Week6 A2023 multifonctionnalité-EN

DDI8003 – Week 6
Multifunctional systems
Manuele Margni (Class built with Pascal Lesage)
Syllabus
2
Syllabus
The midterm oral presentations of the LCA projects will be given in two separate rooms
- Room M-2101 co-modal (on-site and zoom)
Invited groups: « to be defined »
- Room J-1035 (PAVILLON : J. Armand Bombardier) on-site only

Invited groups: « to be defined »
3
4
Ecoinvent (inventory database)
5
ecoinvent – A brief history
90s Demand for LCI data growing in Switzerland
Data development in many institutions
Non-centralized data, sometimes contradictory
No one institution is sufficiently strong to maintain a complete LCI-DB of
good quality
1997 Creation of the NGO « Centre ecoinvent »: ETH Zurich, EPF Lausanne, PSI,
EMPA, ART
Financed by the Swiss government
Goal: create and maintain the ecoinvent LCI-DB
Harmonisation
Characteristics aimed for: scientifically solid and transparent data
All revenues reinvested in the LCI-DB
1999-2007 ecoinvent 1.0-1.3 (≈2500 datasets)
Integration in main software
Starts gaining momentum (particularly in research community)
2007-2013 ecoinvent 2.0-2.2 (≈4000 datasets)
Work started on ecoinvent v3
2013-now ecoinvent v3.9.1b (> 15000 datasets)
6
ecoinvent : some characteristics
Transparency
• Datasets available in disaggregated version (gate-to-gate) and
aggregated (pre-calculated inventories)
• Exhaustive documentation (datasets, hypotheses)
ecoinvent version ecoinvent version

« Unit process » (U) « System process » (S)
Note: some datasets are only available in
« S », e.g. plastics, for confidentiality
7 reasons imposed by industry associations
ecoinvent : some characteristics
Comprehensiveness
• Elementary flows
• All extraction/emissions included, even if the substances are not yet
associated to a characterisation factor
• Intermediary flows
• Effort to cover as many intermediary flows as possible (no exclusion
criteria)
• Infrastructure, transport, packaging, etc.
• Industry sectors
• Extraction of resources, agriculture, production and transformation of
materials, energy, end of life, transport
• Geography
• More countries, especially since v3 – noted with abbreviations
Data integration on uncertainty of flows
8
ecoinvent v3
9
ecoinvent v3
1. Distinction between the nomenclature of products (flows) and

processes (activities)
Aluminium
production, Aluminium
Söderberg
Aluminium
prodution, pre- Aluminium
cooked anodes
Not the same name
10
ecoinvent v3 – Technical differences
2. Distinction between processes of transformation and market
• Transformation process, inputs ≠ outputs
Production of
automotive
Aluminium an automotive
component
component
11
ecoinvent v3 – Technical differences
2. Distinction between processes of transformation and market
• Market process, inputs = outputs
Production of
Aluminium Production of a
aluminium, A
Aluminium automotive
Production of component
Aluminium
aluminium, B
Production of Market for
Aluminium
aluminium, C aluminium
Production of
Aluminium
aluminium, D
Transportation Losses
Transport
by lorry
12
ecoinvent v3 – How it works
Automatic link between suppliers and markets, based on location
Electricity
Electricity
production, A
Electricity
Electricity
Electricity
production, B
Electricity Market for
Electricity
production, C electricity
Electricity
Electricity
production, D
Transportation Transport Losses
by lorry
Region x Region x
13
Automatic link between producers and markets, based on location
Production of
Electricity an automotive
component
Region x
Market for
electricity
Region x
14
And therefore automatic links between producers and suppliers,

based on location
Electricity
Electricity Prodution of
production, A
Electricity automotive
Electricity component
Electricity
production, B
Region x
Electricity Market for
Electricity
production, C electricity
Electricity
Electricity
production, D
Transportation Losses
Transport
by lorry
Region x Region x
15
And therefore automatic links between producers and suppliers,

