are the following:
 transportation of 1 kg cargo over a distance of 100 km for truck processes
 1 vehicle-kilometre for passenger car processes. In the case of a car, the manufacturing and
end of life phases can be connected to the utilisation model.
Adaptable parameters in the datasets are: distance, utilisation ratio, share of road categories (ur-
ban/rural/motorway), required sulphur content and share of biogenic CO2 in fuel and total payload (total pay-
load only applies to trucks).
Due to the fact that transportation processes are very specific for each situation, these processes are deliv-
ered as parameterised processes for individual adaptation.
Calculation of emissions
The basis for the emission calculation for both trucks and passenger cars is emission factors from literature
[HBEFA 2010].
With the assumption that the utilisation ratio behaves linearly (see [Borken et al 1999]), the Emissions
Factors (EF) [g/km] for 1 kg of cargo can be calculated with the following equation:
EFempty  ( EFloaded  EFempty )  utilisatio n  g 
Emission   km  kg 
payload  1 000  utilisatio n  
EFempty Emission factor for empty run [g/km]
EFloaded Emission factor for loaded run [g/km]
utilisation Utilisation ratio referred to mass [-]
payload Maximum payload capacity [t]

The payload and utilisation ratios are variable parameters, which can be set individually by the dataset user.
The total emissions for each pollutant refer to 1 kg cargo (truck) and 1 km (passenger car) and the transporta-
tion distance is calculated based on the driving share (urban: share_ur / rural: share_ru / motorway:
share_mw), the specific emissions (urEm, ruEm, mwEm) in [g/(km*kg)] and the distance [km].
Equation for trucks:
TotalEmission X  share _ mw  mwEm   share _ ru  ruEm   share _ ur  urEm   distance
x Index for a specific pollutant [-]
share_mw Driving share on motorway [%]
mwEm Motorway specific emissions [g/(km*kg)]
share_ru Driving share on interurban road [%]
ruEm Interurban specific emissions [g/(km*kg)]
share_ur Driving share on urban road [%]
urEm Urban road specific emissions [g/(km*kg)]
distance Driven distance [km]

Equation for passenger cars:

78
 System Modelling Features

TotalEmission X  share _ mw  mwEm   share _ ru  ruEm   share _ ur  urEm 

x Index for a specific pollutant [-]
share_mw Driving share on motorway [%]
mwEm Motorway specific emissions [g/(km*kg)]
share_ru Driving share on interurban road [%]
ruEm Interurban specific emissions [g/(km*kg)]
share_ur Driving share on urban road [%]
urEm Urban road specific emissions [g/(km*kg)]
For CO2 emissions the calculations are based on the emission factors according to the previous equations,
where a constant relation of 3.175 kgCO2/kgfuel for fuel consumption is assumed. A medium density of
0.832 kg/l (diesel), results in 2.642 kgCO2/ldiesel, and a medium density of 0.742 kg/l (gasoline), results in
2.356 kgCO2/lgasoline. Due to biogenic shares in today’s fuel, the possibility is given to select the share of biogen-
ic CO2 emissions of the total CO2 emissions.
For sulphur dioxide, a complete stoichiometric conversion of the sulphur contained in the fuel and of oxygen
into SO2 is assumed. The sulphur content in the fuel is a variable parameter, which can be set individually by
the user.
S + O2 → SO2

x _ ppmS kgS 64 g SO2 kg  kgSO2 

EF _ SO2    fuel _ consumption Diesel  
1 000 000 kg fuel 32 g S kgCargo  kgCargo 
EF_SO2 Emission factor for SO2

x_ppms Mass share in fuel

The emission factor for laughing gas (nitrous oxide, N2O) is assumed to be constant for each emission class
and each category of driving road. The emission factor for ammonia (NH3) is set as constant throughout all
categories.
The following systems and emissions are excluded:
 Vehicle production (for passenger car integration is possible due to existing valuable flow)
 Vehicle disposal (for passenger car integration is possible due to existing valuable flow)
 Infrastructure (road)
 Noise
 Diurnal losses and fuelling losses
 Evaporation losses due to Hot-Soak-Emission
 Oil consumption
 Cold-Start Emissions
 Emissions from air conditioner (relevance < 1% see [SCHWARZ ET AL 1999])
 Tire and brake abrasion

79
 System Modelling Features

Representativeness
Concerning representativeness, the emission classes from “Pre-Euro” to “Euro 5” are covered. The technolo-
gies are representative throughout Europe and can be adapted for worldwide locations with a few restrictions.
There is a need to identify the corresponding emission classes.
The referring locations are Germany, Austria and Switzerland. Due to the similarity of the vehicle structure
and the same emissions limit values, the models are representative for the entire EU. With a few restrictions
the model can be assigned to other countries worldwide. Attention should be paid to the fact that the impreci-
sion increases with the increase of the deviation of the vehicle structure as the basis. The road categories and
the utilisation behaviour also affect imprecision. An adaptation can be carried out by setting the driving share
(mw/ru/ur), as well as the utilisation ratio and sulphur content in the fuel, for individual conditions.
The reference year of the dataset is 2011, that data is representative for the period of 2010 to 2016.
Modification of the age structure of vehicles for each emission class leads to changes of the emission profile.
The validity of the dataset is given for about five years (until 2016). Prognoses in [HBEFA 2010] based on
comprehensive time series report that there is no change of emission profiles within a certain size class,
emissions class or road category. Only the different composition of the total vehicle fleet results in some
changes between 2010 and 2016.

Air Transport
The functional unit of air transportation processes is the transportation of 1 kg cargo over a distance of
2500 km. Adaptable variable parameters in the parameterised datasets (with default setting) are: distance
(2500 km), utilisation ratio (66%), sulphur content of fuel (400 ppm) and share of biogenic CO2 (0%). Three
payload capacity categories (22 t / 65 t / 113 t) are addressed based on technical parameters and properties
of A320 / A330 / B747 aircraft.
Inputs: Kerosene and cargo.
Outputs: Cargo and combustion emissions (carbon dioxide, carbon monoxide, methane, nitrogen oxides,
NMVOC, sulphur dioxide, dust)
Not included in the datasets are plane production, end-of-life treatment of the plane and the fuel supply chain
(emissions of exploration, refinery and transportation).
The fuel supply dataset (kerosene) must be linked with the dataset.
The foundation of the data is specifications for A320 / A330 / B747 aircraft, as well as the Third Edition of the
Atmospheric Emission Inventory Guidebook [EMEP/CORINAIR 2002].
Rail Transport
Rail transport processes cover transportation of bulk commodities or packaged goods via light, average and
extra-large diesel and/or electric cargo train. The functional unit is the transportation of 1 kg cargo over a
distance of 100 km. Variable parameters (with default setting) are: distance (100 km), utilisation (40 %) and
for diesel trains the sulphur content of fuel (10 ppm) and share of biogenic CO2 (5 %).
Inputs: Diesel/electricity and cargo
Outputs: Cargo and for the diesel train also combustion emissions

80
 System Modelling Features

Train production, end-of-life treatment of the train and upstream processes for fuel/electricity production are
not included in the dataset.
The fuel/electricity supply dataset must be linked with the dataset.
The datasets are mainly based on literature data. [ECOTRANSIT2010], [IFEU 2010A]

Ship Transport
Ship transport processes cover transportation of various goods via several inland, coastal and ocean-going
vessels. The functional unit is the transportation of 1 kg of cargo over a distance of 100 km. Variable parame-
ters (with the default setting) are: distance (100 km), utilisation (65% for inland vessels and 48% for ocean-
going vessels), sulphur content of fuel (50 ppm for inland vessels up to 2.7% for ocean-going vessels) and
share of biogenic CO2 (5% for inland vessels and 0% for ocean-going vessels).
Inputs: Fuel and cargo
Outputs: Cargo and combustion emissions (carbon dioxide, carbon monoxide, methane, nitrogen oxides,
NMVOC, particulate matter PM 2.5, sulphur dioxide)
Vessel production, end-of-life treatment of the vessel and the fuel supply chain (emissions of exploration,
refinery and transportation) are not included in the dataset.
The datasets are mainly based on literature data from the International Maritime Organization [IMO 2009] ],
technical information [VBD 2003]Fehler! Verweisquelle konnte nicht gefunden werden., emission data
from the European Energy Agency [EMEP/CORINAIR 2006] and the Intergovernmental Panel on Climate
Change [IPCC 2006].

Transport of fluids in pipelines

The LCI dataset should be used for LCI/LCA studies where fluids must be transported via pipeline over a
longer distance. The dataset allows individual settings of the variable parameters. The following parameters
are variable (default settings): utilisation ratio (28%) and distance (100 km). Default values of the variable
parameters must be checked and adjusted for individual use. The dataset does not include the energy supply
route. Therefore the energy supply dataset (electricity) must be linked with this dataset.
The pipeline transport processes can be used to model transportation of fluids in continuous working pipe-
lines. Some representative diameters (0.4 to 1 m) and gradients of pipelines are analysed, because many
variations are possible. The specific energy consumptions as a function of the utilisation ratio are determined
from four basis formulas. The different energy consumption of different diameters over the utilisation ratio can
therefore be calculated. The average utilisation ratio is approximately 28%. Two ranges of diameters and two
different gradients are shown. Additionally, an average pipeline was calculated. The transported kilometres
and the mass of the cargo are known, so the energy consumption in MJ of electricity can be calculated. The
distance and the mass of the transported cargo must be entered by the user. Different pipelines can be cho-
sen (varying the gradient and diameter). The energy consumption is calculated per ton cargo.
Inputs: Cargo and electric power
Outputs: Cargo

81
 System Modelling Features

Not included in the datasets are pipeline production, end-of-life treatment of the pipeline and the electricity
supply chain.
The main source of data is the energy consumption study for transportation systems of the RWTH Aachen
[RWTH 1990].

Other Transport
Other transport consists of excavators for construction works and mining activities, as well as mining trucks.
The functional unit is the handling of 1 t of excavated material. Vehicle performance, load factor, fuel con-
sumption, emission factors, sulphur content of fuel and other technical boundary conditions can be individually
adapted via variable parameters. The predefined parameter settings represent an average performance of the
vehicle.
Inputs: Diesel and excavated material
Outputs: Excavated material and combustion emissions due to engine operation, including regulated emis-
sions (NOx, CO, Hydrocarbons and Particles), fuel-dependent emissions (CO2, SO2, benzene, toluene and
xylene) and others such as CH4 and N2O
Not included in the datasets are vehicle production, end-of-life treatment of the vehicle and the fuel supply
chain.
The datasets are mainly based on vehicle-specific technical data, as well as averaged literature data for emis-
sion profiles from the European Energy Agency [EMEP/CORINAIR 2006B].

4.6.3 Mining, metals and metallurgy

Primary metals are sourced from metal ores containing several different metal components. The production of
a certain metal is therefore typically accompanied by the production of metallic and non-metallic co-products,
e.g. nickel production with cobalt, other platinum group metals and sulphuric acid.
To calculate the Life Cycle Inventory of a single metal, the multifunctionality between product and co-products
must be addressed. Allocation is often the only suitable way to deal with these highly complex co-production
issues in a way that the technical circumstances are properly reflected. The choice of an appropriate alloca-
tion key is important because the metals and other valuable substances contained in ores are very different
concerning their physical properties and value.
For metals with different economic values (e.g. copper production with gold as a co-product), the market price
of the metals is a suitable allocation factor. In order to maintain consistency in environmental impacts as mar-
ket values vary, average market prices over several years (e.g. 10-year market averages) are used. Usually
the market price for concentrate or metal ore cannot be easily determined and in this case the market price is
“derived” based on the metal content.
For other non-metallic co-products, such as the co-products sulphur, benzene, tar of coke for integrated
steelwork creation, other allocation factors are applied, such as the net calorific value.
The metal datasets represent cradle-to-gate datasets of the actual technology mix, e.g. a region-specific mix
of pyro-metallurgical and hydrometallurgical processes for the production of non-ferrous metals, covering all
relevant technical process steps along the value chain, including mining, beneficiation (ore processing includ-

82
 System Modelling Features

ing jaw crushing, milling, Dense Media Separation, Heavy Media Separation (HMS)), smelting (e.g. rotary kiln,
flash furnace, blast furnace, TSL furnace, electric arc furnace), magnetic separation or leaching and refining
(chemical or electro).
The LCI modelling of the process steps mining and beneficiation considers the composition of the mined ore
bodies and the related metal-, process- and site-specific recovery rate, e.g. mill recovery rates within copper
production could be Cu (90%), Mo (75%), Ag (70%) and Au (70%).
Under the assumption that tailing dams include a lining system where water is captured and put back in set-
tling dams or water treatment facilities for reuse, the tailing dam emissions are considered as water losses
through evaporation of the tailing dam.
Metal Recycling
Considering and evaluating the potential and benefit of metal recycling in LCA depends on the specific char-
acteristic of the data system (e.g. field of application, question to be answered, goal & scope). The following
principles are to be taken into account in setting up the life cycle system as the basis for a suitable and repre-
sentative database for metals:
1. Market situation: According to the specific market situation, the metal production of the system un-
der study can be characterised as primary metal production, secondary metal production or the mar-
ket mix from possible primary and secondary production routes.
2. Upstream burden and downstream credit: For metals recovery, the end of life consideration cover-
ing the recycling of metal (downstream credit) turns into an upstream consideration (upstream bur-
den) from the viewpoint of the product system consuming the recovery metal. Chapter [Link] Alloca-
tion procedure in ISO 14044 [ISO 14044 : 2006] requires that allocation procedures must be uniform-
ly applied to similar inputs and outputs of the product system under study, i.e. the use of recovered
metal within a product system (=input) is to be treated equally from a methodological point of view to
metal recovery from a product system (=output). Often this requirement is met by considering only
the net amount of recovered metal to credit for metal recovery. The net amount of recovered metal is
specified by the difference in the amount of metal recovery at the end of life of a product, as well as
the use of recovered metal for production of the product system considered. This procedure is justi-
fied as only the metal loss over the complete product life cycle that is to be taken into account. Nev-
ertheless, in doing so, the differences between the single life cycle phases (production, use and end
of life) will be obliterated.
3. 100% primary / 100% secondary production routes: It should be noted for Life Cycle Inventory
modeling that in actual metal production a 100% primary or a 100% secondary route is not always
given.
4. Definition of key parameters: A mutual understanding of the definitions and terms, e.g. Recycling
rate in LCA = “Ratio of amount of material recycled compared to material introduced in the system in-
itially” is highly important.
5. End of Life scenario/situation “versus” End of Life methodology/approach: It is necessary to
distinguish between the End of Life scenario describing the recycling situation at products’ End of
Life, e.g. recycling into the same product system, no change in inherent material properties, and the
(modelling) approaches/methodologies applied to consider and describe the resulting effects within
LCA.

83
 System Modelling Features

In LCA practice various methodological approaches to consider the recycling of products at their End of Life
phase within LCA are applied. Aspects to be considered in selecting the appropriate End of Life approach are:
ISO-conformity, mass and energy balance, reflection of optimization and reality, data availability, transparen-
cy, easy communication and understanding, field of application and fairness (to any material or product appli-
cation).
A harmonised and consistent description and discussion of these approaches can be found in PFLIEGER/ILG
200715.

4.6.4 Chemistry and plastics

Chemical and plastic products are key players toward environmental performance for two reasons: Chemical
and plastic production uses substantial amounts of energy and resources but the resulting products help to
save substantial amounts of energy or reduce environmental burden in suitable applications. Chemical and
plastic products therefore provide an important foundation for many other industrial fields and products. In
electronics, automotive and construction chemicals and plastics are used in various systems as input materi-
als. It is therefore important to achieve a level of high engineering quality in the modelling of the processes in
these fields.
Primary data collection and/or industrial feedback or validation of the information used, are the best choice.
With specific engineering knowledge, data for chemical plants and operations can be developed with second-
ary information, thus making industry/expert feedback and validation even more important.
Data development of chemical processes follows a defined route in GaBi database work.
1. Information about current technologies is collected
2. Checking relevancy for the given geographical representation
3. Defining the name of the reaction route(s). There is often more than one, even with the same reac-
tants.
4. Defining related stoichiometric equations
5. Defining suitable yields
6. Drawing a process flow sheet
7. Setting up the unit process network and the system
A validation or benchmark of the secondary data with existing data or is done.
Modelling
For each material several different processing technologies are often available. For example, for the produc-
tion of polypropylene, “polymerisation in fluidised bed reactor” and “vertical stirred reactor” is both technolo-
gies that are applied. For each relevant technology an individual process model is created.
Chemical and plastics production sites are often highly integrated. Modelling a single substance product chain
is possible by isolating integrated production lines. The following figure gives a simplified overview for im-
portant organic networks.

