Carbon Footprint Methodology

CARBON FOOTPRINT
MEASUREMENT
METHODOLOGY REPORT
Centre for Water and Waste Technology

University of NSW
Dr Sven Lundie, Matthias Schulz, Dr Greg Peters
in co-operation with
Scion
Dr Barbara Nebel
AgResearch
Dr Stewart Ledgard
for
Fonterra Co-Operative Group Limited
12 January 2009
DOCUMENT STATUS RECORD

Project Title: Carbon Footprint Measurement
Client: Fonterra Co-operative Group Limited
Document Title Methodology Report
Document File Name: 081204_CWWT SCION AGRESEARCH
Methodology for carbon footprint
Disclaimer:
1. The Centre for Water and Waste Technology has taken all reasonable steps to ensure that the information
contained in this publication is accurate at the time of production. In some cases, we have relied on
information supplied by the client.
2. This report has been prepared in accordance with good professional practice. No other warranty,
expressed or implied, is made as to the professional advice given in this report.
3. The Centre for Water and Waste Technology maintains no responsibility for the misrepresentation of
results due to incorrect use of information contained within this report.
4. This report should remain together and be read as a whole.
5. This report has been prepared solely for the benefit of the client listed above. No liability is accepted
by the Centre for Water and Waste Technology with respect to the use of this report by third parties
without prior written approval.
Copyright:
The Centre for Water and Waste Technology, Scion and AgResearch
2 CARBON FOOTPRINT METHODOLOGY REPORT

Table of Contents
1 Executive Summary 6
CONTENTS
CONTENTS
CONTENTS 2 Overview of the project 8
CONTENTS
CONTENTS 3 Methodology for calculating the carbon footprint 9
3.1 Standards and references 9
3.1.1 Existing ISO Standards related to carbon footprint assessment
3.1.2 Carbon Trust BSI DEFRA initiative
3.1.3 IPCC guidelines
3.2 General overview 11
3.3 life cycle assessment calculation procedure 12
3.4 Carbon footprint calculation 12
3.5 Methodology for the farm 12
3.5.1 Overview of published studies on carbon footprint and life cycle
assessment for dairy farm systems to the farm-gate stage
3.5.2 Goal and scope: functional unit, system boundaries, allocation rules
3.5.3 Inventory data
3.5.4 Carbon footprint calculation
3.6 Methodology for processing 21
3.6.1 Goal
3.6.2 Scope
3.6.3 Functional unit
3.6.4 System boundary
3.6.5 Life cycle inventory data
3.6.6 Allocation procedure
3.7 Methodology for distribution 27
3.7.1 Goal
3.7.2 Scope
3.7.4 System boundaries
4 References 31
FOR FONTERRA CO-OPERATIVE GROUP 2009 3

Index to Figures
Figure 1: Life cycle assessment framework as described in ISO standards 14040 and 14044 9
CONTENTS
CONTENTS Figure 2: System boundaries of the carbon footprint calculation with mass and energy flows 11
CONTENTS
CONTENTS
CONTENTS Figure 3. Simplified flowchart of the grass-to-farm-gate NZ production system 14
Figure 4: Flow chart of "grass-t-farm-gate" life cycle for the range of milk suppliers 15
Figure 5. System boundaries of product carbon footprints 22
Figure 6: System boundaries of distribution carbon footprints 28
index to Tables
Table 1. Districts included in each regionally-adjusted category for this study. 17
Table 2. Technical description of Northland, Waikato, Bay of Plenty, Taranaki,

Lower North Island, Westland, Marlborough + Canterbury, Otago + Southland
and weighted NZ average dairy farms (year 2004/2005). 18
Table 7. Key emission factors and parameters according to the IPCC-based methodology 20
Table 8: Industry-specific physico-chemical allocation matrix 25

Abbreviations
A M F
AMF
BSi
CF
Anhydrous milkfat
British Standards Institution
Carbon footprint
CO2-eq Carbon dioxide equivalents
CWWT Centre for Water and Waste Technology
Defra Department for Environment, Food and Rural Affairs
DM Dry matter
EEIO Environmentally extended inputoutput
FU Functional unit
GHG Greenhouse gas
GWP Global Warming Potential
IPCC Intergovernmental Panel on Climate Change
LCA Life Cycle Assessment
LCI Life Cycle Inventory
LCIA Life Cycle Impact Assessment
LDPE Low density polyethylene
LLDPE Linear low density polyethylene
LUC Land use change
MP Milk powder
MPC Milk protein concentrate
NZ New Zealand
PAS Publicly Available Specification
TEU 20 ft equivalent unit (shipping container)
WPC whey protein concentrate

1 Executive Summary
Fonterra has demonstrated a proactive stance on environmental management by commissioning independent

analysis of the carbon footprint of five key products throughout their value chain. Carbon footprint is a
more recent term for global warming potential and refers to the total greenhouse gas emissions associated
with a product or service. Emissions of different individual greenhouses gases are converted into global
warming potential and expressed in the common unit of CO2-equivalents.
The global warming potential reflects the atmospheric heat absorption capacity of individual greenhouse
gases, most importantly methane (CH4), nitrous oxide (N2O) and carbon dioxide (CO2) over a 100 year time
horizon.
A consortium of three independent research organisations (The University of New South Wales, AgResearch
and SCION) were commissioned to calculate the carbon footprint (i.e. perform a greenhouse gas life cycle
assessment) for butter, milk powder (MP), milk protein concentrate (MPC), cheese and caseinate across
their complete supply chains. This comprises all processes within the system starting with the on-farm
and farm-related inputs where the raw milk is produced through to the finished product leaving the dairy
processing sites and being transported to overseas destinations. The farming operations were based on
comprehensive data from 2004/2005 and both processing and distribution data were obtained for the
2006/2007 financial year.
A robust and globally accepted methodology had to be applied that represents current worlds best practice
and science in this area. All three research organisations are working at the forefront of methodological
developments in environmental system analysis, particularly life cycle assessment. The University of New
South Wales and SCION participate in several national and international processes for methodological
standardisation of assessment methods. AgResearch coordinates a NZ review group and contributes to the
Agricultural Working Group of the UK-based Carbon Trust methodology.
Several standards serve as guidelines on how to perform a carbon footprint as part of a life cycle assessment.
ISO 14040 and ISO 14044 (2006) are key standards which describe how the carbon footprint can be
quantified at a product or service level. The recently published PAS 2050 methodology, an initiative from
Carbon Trust, DEFRA and BSI in the UK, is aimed at industry users who are interested in a more consistent
method for assessing the carbon footprint of products. The main goal of PAS 2050 is to provide a
common basis for the comparison and communication of results. PAS 2050 builds on previously published
standards but specifies certain technical requirements with regards to the carbon footprint measurement.
The methodology used in the project undertaken for Fonterra is broadly consistent with PAS 2050 carbon
footprint assessment specifications published in October 2008.
As specified in PAS 2050, the calculation of the carbon footprint of Fonterras dairy products includes
greenhouse gas emissions throughout the supply chain arising from the following sources:
Emissions resulting from processes, including chemical reactions, livestock and other agricultural
processes, refrigerant loss and other emissions sources,
Consumption of energy that has greenhouse gas emissions associated with it,
Consumption of energy carriers that were themselves created using processes that have greenhouse
emissions associated with them (e.g. electricity), and
Wastes that produce greenhouse gas emissions.
Indirect greenhouse gas emission offset mechanisms are not included at any point in the supply chain. In
cases where Fonterra uses renewable energy directly, the benefits are included in the calculations.
The carbon footprint was calculated using the most recent figures as published by the Intergovernmental
Panel for Climate Change in kg CO2-equivalent terms, i.e. with multiplication factors of CO2 1, N2O 298,
CH4 25.
As in many other life cycle assessment studies, allocation procedures play a critical role in determining the
final carbon footprint result. Allocation refers to the rules applied to partitioning the inputs and outputs of
a process or product between the product under study and other products.
For the farming operations a physical allocation approach based on biological causality was adopted. This
method was used to define the physiological feed requirements of the animal to produce milk and meat,
and specific inputs and emissions were then calculated for the milk and meat co-products (consistent with
the highest priority method in PAS 2050). This was calculated to give an allocation ratio for milk and meat
equivalent of 86%:14%.

1 Executive Summary (continued)
With regards to the manufacturing processes, a physico-chemical allocation matrix was used. This type
of allocation matrix was developed specifically for the dairy industry in Australia and reflects the actual
relationships in the processing of dairy products at a high accuracy. This approach is consistent with the
current PAS 2050 specification as well as with other standards in carbon footprint and life cycle assessment
methodology.
With respect to the distribution of dairy products from the factory to overseas destinations, no allocation
procedures were necessary since all emissions could be directly linked to the five individual products.
At the grass-to-farm-gate stage, the system boundary covered the extraction, transportation and use of all
raw materials associated with the dairy farm and with land used to grow dairy replacement animals and
supplementary feed sources. Emissions of methane and nitrous oxide (predominantly from dairy cows and
replacement animals) on- and off-farm were calculated using the IPCC-based NZ greenhouse gas Inventory
methodology. At the farm-gate-to-market-port stage, the system boundary covered all material and energy
flows and their associated emissions required for manufacturing and transportation of the products to the
destination markets.

2 Overview of the project
Fonterra commissioned the Centre for Water and Waste Technology (CWWT) at the University of New South
Wales and AgResearch to measure the carbon footprint of a range of Fonterras dairy products. CWWT is
co-operating with SCION, a New Zealand Crown Research Institute.
The project measures Fonterras carbon emissions across its complete supply chain which consists of the
following distinct modules:
Module 1 - On-farm: covers inputs and outputs related to the production of milk from the farming
operation up until it leaves the on-farm milk vat,
Module 2 - Processing: includes the transportation of milk from the on-farm milk vat; the complete
manufacturing process including packaging and storage at the factory site; through to the product
loaded onto transport for delivery, and
Module 3 - Distribution: includes measuring the carbon emissions caused by the transportation of
the product from the manufacturing site, to the warehouse and its shipping to key destinations
internationally.
This report outlines the methodological approach applied in relation to Fonterras carbon footprint
measurement. Section 3.1 will present the different standards and references available, Section 3.2 will
give a general overview of the methodological issues followed by specific life cycle assessment and carbon
footprint calculation procedures (see Sections 3.3 and 3.4). Subsequently, the particular methodological
items referring to processes on the farm (see section 3.5), processing (see section 3.6) and distribution (see
section 3.7) will be described.