based on location
• Some important details:
• Default region = global (GLO)
• When a region < GLO is defined, another region « rest of
world» (RoW) is created
• The more regions created, the more the production
volume of RoW is reduced
16
A group of disaggregated unit processes, many system models
Attributional model
- Average data
- Economic allocation
- “Allocation to the point of
substitution” for recycling
Consequential model
- Marginal data
- System boundary expansion
Non-linked Cut-off model

datasets In your projects,
- Average data
by default
- Economic allocation
- Cut-off for recycling
17
Attributional model
Electricity PVA/PVtotal
Electricity
production, A
Electricity
Electricity PVB/PVtotal
Electricity
production, B
Market for
Electricity PVC/PVtotal electricity
Electricity
production, C
Electricity PVD/PVtotal
Electricity
production, D
18
Consequential model
Electricity 0%
Electricity
production, A
Electricity
Electricity 0%
Electricity
production, B
Market for
Electricity 0% electricity
Electricity
production, C
Electricity 100%
Electricity
production, D
Marginal technology, a default

technology level is applied or
could be imposed
19
Resources
• Definitions: http://www.ecoinvent.org/support/glossary/glossary.html
• Frequently asked questions:
http://www.ecoinvent.org/support/faqs/faqs.html
• Short presentation: http://www.ecoinvent.org/support/documents-and-
files/documents-and-files.html
• Webinar: https://www.youtube.com/channel/UC4Ma__pX9r1-
DoNi2D2Fd6g
20
Multifunctionality
21
Recap: System boundary
• In theory, the only economic flows that cross the system boundary in
the study are directly associated to the function (final demand flow)
• The other economic flows are entirely consumed by the other unit
processes
• All other flows that cross the system boundary are elementary flows
22
Multifunctional processes – the example of a cow
4 m2
20 000 liters
5 tons 1 cow
200 kg
23
Multifunctional processes
20 000 litres
5 tonnes 1 cow
5 tonnes of fertilizer for the milk
25
4 m2
5 tonnes 1 cow

+ 5 tonnes of fertilizer for the leather
26
5 tonnes 1 cow

+ 5 tonnes of fertilizer for the leather
+ 5 tonnes de fertilizer for the meat 400 kg
= 15t!! à ERROR!
27
2 m2
20 000 litres
5 tonnes 1 cow
400 kg
28
• When a process provides more than one function (co-products,

multiple services, etc.) we need to « distribute » the inputs and
outputs to each of the functions to avoid double-counting.
• These processes are omnipresent:
Thermal electric power: Sawmill: Production of NaOH by Transport: several

Electricity + heat construction wood, electroysis of an NaCl products transported
wood chips solution:
NaOH, Cl2, H2
29
• When a process provides more than one function (co-products, multiple

services, etc.) we need to « distribute » the inputs and outputs to each of
the functions to avoid double-counting.
• These processes are omnipresent:
Wheat cultivation: straw, wheat grain
Raising cows: meat, milk, leather
Thermal electric power plant: electricity, heat
Production of NaOH by electroysis of an NaCl solution: NaOH, Cl2, H2
Transportation: many products transported in the same truck
Refining petroleum: carboreactor (aviation fuel), gasoline, diesel,
propane, etc.
Sawmill: construction wood, wood chips for paper production, wood
waste for energy production
30
Why multifonctionality set a problem?
Problem of « double counting »