15
[Link]
Metalle_S01_v02_2007.pdf

84
 System Modelling Features

Figure 4-6: Excerpt of the organic network16 considered in GaBi

To avoid inappropriate isolation measures it is essential to have engineering and technical information to
accurately model those systems.
A well-arranged online overview of important parts of the chemical network is given on the Plastics Europe
Homepage17.
In case of chemical and plastics, it is not meaningful to apply generic modules because the technology speci-
fications differ significantly. Country-specific consumption mixes are useful, because chemical and plastic
products are traded worldwide, meaning that a chemical or plastic material which is provided in a certain
country can be imported from other countries. For the creation of country-specific models, see Chapter 4.2.
Chemical processes often have a co-product system. Unit process isolation (subdivision) is preferable in this
case. If it is not possible, energy products (e.g. fuel gases or steam) are substituted. For remaining by-
products, allocation is applied. If all products and by-products have a calorific value, the allocation key energy
is often used, because it is a good representation of value and upstream demand.
Waste and/or wastewater are always treated (landfill, incineration and/or wastewater treatment) if treatment
pathways are obvious. The treatment technology (landfill or incineration or wastewater treatment) is selected
according to the country-specific situation or individual information.
Production and consumption mix
As the users of the dataset are not always able or willing to determine the exact technology for the production
of their upstream materials, a representative production mix or consumption mix is also provided. The share of

16
Acknowledgements to Dr. Manfred Schuckert for introducing the organic network thinking in the early 90s into GaBi. Still
not broadly considered in the complete LCA community.
17
[Link] (checked 03.11.2011)

85
 System Modelling Features

production or consumption was determined, separately from the dataset for each relevant technology. For
chemicals with different possible production routes, the technology mix represents the distribution of the pro-
duction mix of each technology inside the reference area.
For example, the production of standard polypropylene in the different regions is based on different polymeri-
sation technologies, including the fluidised bed reactor and the vertical stirred reactor. For standard polypro-
pylene the main process models are mixed according to their share in industrial applications with an average
polypropylene dataset.
The consumption mix considers the material trade. Figure 10 shows an example of a mix for the consumption
of epoxy resin in Germany for the reference year 2011. The epoxy resin, which is consumed in Germany, is
produced in Germany (53.4%), Switzerland (20.3%), the Netherlands (9.1%), Italy (8.5%), Spain (4.5%) and
Belgium (4.2%), as seen in the following example.

Figure 4-7: Consumption mix of Epoxy resin in Germany

Technology aspects
A suitable technology route is important for the proper modelling of chemical data. Technological differentia-
tions in GaBi chemical process modelling are considered for different technology routes such as:
 Chlorine and NaOH (amalgam, diaphragm, membrane technology)
 Methanol (combined reforming stand alone and integrated)
 Steam Cracking (gas to naphtha input shares and related product spectrum)

86
 System Modelling Features

 Hydrogen peroxide (SMA and Andrussow process)

 Hydrogen (steam reforming natural gas/fuel oil via synthesis gas, cracking/refinery by-product)
 Oxygen/nitrogen/argon (liquid or gaserous)
 Sulphuric acid (refining desulphurisation, fertiliser production, secondary metallurgy)
 Hydrochloric acid (primary, from epichlorohydrin synthesis, from allylchloride synthesis, from
methylene diisocyanate synthesis, from chlorobenzene synthesis)
 Benzene, toluene and xylene (from reformate or pyrolysis gas or dealkylation or by-product sty-
rene)
 Acetone (via cumene or isopropanol)
 Hexamethylemediamine (via adipic acid or acrylonitrile)
 Titan dioxide (sulphate and chloride process)
 Caprolactam (via phenol or cyclohexane)
 Ethylene oxide (via O2 or air)
The correct technology route for the right process chain can be decisive. PE INTERNATIONAL’s knowledge is
constantly updated according to the latest developments in the chemical industry, including from being open
to feedback and constructive comments while keeping the chemical networks up-to-date.
By-product handling
Methodological tools such as allocation or substitution open up ways to cope with any by-products. Technical
reality guides GaBi modelling, first and foremost, before methodological choices are made. Prominent by-
products are:
 steam (often not at a level of pressure that is directly compatible to the necessary input level)
 fuel gases
 various inorganic or organic acids
 purge or impure side streams
 unreacted monomers
 various salts
In GaBi chemical modelling the use or fate of by-products is investigated. Often chemical sites have a steam
system with various feeds and withdrawing points with different temperature and pressure levels, which
makes substitution of proper temperature and pressure level a suitable approach to handle the overall benefit
of the by-product steam for the entire plant.
Fuel gases can often be used in firing or pre-heating the reaction within the plant, to reduce the use of primary
sources. Related emissions are taken into account.
Acids are often sold. Allocation takes into account that those extracted acids must be cleaned, purified, diluted
or concentrated.

87
 System Modelling Features

Purge and impure side streams or unreacted monomers are often cycled back into the process after cleaning,
distillation or purification.
Proper methodological handling and technical modelling based in fact are important..
Polymer modelling
Aside from the aforementioned topics of consistent mass and energy balances and the correct technology
route, another aspect of polymer modelling should be mentioned: There is a difference between polymer
granultate/resin, polymer compound and polymer part.

crude oil rock salt

raw material
hydrogen
extraction
chlorine-alkali-
energy electrolysis

refinery /
water steam cracker chlorine conditioning
ethylene processes

auxiliary materials chlorination

and additives
vinyl chloride
stabilisers,
plasticisers, polymerisation
fillers
production
energy PVC processes

injection
extrusion calendering blow moulding
moulding

e.g. profiles, e.g. plates e.g. parts e.g. bottles

cables, pipes and films

Figure 4-8: Example of PVC resin - compound- part

As compounds can be produced and used in thousands of specific recipes, GaBi primarily provides granulate
data, which can be used individually to add additives to produce individual compounds and to set up individual
polymer part data.

4.6.5 Construction
The construction sector uses extensive quantities of natural resources, raw materials and energy. Within the
European Union, the construction sector is responsible for a share of 10% of the gross domestic product
(GDP) and creates about 7% of the total employment. Considering their entire life cycle, buildings and con-

88
 System Modelling Features

struction products are responsible for the consumption of approximately 40% of the total European energy
consumption, as well as for the consumption of approximately 40-50% of natural resources.
The anthropogenic material flows caused by the life-cycle of buildings contribute through many environmental
categories to the impact potentials. In order to describe a building during the entire life-cycle, various infor-
mation concerning the depletion of mineral resources (mining and production of building materials), depletion
of energetic resources and release of pollutants (construction material production and transport, energy sup-
ply of production and during utilisation of the building), land use (a quarry and surface sealing by the building)
and waste treatment (construction, use, renovation, demolition) is required.
To structure these datasets, the life cycle is systematically divided into several unit-processes, respectively
forming a chain, becoming a network that represents the mass and energy flows caused by a building from
cradle to grave.

Figure 4-9: Schematic life cycle of a building

Every construction material is produced in order to fulfil a function within building or construction. Accordingly,
analysing individual construction materials should not be done without employing a functional unit that consid-
ers the construction material’s purpose or without considering where it is intended to be used. The functional
unit should always include the performance of a material within a building structure. Simple comparisons on
the basis of mass are misleading.
The background data (e.g. transport, energy supply) used to model the production of construction materials
must be comparable. It will be true for system boundaries and methodological key points (such as cut-off-
criteria and allocation rules), and may influence the result considerably. For construction materials the con-
sistent GaBi background system is used.

89
 System Modelling Features

The GABI 2012 database [GABI 2012] for construction materials covers the most relevant construction materi-
als, as well as more specialised materials used in the construction of buildings, roads or subsurface construc-
tions. It is divided into mineral products (including concrete and concrete products, bricks, sandlime, natural
stones, as well as mineral insulation materials such as rockwool and glass wool), metals (construction), poly-
mers (for construction, including insulation materials such as PUR, EPS or XPS), wood for construction, ce-
ment and gypsum/mortar products and coatings and paints. The database also contains several ready-to-use
building components such as windows with different dimensions and frame types. These windows are based
on a generic, parameterised window model that is capable of “assembling” windows by adjusting parameters.
Such a window model allows for the efficient generation of additional windows, if required.
As stated above, the life cycle inventories of construction materials are – similar to the underlying construction
materials themselves – set up in order to meet a functional demand within a building or other construction and
therefore life cycle analyses in the construction sector must consider the intended function. At the LBP-GaBi
and PE working group, a generic building model has been developed in order to meet the demand for analys-
ing construction materials, as well as construction elements and entire buildings, within the respective context.
This building model served as the methodological basis for the life cycle analysis of the European residential
buildings stock and, since then, has constantly been undergoing further development in order to meet the
needs of building planners, architects and engineers to assess the life cycle performance of existing or
planned buildings. The building model contains not only the construction and frame of the building, but also
heating, cooling and technical appliances.
One special feature in the construction sector is the use of a ‘recycling potential.’ The recycling potential quan-
tifies the environmental burdens that can be avoided by the use of recycled materials in comparison to the
production of new materials. Since metals currently present the highest re-use rate among other construction
materials, they are used as example to explain this concept below.

4.6.6 Renewables
Basic approach of the model
Due to the determinant influence of environmental conditions being variable in time and space at a high spa-
tial heterogeneity of site conditions, agrarian systems belong to the most complex production systems.
Due to the inherent complications characterising an agricultural system, a nonlinear complex agrarian model
was used for plant production (developed by PE and the University of Stuttgart, Germany), this model covers
a multitude of input data, emission factors and parameters.
One significant advantage of the model is its worldwide usability. With affordable operating efforts the model
provides consistent and very accurate results for various agricultural and plantation products and differentiat-
ed adjustable farming practises.
General information on the foreground system
Agrarian systems belong to the most complex production systems within LCA due to their dependence on
environmental conditions that are variable in time (e.g. within a year, from year to year) and in space (e.g. by
country, region, site conditions). The following factors contribute to the complexity of agricultural modelling:
 The variety of different locations

90
 System Modelling Features

 Small scale soil variability within locations

 The large number of farms
 The variety of agricultural practises
 Technically, no determined border to the environment
 Complex and indirect dependence of the output (harvest, emissions) from the input (fertilizers,
location conditions)
 Variable weather conditions within and between different years
 Variable pest populations (insects, weeds, disease pathogens)
 Different crop rotations
System boundary for renewable model
The model includes cradle-to-gate burdens of all relevant input materials for the cultivation process itself
(commercial fertilizer including lime, organic fertilizer, pesticides, seeds, including their production and
transport). The model includes the cradle-to-gate emissions of fuel consumed, as well as the direct emissions
to air from combustions, in the field for operations fuel. The model includes irrigation where it is relevant and
excludes agricultural infrastructure and farm buildings. All relevant processes taking place in the area under
cultivation including emissions into air and ground water (lower limit of rooted soil zone) are considered.
Heavy metals remaining in soil are considered as emissions in soil. Integration of erosive loss of Norg (organic
nitrogen) and Corg (organic carbon) as well as of nutrients (e.g. phosphorus) in water is considered.
Time reference
For annual crops a cultivation period is considered beginning immediately after the harvest of the preceding
crop and ending after harvest of the respective cultivar. In the case of perennial cultivars (i.e. plantations), the
process starts with plantation preparation (ground clearing) and lasts until the respective cultivar must be
cleared for further uses.
Nitrogen cycle
Nitrogen plays a fundamental role for agricultural productivity and is also a major driver for the environmental
performance of an agricultural production system. For these reasons it is essential to evaluate all relevant
nitrogen flows within, to and from the agricultural system. PE’s agriculture model accounted for the nitrogen
cycle that occurs in agricultural systems.
The agriculture and plantation model consists of several sub-models. These are more or less interconnected,
and therefore, modelled together in just one process.
In principle the model considers two themes:
 Allocation of fertilizer on crop rotations (allowable external fertilizer needs or profit of the main
crop)
 The N losses as NO3 into the groundwater and NH3, N2O and N2 into the air and on soil erosion
as NO3- and Norg18 in surface waters during the cultivation of the main crop

18
Organic Nitrogen

91
 System Modelling Features

The model ensures that nitrogen emissions are consistent for the cultivated species. Specifically the model
includes emissions of nitrate (NO3-) in water and nitrous oxide (N2O), nitrogen oxide (NO) and ammonia (NH3)
into air. The model ensures that emissions from erosion and nutrient transfers within crop rotations are mod-
elled consistently. The figure below shows the most important nitrogen flows illustrated by an intensive cultiva-
tion system of an example grain.

Figure 4-10: Nitrogen in the agrarian system

The different N-based emissions were calculated as follows:

 NH3 emissions to air from organic fertilizers were adapted from the model of (BRENTRUP, F. ET.
AL. 2000.) and modelled specifically for the cropping system dependent on the fertilizer-NH4
content, the soil-pH, rainfall and temperature. NH3 emissions to air from mineral fertilizers were
also adapted from (BRENTRUP, F. ET. AL. 2000) and modelled specifically for the cropping sys-
tem dependent on the kind of fertilizer and the soil pH.
 N2 emissions to air occur from complete denitrification. N2 emissons were also taken into con-
sideration to determine the nitrate leaching potential.
 NO emissions to air occur from partial denitrification.
 N2O emissions to air occur from partial denitrification. Indirect N2O emissions are calculated
from the shares of N leached/runoff and volatilised by using emission factors.
 Norg, NO3- and NH4 emissions to water occur due to erosive surface run-off.
 NO3- emission to groundwater is calculated based on the rest of the incoming N not occurring
as gaseous losses or in harvests, litter, unused extractions from the site, storage in soil. De-
pending on the leakage water quantity during the time period evaluated, an increasing part of
this remaining N was calculated as leached nitrate.
Besides nitrogen-based emissions to water and air, phosphorus emissions are taken into consideration.
Phosphorus is a stable compound which is not leached to groundwater significantly but can be washed out

92
 System Modelling Features

with surface runoff of soil to surface water. The reason for this is that the washed out amount of soil contains a
portion of phosphorus which causes an eutrophication effect in the water bodies.
Carbon modelling
Carbon-based emissions such as CH4, CO and CO2 are considered in foreground and background datasets.
Background datasets include emissions resulting from the production of fertilizer, pesticides, electricity and
diesel, while foreground datasets contain emissions such as CO2 due to combustion of fossil fuels by the
tractor or irrigation engines and application and decomposition of urea fertilizer in the soil.
Soil carbon is another potential source or sink of carbon dioxide. Soil carbon balances are used to describe
any increase or decrease in soil organic carbon (SOC) content caused by a change in land management, with
the implication that increased/decreased soil carbon (C) storage mitigates or increases climate change. Limi-
tations of C sequestration for climate change mitigation include the following constraints: (i) the quantity of C
stored in soil is finite, (ii) the process is reversible and (iii) even if SOC is increased there may be changes in
the fluxes of other greenhouse gases, especially nitrous oxide (N2O) and methane. Due to the strong scepti-
cism present in current literature, the large amount of variables included in soil carbon sequestration is not
considered within the scope of this study.
Aside from emissions, positive effects (sinks) due to natural conversion of gases in the soil were considered.
Gaseous sinks are related predominantly to the methane depression function of natural soils due to their
oxidising and microbial transformation of methane.
The biogenic CO2 sequestered in the cotton plant and its fibre is directly accounted for in the inventory as an
input or uptake of carbon dioxide, which is treated as a negative emission of carbon dioxide to air.
The complete bio-carbon modelling across the entire GaBi database content was checked and wherever
divides into bio-carbon and fossil carbon could be identified, it was updated.
Water in the renewable modelling
Water use is modelled based on the calculations of Pfister (PFISTER2011). A generic water model allows the
selection of different plant water requirements and irrigation regimes depending on the specific regional condi-
tions (e.g. precipitation, irrigation demand and irrigation technique). For details please see the reference "En-
vironmental Impact of Water Use in Global Crop Production".
Land use in the renewable modelling
Based on site-specific soil and climate parameters such as soil type, humus content, summer precipitation or
evapotranspiration, implications of different land use types have been modelled for both transformation and
occupation.
A set of indicators related to ecosystem functions has been defined to model land use aspects in LCA and
incorporate them into the software: Erosion Resistance, capability of soil to prevent soil loss; Physicochemical
Filtration, ability of soil to absorb dissolved substances from the soil solution to prevent pollutants from enter-
ing the soil matrix (characteristic value: cation exchange capacity); Mechanical Filtration, mechanical ability of
soils to clean a suspension through the binding of pollutants on soil particles; Groundwater Replenishment,
capacity to regenerate groundwater; Biotic Production: ability of the ecosystem to produce biomass.