3 Methodology for calculating the carbon footprint
iso 3.1 Standards and references

3.1.1 Existing ISO Standards related to carbon footprint assessment
Several ISO standards relate to the assessment of a carbon footprint at different systems levels.
The carbon footprint can be quantified either at a product or service level as in life cycle assessment
described by ISO standards 14040 and 14044 (2006) or at an organisation or company level as
described in ISO standard 14064-1 (2006).
When assessed at the product level, the carbon footprint is one indicator amongst a basket of
indicators (e.g.: eutrophication; land use; toxicity and energy use) that are quantified as part of life
cycle assessment. Life cycle assessment consists of four steps as described in Figure 1: the definition
of the goal and scope of the study, the inventory analysis, the life cycle impact assessment (LCIA)
and the interpretation.
LIFE CYCLE ASSESSMENT

FRAMEWORK
GOAL AND
SCOPE
DEFINITION
DIRECT APPLICATIONS:
- Product development and improvement
- Strategic planning
INVENTORY
INTERPRETATION
ANALYSIS - Public policy making
- Marketing
- Other
IMPACT
ASSESSMENT
Figure 1: Life cycle assessment framework as described in ISO standards

14040 and 14044
In the goal and scope phase, the purpose of the study, its scope (geographic, temporal, and
technological), the studied function and corresponding system are defined. The level of detail and
data quality requirements of a life cycle assessment can vary significantly depending on its particular
goal and the intended use of the study. The scope phase should define the type of critical review
desired. Depending on the scope of the study, the critical review can be carried out by internal or
external experts or by a panel of interested parties.

In the inventory analysis the resources consumed and the emissions to the environment are quantified
at all stages of the life cycle of the product studied from the extraction of resources, through to
the production of materials, product parts and the product itself, and the use of the product, to its
o
reuse, recycling or final disposal (Guine et al., 2002).
is
For each environmental impact in the life cycle impact assessment stage, a characterisation model
is used to convert the inventory data contributing to this impact, into indicator results. This is done
by multiplying the emissions of each substance by a characterisation factor for each impact category
to which it may potentially contribute. Characterisation factors are substance-specific, quantitative
representations of the additional environmental pressure per unit emission of a substance. For a
carbon footprint, these characterisation factors are called global warming potentials and should be
based on IPCC guidelines.
The interpretation phase is the final phase of the life cycle assessment procedure where the results
are summarised and discussed as a basis for conclusions, recommendations and decision-making in
accordance with the goal and scope definition (ISO 14040, 2006).
The ISO standards for life cycle assessment do not actually describe a single method but allow
organisations some flexibility. Application of life cycle assessment methodology in general and
especially to food and farming systems is still the subject of on-going research and debate to achieve
more harmonised and accurate practice across all practitioners even if the data quality requirements
need to be adapted to the goal and scope of each study.
The international and national IPCC guidelines represent the most formalised and widely accepted
references for the actual quantification of the greenhouse gas emissions from a system. Existing
databases such as Eco-invent and life cycle assessment software tools such as GaBi or Simapro are
essential tools for the implementation and accuracy of life cycle assessment/carbon footprint studies.
IPCC guidelines will be more widely introduced later in this chapter and the actual assessment
method for the carbon footprint of dairy products will be described specifically for each module in
chapters 3.5, 3.6 and 3.7.
3.1.2 Carbon Trust BSI DEFRA initiative

As already mentioned the implementation of the life cycle assessment methodology is specific to
the goal and scope of each study and requires specific assumptions and data. The current initiative
from the Carbon Trust, DEFRA and BSI in the UK is targeted at applying life cycle assessment over
a wide range of products in a consistent manner for industry users, focusing only on the carbon
footprint indicator.
The goal and scope for this indicator has specific key assumptions (system boundaries, allocation
rules) and also requires some arbitrary assumptions and simplifications to get harmonised results
across all product categories. The Carbon Trust method PAS 2050, published October 2008, for
instance, distinguishes a business-to-consumer assessment and a business-to-business assessment.
The business-to-consumer assessment includes the emissions from the full life cycle of the product
(cradle-to-grave) whereas the business-to-business assessment includes the greenhouse gas
emissions released up to and including the point where the input arrives at a new organisation
(cradle-to-gate) (PAS 2050, 2008). The Carbon Trust method is therefore industry or production
focused and is therefore a relevant reference for Fonterra to analyse the carbon footprint of their
dairy products from cradle to its shipping to key destinations internationally. The methods used for
this project are broadly consistent with PAS 2050.
3.1.3 IPCC guidelines

The Intergovernmental Panel on Climate Change (IPCC) is responsible for assessing the science
related to climate change at a global level. Countries which ratified the Kyoto Protocol committed
themselves to individual, legally-binding targets to limit or reduce their greenhouse gas emissions.
The goal for all countries which ratified the Kyoto Protocol is to reduce aggregate emissions by at
least 5 percent below 1990 levels in the commitment period 2008 to 2012. New Zealands target
in this group of nations is 100 percent of the level in 1990. To achieve their targets, those countries
must put in place domestic policies and measures to address emissions, and are required to report
their greenhouse gas inventory annually at the country level.
IPCC guidelines provide methodologies for estimating national inventories of anthropogenic
emissions by sources and removals by sinks of greenhouse gases. Therefore the boundaries of the

studied system in those inventories correspond to the boundaries of the country and do not follow
a life cycle approach.
The 2006 IPCC guidelines generally provide advice on estimation methods at three levels of detail,
o
from tier 1 (the default method) to tier 3 (the most detailed method).
is
The advice consists of mathematical specification of the methods, information on emission factors
or other parameters to use in generating the estimates, and sources of activity data to estimate
the overall level of net emissions. The provision of different tiers enables focus of efforts on the
emissions and removals that contribute most significantly to the total greenhouse gas emissions
(IPCC, 2006). Although the purpose and system boundaries of IPCC guidelines are different, these
emission factors and methods to estimate greenhouse gas emissions are also relevant for life cycle
assessment studies.
IPCC guidelines provide default emission factors for greenhouse gas emissions from all sectors, it
also encourages the use of country-specific factors if the methods used to calculate them can be
defended and withstand international peer-review. Consequently, specific research has been carried
out in New Zealand and progressively integrated into IPCC-NZ guidelines for the main sources of
GHG emissions, being nitrous oxide emission from nitrogen excreted from grazing animals and
methane due to enteric fermentation (MfE, 2006).
3.2 General overview

In life cycle assessment methodology usually all inputs and outputs from the system are based
on the cradle-to-grave approach. This means that inputs into the system should be flows from
the environment without any transformation from humans and outputs should be discarded to
the environment without subsequent human transformation (ISO 14040, 2006). Each system
considers upstream processes with regard to the extraction of raw materials and the manufacturing
of products being used in the system and it considers downstream processes as well as all final
emissions to the environment.
In this study the system boundary encompasses all processes within the system starting with the
on-farm and farm-related inputs where the raw milk is produced through to the finished product
leaving the dairy processing sites and being transported to overseas destinations. This business-to-
business assessment is broadly consistent with the recently published Carbon Trust methodology
PAS 2050:2008 and is referred to as a cradle-to-gate assessment (PAS 2050, 2008). The system
boundaries and flow diagram of the entire life cycle of Fonterras dairy products selected for this
study are presented in Figure 2.
SYSTEM BOUNDARY Chemicals Ingredients Water Energy (fuel,themal,

electric energy)
to all processes within

Milk Milk the system
FARM TRANSPORTATION
PROCESSING WASTE
PLANT MANAGEMENT
PRODUCTION Packaging
WASTE OF PACKAGING
MANAGEMENT MATERIAL material
Product and Packaging material
Distribution in NZ
Product and Packaging material
Storage
Port handling and cooling
International
transport
(overseas ports)
Figure 2: System boundaries of the carbon footprint calculation with mass and energy flows

As described in PAS 2050, the calculation of the carbon footprint of Fonterras dairy products
includes emissions throughout the supply chain arising from the following sources:
Releases resulting from processes, including chemical reactions, livestock and other
o
agricultural processes, refrigerant loss and other emissions sources,
is
Consumption of energy that has greenhouse emissions associated with it,
Consumption of energy carriers that were themselves created using processes that have
greenhouse gas emissions associated with them (e.g. electricity), and
The detail of those emissions and their calculation is given for each module specifically.
As required by the PAS 2050, indirect greenhouse emission offset mechanisms are not included
at any point in the supply chain. Where Fonterra uses renewable energy directly, the benefits are
included in the calculations.
3.3 Life cycle assessment calculation procedure

In principle, three life cycle assessment calculation procedures can be applied for calculating the
carbon footprint of a product: 1) process-based life cycle assessment approach, 2) an input-output-
based life cycle assessment approach and 3) hybrid approach combining process analysis with input-
output analysis (see Suh et al, 2004 for details).
The first approach can be described as accurate on a process level, but it has truncation errors as it
does not take into account all processes of the supply chain of a product.
The second approach overcomes the problem associated with the truncation error by taking into
account the entire supply chain, but it lacks of precision at a process level. The hybrid approach
ideally combines the advantages of both approaches, however, such a tool is currently not available
to be used in a commercial environment in New Zealand.
Therefore, a choice needed to be made between a process - or input-output-based approach. The
project team decided to take a very detailed process-based approach because it allows quantification
of the carbon footprint of processes at a very high level of precision. PAS 2050 is also in line with
this approach. This is of overarching importance because some of the processes, such as emissions
from feedstock, may have a significant contribution to Fonterras carbon footprint. Accuracy at the
process level outranks by far the advantage of the input-output approach (i.e. covering the complete
supply chain).
3.4 Carbon footprint calculation

After discussion with all partners of this project, the carbon footprint (equivalent to global warming
potential) for a 100 year time horizon (GWP100) was calculated according to the most recent IPCC
reference in kg CO2-equivalent, i.e. with multiplication factors of CO2 1, N2O 298, CH4 25. Global
warming potential corresponds to the impact of emissions on the heat radiation absorption of the
atmosphere.
3.5 Methodology for the farm

3.5.1 Overview of published studies on carbon footprint and life cycle
assessment for dairy farm systems to the farm-gate stage
In the published carbon footprint and life cycle assessment studies on dairy systems, the goal and
scope of the studies are properly defined as well as the studied system and system boundaries. The
allocation rules used at the farm gate level between milk and meat are diverse and include:
no allocation at all (Haas et al., 2001),
allocation based on biological causality (feed requirement for each function) (Cederberg and
Mattsson, 2000; Williams et al., 2006; Basset-Mens et al., 2008),
economic allocation, which is the most used across all studies, and is based on the economic
returns for the different products (Cederberg and Flysj, 2004; Berlin, 2002; Eide, 2002;
Hospido et al., 2003 & 2004; Casey and Holden, 2005; Thomassen et al., 2007).
The design of the studied farm systems and their technical data are based on a range of approaches
from the survey of a sample of farms (studies have used between 1 and 11 farms per production
type) to the use of national statistics and database. Some studies have compared different types
of production such as organic versus conventional or conventional intensive versus conventional
extensified. Conversely, in other studies, a representative weighted average of all types of production
at the national level is calculated (Williams et al. 2006).