Only one intermediary flow can cross the system boundary
Multifunctionality problem
Multifunctional
process
Coproduct B,
towards another
product system
Coproduct A, in
product system
studied
Problem: What proportion of impacts associated to the inputs and

outputs of the multifunctional process should be attributed to coproduct A?
32
• We saw that the only economic flow that may cross the system
boundary is that which is directly associated to the functional unit
• With multifunctional processes, supplementary economic flows cross
the system boundary: the system has more than one function
33
• We saw that the sole economic flow that may cross the system
boundary is that which is directly associated to the functional unit
• With multifunctional processes, we will also have non-inversible
A matrices (not square)…
pcogen pGN
Electricity (kWh) 1 -2,5
Natural gas (MJ) -10,5 1000 A
Heat (MJ) 1,8 0
Raw natural gas (m3) 0 -25
CO2 (kg) 0,5 10
B
SO2 (kg) 0,25 0,01
34
How to solve multifunctionality in LCA?
Subdivision
System
expansion
Underlying
physical
relationship
Allocation
Multifunctional processes and ISO 14044
ISO 14044, section 4.3.4.2: The study must identify the processes
shared with other product systems and follow the step-by-step
procedure presented here:
Step 1: Avoid allocation as much as possible by:

• Subdividing the unit process by creating two sub-processes or
more;
• Extend the system boundary of the product system
Step 2: Allocate the inputs and outputs in a way that reflects

underlying physical relationships existing between them.
Step 3: Allocate inputs and outputs in a way that reflects other

mutual relationships (mass, economic, etc.)
ISO calls them steps, but they are in fact 4
mutually exclusive options
36

Subdivision
more;
• ExtendExtension
the system boundary of the product system
Step 2: Allocate Allocation

the inputs and outputs in a way that reflects
– Underlying physical relationship

Allocation
mutual relationships (mass, – Other relationship
economic, etc.)
37

more;


38
ISO 14044 Procedure– Step 1a: Subdivision
Multifunctional
process
Coproduct B,
towards another
product system
Coproduct A, in
product system
studied
Step 1.1: Divide the process in sub-processes
39
Monofunctional Monofunctional
process A process B
Product A, in Product B, in
product system product system
studied studied
Step 1.1: Divide the process in sub-processes
Note: only applicable if the process isn’t actually multifunctional
40
• Example: LCA of cakes.

Factory
We have data for all inputs

Cakes, used in the Cookies, not used in
and outputs of a factory product system the product system
producing two products: analyzed analyzed
cakes and cookies.
The data collection could be

refined to distinguish the
processes in the factory « Cake » « Cookie »
that are used for the cake process process
and those used for the

cookies. Product A, used in Produit B, not used in
the product system the product system
analyzed analyzed
41

more;


42
ISO 14044 Procedure– Step 1b: Expanding the system boundary
Multifunctional
process
-
Coproduct B
+
Coproduct A, in the
product system Monofunctional
analyzed process B
Step 1.2: Expand the system boundary
Note: Only applicable if there is a process producing a functionally

equivalent product to product B
43
• In this approach, the product system is expanded to include avoided unit

processes
• LCA on function A, product system A and B
• We take into account the avoided (negative) autonomous system, that
represents an alternative way to produce B
à substitution, displacement, avoidance
• We consider then that the expanded system is equivalent to a product
system that doesn’t have function B as an output
New scope
(Multifunctional) Autonomous Monofunctional

system - system avoided = system
Function A Function B Function B Function A
44
To use system expansion,

we need to be able to
identify an alternative way
to produce the co-product
that doesn’t interest us.
45
- =
pcogen pGN pchaleur

Electricity (kWh) 1 -2,5 0
Natural gas (MJ) -10,5 1000 -1.05
Heat (MJ) 1,8 0 1 Aàsquare!
Raw natural gas (m3) 0 -25 0
CO2 (kg) 0,5 10 0,06
SO2 (kg) 0,25 0,01 6e-7 B
46

more;