93
 System Modelling Features

Electricity and thermal energy in the renewable modelling

The electricity (and thermal energy as by-product) used is modelled according to the individual country-
specific situation. The country-specific modelling is achieved on multiple levels. First, the individual power
plants in service are modelled according to the current national grid. This modelling includes net losses and
imported electricity. Second, the national emission and efficiency standards of the power plants are modelled.
Third, the country-specific fuel supply (share of resources used, by import and/or domestic supply) including
the country-specific properties (e.g. element and energy contents) are accounted for. Fourth, the import,
transport, mining and exploration processes for the energy carrier supply chain are modelled according to the
specific situation of each power-producing country. The different mining and exploration techniques (emis-
sions and efficiencies) in the different exploration countries are accounted for according to current engineering
knowledge and information.
Transport in the renewable modelling
If transportation is present, all relevant and known transport processes are included. Overseas transport,
including rail and truck transport to and from major ports for imported bulk resources are included, if relevant.
Furthermore, all relevant and known pipeline and/or tanker transport of gases and oil imports are included.
Fuels in the renewable modelling
Coal, crude oil, natural gas and uranium are modelled according to the specific import situation (see electrici-
ty). Refinery products: Diesel, gasoline, technical gases, fuel oils, basic oils and residues such as bitumen are
modelled via a country-specific, refinery parameterised model. The refinery model represents the current
national standard in refinery techniques (e.g. emission level, internal energy consumption) as well as the
individual country-specific product output spectrum, which can be quite different from country to country. The
refinery products used show the individual country-specific use of resources. The supply of crude oil is mod-
elled, again, according to the country-specific crude oil situation with the respective properties of each re-
source.

4.6.7 Electronics
The distinct characteristics of electronic and electro-mechanic components are complexity, sizeable numbers
and the variety of part components. Considering the existing part components, more than 10 million compo-
nents can be counted. An electronic subsystem (e.g. PWB - Printing Wiring Board) is often equipped with
several hundreds of different components.
The demand exists to make datasets for electronic components available, since electronics are applied in
various fields such as automotive, houses, consumer products, and information and communication systems.
It is currently not possible from a timeframe and resource perspective to create an individual dataset for each
of the 10 million electronic components. The challenge here is selection, which datasets to utilise, how to deal
with the vast amount of parts and how to reduce the numbers of datasets by providing the representativeness
of those datasets.
In order to make a statement about the representativeness of an electronic component, the whole scene must
be understood. The extensive experience of the electronics team at PE INTERNATIONAL facilitates repre-
sentative component determination, after having analysed hundreds of electronic boards and al-
ways/often/rarely-used components and their applications. Knowledge of often-used materials and most sig-

94
 System Modelling Features

nificant steps of component manufacture are also important. The identification of significant manufacturing
steps is supported by other technical fields. If data are not directly acquired from the electronics supply chain,
either similar technical processes or comparable technical fields in which the identified manufacturing pro-
cesses have been applied, supporting the determination of the relevant environmental impact. Only the inter-
action of all three conditions: experience, knowledge about similar processes and knowledge concerning the
market situation, make the identification of relevant and representative components with their technologies
and materials possible.
Even though not all electronic components can be judged according to their representativeness, the most
relevant causes of environmental potentials from groups of similar electronic components can be identified,
after the investigation of a certain amount of products. For example the difference in environmental impacts is
possible to identify between semiconductors and resistors, or between active components (e.g. semiconduc-
tors, diodes and discrete transistors) and passive components (e.g. capacitors, resistors, inductions), or even
by comparing different types of technologies (e.g. SMD (surface mount device) or THT (through hole technol-
ogy)). The more knowledge is gained, the better and easier it is to identify which fields and components of
electronic products cause significant and less significant environmental impacts.
In order to model representative electronic products, subsystems or components, environmental knowledge
and availability of huge numbers of materials are necessary, such as metals, plastics and ceramics, since
electronic products can consist of most elements in the periodic table. Additionally, a broad range of many
technical manufacturing processes and their environmental causes are necessary to know, such as sputter-
ing, lacquering, sintering, winding, soldering, clean room condition, etching, electrolysing, vacuum metal dis-
persion and many more.
As a result a list of electronic components covers this vast milieu. Its representatively is distinguished by vari-
ous specifications related to their function, size, housing types, material content and composition, as well as
mounting technology.
Clearly-structured nomenclature including all required information for component specification ensures the
intended use of available datasets:
Examples for dataset nomenclature:
Capacitor Al-capacitor SMD (300mg) D6.3x5.4 Diode power THT DO201 (1.12g) D5.3x9.5

Function Technology Mounting Mass per Dimensions Function Mounting Housing/ Mass Dimensions
technology piece technology technology

For representative LCI models of electronic assemblies and systems, like populated printed wiring boards, the
following modelling principles are applied:

 Electronic components are modelled according to component-specific properties, e.g. function,

case type, size, number of pins, die size, SMD/THT.
 Electronic components are modelled according to a functional unit “Number of pieces.”
 In the event that a dataset representing a component to be modelled is not available in the Ga-
Bi database, , informed assumptions are made by choosing electronic components that are
most similar, and related to housing types, function and production processes. A component
scaling tool is available to support such a selection process.

95
 System Modelling Features

Printed wiring boards (PWB) are mainly modelled by area (functional unit) due to fact that PWB dimensions
and number of layers are the most sensitive parameters for PWB-related environmental impacts and primary
energy use.
Modelling
Based on the necessity to model and assess electronic systems with justifiable effort, the electronics team of
PE INTERNATIONAL developed the modular system called Generic Modules system. The target is to estab-
lish a Generic Module for each group of electronic components, e.g. resistors, ceramic capacitors or sub-
strates.
The model based on Generic Modules of a typical electronic system follows a hierarchical structure. The sys-
tem is divided into several subsystems. The subsystems themselves are modelled on the basis of the Generic
Modules, as presented in Figure 4-11.

Figure 4-11: Creation of a model for an electronic product - modular structure via Generic
Modules

Technical systems form the basis for highly-flexible modules. With few variable parameters such as size,
number of layers and type of finishing in the case of a PWB, these modules can be adapted to a specific
product or system under consideration.
After the determination of the representative components and their relevant technologies, for typical electronic
subsystems, a Generic Module is created: housing, substrate, connection system, electronic components and
electro-mechanical parts:
Housing: Typical housings are made by injection-moulding of plastics (e.g. PC/ABS) or are metal housings
(e.g. from aluminium die casts or steel sheets). The models contain all relevant preliminary process steps. For
plastic housings it is crude oil extraction, production of plastic granulate and the injection moulding itself, in-
cluding the respective demand for auxiliaries, energies and transport in each process step.

96
 System Modelling Features

Substrate: The substrate is the PWB without components or the connection system. PWBs are modelled
according to the number of layers, size, weight and composition (e.g. content of copper, glass fibres, TBBA or
Au/Ni finishing). If this information is not available, pre-defined average compositions may be used as de-
scribed above.
Connection system: Usually solder pastes, formerly mainly SnPbAg and now typically lead-free solders, are
used based on a number of varying metal solder elements.
Electronic components: An extensive database containing the material contents of the main groups of compo-
nents such as resistors, capacitors, coils, filters, transistors, diodes and semiconductors are available. Seeing
as millions of different components may be contained in electronic products, they are reduced to several rep-
resentative components and are constantly updated and extended.
Electro-mechanical and other parts: This subsystem contains models of switches, plugs, heat sinks or shield-
ing and other non-standard parts such as displays, keys or sensors.
The Generic Modules are adapted via variable parameters. The significant functional units used depend on
the subsystem, e.g. piece for components, area for boards and assembly lines, kilograms for solders and
electro-mechanics.
The GaBi database contains aggregated datasets for components, which are based on the above-described
Generic Modules. Further datasets can be set up easily using the Generic Modules.

4.6.8 Recycling or End-of-Life measures

Resource conservation and keeping valuable materials in the technical life-cycles are relevant aspects in
analyzing the environmental performance of many materials.
After the life cycle phases of production and use/maintenance several options exist concerning the further
application of used materials and products (like recycling, recovery and disposal or any share of each) or
offsetting their secondary value.
According to ISO only elementary flows (plus the product flows) describe a Life Cycle Inventory. Secondary
materials such as scrap (like metal scrap or glass cullet) represent non-elementary flows and are linked to
previous or subsequent product life cycles. Within a LCA study these flows are typically modeled following
methodological approaches such as cut-off approach, closed loop approach, open loop approach and value-
corrected substitution approach.
Within the GaBi databases [GABI 2012] the cradle-to-gate data for metals (or container/float glass) still list the
externally supplied secondary material inputs (e.g. carbon steel scrap sourced from merchants or other steel-
works), if given and of significance regarding the overall environmental performance. This allows the user of
the dataset to apply the methodological approach of choice to analyze in detail the potential/benefit of recy-
cling along the life cycle of a product. Example life cycle models are provided within the GaBi databases for
user guidance [GABI 2012].
In cases where an input or output of a secondary material is of no or very low relevance regarding the envi-
ronmental life cycle performance of a material or product, the modeling of secondary material inputs or out-
puts is completed, using the “value of scrap” approach, to avoid misinterpretation19.

19
The possible (small) error made introduces much less uncertainty than the potential (large) error to be made, if left un-
treated.

97
 System Modelling Features

The “value of scrap” approach addresses the question of how to deal with the recycling of metal scrap in
LCA/LCI. The principle idea behind the approach is to define the Life Cycle Inventory of metal scrap, describ-
ing the “value of scrap.”
The “value of scrap” is defined as the difference in LCI of the (theoretical) 100% primary and 100% scrap
production routes in metal production, considering the process yield of the recycling step.
Datasets provided with GaBi with the “value of scrap” are carbon steel scrap by World Steel Association
(worldsteel) and stainless steel scrap by the European Steel Association (EUROFER).
Furthermore, we provide datasets on “value corrected substitution”. The intent is to apply a value-corrected
credit for the substitution of metals in open-loop recycling situations where the inherent properties of the mate-
rial have been changed in the sense of down cycling. To apply the dataset, connect the EoL scrap flow (after
collection and separation, but before remelting) to the input of this process flow of the type [Waste for recov-
ery]. Then connect the primary metal dataset to be substituted, to the negative input flow of the type [Metals].
The negative input applies the appropriate credit for the scrap class stated in the process name (e.g., alumi-
num auto fragments, baled used beverage can, etc.). The parameter for the price ratio represents the ratio
between the scrap class and the LME primary metal price, which may be changed by the user, if necessary,
using the referenced sources.
Recycling
Two general different recycling cases can be found in LCA discussion: Closed loop recycling and open loop
recycling.
Closed loop recycling involves the recycling, recovery or reuse of material in a quasi-identical second use,
including the respective demand to do so.
Open loop recycling corresponds to the conversion of material from one or more products into a new product
or other application, involving a change in the inherent properties of the material itself (often with quality deg-
radation).
Recycling can be understood as allocation between different life cycles. Time must be taken into account for
durable products and the current situation of production must be separated from that of future recycling op-
tions and possibilities. For production, the current market situation must be assessed (ratio of primary material
to recycled material in current production). In parallel, the recycling potential reflects the gross “value” of the
product that principally exists in EOL. The net recycling potential reflects the current secondary material use in
the market situation (deducted from the theoretical “value”).
In the GaBi databases current secondary material use and recycling rates are modelled according to the indi-
vidual commodity or material and the respective market situation. Please see the specific data and chapters
below for details. GaBi focuses on consistency of recycling and end-of-life processes like incineration, landfill
and wastewater treatment with all other life-cycle stages. Three generic models were therefore generated:
3. Waste incineration model
4. Landfill model
5. Wastewater treatment model
These models follow the general rules of the modelling principles. All models represent standard technologies
and are based on parameterised unit processes. For the generation of datasets (e.g. DE: Landfill for inert

98
 System Modelling Features

matter), the models are specified according to the conditions as outlined in the dataset documentation. In-
cluded are country or region-specific background datasets, country or region-specific process efficiencies and
specific input information about the characteristics of waste and wastewater.
Incineration model
The incineration model is defined based on the treatment of average municipal solid waste (MSW). The ther-
mal treatment of a single waste fraction like paper or plastic or even specific wastes like Polyamide 6 is not
actually done in a waste-to-energy (WtE) plant. The model and settings for the average MSW allow the envi-
ronmental burden (emissions and also resource consumption of auxiliaries), energy production, as well as the
credits (metal scrap recovery) to be attributed to a single fraction or specific incinerated waste within a stand-
ard MSW. The following figure gives an overview of the first level of the GaBi incineration model.

Figure 4-12: Exemplary incineration model with in GaBi (here average European domestic
waste treatment with dry offgas cleaning)

The output of energy products (electricity and steam) leaving the product system is dependent on the heating
value of the specific input and the internal consumption of energy necessary to treat the specific waste. The
internal energy consumption is calculated based on the elementary composition of the specific input (e.g.
energy demand for flue gas treatment) and standard values (e.g. handling of waste before incineration). The
gross energy efficiency and the share of produced electricity and steam is taken from the country-/region-
specific average WtE plant for municipal solid waste (MSW) in Germany or Europe.

99
 System Modelling Features

Opening up the core plan “incineration/SNCR/Boiler/Off-gas treatment” of the previous figure will show further
detail of the GaBi incineration model.

Figure 4-13: Details of incineration and dry offgas cleaning in GaBi incineration model

The incineration model was set-up to account for two technologies (wet and dry off-gas treatment) and verified
with measured data from a number of German and European incinerators, as well as data from literature. The
heating value of the input can be specified or calculated based on the elementary composition of the input.
The material flow in the plant is calculated using individual transfer coefficients for every element and stage of
the incinerator. The transfer coefficients for the final release of the flue gas to the atmosphere is verified and
adapted with literature data and real plant data of European and WtE plants.
For input specification in the model, the following elements and compounds can be used: Ag, Al, AlOx, As,
ash, Ba, Br, C_Carbonate (inorganic carbon), C_HC (fossil carbon), C_HB_Bio (biogenic carbon) Ca, Cd, Cl,
CN, Co, Cr, Cu, F, Fe, H, H2O, Hg, J, K, Mg, Mn, N, Na, NH4, Ni, O, P, Pb, S, Sb, SiO2, Sn, SO4, Ti, Tl, V,
Zn.
The modelled emissions to air in the flue gas of the incinerator are: As, Ba, Cd, Co, CO, CO 2 (fossil and bio-
genic), Cr, Cu, dioxins, HBr, HCl, HF, HJ, Hg, Mn, N2O, NH3, Ni, NMVOC, NOx, particles, Pb, Sb, Sn, SO2, Tl,
V, Zn. Most of the emissions leaving the system are input-dependent. That means there is a stoichiometrical
correlation between input and output. Other emissions are a function of the technology utilised and therefore
independent of the specific input. The input-dependent emissions are linear to the elementary composition of
the waste. The technology dependent emissions are constant in a specific range. Input-dependent parameters

100
 System Modelling Features

are the emissions CO2, HCl, HF, SO2 caused by the relevant input of these elements. The amounts of slag,
boiler and filter ash produced, as well as recovered ferrous metal scrap, are also input-dependent. Technolo-
gy dependent parameters are CO, VOC and dioxin emissions.
Ashes and filter residues which are dumped in specific hazardous waste underground dumps – as in the 2011
version – but are accounted for as “hazardous waste (deposited)” are to acknowledge EPD best practise.
The datasets already include the credits given for the recovery of ferrous metal scrap.
Landfill model
The elementary and system flows to and from the landfill site are allocated to the elementary content in the
waste input. The amount of generated landfill gas is calculated based on the organic carbon content in the
waste input and represents an average landfill gas composition.
The input of auxiliaries for the landfilling of one kilogram of waste is partially constant for all types of wastes
(e.g. energy for compacting, materials for the landfill construction) and partially dependent on the elementary
composition of the waste (e.g. ferric chloride for the treatment of leachate). The inert landfill sites do not gen-
erate landfill gas, nor is the leakage technically treated before going to the receiving water.
Landfill gas losses/flare and recovery ratios were checked and adapted to reflect the latest information.

Figure 4-14: Exemplary landfill model in GaBi (here commercial waste composition for certain
geographic example regions)

The landfill model is parameterised to allow the generation of different datasets according to the waste input
and region/country specific details. Important parameters and parameter sets:
 elementary composition of the disposed waste
 different technologies for the sealing and cap (layers)
 differing surrounding conditions (e.g. precipitation)

101
 System Modelling Features

 rates and treatment routes of collected landfill gas and CHP efficiencies and rates (combined
heat and power production)
 rates of leakage collection and treatment efficiencies (COD and AOX)
 transfer coefficients to describe the fate of elements over a period of 100 years
The waste input can be specified by its elementary composition (27 elements) and additional waste-specific
information (e.g. inert substances content, non-degradable carbon and nitrogen content).
The model of the landfill body calculates, based on the element specific transfer coefficients, the input de-
pendent amount of substances and elements going to leakage collection, landfill gas and soil.
The amount and types of materials for the cap and sealing of the landfill site are adapted to specific situations
(background processes, thickness of layers rates of leakage collection), where relevant and applicable.
The collected leakage is either going to a technical treatment (to minimise the organic compounds in the
wastewater) or directly to the receiving water (landfill site for inert waste). In case of technical treatment of the
leakage, the generated sludge is dried and disposed of in an underground deposit.
Part of the landfill gas is collected and either flared or used to produce electricity or both electricity and heat.
The uncollected landfill gas is directly released to the atmosphere. The share of the different treatment route
of landfill gas can be adjusted to the country or region-specific situation. For simplification reasons, the landfill
gas composition only represents the average useable landfill gas. The amount depends on the organic carbon
content in the waste composition and the assumed degradation over 100 years.
Wastewater treatment model
The elementary and system flows to and from the wastewater treatment plant are allocated to the elementary
content in the wastewater input.
The wastewater treatment represents an average/typical wastewater treatment from industrial processes. It
contains mechanical, biological and chemical treatment steps for the wastewater (including precipitation and
neutralisation), and treatment steps for the sludge (thickening, dewatering). The outflow goes directly to the
receiving water (natural surface water).