In published studies, details of the inventory of greenhouse gases are rarely fully described and a
range of approaches can be found. Some studies applied the methods used in their national GHG
inventories (Williams et al. 2006), while other studies used a mix of specific references on emissions
o
factors for dairy farm systems in their countries and IPCC guidelines (Cederberg and Flysj, 2004;
s
Casey and Holden 2005; Thomassen et al., 2007).
i
Some studies apparently used quite simplified inventory covering only the key aspects such as
methane emissions from cows (Hospido et al., 2003 and 2004). Williams et al. (2006) and
Thomassen et al. (2007) appear to have carried out an exhaustive inventory but it is difficult to
check the completeness of the GHG inventory for most studies since not all components of the
GHG inventory were mentioned. For instance, in Cederberg and Flysj (2004), the direct nitrous
oxide emission from N fertiliser and the indirect nitrous oxide emission due to leaching were not
mentioned. Casey and Holden (2005) did not mention indirect nitrous oxide emissions and the
methods used by Haas et al (2001) and Eide (2002) to estimate methane and nitrous oxide emissions
were not described.
The level of discrepancy in the methods used and the degree of completeness of the published
greenhouse gas inventories is uncertain. A more in-depth study with all interested parties is required
to harmonise as much as possible the methods used and guarantee more accurate comparisons.
3.5.2 Goal and scope: functional unit, system boundaries, allocation rules
The goal of module 1 is to provide Fonterra with an assessment of the grass-to-farm-gate carbon
footprint for a range of NZ dairy farm systems. This goal is motivated by Fonterras commitment to
sustainability and the increased global awareness of the detrimental effect of greenhouse gases on
the environment.
The scope of the current study covers estimation of the carbon footprint for a significant range of
dairy farms supplying Fonterra depending on key regional differences across New Zealand.
A wide range of milk suppliers were analysed in this study for the 2004-2005 season, using
information mainly based on the DairyNZs1 ProfitWatch survey of farms and LIC2 (2005) statistics.
The farm types included:
All regional average dairy farms across New Zealand,
An NZ average dairy farm, designed by weighting (on an on-farm ha basis) regional average
data.
The functional unit of the study is one kilogram of milksolids (milkfat plus protein).3
The system boundaries (see Figures 3 and 4) were set up from grass-to-farm-gate, which included:
Production of milk on-farm including on-farm pasture production (methane and nitrous oxide
from animals), cow management (diesel, petrol) and milk extraction, farm dairy effluent
management and water supply (electricity),
Production of supplementary feed,
Off-farm pasture production for the dairy cow replacements and any cows grazed off over
winter, and
Production and delivery of crop and pasture inputs.
1. New Zealand dairy research and extension organisation.

2. LIC Livestock Improvement Corporation which provides farm management information, herd testing, artificial breeding
services, DNA analysis, farm advisory extension service, research to improve farm productivity.
3. Milksolids is milksolid that contains milk fat plus protein. Farmers get paid on the basis of milksolids supplied. Milk solids
(two words as opposed to one) is milk protein, plus fat and other elements in milk.

Input Off-farm level Farm level
production
o
Off-farm
Herd management
s
grazing
i
heffers
EXTRACTION
AND Fertilizers Meat
DELIVERY
OF RAW Electricity
MATERIALS Pasture production
AND ENERGY
RESOURCES Diesel Milk
Extraction 1 kg MS
Off-farm
maize
Farm dairy effluent
silage
Figure 3. Simplified flowchart of the grass-to-farm-gate NZ production system
This should account for at least 99% of the likely life cycle emissions from grass-to-farm-gate thereby
meeting one of the key requirements of the PAS 2050.
Capital was excluded in calculations, as specified in the PAS 2050 (2008). In keeping with
specifications in the PAS 2050 (2008), land use change was included by accounting for any changes
in land use from January 1990 to 2005 (the year of data used in the study). This used relevant land
use change factors from the NZ Greenhouse Gas Inventory.
Impacts were allocated between the co-products milk and meat according to a biological causality,
which was based on the physiological feed requirements of the animal to produce milk and meat
(Basset-Mens et al., 2008). This accounted for all specific inputs and greenhouse gas emissions
associated with the animal growth phase (from birth to mature live-weight) within the whole-farm
system, which were allocated to meat production. Thus, it met the first priority for allocation in
the PAS 2050 (2008). It equated to a relative allocation between milk and meat of 86%:14%. This
methodology is consistent with the highest priority method in PAS 2050 (2008) and was preferred
over economic allocation which can vary over time due to changes in prices for milk and meat
products.

EXTRACTION OF RAW MATERIALS
P rock Sulphur KCI Natural gas Crude Oil Natural
o
(North Africa) (Canada) (Canada) (NZ/Indonesia) (overseas) gas/coal
s
(NZ)
i Road
& sea
transport
SSP
(NZ)
Fertiliser manufacturing
KCI
Urea
(NZ/Indonesia)
Road
& sea
transport
Road Electricity
Lime Petrol &
& sea
(NZ) Diesel NZ
transport
FEED SUPPLEMENTS FROM NZ
Maize Pasture Barley grain Road

silage silage (Sc. 6-8) transport
(Sc. 1-5) (all sc.)
Road, rail
Road
& sea
transport
transport
FARM
Farm dairy
effluent
Pasture Cow Milk

production management extraction
Road Milksolids
transport
{
Wintering-off Off-farm Meat

Pasture (all scenarios grazing of
except sc. 8) replacements
Swedes/kale (sc. 8) (all scenarios)
Scenarios legend:
Sc. 1: Northland Sc. 5: Lower North Island
Sc. 2: Waikato Sc. 6: Westland
Sc. 3: Bay of Plenty Sc. 7: Marlborough + Canterbury
Sc. 4: Taranaki Sc. 8: Otago, Southland
Figure 4. Flow chart of grass-to-farm-gate life cycle for the range of milk
suppliers.
3.5.3 Inventory data
3.5.3.1 Regional average and New Zealand weighted average dairy farms
LIC statistics (LIC, 2005) and regional ProfitWatch data provided by DairyNZ were
o
the two main sources of data used to design an exhaustive and non-overlapping
is
range of regional scenarios for the season 2004-2005. The data was aggregated
in order to match LIC regional categories and ProfitWatch regional categories (Table
1). Cow numbers, milk production, milk quality and size of farms were derived
from LIC statistics while replacement rate, cow weight and key input data such as
fertiliser-nutrients and feed supplements were obtained from the regional ProfitWatch
database.
The dry matter intake by animals was estimated by using the NZ inventory model of
Clark et al. (2003) for both dairy cow and replacement animals. Technical data for the
eight regional NZ dairy farms and the weighted NZ average dairy farm are presented
in Table 2.
For each region, specific data on wintering-off, stock replacement and brought-
in feed was collected. The practices of farmers in each region and distances of
transportation for animals grazed-off farm were based on a survey of consulting
officers from DairyNZ and experts from PGG Wrightson (PGGW). Data on the
distances of fertiliser transportation for each region were provided by experts from
fertiliser companies (Ballance and Ravensdown). Transportation distances for
brought-in-feed supplements were obtained from experts from Pioneer (maize
marketing company), DairyNZ and PGGW. Distances for transport of animals grazed-
off farm were collected by surveying different experts from the consulting officers of
DairyNZ and PGGW.
The fuel consumption for all agricultural components including cow management,
pasture production, supplementary feed production and delivery, was calculated
from the analysis of all single operations needed specifically for each scenario and
parameterised in the life cycle assessment model (SIMAPRO). Electricity consumption
was calculated as a function of cow numbers based on an NZ study by Sims et al.
(2005) and as a function of irrigation based on a summary of types of irrigation
systems, mm irrigation water applied and typical depth of pumping.
New Zealand Regions (source Ministry of the Environment)
NORTH ISLAND
Northland
Auckland
Waikato
Bay of Plenty
Gisborne
Hawkes Bay
Taranaki
Manawatu/Wanganui
Wellington
SOUTH ISLAND
Nelson
Marlborough
Tasman
West Coast
Canterbury
Otago
Southland

is
Table 1. Districts included in each regionally-adjusted category for this study.
Districts included
Northland
Far North
Whangarei
Kaipara
Rodney
Waikato
Manukau
Papakura
Franklin
Sth Auckland
Bay of Plenty
Bay of Plenty
Central Plateau
East Coast
Taranaki
Taranaki
Lower North
Island
Hawkes Bay
Wellington
Wairarapa
Westland
West Coast
Tasman
o
Marlborough +
Canterbury
Marlborough
Kaikoura
Nth Canterbury
Sth Canterbury
Otago +
Southland
Dunedin
Clutha
Southland
Western Uplands Waitaki
FOR FONTERRA CO-OPERATIVE GROUP 2009

17
18
is
Table 2. Technical description of Northland, Waikato, Bay of Plenty, Taranaki, Lower North Island, Westland, Marlborough + Canterbury,
o
Otago + Southland and weighted NZ average dairy farms (year 2004/2005).
All per on-farm ha unless Northland Waikato Bay of Taranaki Lower North Westland Marlborough Otago + Weighted
specified Plenty Island + Canterbury Southland NZ average
On-farm area, ha 114 95 130 87 119 143 188 171 123
CARBON FOOTPRINT METHODOLOGY REPORT