47
ISO 14044 Procedure – Step 2: underlying physical relationships
Step 2: Allocate between products in a way that reflects

underlying physical relationships existing between them
Note: only applicable if we can vary the co-products

independently (combined production)
48
à Applies to combined productions

• This method allows us to determine how the inputs and outputs are
affected by an increase in production of one of the co-products.
• This information is used to estimate the inputs and outputs associated to
one product only
• Coming back to LCA example of the factory producing cakes and cookies:
If it’s impossible to simply attribute the different unit processes inside
the factory to those producing cakes, we could vary the quantity of
cakes produced: this would affect the inputs and outputs of the
factory
In this way, we could determine which inputs are directly attributable
to cakes
• Not be be confused with applying a physical property like mass!
49
Day 1: 5000
cookies Day 2: 10000
Flour Factory
cakes Day 1: 1000
Day 1: 600,000kg Day 2: 500
Day 2: 450,000kg
How much flour do I need per cookie (#flour/cookie) ? Per cake (#flour/cake) ?
For day1 (or 2) : #flour = #cookies*#flour/cookies + #cakes*#flour/cakes

For day 1- day2: ∆flour=∆cookies*#flour/cookies + ∆cakes*#flour/cakes
à #flour/cake=500
à #flour/cookie=20
50

• Subdividing the unit process by creating two sub-processes or more;


51
ISO 14044 Procedure – Step 3: Other relationship
Monofunctional
α process A
Coproduct A, in product system studied
α+β=1
Monofunctional
process B
β
Step 3: Attribute according to another mutual relationship

(economic value, mass, surface, exergy, etc.)
Note: Results in a outcome that only works for the chosen

relationship) . Useful for joint production.
53
à Applies to joint productions
• The inputs and outputs are divided between the different products
• According to a physical property (e.g. mass, energy, surface)
• According to economic value of the co-products
• According to another valuable relationship
• Causality principle : the factors used to allocate a flow to a product

should reflect the degree of responsability of the product for generating
that flow
54
• Allocation example
0,4 kg phosphorus to water
40 tkm of tractor 800 kg of grain
Cultivation of
4 kg fertilizer wheat
200 kg of straw
• Choice of allocation base: mass.

• The two products are measured by their mass
55
• Allocation example
Cultivation of
200 kg of straw
è Re-calculate the quantities of inputs/outputs for the production

of 1kg of grain using mass allocation
56
• Allocation example: mass
Cultivation of
200 kg of straw
• Total mass produced: 800 + 200 = 1000 kg

• Allocation factor for grain = 800/1000 = 80%
• Allocation factor for straw = 200/1000 = 20%
57

0,4*80% = 0,32 kg
phosphorus to water
40*80% = 32 tkm of 800 kg of grain

tractor
Production of
4*80% = 3,2 kg wheat grain
fertilizer
0,4*20% = 0,08 kg
phosphorus to water
40*20% = 8 tkm of
tractor Production of
4*20% = 0,8 kg straw
fertilizer 200 kg of straw
58

0.4 g phosphorus to
water
0.04 tkm of tractor 1 kg of grain
Production of
4 g fertilizer wheat grain
0.4 g phosphorus to
water
0.04 tkm of tractor

Production of
4 g fertilizer straw
1 kg of straw
59
• Example of economic allocation

Let’s take the same process, but let’s use an economic allocation
Let’s suppose the following prices:
• Wheat grains: 0,25$/kg
• Straw: 0,1$/kg
60

Let’s take the same process, but let’s use an economic allocation
Let’s suppose the following prices:
• Wheat grains: 0,25$/kg
• Straw: 0,1$/kg
• Total revenue: 800 kg * 0,25 $/kg + 200 kg * 0,1 $/kg = 220$

• Allocation factor for grain = 200/220 = 91%
• Allocation factor for straw = 20/220 = 9%
61
• Allocation example: economic

0,4*91% = 0,364 kg
phosphorus to water
40*91% = 36,4 tkm 800 kg of grain

tractor
Production of
4*91% = 3,64 kg wheat grain
fertilizer
0,4*9% = 0,036 kg
phosphorus to water
40*9% = 3,6 tkm

tractor Production of
4*9% = 0,36 kg straw
fertilizer 200 kg straw
62
• Allocation example: economic

0.455 g phosphorus to
water
0.0455 tkm tractor 1 kg of grain
Production of
4.55 g fertilizer wheat grain
0.18 g phosphorus to
water
0.018 tkm tractor