102
 System Modelling Features

Figure 4-15: Exemplary wastewater treatment model in GaBi (here municipal wastewater for
German circumstances)

The process steps take average elimination and transfer coefficients into account. The sewage passes
through the bar screens for rag removal. In this section automatic bar screen cleaners remove large solids
(rags, plastics) from the raw sewage. Next, the sewage is transported to the grit tanks. These tanks reduce
the velocity of the sewage so heavy particles can settle to the bottom. In the separator suspended particles
such as oils, fats are removed. The settlement tank can remove the larger suspended solids. FeSO4, and
Ca(OH)2 are used as precipitant agents in the mixing tank to remove metals. Ca(OH)2 and H2SO4 regulate the
pH value. The primary clarifiers remove the suspended solids from the mixing tank prior to discharge to the
aeration tanks. The aeration tanks provide a location where biological treatment of the sewage takes place.
The activated sludge converts organic substances into oxidised products, which are settled out in the second-
ary clarifiers. Phosphoric acid is used as nutrient for micro-organisms. The cleared overflow in the secondary
clarifiers goes to a natural surface water body (stream, river or bay). The settled solids, from the settlement
tank, the primary clarifiers and secondary clarifiers, are pumped to the primary thickener where the solids are
thickened (water content of the thickened sludge is 96%). The sludge is pumped to filter presses for dewater-
ing, which use chemical flocculants to separate the water from the solids (water content of the dewatered
sludge is 65%). In this dataset sludge for agricultural application is produced. For this reason the sludge is not
dried and supplied after dewatering. The output is wet sludge (dry content is 35%) containing N, P2O5 and
K2O according to statistics and calculations which is included in the plan for the given fertilizer credit.

103
 Review, documentation and validation

5 Review, documentation and validation

Data that is officially published in publications or a web page is not sufficient proof of its quality. Even if pro-
fessional review processes are in place for journal publications, the scientific quality of the article or paper can
be proven, and the “correctness” of the underlying data cannot be validated in most cases. Even if it is easier
for the user to simply “cite” a data source, a validation or verification routine for the data is essential.
There is presently no specific ISO standard in existence for data quality reviews. The existing ISO standards
ensure quality and consistency of LCA reporting.

5.1 Review procedures and check routines

The core principle of PE is to provide quality information. PE has therefore set up a review and validation
procedure within its GaBi Database concept and management scheme based on the four quality check layers:
 Internal entry quality checks
 Internal resulting quality checks
 External resulting non-public quality checks
 External resulting public quality checks
See Chapter 2.1 for more details. It is important to base the review of data and databases on ISO principles
accompanied by practical experiences in data collection, data set-up, database maintenance and updates in
industrial practises. Plausibility and technical routines in GaBi raw data20 and process data handling are the
main instruments to avoid, detect and reduce errors.
These routines support data collection and systematic error identification in inventories by understanding the
underlying technical process and being able to identify potentially incorrect or missing values and flows (con-
spicuous values, type faults, conversion/unit errors).

5.1.1 Technical information and documentation routines in GaBi

The checklist for the collected data and resulting unit process information, which is documented either on plan
system level, in the unit process or in the resulting aggregated process:
 Data source (reproducibility), reliability of the sources, representativeness of the sources
 Technical conditions (state of the art, conventional process, established process, pilot plant, la-
boratory operation)
 Process integration: Stand-alone process or integrated into a large facility
 Calculation method (average, specific)
 Technically relevant process steps are represented on plan system level
 Types and quantity reactant/product
 Efficiency/stoichiometry of chemical reactions; monitoring of the rate of yield
 Types and quantity of by-products, wastes or remaining and its fate

20
Raw data is any data or metadata needed so set up an LCA dataset

104
 Review, documentation and validation

 Emissions spectrum (relation between in- and outputs, comparison to similar processes)
 Types and quantity of circulating flows (purge, monomers, production recycling material)
 Auxiliary material and utilities
 Input chemicals and substances for end of pipe measures (lime, NH3)
These technical information points help to identify gaps and enable balance checks and plausibility checks.

5.1.2 Important material and energy balances

The following balance checks are done with any unit process and plan system, to trace and eliminate gaps
and errors.
 Energy balance: net or gross calorific value (sum of renewable and non-renewable)
 Mass balance (what goes in must come out)
 Element balance: often C or metal content (also check for raw material recovery)
 Reaction equations

5.1.3 Plausibility of emission profiles and avoiding errors

The basic principle is to avoid too high and too low values and/or missing emissions. The plausibility and error
checking must therefore not only take place on the process level but also on the plan and supply chain level.
There are typical emissions for typical industrial operations for each type of process. These indications are
used to monitor and compare similar processes. Knowing the frequent error sources is the best way to man-
age and avoid them.
Data entry with the wrong comma/point setting (factor 10, 100, 1000) results in figures that are too high or too
low. New or updated data in GaBi is double checked, individually by the data developer with existing or com-
parable datasets, and in the case of bigger data volumes, automatically (“GaBi process comparison tool”) by
routine checks of the relevant impacts with the predecessor.
Another error source is data entry with wrong units:
 mg – µg or kg – t leads towards factor 1000 / 0.001 error
 MJ – kWh leads towards factor 3.6 / 0.28 error
 BTU – kWh leads towards factor 1000 / 0.001 error
 BTU – MJ leads towards factor 3000 / 0.0003 error
GaBi supports the avoidance of this error by offering automatic unit conversion.
If the emissions or impacts appear to be surprisingly low, the following checks are undertaken in GaBi data-
base work:
 connection of significant processes back to the resource (aggregated dataset or plan system of
upstream processes)
 modelling of fuels only, omitting combustion emissions in the unit process (thermal energy or
emission modelling)

105
 Review, documentation and validation

 transports are modelled but not adjusted to the correct distances

 unsuitable substitution used
 wastewater impacts not modelled (wastewater leaves untreated)
 burden free entry of secondary materials into the life cycle phase
 CO2 balance not addressed (renewable), CO2 intake or emission not/wrongly considered
If the emissions or impacts appear to be surprisingly high, the following checks are undertaken in GaBi data-
base work:
 by-products not substituted or allocated
 system expansion not suitable (loss of focus or function added in unsuitable way)
 useful energy output (e.g. steam) not considered correctly
 waste treatment or wastewater treatment overestimated, scrap input modelled as pure primary
route (sector-specific)
 CO2 balance not addressed (renewable), CO2 intake or emission not/wrongly considered
Plausibility and error checks are critically discussed and optimised in data-related projects with industrial cus-
tomers and respective critical reviewers of our work, with our academic cooperation partners, LBP- University
of Stuttgart and Fraunhofer IBP, as well as with independent testing and certification partners.

5.2 Documentation
Documentation is essential in order to assure reproducibility and transparency of the datasets, as well as to
clarify the scope of the datasets and the possible applications.
In GaBi documentation, recommendations to mandatory and optional information, which are either based on
international standards such as ISO 14040, ISO 14044 and ELCD or on the experience of PE INTERNA-
TIONAL and LBP- University of Stuttgart. The requirements of ISO 14040 [ISO 14040 : 2006] and 14044 [ISO
14044 : 2006] are considered.
The metadata documentation of the datasets in "GABI 2012 database [GABI 2012]” is based on the documen-
tation recommendations of the "International Reference Life Cycle Data System" [ILCD 2010] Handbook of the
European Commission's European Platform on Life Cycle Assessment, while – due to the dynamic nature of
the topic database harmonisation [see UNEP/SETAC 2011] – not strictly meeting them in each and every
case.
Please see the individual GaBi documentation [GABI 2012] in the respective LCI processes of the GaBi data-
base (example of documentation is shown in Chapter 5.2.3) or on the GaBi Webpage [Link]
[Link].

5.2.1 Nomenclature
Consistent nomenclature is an essential aspect of the database quality. Any database object including impact
characterisation factors or flow characteristics like calorific values, flows, processes and plan systems must be
properly named.

106
 Review, documentation and validation

Flow and process names are especially important. The flows and processes in GaBi are arranged in a hierar-
chy for storage.
The flow hierarchy is structured according to technical aspects (for non-elementary flows and resources) and
according to emission compartments air, water and soil.
In general all relevant LCI elementary flows (resources and emissions) in GaBi are pre-defined. Therefore the
number of elementary flows that must be newly-defined by the user is few to none.
If a new process or new flow is created because it is not available in the database, consistency with existing
processes or flows is kept.
In the GaBi database flows and processes are biunique, which is an important basis of consistency and a
prerequisite for data exchange.

5.2.2 Documentation of Flows

The documentation of flows is an important component of the inherent documentation of processes and LCI
results. Flow documentation is an integral part due to the direct influence of the flow properties to the results
of LCI and LCIA.
Flows in GaBi are (if suitable) documented by:
 Reference quantity
 Synonyms of the main flow name
 CAS number
 Sum formula
 Region or location of the flow, e.g. region Western Europe
 Field for general comments to add further information
Information for the flow such as synonyms and CAS number are documented in GaBi according to ILCD (see
Figure 4-11).

5.2.3 Documentation of LCI process data

The documentation of the LCI datasets in GaBi covers relevant technical and supply chain information that is
necessary to understand the technological basis and background of the modelled system. Further multiple
metadata are given to enable the further use within important documentation schemes like ILCD, EPDs and
EcoSpold. For further details see the documentation tab in each dataset.

107
 Review, documentation and validation

Figure 5-1: Example documentation in GaBi (excerpt) [GABI 2012]

108
 Review, documentation and validation

5.3 Validation
The validation procedures of GaBi databases are implemented on different levels.
1. Consistency and Completeness of database objects
Consistency of flows and completeness of the necessary flow characteristics are validated internally at PE,
following standard routine. PE provides several different databases consistent to our own databases. Rou-
tines and technical tools exist therefore to trace and identify possible errors and ensure consistency, com-
pleteness and biunique database entries.
2. Content on technical process level
The technical content is constantly validated in LCA work with GaBi data by related industry experts, branch
experts or process operators. Validating technical content of datasets needs technical understanding. If com-
panies provide data, PE validates the data (because it must fit in detail and consistency to the surrounding
system) and, depending on the type and purpose of the data, LBP University of Stuttgart or a third-party vali-
dator or reviewer is involved.
3. Methodological LCI approach
Methodological LCI approaches in GaBi databases are based on relevant standards and reference works, and
are presented and discussed in and benchmarked against different academic, political and professional
frameworks (like e.g. ILCD 2010, NETZWERK2011, PLASTICSEU 2011, UNEP/SETAC 2011, ISO 21930:2007)
to ensure acceptance and applicability. A validation of methodological approaches is constantly conducted in
the context of the use of GaBi data and process chain details within the given framework and the respective
critical reviews of studies which utilise the databases.
4. Methodological approach LCIA
New impact methods in GaBi are implemented preferably by involving the respective LCIA method develop-
ers, to implement the given method in the most suitable way. This implementation includes proactive critical
discourse between scientific detail and practical applicability. The validation of the method is preferably con-
ducted jointly by the developers and PE.
5. Content on LCI and LCIA level
In many LCA projects reviews are undertaken and the background data (chains) are reviewed and discussed
with the project group and with the reviewer. We grant reviewers access to the background systems under
bilateral agreements. PE INTERNATIONAL studies, GaBi results and dataset benchmarks are often publicly
discussed in external field tests or in comparisons. A broad user community is constantly using, comparing,
benchmarking, screening and reviewing GaBi data and data results, which are published in various channels.
User feedback is collected and incorporated into the database management routine.

109
 Literature

6 Literature

BAITZ 2002 Baitz, M.: DIE BEDEUTUNG DER FUNKTIONSBASIERTEN CHARAKTERISIERUNG VON FLÄCHEN-
INANSPRUCHNAHMEN IN INDUSTRIELLEN PROZESSKETTENANALYSEN. EIN BEITRAG ZUR GANZ-
HEITLICHEN BILANZIERUNG. DISSERTATION. SHAKER (BERICHTE AUS DER UMWELTTECHNIK),
AACHEN, 2002

BECK, BOS, WITT- Beck, T.; Bos, U.; Wittstock, B.: LANCA – CALCULATION OF LAND USE INDICATOR VALUES
STOCK ET AL. 2010 IN LIFE CYCLE ASSESSMENT; 2010

ONLINE: [Link].

BRENTRUP, F. ET. AL. Methods to estimate on-field nitrogen emissions from crop production as an input to
2000 LCA studies in the Agricultural Sector. The International Journal of Life Cycle As-
sessment 5(6) 349-357.

BORKEN ET AL 1999 BORKEN, J.; PATYK, A.; REINHARDT, G. A: BASISDATEN FÜR ÖKOLOGISCHE BILANZIERUNGEN –
EINSATZ VON NUTZFAHRZEUGEN IN TRANSPORT, LANDWIRTSCHAFT UND BERGBAU, BRAUN-
SCHWEIG : VIEWEG, 1999

CML 2001 CML: CML'S IMPACT ASSESSMENT METHODS AND CHARACTERISATION FACTORS. LEIDEN UNI-
VERSITY, INSTITUTE OF ENVIRONMENTAL SCIENCE (CML), 2001

ECO-INDICATOR 95 : ECO-INDICATOR 95: A DAMAGE ORIENTED METHOD FOR LIFE CYCLE IMPACT ASSESSMENT
2000 METHODOLOGY REPORT NR. 199/36 A, PRÉ CONSULTANTS B.V., AMERSFOORT, THE NETHER-
LANDS, 2000

ECO-INDICATOR 99 : ECO-INDICATOR 99: A DAMAGE ORIENTED METHOD FOR LIFE CYCLE IMPACT ASSESSMENT
2000 METHODOLOGY REPORT NR. 199/36 A, PRÉ CONSULTANTS B.V., AMERSFOORT, THE NETHER-
LANDS, 2000

ECOTRANSIT2010 IFEU HEIDELBERG, ÖKO-INSTITUT, IVE / RMCON, ECOTRANSIT WORLD, ECOLOGICAL

TRANSPORT INFORMATION TOOL FOR WORLDWIDE TRANSPORTS: METHODOLOGY AND DATA,
BERLIN, HANNOVER, HEIDELBERG; GERMANY, 2010

EMEP/CORINAIR JOINT EMEP/CORINAIR ATMOSPHERIC EMISSION INVENTORY GUIDEBOOK, EUROPEAN ENVI-

2002 RONMENT AGENCY, THIRD EDITION, COPENHAGEN; DENMARK, 2002.

EMEP/CORINAIR EMEP/CORINAIR EMISSION INVENTORY GUIDEBOOK – 2006, EUROPEAN ENERGY AGENCY,

2006 2006.

EMEP/CORINAIR EMEP/CORINAIR: EMISSION INVENTORY GUIDEBOOK: ROAD TRANSPORT, EUROPEAN ENVI-

2006B RONMENT AGENCY, COPENHAGEN, 2006

EPS 1999 EPS: A SYSTEMATIC APPROACH TO ENVIRONMENTAL STRATEGIES IN PRODUCT DEVELOPMENT

(EPS). VERSION 2000 - GENERAL SYSTEM CHARACTERISTICS. CPM REPORT,1999

EUROSTAT 2012 HTTP://[Link]/PORTAL/PAGE/PORTAL/EUROSTAT/HOME/

110
 Literature

FINAT 2000 FINAT, A.G.: 2ND ETAP FORUM: MARKETS FOR SUSTAINABLE CONSTRUCTION. OPENING
SPEECH. A. GONZÁLES FINAT, DIRECTOR, NEW AND RENEWABLE SOURCES OF ENERGY, EN-
ERGY EFFICIENCY & INNOVATION, DG TRANSPORT & ENERGY, EUROPEAN COMMISSION.
BRUSSELS: JUNE 11TH, 2000

GABI 2012 GABI, PE INTERNATIONAL AG; LBP-GABI, UNIVERSITY OF STUTTGART: GABI SOFTWARE
SYSTEM, LEINFELDEN-ECHTERDINGEN / GERMANY, 2012

GHGPC 2011 GHG PROTOCOL CORPORATE VALUE CHAIN ACCOUNTING AND REPORTING STANDARD,
WRI/WBCSD, 2011

GHGPP 2011 PRODUCT LIFE CYCLE ACCOUNTING AND REPORTING STANDARD, WRI/WBCSD, 2011

GUINÉE ET AL. 1996 LCA IMPACT ASSESSMENT OF TOXIC RELEASES; GENERIC MODELLING OF FATE, EXPOSURE AND
EFFECT FOR ECOSYSTEMS AND HUMAN BEINGS. (NO. 1996/21) CENTRE OF ENVIRONMENTAL
SCIENCE (CML) LEIDEN AND NATIONAL INSTITUTE OF PUBLIC HEALTH AND ENVIRONMENTAL
PROTECTION (RIVM), BILTHOVEN, MAY 1996.