Cow weight, kg 449 455 461 440 456 417 465 487 456
Dairy cows 2.20 2.96 2.74 2.82 2.71 2.28 3.05 2.69 2.76
Replacement rate 0.22 0.21 0.22 0.20 0.24 0.23 0.25 0.27 0.23
Kg milksolids 581 904 864 870 852 695 976 939 856
Butterfat (% mass/mass) 4.68 4.77 4.64 5.12 4.76 4.97 4.66 4.66 4.78
Protein (% mass/mass) 3.49 3.54 3.48 3.73 3.54 3.63 3.56 3.59 3.56
Kg fertiliser N 83 117 127 85 91 105 149 104 109
Kg fertiliser P 37 48 51 52 44 37 36 49 45
Kg DM maize silage* 254 546 544 439 630 0 0 0 426**
Kg DM pasture silage 206 252 244 334 898 503 1012 1430 544
Kg DM barley grain 0 0 0 0 0 147 439 1627 0**
*: for the first five regional scenarios, all feed supplements except pasture silage were assumed to be maize silage while for the three last regional scenarios they were assumed to be
barley grain.
**: Since the first five regions represent about 70% of the total dairy area and production in NZ, it was assumed for the weighted NZ average that all feed supplements (other than
pasture silage) were maize silage.
3.5.3.2 Supplementary feed production, off-farm grazing of replacements and
wintering-off
For all scenarios, an average beef farm was assumed to be used for grazing
o
replacements, based on the MAF Intensive Beef Monitor Farm (Ledgard et al. 2003).
is
The pasture silage was also assumed to come from the beef farm where replacement
animals were grazed. For scenarios using maize silage, it was assumed to be produced
on a typical forage cropping block off the farm. The technical data for maize silage
production was provided by the NZ branch of an international seed company (Ian
Williams, Pioneer, pers. comm.). Current practices were assumed for the use of inputs
for the production of pastures and silages (Basset-Mens et al. 2008). Barley grain was
assumed to be similar to triticale grain and thus a recent life cycle inventory for triticale
grain in the South Island was used. For all scenarios (except Otago + Southland),
wintering-off of dairy cows was assumed to be on pasture while for the Otago +
Southland scenario, cows were assumed to be wintered-off on a brassica (swedes +
kale) crop. Data for pasture land used for wintering off dairy cows was unavailable and
therefore conservative information was used which was similar to that for the average
NZ dairy farm (in practice this is likely to overestimate emissions).
3.5.3.3 Inventory-based greenhouse gas emissions from agricultural land

The inventory of greenhouse gas emissions covering methane (CH4) from enteric
fermentation by cows, CH4 and nitrous oxide (N2O) from excreta deposited on pasture
and from farm dairy effluent, and CO2 emissions from lime and urea application was
based on IPCC (2006) and specific IPCC-NZ methodology (de Klein et al. 2001; Clark
2001; Saggar et al. 2003, MfE 2007).
3.5.3.3.1 Intake model
The animal feed intake model used in estimating CH4 and N2O emissions was
that used in the Tier 2 approach of the NZ IPCC inventory (Clark et al. 2003).
It is a comprehensive model that operates at a monthly time step and utilises
data on livestock numbers, livestock performance and diet quality. Within the
dairy category, the model subdivides population into animal sub-categories
such as dairy cows in milk, heifers from 0 to 1 year old and heifers from 1
to 2 years old. Dry matter intake was estimated by calculating the energy
required to meet the assumed levels of performance (MJ metabolisable energy
(ME) per day) and dividing this value by the energy concentration of the diet
consumed (MJ ME per kg dry matter). For dairy cattle, energy requirements
were calculated using the algorithms presented in the publication Feeding
standards for Australian livestock: Ruminants (CSIRO, 1990). The
intake model from the NZ inventory described above was used for each
scenario by entering data on milk production, milk composition and cow
liveweight and by adjusting the feed quality (metabolisable energy, digestibility
and N concentrations), on a monthly basis to account for all feed supplements
used in addition to pasture. The monthly rates of supplementary feeding were
based on the teams expertise.
3.5.3.3.2 N excreted
The N excreted by animals used the NZ inventory methodology, which was
based on principles in the OVERSEER model (Wheeler et al. 2003). Dry
matter intake was based on the method outlined in the previous section. This
was then multiplied by the average NZ pasture N concentration (from a review
of data for the NZ inventory) to calculate N intake. For pasture silage, maize
silage and barley grain, N concentrations were obtained from the OVERSEER
nutrient budget model (Wheeler et al. 2003) and are based on the mean of
samples submitted to a major NZ laboratory, FeedTech. The N in milk or meat
products (based on the NZ inventory) was subtracted from total N intake in
order to calculate the amount of N excreted. In the IPCC-NZ inventory, 95% of
the N excreted by milking cows is assumed to be applied onto pastures during
grazing and 5% to be stored in an anaerobic lagoon before being reapplied
onto pastures. The same was assumed in this study for all scenarios.

3.5.3.3.3 Methane emissions
Methane emissions from enteric fermentation were calculated from the
product of energy and dry matter intake by animals using the NZ Inventory
o
model and the IPCC-NZ emission factors (Table 3). Methane emissions due
is
to waste management were calculated by multiplying faecal dry matter
(1- digestibility of feed) by specific emission factors according to Saggar et al.
(2003) and MfE (2007) for faecal dry matter deposited on pastures and for
effluent stored in an anaerobic lagoon respectively.
3.5.3.3.4 Nitrous oxide emissions
Direct N2O emissions were calculated by multiplying N inputs by IPCC (2006)
and NZ-specific emission factors corresponding to the fraction emitted to
the atmosphere as N2O (de Klein et al. 2001; Table 3). In particular, the
emission factor for N2O from excreta-N on grazed pasture is 1% in the NZ
context (MfE 2007) compared with a default emission factor of 2% for the
IPCC methodology (IPCC 2006). Considerable research effort has gone into
establishing a country-specific value for this emission factor (de Klein et al.
2003; de Klein et al. 2004; Kelliher et al. 2005). Indirect N2O emissions were
calculated using the IPCC (2006) and IPCC-NZ N source and emission factors,
with the NZ-specific factors developed from research and reviews carried out
by scientists in our project team.
Key emission factors and parameters from IPCC (2006) and the IPCC-NZ
inventory are summarised in Table 3.
3.5.3.3.5 CO2 emissions from lime and urea application
Direct CO2 emissions from soils due to lime and urea application were
calculated according to the default IPCC emission factor (IPCC 2006). The
CO2 absorbed by plants was not taken into account since it is in equilibrium
with losses from the grazing cycle and plant respiration.
Table 3. Key emission factors and parameters according to the IPCC-based methodology
Process Mean References for mean

kg of CH4 emitted per tonne of DM 21.6 Clark (2001)
ingested
kg of N2O-N emitted per kg of N:
Due to N fertiliser use (EF1) 0.01 IPCC (2006)
Due to excreta deposited during grazing 0.01 De Klein et al. (2001) & Kelliher et al. (2005)
(EF3)
Due to excreta stored in an anaerobic 0.001 IPCC (2006)
lagoon (EF3 AL)
Due to atmospheric deposition of NH3-N 0.01 IPCC (2006)
(EF4)
Due to leaching and runoff of NO3-N 0.0075 IPCC (2006)
(EF5)
kg of NO3-N emitted per kg of N excreted 0.07 Thomas et al. (2005)
or N fertiliser applied (FRACLEACH)
kg of NH3-N (and NOx-N) emitted per kg 0.20 IPCC (2006)
of N excreted (FRACGASM)
kg of NH3-N emitted per kg of N fertiliser 0.10 IPCC (2006)
applied (FRACGASF)

3.5.4 Carbon footprint calculation
The same calculation methods for total greenhouse gas emissions (i.e. the carbon footprint) were
used across all partners of the project for the production and delivery of NZ diesel, petrol and
o
electricity. These were provided by SCION. The carbon footprint calculations for the manufacturing
is
and delivery of fertilisers were based on the Eco-Invent database (Frischknecht et al. 2005) adjusted
as far as possible to the NZ situation in terms of distances and technology used.
A brief summary was done of the use of refrigerants (mainly associated with vats for chilling
milk on farm prior to collection) after discussion with a local expert (D. Grey pers. comm.). The
estimate of emissions associated with HFCs and carbon footprint CFCs equated to roughly 0.012
kg CO2-equivalent/kg milksolids. This represented about 0.1% of the total carbon footprint
and was incorporated in the carbon footprint estimates in keeping with the goal of including all
contributors.
The Inventory approach, system boundaries, and sources of GHG emissions were peer-reviewed
by experts from overseas (Dr David Chadwick from UK for CH4 and N2O specifics and Dr Christel
Cederberg for overall life cycle assessment methodology).
3.6 Methodology for processing

Greenhouse gas emissions that are included in the carbon footprint calculations are determined
with reference to Life Cycle Assessment techniques. In this Section relevant methodological issues
are described in order to make the carbon footprint calculation procedure of the processing phase
(module 2) transparent.
3.6.1 Goal
The goal of Module 2 is to provide Fonterra with an assessment of the farm-to-factory- gate carbon
footprint for 5 products which are predominantly sent to overseas markets.
This goal is motivated by Fonterras commitment to sustainability and the increased global awareness
of the detrimental effect of greenhouse gases on the environment.
3.6.2 Scope
All processes are included in this module that cover the transportation of the milk from the on-farm
milk vat, the complete manufacturing process,(including packaging and storage at the factory site),
to the final product loaded on distribution transport for delivery. All processes take place in New
Zealand.
The 5 products included in the scope of this study are:
butter,
milk powder,
milk protein concentrate,
cheese
caseinate.

The functional unit (FU) for each product is 1 tonne product leaving the factory gate.
The functional unit describes the main function fulfilled by a product system and is the basis for all
calculations.
3.6.4 System boundary

In this study the system boundary encompasses all processes within the system starting with
the transport of the raw milk to the processing sites and inter-factory intermediate product
transport,
multi-product manufacturing at 22 production and 2 transfer sites under consideration of all
products even if they are not part of the analysis,
production, delivery and consumption of operating materials, i.e. chemicals, packaging
materials, ingredients,
electrical and thermal energy production on-site and off-site, and

freshwater delivery and wastewater treatment.
The system boundaries and flow diagram of the entire life cycle of Fonterras dairy products selected
for this study are presented in Figure 5.
iso
SYSTEM
BOUNDARY Chemicals Ingredients Water
Energy
(fuel, thermal,
electrical energy)
To all processes
Milk Milk within the system
Farm Transportion
Processing plant Waste

management
Pack.
Waste Production of material
management packaging
material
Prod. + Pack. material
Distribution in NZ
Prod. + Pack. material
Storage
Port handling + cooling
SYSTEM BOUNDARY International transport

FOR PROCESSING (overseas ports)
Figure 5. System boundaries of product carbon footprints
As described in PAS 2050 (2008), the calculation of the carbon footprint of Fonterras dairy products
includes emissions throughout the supply chain arising from the following sources:
Releases resulting from processes, including chemical and ingredients production, refrigerant
manufacturing and losses and other emissions sources,
Usage of energy that has greenhouse gas emissions associated with it,
greenhouse has emissions associated with them (e.g. electricity), and
As also required by PAS 2050, indirect greenhouse gas emission offset mechanisms are not included
at any point in the supply chain. Where Fonterra uses renewable energy directly, the benefits are
included in the calculations.
The carbon footprint calculation excludes greenhouse gas emissions from the manufacture and
ongoing maintenance of capital goods, such as plant machinery, transport equipment, electricity
generating plant, etc., used in the manufacture of the product; and transport of employees to their
normal place of work (PAS 2050 (2007), p 8).