Production of
1.8 g fertilizer straw
1 kg straw
63
• Allocation example: mass vs economic

0.455 0.4 g phosphorus to
water
0.455
0.04 tkm of tractor 1 kg of grain
Production of
4.55
4 g fertilizer wheat grain
0.18 0.4 g phosphorus to

water
0.018
0.04 tkm of tractor
Production of
1.8
4 g fertilizer straw
1 kg of straw
64
• Another example of economic allocation

Metal Transformed metal [kg]
transformation
Vapour [MJ]
process
A factory transforms metal with a process using heat. A heat recovery

machine produces steam that is sold to a factory nearby.
Inconsistent data à economic allocation.
The factory sells metal at a price of 45 $/kg, and steam at 5 $/MJ
Allocation factor for metal: 45 / (45 + 5) = 90% ?
• NO!!!!!! We cannot add the prices, but we could add the revenues.
Over one year, the factory produces 300 tonnes of metal and 800 GJ of
steam
65

The factory sells metal at a price of 45 $/kg, and steam at 5 $/MJ
Over one year, the factory produces 300 tonnes of metal and 800 GJ of
steam
Product Quantity Price Revenues Allocation

factor
Metal 300E3 kg 45$/kg 13500 k$ 77,1%
Steam 800E3 MJ 5$/MJ 4000 k$ 22,9%
66
ISO 14044 Procedure – Step 3: Other relationship - examples
Gasoline [l]
Raw Natural gas [m3]
Chemical products [kg]
Common property?
Machinery
Gravel [tonnes]
Explosives
Diamonds [g]
Combustibles
67
Gasoline [l]
Raw Natural gas [m3]
Chemical products [kg]
à Volumic allocation would make no sense!

à Economic value? Calorific value?
Machinery
Gravel [tonnes]
Explosives
Diamonds [g]
Combustibles
à Diamond exploitation is the root of the process!

68
How to do it?
1. Identify all co-products and the quantities produced

2. Look at the units of the co-products: are they consistent?
3. Is there something common, comparable to all the co-products? Do you think that
the impacts of the co-products would be proportional to this basis? (mass,
volume, energy, surface area, price, time, etc.) We need to think about the process
for answering this question.
4. Choose the imputation basis, name it and justify it.
5. Calculate an imputation factor for each co-product.
6. Apply the factor to each flow (elementary and economic) except for co-product
flows.
7. Exception: are there any flows that you can fully impute to certain co-products?
8. Create a new elementary process for each co-product.
9. Scale up each elementary process to obtain 1 unit of product.
70
1 kWh
électricité
Production 0,5 * 0,67 = 0,33 kg CO2

électricité,
cogénération 0,25 * 0,67 = 0,167 kg SO2
10,5 * 0,67 = 7,0 MJ gaz naturel

1 MJ
1,8 MJ
chaleur
chaleur
Production 0,5 * 0,33 = 0,167 kg CO2 Production 0,093 kg CO2

chaleur, chaleur,
cogénération 0,25 * 0,33 = 0,083 kg SO2 cogénération 0,046 kg SO2
• Two co-products 10,5 * 0,33 = 3.5 MJ gaz naturel 1,9 MJ gaz naturel
• Coherent units, with conversion (MJ, kWh)

• Are they comparable? Let’s say yes...
• Basis of allocation chosen: Energy. We suppose that heat and electricity are both useful to do
work (which is not necessarily true…à exergy would be a better metric)
• Calculating the allocation factors:
Total energy: 1 kWh * (3,6 MJ/kWh) + 1,8 MJ = 5,4 MJ
Allocation factor electricity: 3,6/5,4 = 0.67
Allocation factor heat: 1,8/5,4 = 0.33
• Apply the factors
• Re-normalize, as needed
Multifunctional processes - Allocation
pcogen pGN
Electricity (kWh) 1 -2,5
Natural gas (MJ) -10,5 1000 A
Heat (MJ) 1,8 0
Raw natural gas (m3) 0 -25
CO2 (kg) 0,5 10
B
SO2 (kg) 0,25 0,01
pelect pGN pchaleur