GUINÈE ET AL. 2001 GUINÉE, J. ET. AL. HANDBOOK ON LIFE CYCLE ASSESSMENT - OPERATIONAL GUIDE TO THE
ISO STANDARDS. CENTRE OF ENVIRONMENTAL SCIENCE, LEIDEN UNIVERSITY (CML); THE
NETHERLANDS, 2001.

GUINÉE ET AL. 2002 HANDBOOK ON LIFE CYCLE ASSESSMENT: AN OPERATIONAL GUIDE TO THE ISO STANDARDS;
DORDRECHT: KLUVVER ACADEMIC PUBLSIHERS, 2002.

HAUSCHILD 2003 HAUSCHILD, M. Z.: SPATIAL DIFFERENTIATION IN LIFE CYCLE IMPACT ASSESSMENT - THE EDIP-
2003 METHODOLOGY. GUIDELINES FROM THE DANISH EPA PARTICIPANT(S), TECHNICAL UNI-
VERSITY OF DENMARK, DEPARTMENT OF MANUFACTURING ENGINEERING AND MANAGEMENT,
DENMARK, 2003

HBEFA 2010 HANDBOOK EMISSION FACTORS FOR ROAD TRANSPORT (HBEFA), VERSION 3.1, UMWELT-
BUNDESAMT BERLIN; BUWAL / OFEFP BERN; UMWELTBUNDESAMT WIEN,
HTTP://[Link], BERLIN, BERN, VIENNA / GERMANY, SWITZERLAND, AUSTRIA, 2010

IBU 2011 INSTITUT BAUEN UND UMWELT E.V. (IBU)

IFEU 2010A IFEU HEIDELBERG, ENERGIEVERBRAUCH UND SCHADSTOFFEMISSIONEN DES MOTORISIERTEN

VERKEHRS IN DEUTSCHLAND 1960-2030 (TREMOD, V5), HEIDELBERG; GERMANY, 2010

IMO 2009 INTERNATIONAL MARITIME ORGANIZATION: SECOND IMO GHG STUDY 2009, FINAL REPORT,
APRIL 2009

IKP 2003 INSTITUT FÜR KUNSTSTOFFPRÜFUNG UND KUNSTSTOFFKUNDE DER UNIVERSITÄT STUTTGART,
ABTEILUNG GANZHEITLICHE BILANZIERUNG, 2003

ILCD 2010 ILCD HANDBOOK, GENERAL GUIDE FOR LIFE CYCLE ASSESSMENT - DE-TAILED GUIDANCE,
EUROPEAN UNION, 2010

IMPACT 2002 IMPACT 2002+ : ÉCOLE POLYTECHNIQUE FÉDÉRALE DE LAUSANNE, LAUSANNE, 2002

111
 Literature

ILCD 2011 EUROPEAN COMMISSION-JOINT RESEARCH CENTRE - INSTITUTE FOR ENVIRONMENT AND
SUSTAINABILITY: INTERNATIONAL REFERENCE LIFE CYCLE DATA SYSTEM (ILCD) HANDBOOK-
RECOMMENDATIONS FOR LIFE CYCLE IMPACT ASSESSMENT IN THE EUROPEAN CONTEXT. FIRST
EDITION NOVEMBER 2011. EUR 24571 EN. LUXEMBURG. PUBLICATIONS OFFICE OF THE EU-
ROPEAN UNION; 2011

IPCC 2006 INTERGOVERNMENTAL PANEL ON CLIMATE CHANGE (IPCC) REVISED GUIDELINES FOR NA-
TIONAL GREENHOUSE GAS INVENTORIES FROM 2006

ISRIC WISE 2002 Batjes, N. H. : ISRIC-WISE -- GLOBAL SOIL PROFILE DATA (VER. 1.1). WAGENINGEN., 2002
ONLINE:HTTP://[Link]/ISRIC/.

ISO 14040 : 2006 ISO 14040: ENVIRONMENTAL MANAGEMENT – LIFE CYCLE ASSESSMENT – PRINCIPLES AND
GUIDELINES, 2006.

ISO 14021:1999 ENVIRONMENTAL LABELS AND DECLARATIONS -- SELF-DECLARED ENVIRONMENTAL CLAIMS

(TYPE II ENVIRONMENTAL LABELLING)

ISO 14025:2006 ENVIRONMENTAL LABELS AND DECLARATIONS -- TYPE III ENVIRONMENTAL DECLARATIONS --
PRINCIPLES AND PROCEDURES

ISO 14044 : 2006 ISO 14044, ENVIRONMENTAL MANAGEMENT – LIFE CYCLE ASSESSMENT – REQUIREMENTS AND
GUIDELINES, 2006.

ISO 14020 : 2000 ENVIRONMENTAL LABELS AND DECLARATIONS -- GENERAL PRINCIPLES

ISO 14064-1:2006 GREENHOUSE GASES -- PART 1: SPECIFICATION WITH GUIDANCE AT THE ORGANIZATION LEVEL
FOR QUANTIFICATION AND REPORTING OF GREENHOUSE GAS EMISSIONS AND REMOVALS

ISO 14067 ISO 14067, CARBONFOOTPRINTOFPRODUCTS

ISO 15804 SUSTAINABILITY OF CONSTRUCTION WORKS - ENVIRONMENTAL PRODUCT DECLARATIONS -

CORE RULES FOR THE PRODUCT CATEGORY OF CONSTRUCTION PRODUCTS; GERMAN VERSION
FPREN 15804:2011

ISO 21930:2007 SUSTAINABILITY IN BUILDING CONSTRUCTION -- ENVIRONMENTAL DECLARATION OF BUILDING

PRODUCTS

KREISSIG & KÜMMEL KREIßIG, J. UND J. KÜMMEL: BAUSTOFF-ÖKOBILANZEN. WIRKUNGSABSCHÄTZUNG UND AUS-
1999 WERTUNG IN DER STEINE-ERDEN-INDUSTRIE. HRSG. BUNDESVERBAND BAUSTOFFE STEINE +
ERDEN E.V., 1999

KUPFER 2005 PROGNOSE VON UMWELTAUSWIRKUNGEN BEI DER ENTWICKLUNG CHEMISCHER ANLAGEN,
DISSERTATION, UNIVERSITY OF STUTTGART, 2005

MILA` I CANALS 2007 METHOD FOR ASSESSING IMPACTS ON LIFE SUPPORT FUNCTIONS (LSF) RELATED TO THE USE
OF ‘FERTILE LAND’ IN LIFE CYCLE ASSESSMENT (LCA). JOURNAL OF CLEANER PRODUCTION 15
(2007) 1426-1440

MITCHELL 2003 Mitchell, T.D. : A COMPREHENSIVE SET OF CLIMATE SCENARIOS FOR EUROPE AND THE GLOBE.
TYN CY 1.1 DATA-EST. TYNDALL CENTRE FOR CLIMATE CHANGE RESEARCH, 2003
ONLINE: HTTP://[Link]/~TIMM/CTY/OBS/TYN_CY_1_1.HTML.

112
 Literature

NETZWERK2011 NETZWERK LEBENSZYKLUSDATEN WIRD AM FORSCHUNGSZENTRUM KARLSRUHE, 2011

NF P 01 010 : 2004 FRENCH EPD SYSTEM: NORME FRANÇAISE DÉFINISSANT LE CONTENU ET LE MODE DE
RÉALISATION DE LA FICHE DE DÉCLARATION ENVIRONNEMENTALE ET SANITAIRE DANS LE
CADRE DES PRODUITS DE CONSTRUCTION, DECEMBER 2004.

PAS 2050:2011 SPECIFICATION FOR THE ASSESSMENT OF THE LIFE CYCLE GREENHOUSE GAS EMISSIONS
OF GOODS AND SERVICES, BSI 2011

PFISTER2011 PFISTER, S.; BAYER, P.; KOEHLER, A.; HELLWEG. S. (2011): ENVIRONMENTAL SCIENCE &
TECHNOLOGY 2011 45 (13), 5761-5768 ENVIRONMENTAL IMPACTS OF WATER USE IN GLOB-
AL CROP PRODUCTION: HOTSPOTS AND TRADE-OFFS WITH LAND USE.

PFLIEGER/ILG 2007 PFLIEGER, J.; ILG, R.: ANALYSE BESTEHENDER METHODISCHER ANSÄTZE ZUR BERÜCKSICHTI-
GUNG DES RECYCLINGS VON METALLEN IM RAHMEN DER ÖKOBILANZ, NETZWERK LEBENSZYK-
LUSDATEN ARBEITSKREIS METALLISCHE ROHSTOFFE, UNIVERSITÄT STUTT-
GART/FORSCHUNGSZENTRUM KARLSRUHE, 2007

PLASTICSEU 2011 ECO-PROFILES AND ENVIRONMENTAL DECLARATIONS PLASTICSEUROPE, VERSION 2.0 (APRIL
2011)

RABL AND SPADARO RABL, A. AND SPADARO, J.V. (2004). THE RISKPOLL SOFTWARE, VERSION IS 1.051 (DATED
2004 AUGUST 2004). [Link].

RECIPE 2012 RECIPE MID/ENDPOINT METHOD, VERSION 1.07 JULY 2012; ONLINE:
[Link]

RWTH 1990 HRSG.: RHEIN.-WESTF. TECHNISCHEN HOCHSCHULE AACHEN: SPEZIFISCHER ENERGIEEIN-

SATZ IM VERKEHR, ERMITTLUNG UND VERGLEICH DER SPEZIFISCHEN ENERGIEVERBRÄUCHE,
FORSCHUNGSBERICHT FE NR. 90 247/88, AACHEN 1990.

SCHWARZ ET AL 1999 SCHWARZ, DR. WINFRIED; LEISEWITZ, DR. ANDRÉ: EMISSIONEN UND MINDERUNGSPOTENTIAL
VON HFKW, FKW UND SF6 IN DEUTSCHLAND, IM AUFTRAG DES UMWELTBUNDESAMTES,
FORSCHUNGSBERICHT 29841256, FRANKFURT, 1999

SEPPÄLÄ ET AL. 2006 COUNTRY-DEPENDENT CHARACTERISATION FACTORS FOR ACIDIFICATION AND TERRESTRIAL
EUTROPHICATION BASED ON ACCUMULATED EXCEEDANCE AS AN IMPACT CATEGORY INDICA-
TOR. INT J LCA 11 (6) 403 – 416 (2006)

TRACI 2012 TRACI, TOOL FOR REDUCTION AND ASSESSMENT OF CHEMICAL AND OTHER ENVIRONMENTAL
IMPACTS 2.1 (TRACI), U.S. ENVIRONMENTAL PROTECTION AGENCY, 2012; ONLINE:
HTTP://[Link]/NRMRL/STD/SAB/TRACI/

TRACI 1996 TRACI: THE TOOL FOR THE REDUCTION AND ASSESSMENT OF CHEMICALS AND OTHER ENVI-
RONMENTAL IMPACTS, US ENVIRONMENTAL PROTECTION AGENCY, U.S.A., 1996

UBA BERLIN 2004 UBA BERLIN; BUWAL / OFEFP BERN; UMWELTBUNDESAMT WIEN: HANDBUCH EMISSIONS-
FAKTOREN DES STRAßENVERKEHRS, VERSION 2.1, BERLIN, 2004

UBP 2006 THE ECOLOGICAL SCARCITY METHOD ECO-FACTORS 2006. FEDERAL OFFICE FOR THE ENVI-
RONMENT FOEN. BERN, 2009

UNEP/SETAC 2011 UNEP/SETAC: GLOBAL GUIDANCE PRINCIPLES FOR LIFE CYCLE ASSESSMENT DATABASES – A
BASIS FOR GREENER PROCESSES AND PRODUCTS, 2011

113
 Literature

USETOX 2010 ROSENBAUM, RK ET AL: USETOX - THE UNEP/SETAC-CONSENSUS MODEL: RECOMMENDED CHAR-
ACTERISATION FACTORS FOR HUMAN TOXICITY AND FRESHWATER ECOTOXICITY IN LIFE CYCLE IMPACT
ASSESSMENT. INTERNATIONAL JOURNAL OF LIFE CYCLE ASSESSMENT 13(7): 532-546., 2008;
USETOXTM MODEL 1.01 AND ORGANIC DATABASE 1.01, 2010

VBD 2003 PERSONAL INFORMATION: VERSUCHSANSTALT FÜR BINNENSCHIFFBAU E.V. (VBD), 2003 (RENAMED
2004 IN ENTWICKLUNGSZENTRUM FÜR SCHIFFSTECHNIK UND TRANSPORTSYSTEME E.V. (DST))

WORLD BANK 2012 HTTP://[Link]/DATA/[Link]

114
 Supplement A

Supplement A Description of result and impact categories

This chapter describes the impact assessment methodologies available in GaBi 6 (called quantities in the
GaBi tool). The description is divided into overall impact categories (e.g. global warming, acidification.) and
the approach of each of the available impact methodologies (e.g. CML, Ecoindicator) is described.
Methodologies covering only specific impact categories, e.g. USEtox for toxicity and IPCC for global warming,
are described under each impact category.
The International Reference Life Cycle Data System (ILCD) has published ‘Recommendations for Life Cycle
Impact Assessment in the European context’ which chooses the methodology which has been evaluated as
the best within the impact category [ILCD 2011]. This leads to the set of impact categories in Table J. The
approach of each methodology is described in the appropriate chapter.

115
 Supplement A

Table J: ILCD set of recommended impacts

Supplement A 1 Primary energy consumption

Primary energy demand (PED) is often difficult to determine due to the various types of energy sources. Pri-
mary energy demand is the quantity of energy directly withdrawn from the hydrosphere, atmosphere or geo-
sphere or energy source without any anthropogenic changes. For fossil fuels and uranium, PED would be the
amount of resources withdrawn expressed in their energy equivalents (i.e. the energy content of the raw ma-
terial). For renewable resources, the energy characterised by the amount of biomass consumed would be
described. PED for hydropower would be based on the amount of energy that is gained from the change in
the potential energy of the water (i.e. from the height difference). The following primary energies are designat-
ed as aggregated values:

116
 Supplement A

The total “Primary energy consumption non-renewable,” given in MJ, essentially characterises the gain
from the energy sources: natural gas, crude oil, lignite, coal and uranium. Natural gas and crude oil will be
used both for energy production and as material constituents, such as in plastics. Coal will primarily be used
for energy production. Uranium will only be used for electricity production in nuclear power stations.
The total “Primary energy consumption renewable,” given in MJ, is generally accounted for separately and
comprises hydropower, wind power, solar energy and biomass.
It is important that end use energy (e.g. 1 kWh of electricity) and primary energy are not confused with each
other; otherwise, the efficiency loss in production and supply of the end energy will not be accounted for.
The energy content of the manufactured products will be considered to be feedstock energy content. It will be
characterised by the net calorific value of the product. It represents the still-usable energy content that results,
such as incineration with energy recovery.

Supplement A 2 Waste categories

In GaBi background databases waste is further treated for known waste pathways towards final emissions in
incinerators or landfill bodies, if suitable indications exist (e.g. according to waste directives).
If specific wastes are deposited without further treatment, they are indicated with the addition “deposited.”
If waste treatment routes are unknown, unspecific or not definable, GaBi documents the related specific waste
flow and the specific waste amount with a waste star “*” meaning it can be further treated if the user knows
the specific waste treatment pathway. Categories such as stockpile goods, consumer waste, hazardous waste
and radioactive waste, group those specific waste flows together.

Supplement A 3 Global Warming Potential (GWP)

The mechanism of the greenhouse effect can be observed on a small scale, as the name suggests, in a
greenhouse. These effects also occur on a global scale. The occurring short-wave radiation from the sun
comes into contact with the earth’s surface and is partially absorbed (leading to direct warming) and partially
reflected as infrared radiation. The reflected part is absorbed by greenhouse gases in the troposphere and is
re-radiated in all directions, including back to earth. This results in a warming effect at the earth’s surface.
In addition to the natural mechanism, the greenhouse effect is enhanced by human activities. Greenhouse
gases, believed to be anthropogenically caused or increased, include carbon dioxide, methane and CFCs.
Figure A-1 shows the main processes of the anthropogenic greenhouse effect. An analysis of the greenhouse
effect should consider the possible long term global effects.