Life cycle inventory data were collected in close cooperation with Fonterra. At the beginning of
the project a detailed data collection questionnaire was sent to Fonterra staff. All data the project
CWWT / SCION team requested was provided by Fonterra and additional data sources were
considered if no company specific information was available. Secondary data are used for so-called
background processes, such as refinery of crude oil which is processed in diesel and electricity
generation. Sources for secondary data from internationally agreed datasets are contained in the
life cycle assessment software tool GaBi. Additionally, data from peer reviewed journal articles are
used.

CWWT/SCION staff visited several sites in order to get an impression of the manufacturing practices
and processing technologies in use.
All data was checked by the project team and an independent reviewer with regard to consistency
o
and completeness. The data was then incorporated into an Excel-spreadsheet that built the basis
is
for the carbon footprint calculation in life cycle assessment-software tool GaBi.
All processes are included in this module starting from the transportation of the milk from the
on-farm milk vat, the complete manufacturing process, including packaging and storage at the
factory site and the loading of the final product on distribution transport for delivery. All processes
for the production of the dairy products take place in New Zealand.
The life cycle inventory data, used for calculating the carbon footprint of the processing phase,
contains information on raw milk or other intermediate products processed at 22 manufacturing
sites, inter-factory intermediate product transfer (i.e. a total of 46 intermediate products), all
products manufactured (all types of fat products, milk powders, cheese, caseinate based products,
whey products, concentrated milk), electricity and thermal energy consumption figures, ingredients
and chemicals usage, water consumption figures and refrigerant usage, packaging materials and
waste management. Transport effort is taken into account of raw milk from the farms to the sites,
chemicals and packaging materials including truck size and load efficiencies as well as transport
impacts.
Data used in the carbon footprint measurement for processing are described in more detail below:
Transport of raw milk from the farm to the manufacturing sites is taken into account by
reporting the quantity of diesel required for transportation. Two types of information are
used for calculating transport effort: 1) diesel fuel consumption figures of the transport fleet
as well as 2) transport distances multiplied by the mass transported in order to calculate the
tonne-kilometres (tkm) associated with delivery (Fonterra, 2008). These figures take into
account the type and size of the truck as well as the average load factor (see Mueller & Baas,
2004).
There are large amounts of different intermediate products transported between manufacturing
sites. The logistics between the sites is very complex, since most sites both receive and send
various amounts of different intermediate products. These inter-site intermediate product
transfers have been fully considered since only through these additional milk solids flows
between the sites the actual production of dairy products is ensured.
Final annual production figures of all Fonterra dairy products are included, i.e. fat products
(butter, anhydrous milk fat, fat blend), milk powders (skim, whole and butter milk, nutritional
and speciality powders, MPC42, 56, 70 and 85), cheese (dry and brine salt, mozzarella and
cream cheese types), casein based products (caseinate and casein) and whey products (whey
powder, WPC, whey fractions, lactose, lactalbumin, alamin and ethanol). All dairy products
need to be taken into account for the calculation of the 5 dairy products under study.
In addition, chemical consumption figures are considered for each plant: In total 28 different
types of chemicals are used on-site out of which 4 chemicals (caustic soda, nitric acid, triplex,
sodium hypochlorite) contribute 961% in mass to the total chemical usage. Therefore, only
these 4 chemicals are included in the carbon footprint calculation. Some of these chemicals
are directly allocated to specific products, e.g. caustic and nitric acid to butter production
(Fonterra, 2008). Other chemical consumption figures, e.g. chemicals for raw milk treatment,
can not be directly attributed to one product. These chemicals are allocated according to
the procedure described in Chapter 3.6.6 based on classification as acidic or basic chemicals.
Transport of chemicals is included in the analysis if distances were provided by Fonterra.
Beside the intermediate products, which are fully considered, more than one hundred
different ingredients go into the production of Fonterra dairy products. In most cases, the
amounts of ingredients are rather small and can be neglected. Salt and palm oil are the most
important ingredients and were included in the study. The transport effort has been included
in the analysis if distances were provided (Fonterra, 2008).
Company data are provided for electricity and thermal energy (black coal, natural gas, oil,
LPG and biogas). For all energy carriers LCI data are taken from the GaBi database except
for LPG. LPG is substituted with natural gas.
There are several sites which have on-site co-generation. For these plants detailed information
was provided to the UNSW team with regard to primary energy inputs and electricity and
thermal energy outputs. Primary energy inputs and emissions are allocated to electricity and
steam following the guidelines of WRI WBCSD (2006).

These cogeneration sites generate electricity surpluses which are fed into the New Zealand grid. In
these cases the amount of electricity and its associated emissions are subtracted from the primary
energy inputs of the co-generation plants. The remaining primary energy inputs and emissions are
o
allocated to the products manufactured.
is
The remaining manufacturing sites are using electricity from the NZ grid. On this account the specific
emission profile for the NZ grid mix are utilized in the model.
Electricity consumption of offsite stores are excluded from the analysis because it can not be
attributed to the specific products. This electricity demand equals ~1.5% of the total.
The quantity of packaging materials and their respective material compositions (predominantly
low density polyethylene and linear low density polyethylene) are included in this study.
Transport of packaging materials by road and rail is included if distance were provided by
Fonterra. Some of the packaging material can be allocated directly to dairy products. If that
is not the case, packaging materials are allocated according to Feitz et al. (2005, see also
Chapter 3.6.6).
Nitrogen and carbon dioxide are predominantly used in the packaging process for milk
powder and to a lesser extend for anhydrous milk fat. Hence, these gases were directly linked
to the production of these products.
Data on refrigerants are reported for several sites. The refrigerants are allocated on a mass
basis between butter and cheese (Somers, 2008, pers. comm.).
Water and wastewater quantities are allocated on a milk solids basis to the products.
Information on the wastewater treatment level attained at processing sites has been provided
by Fonterra. The electricity intensity of wastewater treatment depends on the level of
treatment that has been calculated based on Lundie et al. (2005).
Solids waste management and material recycling have been excluded from the analysis due
to 1) their small quantities and 2) difficulties of allocating the waste quantities to individual
products.
3.6.6 Allocation procedure

There are two principle approaches dealing with this multi-functional situation, i.e. system boundary
expansion or allocation methods. In this case, only allocation is described in detail as system
boundary expansion does not seem to be a practical solution for this project. This is because:
1) System boundary expansion generally introduces new multi-functional processes (Guine et
al, 2002); some sort of allocation is still needed in order to collect the necessary background
data.
2) Broadening the system boundaries makes the process of data collection much more
complicated; not only are more data needed, but appropriate data are needed (Curran,
2006).
3) Larger systems run the risk of being less transparent in that there is more detail on how data
were arrived at than can be conveyed conveniently (Curran, 2006).
See also Lundie et al (2007) for more details.
Below the three major allocation procedures are briefly described:
Allocation based on physical properties: Mass, molar flows, energy contents or volume are
physical properties which are used to allocate the inputs and outputs of the product / service
under study. This approach may, but most likely does not reflect the causal relationship (see
also Section 2.2.2. in Lundie et al (2007) for more details). Guine et al (2002) go even
further. They discredit this approach for a lack of justification (as do Huppes and Schneider
1994), as there is no causality involved, e.g. the mass of outputs cannot cause inputs by
physical causation. At best, allocation based on mass or energy can be used as a proxy for
allocation based on economic value (Guine et al 2004).
Physico-chemical allocation: One way of solving the allocation problem is to generate a
product or industry specific physico-chemical allocation matrix which reflects the actual
allocation of resources in a whole plant. Feitz et al (2007) developed systematically an
industry specific physico-chemical allocation matrix for the dairy industry that reflects closely
the causalities in the dairy manufacturing industry. It may be seen as an optimisation
procedure for producing a changed output within the production functions constraints and
the variation in production technologies in the system. In this sense, it uses physico-chemical
relations, as part of production functions. This approach has been developed at a sector
level, linking in to new developments in life cycle assessment.

Feitz et al (2007) recommend that such physico-chemical allocation matrices may be
developed for other industrial sectors which do have similar production processes; for
example, agriculture (e.g. the meat industry); construction (e.g. sand, gravel and other
o
construction materials); mining (e.g. gold and lead) and petrochemical industries (e.g.
s
automotive fuels; see Feitz et al (2007) for details how to calculate such a physico-chemical
i
allocation matrix).
Economic allocation: A systematic approach to allocation has been suggested by Guine et al
(2002). Guine et al (2002) recommend economic allocation as a baseline method for most
detailed life cycle assessment applications, because it seems the only generally applicable
method (Guine et al 2004). This avoids the problem that differences between alternatives
are caused by different allocation methods applied instead of being due to the underlying
reality. This position may seem to contradict the ISO 14044 recommendation that allocation
should preferentially be done on the basis of physical relationships.
There is no universal consensus on which type of allocation should be used for a particular
application. ISO 14044 (2006) and Lundie et al. (2007) recommend allocation based on physical-
properties and physico-chemical relationships. Guine et al. (2002) favour economic allocation, but
only if physico-chemical relationships cannot be established. PAS 2050:2008 recommends dividing
the unit processes to be allocated into two or more sub-processes, which could be done using a
physico-chemical allocation procedure (PAS 2050, 2008).
Fonterra staff and the project team decided to select physico-chemical allocation for two reasons: 1)
this type of allocation matrix was developed particularly for the dairy industry in Australia reflecting
the actual relationships in the processing of dairy products at a high accuracy, and 2) it comes closest
to the recommendations of ISO and Lundie et al. (2007).
Applying this to Fonterra Cooperative Group is very complex. There are 22 manufacturing sites that
produce two or more of the dairy products as listed above. There are only 6 mono-product plants
and there are substantial intersite-intermediate product transfers. Only the intermediate product
transfers ensure that the milk balance closes on a plant level.
Because of this complicated situation the inputs and outputs of all plants need to be allocated on
a plant level if they could not be attributed to individual products (direct attribution to products is
possible for some ingredients, chemicals and some of the packaging materials). Once the inputs and
outputs were allocated, weighted averages per product group were calculated.
The allocation of raw milk is important in dairy product life cycle assessments as this step allocates
upstream dairy farming processes (e.g. feed supplements, fertilisers, herbicides, water, etc.) to
different products. Using the approach developed by Feitz et al. (2005), the amount of raw milk
is assigned on a milk solids basis, which includes fat, protein, lactose and ash, and the degree of
milk solids concentration in the final product (see Table 1 in Feitz et al., 2005). In the course of this
project, NZ specific data became available so that the raw milk allocation to different products could
be done using NZ specific allocation factors. Since the raw milk transport efforts are directly related
to the raw milk content of the different products, these allocation factors are also specific to the
Fonterra situation. For the allocation of all other processes, the allocation factors as determined in
Feitz et al. (2005) were applied.