Electricity (kWh) 1 -2,5 0
Natural gas (MJ) -7 1000 -1,9 A
Heat (MJ) 0 0 1 Square!
Raw natural gas (m3) 0 -25 0
CO2 (kg) 0,33 10 0,093 B
SO2 (kg) 0,167 0,01 0,046
73
How to choose the approach to address multifonctionality?
ISO 14044, section 4.3.4.2: The study must […] follow the step-by-
step the procedure presented here:

more;Subdivision
Extension

Allocation – Underlying physical relationship

Allocation – Other relationship
74
How to choose the approach to address multifunctionality?
According to the
UNEP Life Cycle Initiative (2016) …
Life Cycle Initiative. (2016)

Steps to identify the most
appropriate allocation approach.
Retrieved from: https://www.lifecycleinitiative.org/training-resources/global-
guidance-principles-ggp-on-life-cycle-assessment-data-and-databases/
ISO 14044 Procedure - multifunctionality
• « Attributing impacts » to a co-product is entirely imaginary. In reality,

the co-products will always be manufactured together. For this reason,
allocation is an arbitrary operation subject to debate.
The choice of allocation approach can change the

results and the conclusion of a study!!!
• Like anytime, there are assumptions in LCA, transparency is necessary.

• When in doubt, or if controversy arises on the best method to use,
perform a scenario/sensitivity analysis
76
Multifunctionality and recycling
77
Open-loop recycling
Open-loop recycling has two functions, each associated to its

respective product system
Elimination of waste from

product system A For a study we
can generate
waste that will
be recycled
AND/OR use
recycled
Recycling material
Recycled material used in

product system B
Open-loop recycling
Who is responsible for the impacts of recycling?
Extraction
â Manufacturi Use
Material Disposal A life cycle
ng
Recycling
Extraction
â Manufacturi Use
Material Disposal
ng B life cycle
80
Open-loop recycling
Who is responsible for the impacts of the production of initial virgin material?
Who is responsible for the impacts of final disposal?
Extraction
â Manufacturi Use
ng
Recycling
Extraction
â Manufacturi Use
Material Disposal
ng B life cycle
81
Open-loop recycling
Who takes credits for the avoided impacts of disposal?

Who take credits for the avoided impacts of the production of
virgin raw material?
Extraction
â Manufacturi Use
ng
Recycling
Extraction
â Manufacturi Use
Material Disposal
ng B life cycle
82
Most common approaches
• Recycled content method

• End of life recycling method
• There are more:

• Circular Footprint Formula (CFF)
https://ec.europa.eu/environment/eussd/pdf/Webinar%20CFF%20Circular%20Footprint%20For
mula_final-shown_8Oct2019.pdf
• Method of number of uses (usually used for paper, but not
covered in this course)
• Methods based on the quality of the recycled material
• 50-50 method
• …
83
Example: description
• Let’s assume a product composed of 1kg of material

• Is manufactured from x% virgin material and y% recycled material
• At the end of its life, α% of the collected material goes to final
disposal (e.g. landfill) and β% is recycled
Using recycled material Recycling of residual

as input materials as output
Extraction/
x% 100% α% Disposal
Materials
Manufacturing Use Collection
y% β% Recycling
Recycling
Life cycles of other products

84
Example: “Recycled content” approach
• Recycled content approach (cut-off):

• Recycling is attributed to the life cycle that uses the
recycled material
• Recycling at end of life is simply excluded from the
system (no credits, but no impacts of end of life
treatment either)
Extraction/
x% 100% α% Disposal
Materials
y% β% Recycling
Recycling