117
 Supplement A

The global warming potential is calculated in

carbon dioxide equivalents (CO2-Eq.), meaning Absorption
Reflection
that the greenhouse potential of an emission is UV - radiation

given in relation to CO2. Since the residence Infrared

T

ra
ce
radiation
time of gases in the atmosphere is incorpo-

ga
se
CFCs

si
nt
CO2
rated into the calculation, a time range for the

he
CH4

atmo
assessment must also be specified. A usual

sphe
re
period is 100 years.

Figure A-1: Greenhouse effect

(KREISSIG & KÜMMEL 1999)

IPCC
All LCIA methodologies have GWP factors which have been determined from the International Panel on Cli-
mate Change (IPCC) as the basis of the GWP factors. However, because update schedules are different, two
specific IPCC lists of GWP factors are available in GaBi, as updated in the summer 2012. One includes bio-
genic carbon and one excludes it – see further description of this under ReCiPe 1.07.

CML
CML uses the indices published by the IPCC. Because of the uncertainties in net GWPs for ozone-depleting
gases, these indices have not been included in the baseline method. If these uncertainties can be narrowed
down in further research, net GWPs should be used for ozone-depleting gases. [CML 2001]
The GWPs for 100 years are recommended as the baseline characterisation method for climate change. The
IPCC also provides GWPs for 20 and 500 years. Although 500 years is closer to eternity, CML does not rec-
ommend using the GWPs for 500 years as the baseline, due to growing uncertainties in GWP with increasing
time span. [CML 2001]
CML includes biogenic carbon at the same level as fossil carbon, hence CO 2 uptake has a GWP of -1 kg CO2
eq., and the subsequent release has the factor of 1 kg CO2 eq.

EDIP 2003
The criteria applied in the EDIP methodology to determine if a substance contributes to global warming follow
the IPCC’s recommendation. At one point the EDIP method goes further than the IPCC’s recommendation by
including contribution from organic compounds and carbon monoxide of petrochemical origin, which is de-
graded to CO2 in the atmosphere. CO2 emissions are evaluated for whether they constitute a net addition of
CO2 to the atmosphere, and not what they derive from fossil carbon sources, but rather from biomass, and
simply represent a manipulation of part of the natural carbon cycle. [HAUSCHILD 2003]

Ecoindicator 99
Ecoindicator 99 works with three damage-oriented categories: Human health, ecosystem quality and re-
sources. These categories are subdivided into mid-point indicators falling under human health impact from
climate change which here is considered equivalent to global warming. [GUINÈE ET AL. 2001]
The health-indicator is expressed as the number of Disability-Adjusted Life Years (DALYs), measuring the
total amount of ill health, due to disability and premature death, attributable to specific diseases and injuries.

118
 Supplement A

The methodology document mentions several possible effects from climate change of which three are includ-
ed in the impact classification:
 Exposure to thermal extremes with the outcome of altered rates of heat- and cold-related illnesses
and death
 Effects on range and activity of vectors and infective parasites with subsequent disease incidences
 Sea-level rise, with population displacement and damage to infrastructure, and with the outcome of
an increased risk of infectious disease and psychological disorders
These effects appear in one calculation factor of a number of DALYs per kg of substance emission.

Impact 2002+
The Impact 2002+ methodology operates with the same three damage-oriented impact categories as Ecoindi-
cator 99: Human health, ecosystem quality and resources. However, from the authors' point of view, the mod-
elling up to the damage of the impact of climate change on ecosystem quality and human health is not accu-
rate enough to derive reliable damage characterisation factors. The interpretation, therefore, directly takes
place at midpoint level, making global warming a stand-alone endpoint category with units of kg of CO2-
equivalents. The assumed time horizon is 500 years to account for both short and long term effects. [IMPACT
2002]

ReCiPe 1.07
The ReCiPe methodology operates with both mid-point and end-point indicators:
 End-point: The same three damage-oriented impact categories as Ecoindicator 99 and Impact
2002+; Human health, ecosystem quality, and resources
 Mid-point: 18 mid-point indicators; one of which is global warming.
ReCiPe 2008 was updated to version 1.07, released in July 2012. The researchers are interested in the mar-
ginal effect of adding a relatively small amount of CO2 or other greenhouse gas, and not the impact of all
emissions. With no models readily available, the IPCC climate change equivalence factors from the 2007
report are used as the midpoint characterisation factors.
Furthermore, data was found in literature linking the mid-point indicator to a temperature increase and after-
wards, to effects on ecosystem quality and human health. [RECIPE 201]
ReCiPe excludes biogenic carbon from the calculations, therefore CO2 uptake has the factor 0, as does the
subsequent release of biogenic carbon.
This necessitates an adjustment of the emission factor for biogenic methane release. The argument is that if
we model carbon dioxide uptake which is later released as methane, then we need to have a 1:1 molar car-
bon balance.
We therefore need:
1 mole CO2 = 44 g : 1 mole CH4 = 16 g
44 g CO2 : 16 g CH4
2.75 g CO2 : 1 g CH4

119
 Supplement A

Consider a plant that sequesters 2.75 kg CO2 and this carbon is eventually entirely re-released as 1 kg me-
thane. If we model this system including the sequestered carbon, then the GWP calculation will be as follows:
- Sequestered CO2 = 2.75 kg => -2.75 kg CO2e
- Emission of CH4 = 1 kg => 25 kg CO2e
- Net emission = 25 - 2.75 => 22.25 kg CO2e
Therefore, if we set the sequestered CO2 to zero, we need to give the biogenic CH4 an emission factor of
22.25 kg CO2 eq to have the proper net emission factor.

TRACI 2.1
TRACI was updated to version 2.1 in the summer of 2012. The methodology utilises global warming potentials
(GWPs) to calculate the potency of greenhouse gases relative to CO2, according to latest IPCC publications,
almost identically to the CML methodology. [TRACI 2012]

UBP 2006, Ecological Scarcity Method

The “ecological scarcity” method permits impact assessment of life cycle inventories according to the “dis-
tance to target” principle.
Eco-factors, expressed as eco-points per unit of pollutant emission or resource extraction, are normalised and
weighted according to Swiss national policy targets, as well as international targets supported by Switzerland.
For global warming the Kyoto protocol governs the reduction target, and the IPCC factors translate into the
other greenhouse gases. [UBP 2006]
For the comfort of the user, we applied some frequently used impact methods of “Global Warming Potential”
(like CML and IPCC) with both approaches, including and excluding biogenic carbon flows. If biogenic carbon
as an emission is accounted for, the respective CO2 uptake from air (modelled as resources) is consistently
modelled as well. Before interpreting and communicating results, the user should check for the specific goal,
scope and modelling approach in his application case and choose an appropriate Global Warming Impact
method, including or excluding biogenic carbon flows.

Global Warming Impact methods excluding biogenic carbon

IPCC global warming, excluding biogenic carbon [kg CO2-Equiv.]
CML2001 - Nov. 2010, Global Warming Potential, excl biogenic carbon (GWP 100 years) [kg CO2-Equiv.]
ReCiPe Midpoint (H) - Climate change [kg CO2-Equiv.]
ReCiPe 1.07 Midpoint (H) - Climate change [kg CO2-Equiv.]
TRACI 2.0, Global Warming Air, excluding biogenic carbon [kg CO2-Equiv.]
TRACI 2.1, Global Warming Air [kg CO2-Equiv.]

Global Warming Impact methods including biogenic carbon

IPCC global warming, including biogenic carbon [kg CO2-Equiv.]

120
 Supplement A

CML2001 - Nov. 2010, Global Warming Potential (GWP 100 years) [kg CO2-Equiv.]
CML96, Global warming potential (GWP 100 years) [kg CO2-Equiv.]
CML96, Global warming potential (GWP 20 years) [kg CO2-Equiv.]
CML96, Global warming potential (GWP 500 years) [kg CO2-Equiv.]
I02+ v2.1 - Global warming 500yr - Midpoint [kg CO2-Eq. to air]
TRACI 2.0, Global Warming Air [kg CO2-Equiv.]
EDIP 2003, Global warming [kg CO2-Equiv.]
EDIP 1997, Global warming potential (GWP 100 years) [kg CO2-Equiv.]

Supplement A 4 Acidification Potential (AP)

CML
The acidification of soils and waters occurs predominantly through the transformation of air pollutants into
acids. This leads to a decrease in the pH-value of rainwater and fog from 5.6 to 4 and below. Sulphur dioxide
and nitrogen oxide and their respective acids (H2SO4 und HNO3) produce relevant contributions. Ecosystems
are damaged, so forest dieback is the most well-known impact.
Acidification has direct and indirect damaging effects (such as nutrients being washed out of soils or an in-
creased solubility of metals into soils). But even buildings and building materials can be damaged. Examples
include metals and natural stones which are corroded or disintegrated at an increased rate.
When analysing acidification, it should be considered that although it is a global problem, the regional effects
of acidification can vary. Figure A-2 displays the primary impact pathways of acidification. [GUINÈE ET AL.
2001]
The acidification potential is given in sulphur
dioxide equivalents (SO2-Eq.). The acidification
potential is described as the ability of certain NOX
substances to build and release H+ ions. Cer- SO2
tain emissions can also have an acidification
potential, if the given S-, N- and halogen atoms H2SO44
HNO3
are set in proportion to the molecular mass of
the emission. The reference substance is sul-
phur dioxide.

Figure A-2: Acidification Potential

(KREISSIG & KÜMMEL 1999)

The average European characterisation factors of [CML 2001] are currently recommended as the best availa-
ble practise. Regional factors have not been adopted as the baseline, because it is not always possible, nor
desirable, to consider differences between emission sites in LCA.

121
 Supplement A

It is therefore important that emission site-independent characterisation factors become available, even for
those impact categories for which local sensitivity is important. [GUINÈE ET AL. 2001]
Accumulated exceedance (AE)
This study uses atmospheric models to calculate the deposition of released acidifying and eutrophing sub-
stance per release country and relates this value to the capacity of the receiving soil to neutralize the effects.
The method integrates both the exceeded area and amount of exceedance per kg of released substance
[SEPPÄLÄ ET AL. 2006].

EDIP 2003
Site-generic factors have been established as well as site-dependent factors for 44 European countries or
regions. The acidification factors relate an emission by its region of release to the acidifying impact on its
deposition areas.
The application of the EDIP2003 site-generic acidification factors is similar to the application of EDIP97 fac-
tors which are also site-generic.
The site-generic as well as the site-dependent EDIP2003 acidification potentials of an emission are expressed
as the area of ecosystem which is brought to exceed the critical load of acidification as a consequence of the
emission (area of unprotected ecosystem = m2 UES).
In comparison the EDIP97 acidification potential is expressed as the emission of SO2 that would lead to the
same potential release of protons in the environment (g SO2-Eq.) similar to the CML methodology.
[HAUSCHILD 2003]

Ecoindicator 99
For acidification, eutrophication and land-use the impacts are calculated using the Potentially Disappeared
Fraction (PDF) of species. The PDF is used to express the effects on vascular plant populations in an area.
The PDF can be interpreted as the fraction of species that has a high probability of no occurrence in a region
due to unfavourable conditions. The fate and damage of emitted substances are calculated via computer
models of the Netherlands.

Impact 2002+
The characterisation factors for aquatic acidification are expressed in SO2-equivalents and are adapted from
the EDIP1997 methodology which also corresponds to the approach from CML. [IMPACT 2002]

ReCiPe 1.07
The ReCiPe methodology calculates acidification as the Potentially Disappeared Fraction (PDF) of species in
forest ecosystems on a European scale, which is similar to the Ecoindicator approach. [RECIPE 201]

TRACI 2.1
TRACI 2.1 utilises the existing TRACI methodology for acidification plus some additional substances. The
calculations are performed for US conditions and the reference substance is kg SO2 eq. [TRACI 201]

122
 Supplement A

UBP 2006, Ecological Scarcity Method

The method has adapted CML values as the approach for acidification.

Supplement A 5 Eutrophication Potential (EP)

CML
Eutrophication is the enrichment of nutrients in a certain place. Eutrophication can be aquatic or terrestrial. Air
pollutants, wastewater and fertilisation in agriculture all contribute to eutrophication.
The result in water is an accelerated algae growth, which in turn, prevents sunlight from reaching the lower
depths. This leads to a decrease in photosynthesis and less oxygen production. Oxygen is also needed for
the decomposition of dead algae. Both effects cause a decreased oxygen concentration in the water, which
can eventually lead to fish dying and to anaerobic decomposition (decomposition without the presence of
oxygen). Hydrogen sulphide and methane are produced. This can lead to the destruction of the eco-system,
among other consequences.
On eutrophicated soils an increased susceptibility of plants to diseases and pests is often observed, as is
degradation of plant stability. If the nutrification level exceeds the amounts of nitrogen necessary for a maxi-
mum harvest, it can lead to an enrichment of nitrate. This can cause, by means of leaching, increased nitrate
content in groundwater. Nitrate also ends up in drinking water.
Nitrate at low levels is harmless from a toxico- Air pollution

logical point of view. Nitrite, however, is a reac- NOX N2O

NH3
tion product of nitrate and toxic to humans. The
Fertilisation
causes of eutrophication are displayed in Fig-
ure A-3. The eutrophication potential is calcu-
lated in phosphate equivalents (PO4-Eq.). As
with acidification potential, it is important to NO3 -
NH4+
Waste water
remember that the effects of eutrophication
-3
PO4

potential differ regionally. Figure A-3: Eutrophication Potential

(KREISSIG & KÜMMEL 1999)

All emissions of N and P to air, water and soil and of organic matter to water are aggregated into a single
measure, as this allows both terrestrial and aquatic eutrophication to be assessed. The characterisation fac-
tors in PO4-equivalents, NO3-equivalents and O2-equivalents are all interchangeable, and PO4-equivalents are
used. [GUINÈE ET AL. 2001]

Accumulated exceedance (AE)

This study uses atmospheric models to calculate the deposition of released acidifying and eutrophing sub-
stance per releasing country and relates this to the capacity of the receiving soil to neutralize the effects. The

123
 Supplement A

method integrates both the exceeded area and amount of exceedance per kg of released substance [SEPPÄLÄ
ET AL. 2006].

EDIP 2003
The EDIP 2003 methodology distinguishes between aquatic and terrestrial eutrophication.
Aquatic eutrophication
The aquatic inputs are atmospheric deposition of nitrogen on soil and coastal seas, phosphorus and nitrogen
supply to agricultural soils, phosphorus and nitrogen discharged with municipal wastewater. A computer mod-
el (CARMEN) calculates transport of the inlet nutrients to surface water.
The nitrogen and phosphorus sources have been allocated to each grid-element on the basis of the distribu-
tion of land uses in the given grid-element (arable land, grassland, permanent crops, forest, urban area, inland
waters).
The transport of nutrient by rivers to sea is modelled assuming fixed removal rates of N and P in freshwater
systems. [HAUSCHILD 2003]
Terrestrial eutrophication
Site-dependent factors have been established for 44 European countries or regions. The eutrophication fac-
tors relate an emission by its region of release to the acidifying impact on its deposition areas.
The site-generic terrestrial eutrophication factors are established as the European average over the 15 EU
member countries in the EU15 plus Switzerland and Norway, weighted by the national emissions. The site-
generic as well as the site-dependent EDIP2003 acidification potentials of an emission are expressed as the
area of ecosystem whose inclusion exceeds the critical load of eutrophication as a consequence of the emis-
sion (area of unprotected ecosystem = m2 UES). [HAUSCHILD 2003]

Ecoindicator 99
For acidification, eutrophication and land-use the impacts are calculated using the Potentially Disappeared
Fraction (PDF) of species. The PDF is used to express the effects on vascular plant populations in an area.
The PDF can be interpreted as the fraction of species that has a high probability of no occurrence in a region
due to unfavourable conditions. The fate and damage of emitted substances are calculated via computer
models of the Netherlands. [ECO-INDICATOR 99 : 2000]

Impact 2002+
Midpoint characterisation factors (in kg PO43--equivalents) are given for emissions into air, water and soil with
characterisation factors taken directly from CML.
No aquatic eutrophication damage factors (in PDF·m2·yr/kg emission) are given because no available studies
support the assessment of damage factors for aquatic eutrophication. [IMPACT 2002]

ReCiPe 1.07
ReCiPe operates with both mid-point and end-point indicators.