For example, process energy coefficients for a plant that only produced milk powders (the cream
was sent off site) was then used to update the butter process energy coefficients for a plant that
only produced butter and milk powders. Once an estimate for butter was obtained, this was used
o
to determine the AMF process energy coefficients in a plant that produced milk powders, butter
s
and AMF, and so on. By applying this procedure for 17 plants for many different products (using
i
the same level of technology), a very good estimate of the average process energy figures (and
resource efficiency coefficients) for different products could be obtained. The coefficients could be
further refined by using an approach similar to the residual allocation system (RAS) method, used
for optimising input coefficients in input-output tables (Stone, 1963; Bacharach, 1970; Parikh, 1979;
van der Linden & Dietzenbacher, 2000).
The allocation matrix given in Table 4 enables the allocation of the primary resources for any
combination of dairy products based on physico-chemical principles. By normalising the coefficients
to milk powder, for example, the relative resource use for each product is maintained and the
allocation matrix is capable of accounting for differences in overall plant efficiency (e.g. using less
efficient coal rather than natural gas for thermal process energy). Normalising the average resource
data also ensures company confidentiality. The percent allocation is determined by multiplying the
annual production of a product by its unique coefficient (or allocation factor; AF given in Table 4 and
then dividing by the sum of all products multiplied by its specific AF, i.e.
Allocation(%)i = Productioni AFi / Productionij AFij
with: i = product under investigation and
j = all products that are produced at the manufacturing site
The determined percentage allocation is multiplied by the input or output flow of interest, e.g. fuel
use or wastewater emission.
In this particular case the milk balances for Fonterra dairy processing plants are complicated due
to the transfer of intermediate products between factories. To account for these inter-factory
transfers, intermediate products going into factories were considered an additional milk solids input
reflecting the particular milk solids concentration of the respective intermediate product. Therefore,
factory inputs are based on the sum of both intermediate product and raw milk milk solids. Some
intermediate products, can only be used for the manufacturing of certain products (e.g. cream for
the production of butter, anhydrous milk fat (AMF) and other fat products or skim milk for the
production of skim milk powder). Hence, this kind of intermediate product is only allocated to the
respective products and the equivalent quantity of final product is deducted from the total amount
of product manufactured at the plant. For example, Site Kauri produces 22,422 tonnes of butter and
15,729 tonnes of AMF per year. In addition to its raw milk inputs, Kauri also receives 30,841 tonnes
of cream from other sites which represents an additional milk solids input of 14,323 tonnes (milk
solids content of cream = 46.4%). Since the butter vs. AMF production ratio at Kauri is 59% vs.
41%, the milk solids coming from the cream are distributed accordingly. As a result, Kauri produces
10,266 tonnes of butter and 5,905 tonnes of AMF from its cream imports. The remaining amount
of butter is produced from incoming raw milk milk solids.
For other intermediate products (e.g. milk concentrate or whey) this direct allocation procedure was
not necessary since these intermediate products can be used for the whole Fonterra product range.
Consequently, the milk solids from these types of intermediate products were added to the milk
solids coming from the plants raw milk intake.
The handling of the intermediate products as described above was an important distinction because
otherwise the raw milk would be distributed over all products, whereas in reality it is primarily used
for particular products only and other products have high intakes of intermediate products. The
net effect would be an under-prediction of the raw milk required for products which dont require
intermediate product inputs.
In this study the allocation is carried out for butter, milk powder, milk protein concentrate, cheese
and caseinate. While allocation factors are available for butter, milk powder and cheese, similar
products needed to be selected for milk protein concentrate and caseinate. In consultation with
Fonterra staff, whey protein concentrate/lactose was chosen for milk protein concentrate and whey
powder for caseinate (Barnett, 2008, pers. comm.). With regards to milk protein concentrate, the
specific thermal energy requirements were taken from the Fonterra Engineering Cost Models and
allocated directly to MPC on a product basis (Barnett, 2008, pers. comm.).

The resulting final factory balance is as follows:
INPUTS OUTPUTS
iso
Raw milk in (t) Products out
x x
13.82% milk solids products individual milk
+ solids content
intermediate product = +
imports (t) intermediate product
x exports
respective % milk solids x
respective milk solids
content
3.7 Methodology for distribution
3.7.1 Goal
The goal of module 3 is to provide Fonterra with an assessment the carbon footprint of the
distribution of its products from the manufacturing sites to the destination markets.
3.7.2 Scope
The finished goods supply chain commences on loading of the finished product onto distribution
transport at the weighted average manufacturing site and continues until the defined destinations,
i.e. butter to Zeebrugge (Belgium), Arhus (Denmark) and Damietta (Egypt), milk powder to Manila
(Philippines), Colombo (Sri Lanka), Port Kelang (Malaysia), Puerto Cabello (Venezuela), Apapa
(Nigeria) and Toronto (Canada), MPC to Philadelphia (USA), cheese to Zeebrugge (Belgium) and
Yokohama (Japan), and caseinate to Hamburg (Germany).

The Functional unit (FU) for each product is 1 tonne product leaving the factory gate.
3.7.4 System boundaries

In this study the system boundary encompasses all processes within the system starting with
the transport by truck and train of the finished products from NZ manufacturing sites to a
store (port or non port),
storage of products including energy (electricity and fuel) and chemical at the warehouse,
port handling in New Zealand and at the destination country, and
sea transport to the destination country.

The system boundaries and flow diagram of the distribution phase of Fonterras dairy products
selected for this study are presented in Figure 6.
iso
transport
manufacturing site
(truck or train)
port store
transport
port store
(truck or train)
transport
(truck or train)
diesel
port handling
shipping
port handling
Figure 6: System boundaries of distribution carbon footprints
As described in PAS 2050 (2008) the calculation of the carbon footprint of Fonterras dairy products
includes emissions throughout the entire logistic chain arising from the following sources:
Releases resulting from products handling processes,
Usage energy that has greenhouse gas emissions associated with it (transportation etc.),
and
greenhouse emissions associated with them (e.g. electricity).
As required by the PAS 2050 as well, indirect greenhouse gas emission offset mechanisms are not
included at any point in the supply chain. No renewable energy directly is used by Fonterra; hence
no benefits thereof are included in the calculations.
The carbon footprint calculation excludes greenhouse gas emissions from the manufacture and
ongoing maintenance of capital goods, such as plant machinery, transport equipment, electricity
generating plant, etc., used in the manufacture of the product; and transport of employees to their
normal place of work (PAS 2050 (2007), p 8). This is in line with other studies, such as Lundie et
al (2001).

Life cycle inventory data were collected in close cooperation with Fonterra. At the beginning of
the project a detailed data collection questionnaire was sent to Fonterra staff. All data the project
CWWT / SCION team requested was provided by Fonterra, additional data sources were considered
if no company specific information was available.

All data was checked by the project team and an independent reviewer with regard to consistency
and completeness. The data was then incorporated into an Excel-spreadsheet that built the basis
for the carbon footprint calculation in life cycle assessment-software tool GaBi.
o
Data used in the carbon footprint measurement for distribution are described in more detail
is
below:
All products have been manually aggregated for this analysis. The caseinate aggregation
includes calcium caseinate and sodium caseinate. The milk protein concentrate aggregation
includes Milk Protein Isolates (MPI) etc. Butter aggregation does not include anhydrous milk
fat or other oils.
All volumes are in metric tonnes.
he average metric tonnes per TEU (20ft equivalent unit) for the products in scope are 15.6
mt for Milk Powder, MPC and caseinate and 18 mt for cheese and butter.
All calculations are provided per mt rather than per container.
It is presumed that all products leaving sites leave either on pallets or in containers.
The domestic distances were determined with the help of Google maps. Depending on the
different distances between destinations in NZ (factory to port store or non-port store; port
store to port or other non-port store; non-port store to port store) for each product, average
distances were calculated. These were then weighted according to the volume or product
travelling on each corridor. Uncertainty may be introduced in some cases by port store to
port transportation. These moves are typically less than 5 km and thus negligible relative to
the much larger distances included in the model.
No product refrigeration is required for domestic transportation. The insulation capability of
the containers is sufficient for short domestic transportation ensuring that the temperature
within the container remains cold enough.
Transportation to the domestic customer is excluded.
Movements from factories have not been split further into movements to stores versus
movements to ports since an average distance was calculated.
All the distances are weighted.
Trucking is calculated based on modified information from Mller and Baas (2004) assuming
a split of 12% highway, 18% urban and 70% rural4. Data regarding trucks with a capacity
of 34-40 tonnes are taken from the GaBi life cycle assessment database.
For transport by train an average utilisation rate of 50% is assumed, while the rate is 60%
for trucks (Fonterra, 2008).
The average power demand of a cooling unit for a 40ft reefer container has been calculated
with 3.1 kW based on (Thermoking, 2008; Climatecontrol 2008).
No air freight takes place.
3.7.5.1 Storage of products
The on-site storage of products is included in the total annual energy consumption figures
which in turn are included in the carbon footprint calculation. In 2007, 30% of the butter
make and 34% of the cheese make were held off site (Jones 2008). These emissions were
not covered in the calculation for the following reasons: Warehouse capacity data and various
utilisation rates for different time periods were provided by Fonterra but no information was
available on the energy consumption of the warehouses. Public data on product storage at
warehouses is very rare. Lundie et al. (2004) calculated some generic energy consumption
figures for storage at the supermarket. However, these figures appear to be an overestimate
compared to the situation in commercial warehouses. Hence, energy consumption of off-site
warehouse is not considered in this study due to its poor data quality.
3.7.5.2 Port handling
Containers are stored at the harbour for approximately 7 days (point of departure and
destination). During this period of time cooling is required for butter and cheese. No power
requirements are available for the harbours in New Zealand nor for any destination harbour.
Therefore, we calculated the average power demand of 3.1 kW for a 40ft reefer container
(Thermoking, 2008; Climatecontrol 2008) with an assumed utilisation rate of 60% resulting
in an electrical power demand of 8.7 kWhel.
4. Travel on gravel roads (i.e. 35%) are classified as rural transportation.