85
Example: “End of life recycling” approach
• End of life recycling approach

• Materials used by the life cycle are assumed as virgin,
regardless of the recycled content
• Recycling at the end of life avoids producing raw
materials (substitution, system boundary expansion)
Extraction/
100% 100% α% Disposal
Materials
0% β% Recycling
Recycling
β%
Extraction/Materials
avoided (negative)
86
Main approaches – what do they imply?
Recycled content approach:

Boosts use of recycled materials
• If impacts of recycling < impacts of initial production
of virgin materials
Encourages (though weak) recycling at the end of life
• Recycled material isn’t associated to impacts of final
disposal
87
Main approaches – what do they imply?
End of life recycling approach:

Doesn’t encourage the use of recycled materials
• All input materials are associated to impacts of virgin
materials, regardless of their recycled content
Encourages recycling at end of life (Design for
Recycling)
• Recycling à « credit »
• (must obviously take into account the modification of
inherent properties of the materials)
Discourages dissipation of materials
• If they can’t be collected, it’s impossible to recycle
them
88

Week6 A2023 multifonctionnalité-EN

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Week6 A2023 multifonctionnalité-EN

Uploaded by

Copyright:

Available Formats

DDI8003 – Week 6

- Room J-1035 (PAVILLON : J. Armand Bombardier) on-site only

ecoinvent version ecoinvent version

1. Distinction between the nomenclature of products (flows) and

Not the same name

2. Distinction between processes of transformation and market

• Transformation process, inputs ≠ outputs

2. Distinction between processes of transformation and market

• Market process, inputs = outputs

Automatic link between suppliers and markets, based on location

Automatic link between producers and markets, based on location

And therefore automatic links between producers and suppliers,

And therefore automatic links between producers and suppliers,

A group of disaggregated unit processes, many system models

Non-linked Cut-off model

Marginal technology, a default

5 tonnes of fertilizer for the milk

5 tonnes of fertilizer for the milk

5 tonnes of fertilizer for the milk

• When a process provides more than one function (co-products,

Thermal electric power: Sawmill: Production of NaOH by Transport: several

• When a process provides more than one function (co-products, multiple

Problem of « double counting »

Problem: What proportion of impacts associated to the inputs and

Step 1: Avoid allocation as much as possible by:

Step 2: Allocate the inputs and outputs in a way that reflects

Step 3: Allocate inputs and outputs in a way that reflects other

Step 1: Avoid allocation as much as possible by:

Step 2: Allocate Allocation

Step 3: Allocate inputs and outputs in a way that reflects other

Step 1: Avoid allocation as much as possible by:

Step 2: Allocate the inputs and outputs in a way that reflects

Step 3: Allocate inputs and outputs in a way that reflects other

Step 1.1: Divide the process in sub-processes

Step 1.1: Divide the process in sub-processes

Note: only applicable if the process isn’t actually multifunctional

• Example: LCA of cakes.

We have data for all inputs

cakes and cookies.

The data collection could be

and those used for the

Step 1: Avoid allocation as much as possible by:

Step 2: Allocate the inputs and outputs in a way that reflects

Step 3: Allocate inputs and outputs in a way that reflects other

Step 1.2: Expand the system boundary

Note: Only applicable if there is a process producing a functionally

• In this approach, the product system is expanded to include avoided unit

(Multifunctional) Autonomous Monofunctional

Function A Function B Function B Function A

To use system expansion,

pcogen pGN pchaleur

Step 1: Avoid allocation as much as possible by:

Step 2: Allocate the inputs and outputs in a way that reflects

Step 3: Allocate inputs and outputs in a way that reflects other

Step 2: Allocate between products in a way that reflects

Note: only applicable if we can vary the co-products

à Applies to combined productions

For day1 (or 2) : #flour = #cookies*#flour/cookies + #cakes*#flour/cakes

Step 1: Avoid allocation as much as possible by:

Step 2: Allocate the inputs and outputs in a way that reflects

Step 3: Allocate inputs and outputs in a way that reflects other

Coproduct A, in product system studied

For day1 (or 2) : #flour = #cookies#flour/cookies + #cakes#flour/cakes