124
 Supplement A

Mid-point indicators are divided into freshwater and marine eutrophication. At the freshwater level, only phos-
phorous is included and at the marine level, only nitrogen is included. It can be written as the marginal con-
centration increment in tn/km3 in exposed aquatic system per marginal increase of emission rate in tn/yr,
hence with the unit yr/km3. This is the amount supplied per kg of pure nitrogen or phosphorus emitted. When
included in GaBi this value is then converted into phosphorus and nitrogen equivalents for the emitted sub-
stances21.
As an endpoint, ReCiPe operates with species loss in freshwater on a European scale. [RECIPE 2012]

TRACI 2.1
The characterisation factors of TRACI 2.1 estimate the eutrophication potential of a release of chemical con-
taining N or P to air or water relative to 1 kg N discharged directly to surface freshwater, therefore with the unit
kg N eq. [TRACI 201]

UBP 2006, Ecological Scarcity Method

The “ecological scarcity” method permits impact assessment of life cycle inventories according to the “dis-
tance to target” principle.
Eco-factors, expressed as eco-points per unit of pollutant emission or resource extraction, are normalised and
weighted according to Swiss national policy targets, as well as international targets supported by Switzerland.
For acidification this is a 50% reduction target in Rhine catchment according to the OSPAR Commission.
[UBP 2006]

Supplement A 6 Photochemical Ozone Creation Potential (POCP)

CML
Despite playing a protective role in the stratosphere, ozone at ground level is classified as a damaging trace
gas. Photochemical ozone production in the troposphere, also known as summer smog, is suspected to dam-
age vegetation and material. High concentrations of ozone are toxic to humans.
Radiation from the sun and the presence of nitrogen oxides and hydrocarbons incur complex chemical reac-
tions, producing aggressive reaction products, one of which is ozone. Nitrogen oxides alone do not cause high
ozone concentration levels.
Hydrocarbon emissions occur from incomplete combustion, in conjunction with petrol (storage, turnover, refu-
elling) or from solvents. High concentrations of ozone arise when temperature is high, humidity is low, air is
relatively static and there are high concentrations of hydrocarbons. Today it is assumed that the existence of
NO and CO reduces the accumulated ozone to NO2, CO2 and O2. This means that high concentrations of
ozone do not often occur near hydrocarbon emission sources. Higher ozone concentrations more commonly
arise in areas of clean air, such as forests, where there is less NO and CO (Figure A-4).

21
The emissions to agricultural soil should be multiplied with the fertilizer factors in ReCiPe main report.

125
 Supplement A

In Life Cycle Assessments photochemical

Hydrocarbons
ozone creation potential (POCP) is referred to Nitrogen oxides

in ethylene-equivalents (C2H4-Eq.). During Dry and warm

climate
analysis it is important to note that the actual Ozone
Hydrocarbons
ozone concentration is strongly influenced by
the weather and by the characteristics of local
Nitrogen oxides
conditions.
Figure A-4: Photochemical Ozone Creation Potential (
KREISSIG & KÜMMEL 1999)

The most recent POCP factors are still the ones used for the original CML methodology with only a few
adjustments. [GUINÈE ET AL. 2001]

EDIP 2003
The EDIP2003 characterisation factors for photochemical ozone formation have been developed using the
RAINS model which was also used for development of characterisation factors for acidification and terrestrial
eutrophication. Site-generic factors have been established, in addition to site-dependent factors for 41 Euro-
pean countries or regions. The photochemical ozone formation factors relate an emission by its region of
release to the ozone exposure and impact on vegetation or human beings within its deposition areas.
[HAUSCHILD 2003]

Ecoindicator 99
In Ecoindicator 99 the POCP expresses the incremental ozone concentration per incremental emission for
specific VOC species normalised by the ratio for ethylene, equivalent to the CML approach. This is then calcu-
lated further via epidemiological studies to yield the end-point indicator of Disability Adjusted Life Years
(DALYs). [ECO-INDICATOR 99 : 2000]

Impact 2002+
Photochemical oxidation (damage in DALY/kg emissions) is taken directly from Eco-indicator 99. Midpoints
are given relative to air emissions of ethylene equivalent to CML. [IMPACT 2002]

ReCiPe 1.07
The dynamic model LOTOS-EUROS was applied to calculate intake fractions for ozone due to emissions of
NOx.
The mid-point characterisation factor for ozone formation of a substance is defined as the marginal change in
the 24h-average European concentration of ozone (in kg/m3) due to a marginal change in emission (in
kg/year) expressed as NMVOC-equivalents.
The end-point indicator is human health expressed as DALYs. [RECIPE 201]

126
 Supplement A

TRACI 2.1
Impacts of photochemical ozone creation are quantified using the Maximum Incremental Reactivity (MIR)
scale. This scale is based on model calculations of effects of additions of the VOCs on ozone formation in
one-day box model scenarios representing conditions where ambient ozone is most sensitive to changes in
VOC emissions. The emissions are normalised relative to ozone (O3-equivalents). [TRACI 201]

UBP 2006, Ecological Scarcity Method

Eco-factors, expressed as eco-points per unit of pollutant emission, are normalised against the entirety of
Switzerland and weighted according to Swiss national policy targets. For POCP the target value is the aver-
age of three values [UBP 2006];
 Swiss Federal Air Pollution Control Ordinance’s ambient limit values for ozone
 The Swiss air pollution control strategy stipulates a reduction to the level of 1960 as a minimum tar-
get for NMVOCs
 The environment ministers of Germany, Liechtenstein, Switzerland and Austria adopted a declaration
setting the target of reducing NMVOC emissions by 70-80% from the level of the 1980s.

Supplement A 7 Ozone Depletion Potential (ODP)

Ozone is created in the stratosphere by the disassociation of oxygen atoms that are exposed to short-wave
UV-light. This leads to the formation of the so-called ozone layer in the stratosphere (15-50 km high). About
10% of this ozone reaches the troposphere through mixing processes. In spite of its minimal concentration,
the ozone layer is essential for life on earth. Ozone absorbs the short-wave UV-radiation and releases it in
longer wavelengths. As a result, only a small part of the UV-radiation reaches the earth.
Anthropogenic emissions deplete ozone. This is well-known from reports on the hole in the ozone layer. The
hole is currently confined to the region above Antarctica; however further ozone depletion can be identified,
albeit not to the same extent, over the mid-latitudes (e.g. Europe). The substances which have a depleting
effect on the ozone can essentially be divided into two groups; the chlorofluorocarbons (CFCs) and the nitro-
gen oxides (NOX). Figure A-5 depicts the procedure of ozone depletion.
One effect of ozone depletion is the warming of the earth's surface. The sensitivity of humans, animals and
plants to UV-B and UV-A radiation is of particular importance. Possible effects are changes in growth or a
decrease in harvest crops (disruption of photosynthesis), indications of tumours (skin cancer and eye diseas-
es) and a decrease of sea plankton, which would strongly affect the food chain. In calculating the ozone de-
pletion potential, the anthropogenically-released halogenated hydrocarbons, which can destroy many ozone
molecules, are recorded first. The Ozone Depletion Potential (ODP) results from the calculation of the poten-
tial of different ozone relevant substances.

127
 Supplement A

A scenario for a fixed quantity of emissions of UV - radiation

a CFC reference (CFC 11) is calculated, re-
sulting in an equilibrium state of total ozone
reduction. The same scenario is considered Stratosphere
15 - 50 km Absorption Absorption
for each substance under study where CFC
11 is replaced by the quantity of the sub- CFCs
stance. This leads to the ozone depletion Nitrogen oxide
potential for each respective substance, which
is given in CFC 11-equivalents. An evaluation
of the ozone depletion potential should take
into consideration the long term, global and
partly irreversible effects.

Figure A-5: Ozone Depletion Potential

( KREISSIG & KÜMMEL 1999)

In CML the ODPs published by the World Meteorological Organisation (WMO) from 2002 are used. [GUINÈE
ET AL. 2001]

EDIP 2003
The EDIP factors are calculated via the same principle as CML. [HAUSCHILD 2003]

Ecoindicator 99
The fate of CFC11 was modelled and used to estimate the fate of other substances. Standard ODPs are used
to relate this to reduction in ozone. The increase in UV radiation was then used to estimate the increase in
eye cataract and skin cancer which is finally expressed as Disability Adjusted Life Years (DALYs). [ECO-
INDICATOR 99 : 2000]

Impact 2002+
Midpoints (kg CFC-11-Eq. into air/kg emission) have been obtained from the US Environmental Protection
Agency Ozone Depletion Potential List. The damage factor (in DALY/kg emission) for the midpoint reference
substance (CFC-11) was taken directly from Eco-indicator 99. Damage (in DALY/kg emission) for other sub-
stances has been obtained by the multiplication of the midpoints (in kg CFC-11- Eq. into air/kg emission) and
the CFC-11 damage factor (in DALY/kg CFC-11 emission). [IMPACT 2002]

ReCiPe 1.07
The ODPs from Ecoindicator are used as equivalency factors, characterising substances at the midpoint level.
As an end-point indicator, only damage to human health (skin cancer and cataracts) is addressed because
uncertainty regarding other areas of protection was considered too large. In a new approach the fate of a
marginal increase of emission of ozone depleting substances and the resulting worldwide increase of UVB
exposure is evaluated, taking into account population density, latitude and altitude. For characterisation of

128
 Supplement A

damage, protective factors are accounted for, such as skin colour and culturally-determined habits such as
clothing. [RECIPE 201]

TRACI 2.1
Within TRACI 2.1, the most recent sources of ODPs from WMO (World Meteorological Organization) are used
for each substance. [TRACI 201]

UBP 2006, Ecological Scarcity Method

The Swiss Chemicals Risk Reduction Ordinance prohibits the production, importation and use of ozone- de-
pleting substances. Exemptions regarding importation and use are presently only in place for the maintenance
of existing HCFC refrigeration equipment and for the recycling of HCFC refrigerants with a transitional period
lasting until 2015.
The primary stocks formed in building insulation materials will continue releasing considerable amounts. No
critical flow can therefore be derived directly from the wide-ranging ban on the consumption of ozone-
depleting substances.
The tolerated emissions are taken as the basis for determining the critical flow. As the exemptions for HCFC
use in existing refrigeration equipment terminate in 2015, the anticipated emissions in 2015 are used as the
critical flow (the target). The current emissions are estimated to calculate the ecofactor.
Standard ODPs are used to convert this ecofactor to other ozone-depleting substances. [UBP 2006]

Supplement A 8 Human and eco-toxicity, USEtox

USEtox is a scientific consensus model developed by those behind the CalTOX, IMPACT 2002, USES-LCA,
BETR, EDIP, WATSON and EcoSense.
In 2005, a comprehensive comparison of life cycle impact assessment toxicity characterisation models was
initiated by the United Nations Environment Program (UNEP)–Society for Environmental Toxicology and
Chemistry (SETAC) Life Cycle Initiative, directly involving the model developers of CalTOX, IMPACT 2002,
USES-LCA, BETR, EDIP, WATSON and EcoSense.
The main objectives of this effort were (1) to identify specific sources of differences between the models’ re-
sults and structure, (2) to detect the indispensable model components and (3) to build a scientific consensus
model from them, which represent the recommended practise.
Based on a referenced database, it has now been used to calculate CFs for several thousand substances,
and forms the basis of the recommendations from UNEP-SETAC’s Life Cycle Initiative regarding characterisa-
tion of toxic impacts in life cycle assessment.
The model provides both recommended and interim (not recommended and to be used with caution) charac-
terisation factors for human health and freshwater ecotoxicity impacts.

129
 Supplement A

USEtox calculates characterisation factors for

human toxicity and freshwater ecotoxicity via
Halogenorganic
three steps: environmental fate, exposure and compounds
effects. Heavy metals
DCB PCB
The continental scale of the model consists of
PAH
six compartments: urban air, rural air, agricul-
tural soil, industrial soil, freshwater and Air

coastal marine water. The global scale has the Food

same structure, but without the urban air.
The human exposure model quantifies the Products
increase in amount of a compound transferred
into the human population based on the con- Figure A-6: Human Toxicity Potential
centration increase in the different media.
Human effect factors relate the quantity taken
in to the potential risk of adverse effects in Halogenorganic
humans. It is based on cancerous and non- compounds
Heavy metals PCB
cancerous effects derived from laboratory DCB

studies. PAH

Effect factors for freshwater ecosystems are

based on species-specific data of concentra-
tion at which 50% of a population displays an
effect.
Biosphere
The final characterisation factor for human (Terrestrial ecosystem)
toxicity and aquatic ecotoxicity is calculated by
summation of the continental- and the global- Figure A-7: Terrestrial Eco-Toxicity Potential
scale assessments.
The characterisation factor for human toxicity
is expressed in comparative toxic units Halogenorganic
(CTUh), providing the estimated increase in compounds
Heavy metals PCB
morbidity per unit mass of a chemical emitted DCB
(cases per kilogram). PAH

The characterisation factor for aquatic ecotox-

icity is expressed in comparative toxic units
(CTUe) and provides an estimate of the poten-
tially affected fraction of species (PAF) inte-
grated over time and volume per unit mass of
a chemical emitted (PAF m3-day/ kg). Biosphere
(Aquatic ecosystem)
[USETOX 2010]
Figure A-8: Aquatic Eco-Toxicity Potential

130
 Supplement A

CML
The CML toxicity calculations are based on fate modelling with USES-LCA. This multimedia fate is divided
into 3% surface water, 60% natural soil, 27% agricultural soil and 10% industrial soil. 25% of the rainwater is
infiltrated into the soil.
The potential toxicities (human, aquatic and terrestrial ecosystems) are generated from a proportion based on
the reference substance 1.4-Dichlorbenzol (C6H4Cl2) in the air reference section. The unit is kg 1.4-
Dichlorbenzol-Equiv. (kg DCB-Eq.) per kg emission [GUINÉE ET AL. 2002].
The identification of the toxicity potential is rife with uncertainties because the impacts of the individual sub-
stances are extremely dependent on exposure times and various potential effects are aggregated. The model
is therefore based on a comparison of effects and exposure assessment. It calculates the concentration in the
environment via the amount of emissions, a distribution model and the risk characterisation via an input-
sensitive module. Degradation and transport in other environmental compartments are not represented.
[GUINÈE ET AL. 2001]

EDIP 2003
Toxicity impacts from EDIP 2003 are no longer included in GaBi, as the EDIP methodology has shifted to
using the USEtox methodology to assess toxicity impacts.

Ecoindicator 99
For the fate analysis of carcinogenic substances causing damage to Human Health and ecotoxic substances
causing damage to Ecosystem Quality, the European Uniform System for the Evaluation of Substances
(EUSES) is used. Different environmental media (air, water, sediment, and soil) are modelled as homogene-
ous, well-mixed compartments or boxes.
Substances that cause respiratory effects are modelled with atmospheric deposition models and empirical
observations.
The damage, expressed as the number of Disability-Adjusted Life Years (DALYs), measures the total amount
of ill health, due to disability and premature death, attributable to specific diseases and injuries. [ECO-
INDICATOR 99 : 2000]

Impact 2002+
Impact 2002+ expresses toxicity in a total of four mid-point impact categories; human toxicity (carcinogen and
non-carcinogen effects), respiratory effects (caused by inorganics), aquatic ecotoxicity, and terrestrial ecotoxi-
city.
Damages are expressed in Disability-Adjusted Life Years for human effects and Potentially Disappeared Frac-
tion (PDF) of species for ecotoxic effects. [IMPACT 2002]

131
 Supplement A

ReCiPe 1.07
The characterisation factor of human toxicity and ecotoxicity is composed of the environmental persistence
(fate) and accumulation in the human food chain (exposure), and toxicity (effect) of a chemical. The ReCiPe
method uses an update of the model used in the CML methodology referred to as USES-LCA 2.0.
The potential human toxicity and three categories of eco-toxicity (freshwater, marine and terrestrial) are ex-
pressed as mid-point indicators relative to 1.4-Dichlorbenzol (kg DCB-Eq.).
The end-point indicators are expressed in DALYs for human toxicity and species loss for ecotoxicity. [RECIPE
201]

TRACI 2.1
The TRACI 2.1 methodology has adopted the approach of the USEtox model combined with some additions
from the original TRACI methodology. [TRACI 201]

UBP 2006, Ecological Scarcity Method

The method has developed ecopoints per kg-emitted substance for only a limited amount of substances. [UBP
]

Supplement A 9 Resource depletion

CML
The abiotic depletion potential (ADP) covers some selected natural resources as metal-containing ores, crude
oil and mineral raw materials. Abiotic resources include raw materials from non-living resources that are non-
renewable. This impact category describes the reduction of the global amount of non-renewable raw materi-
als. Non-renewable means a time frame of at least 500 years. The abiotic depletion potential is split into two
sub-categories, elements and fossil.
Abiotic depletion potential (elements) covers an evaluation of the availability of natural elements like minerals
and ores, including uranium ore. The reference substance for the characterisation factors is antimony. Two
calculations of ADP (elements) from CML are integrated in GaBi5, one based on ultimate ultimate resources
(i.e. the total mineral content in the earth crust) and one based on what is evaluated as being economically
feasible to extract. The latter version is recommended by ILCD.
The second sub-category is abiotic depletion potential (fossil), which includes the fossil energy carriers (crude
oil, natural gas, coal resources). MJ is the respective unit. [GUINÈE ET AL. 2001]

EDIP 2003
The former EDIP methodology, EDIP 1997, contained a resource category consisting of 87 resource quanti-
ties (minerals and fossil resources) without any classification or characterisation. This category is omitted in
the EDIP 2003 update. [HAUSCHILD 2003]

132
 Supplement A

Ecoindicator 99
The primary assumption in this method is that if the resource quality is reduced, the effort to extract the re-
maining resource increases. Plain market forces will ensure that mankind always exploits the resources with
the highest quality. This means each time a kg of a resource is used, the quality of the remaining resources is
slightly decreased and thus the effort to extract the remaining resources is increased. The damage to re-
sources is measured in MJ of surplus energy which is defined as the difference between the energy needed to
extract a resource now and at some specific point in the future. [ECO-INDICATOR 99 : 2000]

Impact 2002+
Characterisation factors for non-renewable energy consumption, in terms of the total primary energy extract-
ed, are calculated with the upper heating value. It is taken from ecoinvent (Frischknecht et al. 2003).
Mineral extractions in MJ surplus energy are taken directly from Eco-indicator. [IMPACT 2002]

ReCiPe 1.07
The marginal cost increase on the deposit level can be defined as the marginal average cost increase ($/$)
due to extracting a dollar value of deposit (1/$).
From the marginal cost increase factor on the deposit level, the cost increase factor on commercial metal level
is calculated. The mid-point is then related to iron as iron equivalents (Fe-Eq.).