Containers that are shipped to the destination market need to be handled at the port in New
Zealand and at the destination. Containers are transported within the harbour by vancarriers.
These vehicles usually have diesel-electric engines. Diesel consumption by modern vancarriers
o
is about 19 L/hr. Vancarriers are used for loading and unloading trucks and ships as well as
s
for movements of containers within dockyard operations. Depending on deployment, travel
i
distances differ, i.e. ~1 km for un/loading trucks, ~3 km for un/loading ships and ~4-5 km
for dockyard hinterland movements. The average speed depends on the actual model, i.e.
between 25-30 km (Winter, 2008). Assuming an equal split between the three travel distances
an average travel time can be calculated (i.e. 0.2 hr) plus idle time of 0.17 hr (Winter, 2008).
3.7.5.3 Overseas shipping
For the overseas transportation of products to the destination markets Fonterra data was
used. The overseas transport includes various destination countries for milk powder, butter
and cheese, while caseinate and MPC are investigated only for one destination.
The explanations and assumptions are reproduced below. Additional information obtained
from different sources and the GaBi database is listed as well:
Products modelled in our analysis have been assigned to one or more routes, i.e. milk
powder to Manila (Philippines), Colombo (Sri Lanka), Port Kelang (Malaysia), Puerto Cabello
(Venezuela), Apapa (Nigeria) and Toronto (Canada) and butter to Zeebrugge (Belgium), Arhus
(Denmark) and Damietta (Egypt). Caseinate is delivered to Hamburg (Germany), cheese to
Zeebrugge (Belgium) and Yokohama (Japan) and MPC to Philadelphia (USA). MSL (Maersk)
provided assistance to Fonterra in calculating the transport distances. Maersk provided
Fonterra with data regarding the days of steaming and the transship delays.
Jones (2008) consulted some confidential materials to determine that the average sized ship
is approximately 6,000 TEU.
In arriving at the emission levels for ocean freight a days steaming approach has been taken.
For scheduled advertised services the days steaming approach is considered the most accurate
and auditable approach, reflective of actual consumption rather than trying to estimate the
distance between various port pairs, an approach that may be imprecise. The data used in
calculating the emissions also includes an allowance for empty slot legs, this reflects the fact
that frequently capacity and hence size of vessel (and resulting consumption) will be determined
by the primary leg capacity needs, the result on the secondary leg being spare capacity empty
slots. When legs on a service are identified has having empty slots, it has been assumed that
cargo being carried in both directions has to burden a share of emissions associated with the
empty slots, this reflects a conservative approach ensuring emissions are not underestimated.
Other key features of the approach taken for calculating emissions for ocean freight include:
o Known vessel sizes have been used when available e.g. ex New Zealand,
o Averages have been used for the size of subsequent transhipment vessels,
o Fuel consumption per days have been used using publicly available information and
information provided directly by the carrier,
o Load utilisation and hence empty slots legs have been arrived at using published trade
line data from Containerisation International,
o Scheduled at sea days have been used for calculating total fuel consumption,
excluded from the calculation are days at transhipment port, and
o Allowances have been made for the additional consumption associated with
refrigerated cargo.
This approach is different to the one that is usually taken in life cycle assessment studies: In the life
cycle assessment software tools the transport effort is calculated in two steps: 1) quantifying the
tonne-kilometers (tkm) of the FU and 2) selection and adaptation of the most appropriate ship for
transportation.
The approach taken using days steaming and the transship delays is based on company data,
i.e. MSL (Maersk) services, and appears to be more reliable than the typical life cycle assessment
calculation procedure.
Where a shipment is a reefer route (for cheese and butter) separate diesel consumption was
quantified. This figure equals 42 kg diesel per TEU slot per day. This number has only been applied
to days when the ship is steaming.
The backhaul ratio varies on a lane by lane basis (Fonterra, 2008; Containerisation International,
2008).

4 References
Armstrong, D., Knee, J., Doyle, D., Pritchard, K., Gyles, O. (1998) More milk and dollars from
irrigation water: How much milk can a megalitre make?, Natural Resources and Environment,
Tatura, Victoria.
Bacharach M (1970): Biproportional Matrics and Input-Output Change, Cambridge University Press:
Cambridge
Barnett (2008) personal communication between Jim Barnett and Sven Lundie, April 7 2008
Basset-Mens, C., Ledgard, S., Boyes, M., 2008. Eco-efficiency of increasing intensification scenarios
of milk production in New Zealand. Journal of Ecological Economics, in press.
Berlin, J., 2002. Environmental life cycle assessment (life cycle assessment) of Swedish semi-hard
cheese. International Dairy Journal, 12, 939-953.
Blonk, H.; Lafleur, M.; van Zeijts, H. (1997) Screening life cycle assessment on Milk powder:
Appendix 3 of the report: Towards an environmental information infrastructure for the Dutch
food industry: Exploring the environmental information conversion of five food commodities.
Amsterdam, June 25, 1997.
BS EN ISO 14040:2006, Environmental management Life cycle assessment Principles and
framework
BS EN ISO 14044:2006, Environmental management Life cycle assessment Requirements and
guidelines
BS ISO 14064-1:2006, Greenhouse gases Part 1: Specification with guidance at the organisation
level for quantification and reporting of greenhouse gas emissions and removals
Carbon Trust (2007) Carbon Footprint Measurement Methodology. Version 1.3, 15 March 2007
Casey, J.W., Holden, N.M., 2005b. Analysis of greenhouse gas emissions from the average Irish milk
production system. Agricultural systems 86, 97-114.
Cederberg., C., Mattsson, B., 2000. Life cycle assessment of milk production a comparison of
conventional and organic farming. Journal of Cleaner Production, 8, 49-60.
Cederberg, C., Flysj, A., 2004. Life cycle inventory of 23 dairy farms in south-western Sweden.
Research report, Swedish Institute for food and Biotechnology, download 24 March 2005:
www.sik.se/archive/pdf-filer-katalog/SR728(1).pdf.
Clark, H., 2001. Ruminant methane emissions: a review of the methodology used for national
inventory estimations. Report for Ministry of Agriculture and Fisheries, Wellington, New
Zealand.
Climatecontrol (2008) http://www.climatecontrol.wur.nl/NR/rdonlyres/B8868606-36E3-4DB8-BFA3-
6062FB1E7BE1/64679/Questleaflet2008.pdf; accessed July 2008
Containerization International (2008) January 2008
Crone, L. (2003) Information provided by Lenard Crone, Carrier, to Sven Lundie on 15th, 18th and
20th of August 2003
Curran, M. A. (2006) Co-Product and Input Allocation Approaches for Creating Life Cycle Inventory
Data: A Literature Review. In: International Journal of life cycle assessment, OnlineFirst 1-14
De Klein, C.A.M., Sherlock, R.R., Cameron, K.C., van der Weerden, T.J., 2001. Nitrous oxide emissions
from agricultural soils in New Zealand A review of current knowledge and directions for
future research. J. R. Soc. New Zealand, 31:543-574.
Eide, M.H., 2002. Life Cycle Assessment (life cycle assessment) of Industrial Milk Production.
International Journal of life cycle assessment, 7(2), 115-126.
Feitz, AJ; Lundie, S; Dennien, G, Morain M, Jones M. (2007) Generation of an industry-specific
physico-chemical allocation matrix - Application in the dairy industry and implications for
systems analysis. International Journal of Life Cycle Assessment, 12(2)109-117.
Fonterra (2008) data provided by Fonterra staff to the project team
GaBi (2008) Database information in the GaBi 4.3 Professional Database, PE International,
Stuttgart.
Greenhouse gas protocol, 2004: Corporate accounting and reporting standard, WRI and World
Business Council for Sustainable Development
Guine, J.B.; Huppes, G.; Heijungs, R. Int. J. life cycle assessment 2004, 9(1), 23-33.

Guine, J. B.; Gorree, M.; Heijungs, R.; Huppes, G.; Kleijn, R.; de Koning; A., van Oers, L.; Wegener
Sleeswijk, A.; Suh, W.; Udo de Haes, H. Handbook on Life Cycle Assessment. Operational
Guide to the ISO Standards, Kluwer Academic Publishers: Dordrecht, 2002.Stone, R. A.
(1963). Input-Output Accounts and National Accounts. Paris, Organisation for European
Economic Cooperation.
Guine, J.B., Gorre, M., Heijungs, R., Huppes, G., Kleijn, R., de Koning, A., van Oers, L., Wegener
Sleeswijk, A., Suh, S., Udo de Haes, H.A., de Bruijn, H., van Duin, R., Huijbregts, M.A.J., 2002.
Life cycle assessment. An operational guide to the ISO standards. Centre of Environmental
Science, Leiden University, Leiden, The Netherlands.
Haas, G., Wetterich, F., Kopke, U., 2001. Comparing intensive, extensified and organic grassland
farming in southern Germany by process life cycle assessment. Agriculture, Ecosystems and
Environment, 83, 43-53.
Hospido, A., Moreira, M. T., Feijoo, G., 2003. Simplified life cycle assessment of Galician milk
production. International Dairy Journal, 13, 783-796.
Hospido, A., Moreira, M. T., Feijoo, G., 2004. Life cycle inventory of the Galician dairy sector. DIAS
report, Animal Husbandry. Tjele, Denmark, Danish Institute of Agricultral Science: 61, 118-
126.
Huppes G, Schneider F (1994): Procedures European Workshop Allocation life cycle assessment,
Centre of Environmental Science, Leiden University, 24-25 February 1994, CML: Leiden
IPCC, 1997. Revised 1996 IPCC Guidelines for National Greenhouse Gas Inventories: Reference
Manual (3). Chapter 4: Agriculture. Intergovernmental Panel on Climate Change. Paris,
France. Available at: http://www.ipcc-nggip.iges.or.jp/public/gl/invs6.htm.
Intergovernmental Panel on Climate Change (2001) Climate Change 2001: The Scientific Basis
(Cambridge University Press, Cambridge).
IPCC, 2006. 2006 IPCC Guidelines for National Greenhouse Gas Inventories: Volume 4: Agriculture,
Forestry and other Land Use. Intergovernmental Panel on Climate Change. Paris, France.
Available at:.http://www.ipcc-nggip.iges.or.jp/public/2006gl/vol4.htm
IPCC, 2007: Fourth Assessment Report, Climate Change 2007, Cambridge Univ. press. Cambridge,
UK.
ISO International Standard 14040, 2006. Environmental management Life cycle assessment
Principles and framework. International Organisation for Standardisation (ISO), Geneva,
Switzerland.
ISO International Standard 14044, 2006. Environmental management Life cycle assessment
Requirements and guidelines. International Organisation for Standardisation (ISO), Geneva,
Switzerland.
ISO International Standard 14064-1, 2006. Greenhouse gases Part 1 : Specification with guidance
at the organisation level for quantification and reporting of greenhouse gas emissions and
removals. International Organisation for Standardisation (ISO), Geneva, Switzerland.
Jones (2008) Information provided by Nigel Jones, Fonterra Technical Manager, Fonterra, 2008
Keedwell (2008) Information provided by Roger Keedwell, Fonterra, 2008
Kelliher, F.M., de Klein, C.A.M., Li, Z., Sherlock, R., 2005. Review of nitrous oxide emission factor
(EF3) data. Client report prepared for the Ministry of Agriculture & Forestry, Wellington, New
Zealand.
Kurdikar, D., Fournet, L., Slater, S.C., Paster, M., Gruys, K.J., Gerngross, T.U., Coulon, R. 2001.
Greenhouse Gas Profile of a Plastic Material Derived from a Genetically Modified Plant. In:
Journal of Industrial Ecology, Vol.4, No. 3.
LIC, 2005. Dairy statistics 2004-2005. Livestock Improvement Corporation limited, Hamilton, New
Zealand.
Lundie, S., Ciroth, A., Huppes, G (2007) Inventory methods in life cycle assessment: towards
consistency and improvement. Final Report to UNEP / SETAC global Life Cycle Initiative
(accepted for publication)
Lundie, S., Feitz, A. (2004) Analysis of environmental impacts caused by the distribution of retail dairy
products and post-consumer phase of packaging materials for the products market milk,
UHT milk, cheese, butter and yoghurt / dessert. Final report for Dairy Australia.
Lundie, S., Feitz, A., Changsirivathanathamrong, A., Jones, M., Dennien, G., Morain, M. (2001)
Evaluation of the Environmental Performance of the Australian Dairy Processing Industry
using Life Cycle Assessment. Life Cycle Inventory for Milk Powder, Market Milk, Cheese,
Whey, Butter and Dessert/Yoghurt. Final Report to Dairy Research and Development
Corporation, Sydney, December 2001.
MED, 2006. New Zealand Energy data file. Ministry of Economic Development, Wellington, New
Zealand. Available at: http://www.med.govt.nz/ers/en_stats.html
MfE, 2006. New Zealands Greenhouse Gas Inventory 1990-2004. The National Inventory Report and
Common Reporting Format Tables. Ministry for the Environment, Wellington, New Zealand.
Mueller and Baas, 2004: Profile of the heavy vehicle fleet, Update 2004. Transport Engineering
Research New Zealand Ltd. Prepared for the Road Safety Trust
Parikh A (1979) Forecasts of input-output matrices using the R.A.S. Method, Rev Econ Stat 61,
477-481
Publicly Available Specification PAS 2050 (2007) Specification for the measurement of the embodied
greenhouse gas emissions in products and services. Draft by BSi 6 October 2007 (for
circulation).
Publicly Available Specification PAS 2050 (2008) Specification for the assessment of the life cycle
greenhouse gas emissions of goods and services. BSi, October 2008, ISBN 978 0 580 50978
0.
Ramaswamy, V., O. Boucher, J. Haigh, D. Hauglustaine, J. Haywood, G. Myhre, T. Nakajima, G.Y.
Shi, S. Solomon, 2001: Radiative Forcing of Climate Change. In: Climate Change 2001:
The Scientific Basis. Contribution of working group I to the Third assessment report of the
Intergovernmental Panel on Climate Change [Houghton, J.T., Y., Ding, D.J. Griggs, M. Noguer,
P.J., van der Linden, X. Dai, K. Maskell, and C.A. Johnson (eds.)]. Cambridge University Press,
Cambridge, United Kingdom and New York, NY, USA, 881 pp.
Renewable Obligation Order, 2002, Statutory Instrument 2002 No. 914, HMSO
Saggar, S., Clark, H., Hedley, C., Tate, K., Carran, A., Cosgrove, G., 2003. Methane emissions from
animal dung and waste management systems, and its contribution to the national methane
budget A report prepared for the Ministry of Agriculture and Forestry, Wellington, New
Zealand.
Somers (2008) personal communication between Joanne Somers and Sven Lundie, April 28 2008
Suh S., Lenzen M., Treloar G.J., Hondo H., Horvath A., Huppes G., Jolliet O., Klann U., Krewitt W.,
Moriguchi Y., Munksgaard J. & Norris G. System Boundary Selection in Life-Cycle Inventories
Using Hybrid Approaches. Environmental Science & Technology 38 657-664 (2004)
Stone, R. A. (1963). Input-Output Accounts and National Accounts. Paris, Organisation for European
Economic Cooperation.
Stone R (ed) (1963): Input-Output Relationships, 19541966, A Programme for Growth. Volume 3,
Chapman and Hall, London
Thermoking, (2008) http://www.thermoking.com/tk/index.asp accessed July 2008
Thomas, S.M., Ledgard, S.F., Francis, G.S., 2005. Improving estimates of nitrate leaching
for quantifying New Zealands indirect nitrous oxide emissions. Nutrient Cycling in
Agroecosystems, 73: 213-226.
Thomassen, M., van Calker, K.J., Smits, M.C.J., Iepema, G.J., de Boer, I.J.M., Life cycle assessment of
milk production systems in the Netherlands. Agr. Syst., in press.
van der Linden J.A., Dietzenbacher E. 2000. The determinants of structural change in the European
Union: A new application of RAS, Environ Planning A 32, 2205-2229
Wheeler, D.M., Ledgard, S.F., De Klein, C.A.M., Monaghan, R.M., Carey, P.L., McDowell, R.W., Johns,
K.L., 2003. OVERSEER nutrient budgets moving towards on-farm resource accounting.
Proc. New Zealand Grassland Association, 65: 191-194.
Williams, A.G., Audsley, E., Sandars, D.L., 2006. Determining the environmental burdens and
resource use in the production of agricultural and horticultural commodities. Main report.
Defra Research Project ISO 205. Bedford: Carnfield University and Defra. Available on: www.
silsoe.cranfield.ac.uk, and www.defra.gov.uk.
WRI WBCSD GHG Protocol Initiative calculation tool - allocation of GHG emissions from a combined
heat and power plant 2006. A WRI/WBCSD GHG Protocol Initiative calculation tool. http://
www.ghgprotocol.org/
ZSD_FREIGHT_REPORT for Sept 2007. Weighted average.

Carbon Footprint Methodology

Uploaded by

Copyright:

Available Formats

Carbon Footprint Methodology

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Carbon Footprint Methodology

Uploaded by

Copyright:

Available Formats

What standards and references are discussed regarding carbon footprint assessment methodology?

What standards and references are discussed regarding carbon footprint assessment methodology?

What are the main components/steps of the carbon footprint calculation methodology outlined in the document?

What are the main components/steps of the carbon footprint calculation methodology outlined in the document?

CARBON FOOTPRINT

Centre for Water and Waste Technology

2 CARBON FOOTPRINT METHODOLOGY REPORT

FOR FONTERRA CO-OPERATIVE GROUP 2009 3

Figure 5. System boundaries of product carbon footprints 22

Figure 6: System boundaries of distribution carbon footprints 28

Table 1. Districts included in each regionally-adjusted category for this study. 17

Table 2. Technical description of Northland, Waikato, Bay of Plenty, Taranaki,

Table 8: Industry-specific physico-chemical allocation matrix 25

4 CARBON FOOTPRINT METHODOLOGY REPORT

FOR FONTERRA CO-OPERATIVE GROUP 2009 5

Fonterra has demonstrated a proactive stance on environmental management by commissioning independent

6 CARBON FOOTPRINT METHODOLOGY REPORT

FOR FONTERRA CO-OPERATIVE GROUP 2009 7

8 CARBON FOOTPRINT METHODOLOGY REPORT

iso 3.1 Standards and references

LIFE CYCLE ASSESSMENT

- Product development and improvement

Figure 1: Life cycle assessment framework as described in ISO standards

FOR FONTERRA CO-OPERATIVE GROUP 2009 9

3.1.2 Carbon Trust BSI DEFRA initiative

3.1.3 IPCC guidelines

10 CARBON FOOTPRINT METHODOLOGY REPORT

3.2 General overview

SYSTEM BOUNDARY Chemicals Ingredients Water Energy (fuel,themal,

to all processes within

Product and Packaging material

Product and Packaging material

Port handling and cooling

FOR FONTERRA CO-OPERATIVE GROUP 2009 11

3.3 Life cycle assessment calculation procedure

3.4 Carbon footprint calculation

3.5 Methodology for the farm

12 CARBON FOOTPRINT METHODOLOGY REPORT

1. New Zealand dairy research and extension organisation.

FOR FONTERRA CO-OPERATIVE GROUP 2009 13

Figure 3. Simplified flowchart of the grass-to-farm-gate NZ production system

14 CARBON FOOTPRINT METHODOLOGY REPORT

FEED SUPPLEMENTS FROM NZ

Maize Pasture Barley grain Road

Pasture Cow Milk

Wintering-off Off-farm Meat

New Zealand Regions (source Ministry of the Environment)

16 CARBON FOOTPRINT METHODOLOGY REPORT

Western Uplands Waitaki

FOR FONTERRA CO-OPERATIVE GROUP 2009

CARBON FOOTPRINT METHODOLOGY REPORT

3.5.3.3 Inventory-based greenhouse gas emissions from agricultural land

FOR FONTERRA CO-OPERATIVE GROUP 2009 19

Process Mean References for mean

20 CARBON FOOTPRINT METHODOLOGY REPORT

3.6 Methodology for processing

3.6.3 Functional unit

3.6.4 System boundary

FOR FONTERRA CO-OPERATIVE GROUP 2009 21

Processing plant Waste