TRACI 2.1
The abiotic resource depletion in TRACI 2.1 focuses on fossil fuels with an approach taken from Ecoindicator.
Extraction and production of fossil fuels consume the most economically recoverable reserves first, making
continued extraction more energy intensive, hence the unit of MJ surplus energy. [TRACI 201]

UBP 2006, Ecological Scarcity Method

Eco-factors, expressed as eco-points per MJ of energy consumption are used for energy.
Minerals are not included. [UBP 2006]

Land Use, LANCA

Land use is also considered a limited resource. It is integraged in GaBi 6 via 5 indicators: Erosion resistance,
mechanical filtration, physicochemical filtration, groundwater replenishment, and biotic production. The back-
ground is the LANCA tool (Land Use Indicator Calculation Tool) based on country-specific input data and the
respective land use types. A detailed description of the underlying methods can be found in BECK, BOS, WITT-
STOCK ET AL. 2010.

Land Use, Soil Organic Matter (SOM)

SOM (closely related to soil organic carbon, SOC) is basically a balance of the organic matter in soil related to
the anthropogenic use of land for human activity. Initial organic content, as well as an annual balance of the
organic matter in the soil, is necessary to calculate this [MILA` I CANALS 2007]. It is currently not integrated in

133
 Supplement A

GaBi5 as an impact category but can be calculated from LCI datasets as CO2 extracted from atmosphere
minus carbon flows to water, and carbon uptake in products.

Water
Standardisation for the creation of an approach for water footprinting and water use as an impact assessment
category is underway.
All water-related flows of GaBi LCI data are updated to enable consistent, high quality water modelling for
water use assessments and water footprinting according to the upcoming ISO Water Footprint standard, the
Water Footprint Network Manual and other emerging guidelines.
Four new water quantities where implemented to reflect the latest status of best practise in water foot printing
and water assessments.
Total freshwater consumption (including rainwater)
Blue water consumption
Blue water use
Total freshwater use
Furthermore, we added a “Total freshwater consumption (including rainwater)” quantity in the light of the rec-
ommended ILCD methods carrying a characterised value according to the UBP method.

Supplement A 10 Particulate matter formation (PM)

Riskpoll
The Riskpoll model evaluates human health impacts from primary particles emitted directly and from second-
ary particles formed in the air by emitted substances [RABL AND SPADARO 2004]. The reference unit is kg PM2.5
eq.

ReCiPe 1.07
The atmospheric fate was calculated using a combination of the models EUTREND and LOTOS-EUROS
including effects of both primary and secondary particles. The reference unit is kg PM10 eq.

TRACI 2.1
These intake fractions are calculated as a function of the amount of substance emitted into the environment,
the resulting increase in air concentration, and the breathing rate of the exposed population. The increasing
air concentrations are a function of the location of the release and the accompanying meteorology and the
background concentrations of substances, which may influence secondary particle formation. Substances
were characterised using PM2.5 as the reference substance.

134
 Supplement A

Supplement A 11 Normalization

Normalization relates each impact to a reference of a per capita or a total impact for a given area for a given
year. An overview is given in Table K.

Table K: Normalization references

Methodology Impact calculated (year) Area(s) covered

CML 2001 Total impact (2000) World, Europe
ReCiPe 1.07, Ecoindicator Per capita impact (2000) World, Europe
TRACI 2.1 Per capita impact (2006) USA, USA+Canada
EDIP 2003 Per capita impact (1994) Europe
UBP 2006 Per capita impact (various) Switzerland
USEtox Per capita impact
(2004 Europe)
Europe, North America
(2002/2008 North America)

Conversion between CML and ReCiPe is possible using a global population of 6,118,131,162 and a EU25+3
population of 464,621,109 in year 2000 [EUROSTAT 2012][WORLD BANK 2012]. Notably the ‘+3’ countries in
EU25+3 are Iceland, Norway, and Switzerland.

Supplement A 12 Weighting

The weighting attaches a value to each of the normalized values giving a value based importance of each
impact. This can be based on political reduction targets or on the opinions of experts and/or laymen, for ex-
ample.
In 2012 PE INTERNATIONAL sent out a questionnaire worldwide asking experts to value the main environ-
mental impact categories on a 1-10 scale. The total number of respondents were 245 mainly consultants and
academia and mainly from Europe and North America. Figure A-9 below gives an overview of the respond-
ents with the area and colon of each rectangle representing the number of people within each category.

135
 Supplement A

Figure A-9: Response to PE Weighting 2012

The answers from the questionnaires led to the weighting factors in Table L. The weighting factors are linked
to the impact categories of CML and ReCiPe (Global + Europe), and for TRACI 2.1 (Global + North America).
Additionally, the IPCC category for global warming is also included (Global + Europe + North America).

Table L: PE Weighting 2012

Impact Europe North America Global

Acidification 6.2 5.9 6.1
Eco-Toxicity 6.6 7.0 6.8
Eutrophication 6.6 6.6 6.6
Global Warming 9.3 9.5 9.3
Human Toxicity 6.9 7.5 7.1
Ionising Radiation 5.8 5.0 5.7
Ozone Depletion 6.2 6.1 6.2
Particulate Matter Formation 6.5 6.9 6.7
Photochemical Ozone 6.5 6.7 6.5
Resources, ADP elements 6.3 6.1 6.4
Resources, ADP fossil 6.9 6.7 7.0
Resources, Land Use 7.2 7.1 7.2
Water Footprint 7.9 8.4 8.0

136
 Supplement B

Supplement B Background information on uncertainty

The following chapter provides background information on uncertainty issues in LCA.

Aspects of data uncertainty due to variability in supply chains

While Chapter 1 addressed data and model uncertainty assuming that the practitioner has been able to select
the most appropriate or ‘representative’ datasets for the product system under study, this chapter will attempt
to quantify relevant aspects of uncertainty in background data due to its variability concerning technological
and geographical representativeness.
As mentioned in the previous chapter, +/-10% uncertainty appears to be the minimum overall uncertainty,
even if the model is set up with data of high quality containing few errors.
The model’s degree of representativeness regarding supply chains and technology routes depends on the
specific situation under consideration. It varies due to factors including specific supplier companies and geo-
graphical/national import situations.
The correlation between the background data and the specific situation at hand can only be answered
by performing a primary data collection for each specific supply situation and comparing it with the
average situation represented by the background data.
The background data as such may be very precise and of extremely high representativeness within the situa-
tion where it was set up. The goal of this chapter is to estimate possible variations in background data due to
the mismatch between the average and actual supply chain in a specific situation. To achieve this goal two
types of possible misrepresentation introduced by the user of the data are assessed:
 the influence of varying the import/production country
 the influence of varying the technology route in the same country to supply the same material or
substance
The analysis focuses on chemical products and intermediate products.

Disclaimer:
The following analyses are specific to the products and datasets available in the GaBi databases. The
results cannot be generalised to other products or data sources.

Influence of varying import/production country for same technology

The following chemical substances were analysed for their variability with regard to their geography.

137
 Supplement B

Table M: Chemical substance datasets available for various countries in GaBi

Acetic acid from methanol Hydrogen (Steamreforming fuel oil s)

Acetone by-product phenol methyl styrene (from Cumol) Hydrogen (Steamreforming natural gas)

Adipic acid from cyclohexane Maleic anhydride (MA) by-product PSA (by oxidation of xylene)

AH-salt 63% (HMDA via adipic acid) Maleic anhydride from n-butane

Ammonium sulphate by-product caprolactam Methyl methacrylate (MMA) spent acid recycling

Benzene (from pyrolysis gasoline) Methyl methacrylate (MMA) from acetone and hydrogen cyanide

Benzene (from toluene dealkylation) Methylene diisocyanate (MDI) by-product hydrochloric acid, methano

Benzene by-product BTX (from reformatee) Phenol (toluene oxidation)

Caprolactam from cyclohexane Phenol from cumene

Caprolactam from phenol Phosphoric acid (wet process

Chlorine from chlorine-alkali electrolysis (amalgam) Phthalic anhydride (PAA) (by oxidation of xylene)

Chlorine from chlorine-alkali electrolysis (diaphragm) Propylene glycol over PO-hydrogenation

Chlorine from chlorine-alkali electrolysis (membrane) Propylene oxide (Cell Liquor)

Ethanol (96%) (hydrogenation with nitric acid) Propylene oxide (Chlorohydrin process)

Ethene (ethylene) from steam cracking Propylene oxide by-product t-butanol (Oxirane process)

Ethylbenzene (liquid phase alkylation) p-Xylene (from reformate)

Ethylene glycol from ethene and oxygen via EO Toluene (from pyrolysis gasoline)

Ethylene oxide (EO) by-product carbon dioxide from air Toluene by-product BTX (from reformate)

Ethylene oxide (EO) by-product ethylene glycol Toluene by-product styrene

Hexamethylene diamine (HMDA) via adipic acid Toluene diisocyanate (TDI) by-product toluene diamine, hydrochloric
acid (phosgenation)

Hydrochloric acid by-product methylene diisocyanate (MDI) Xylene mix by-product benzene (from pyrolysis gasoline)

These routes were analysed (as available) concerning process boundary conditions in various countries in-
cluding:
Australia (AU), Belgium (BE), China (CN), Germany (DE), Spain (ES), France (FR), Great Britain (GB), Italy (IT), Japan
(JP), Netherlands (NL), Norway (NO), Thailand (TH), United States (US)

The following figure shows the resulting maximum variations of all analysed materials and substances. The
respective technologies are kept constant and only the country of origin is varied. The figure shows the maxi-
mum variability across the various chemicals that have been analysed, as well as the 90% and 10% percen-
tiles.

138
 Supplement B

Two cases were calculated for each route, assuming that the actual location of the supplier is unknown in a
given LCA project. Choosing the dataset with the lowest burden while the one with the highest burden would
have been appropriate (‘choose min’; uncertainty = (min-max)/max) and vice versa (‘choose max’; uncertain-
ty = (max-min)/min). The resulting values are therefore the relative ‘worst-case errors’ possible based on the
datasets considered.
500%

400%

300%

200%

100%

0%

-100%

-200%
PED AP EP GWP POCP
10% percentile -21% -65% -56% -41% -59%
choose min -68% -95% -79% -82% -93%
choose max 209% 1870% 380% 461% 1288%
90% percentile 27% 189% 129% 70% 143%

Figure B-10: Maximum errors regarding randomly chosen geography

Figure B-10 shows that when assuming that the technology route for a certain substance is known and the
specific country of origin route is not, the maximum uncertainty of the related impacts is between -65% and
+189% for 80% of all chemical substances for which different country-specific datasets are available in the
GaBi Database.
When taking the background information of the GaBi MasterDB in to account, the sensitivity concerning the
country of origin appears to be more relevant for process chains where energy and the respective emissions
from energy supply dominate the impacts. In selected cases country-specific emissions or synthesis efficien-
cies and differences in country-specific upstream supply are also relevant.
Influence of varying technology in the same country
The following chemical substances were analysed regarding their variability with regard to their technology
route in the same country.

139
 Supplement B

Table N: Chemical substance datasets available for various technology routes in GaBi

Chlorine from chlorine-alkali electrolysis diaphragm Ethylene-t-Butylether from C4 and bio ethanol
Chlorine from chlorine-alkali electrolysis membrane Hexamethylene diamine via Adiponitrile

Chlorine from chlorine-alkali electrolysis amalgam Hexamethylene diamine via adipic acid

Acetic acid from vinyl acetate Hydrochloric acid primary from chlorine
Acetic acid from methanol Hydrochloric acid by-product allyl chloride
Acrylamide catalytic hydrolysis Hydrochloric acid by-product chlorobenzene
Acrylamide enzymatic hydration Hydrochloric acid by-product epichlorohydrine
AH salt 63% HMDA from adipic acid Hydrochloric acid by-product Methylene diisocyanate
AH salt 63% HMDA from acrylonitrile Hydrogen Cracker
Ammonium sulphate by-product acetone cyanhydrin Hydrogen Steamreforming fuel oil s
Ammonium sulphate by-product Caprolactam Hydrogen Steamreforming natural gas
Benzene from pyrolysis gasoline Maleic anhydride from n-butane
Benzene from toluene dealkylation Maleic anhydride by-product phthalic anhydride
Benzene by-product BTX Maleic anhydride from benzene
Benzene by-product ethine Methyl methacrylate from acetone and hydrogen
cyanide
Butanediol from ethine, H2 Cracker, allotherm Methyl methacrylate spent acid recycling
Butanediol from ethine H2 Steam ref. natural gas, autotherm Oleic acid from palm oil

Chlorodifluoroethane from 1,1,1-Trichloroethane Oleic acid from rape oil

Chlorodifluoroethane by-product Dichloro-1-fluoroethane Phenol by toluene oxidation
Dichlorpropane by-product epichlorohydrin Phenol by-product acetone
Dichlorpropane by-product dichlorpropane Phosphoric acid (54%)
Ethanol catalytic hydrogenation with phosphoric acid Phosphoric acid (100%)
Ethanol hydrogenation with nitric acid Propylene oxide Cell Liquor
Ethylene glycol by-product Ethylene oxide Propylene oxide Chlorohydrin process
Ethylene glycol of Ethene + oxygen via EO Propylene oxide Oxirane process
Ethylene glycol from Ethyleneoxide Toluene from pyrolysis gasoline
Ethylene oxide by-product carbon dioxide Toluene by-product BTX
Ethylene oxide by-product ethylene glycol via CO2/methane Toluene by-product styrene
Ethylene oxide by-product ethylene glycol via CO2/methane Xylene from pyrolysis gasoline
with CO2 use
Ethylene-t-Butylether from C4 Xylene from reformate

The following figure shows the resulting maximum errors across all analysed materials and substances. Here,
the respective countries of origin are kept constant and only the technology route is varied. The figure shows
the maximum errors across the various chemicals analysed, as well as the 90% and 10% percentiles.
Again, two cases were calculated for each country, assuming that the actual technology route of the supplier
is unknown in a given LCA project: choosing the technology-specific dataset with the lowest burden while the
one with the highest burden would have been appropriate (‘choose min’; uncertainty = (min-max/max)) and
vice versa (‘choose max’; uncertainty = (max-min)/min). The resulting values are therefore again the relative
‘worst-case errors’ possible based on the available datasets.

140
 Supplement B

500%

400%

300%

200%

100%

0%

-100%

-200%
PED AP EP GWP POCP
10% percentile -34% -57% -61% -71% -66%
choose min -96% -94% -93% -96% -96%
choose max 2409% 1596% 1332% 2609% 2731%
90% percentile 52% 132% 156% 248% 197%

Figure B-11: Maximum errors regarding randomly chosen technology

Figure B-11 shows that when assuming that the country of origin for a certain substance is known and the
specific technology route is not, the errors of the related impacts falls between -71% and +248% for 80% of
all chemical substances for which different technologies are available in the GaBi Database. Comparing the
values to the ones in the previous part concerning geography, it is fair to state that it is worse to have an un-
defined specific technology route than an undefined country of origin, since all values are higher for the latter.

141