35
GHG emissions from animal food chains Development of a quantification model using Life Cycle Analysis method Pierre Gerber, Sète, 2 March 2010
GHG emissions from animal food chains
Development of a quantification model using Life Cycle Analysis
method
Pierre Gerber, Sète, 2 March 2010
Content
1. Background and Objectives 2. Choice of LCA 3. Methodology4. Data and classification5. Herd model6. Tentative results7. Conclusions and next steps
Background
• Milk: a growing sector, especially in developing countries • driven by income, demography and changing preferences,• among highest growth rate in agriculture commodity• over 80% of production growth in non OECD countries
(OECD-FAO, 2009)
• Climate change• the worst-case ipcc scenario trajectories are being realized• societies are highly vulnerable, with strong differential effects on
people within and between countries and regions.• risk of crossing tipping points• there is no excuse for inaction
(Climate Change: Global Risks, Challenges & Decisions – 2009, Copenhagen)
Dual challenge of food security and climate change mitigation
The aftermath of Livestock’s Long Shadow (18% of global anthropogenic emissions attributed to livestock)• Raised awareness
• academia, general public, policy debate• over simplification and distortion/selection of messages• focus on climate change• basic results stood the test
• Many studies relying on food chain / live cycle approach: available results and databases but fragmented
• Expectations to move on to an action plan: information for public and private decision makers
Objectives and approach
• General objective: inform decision making• Policy makers: climate, agriculture and food security policies• Private sector: benchmarking and identification of mitigation options• Consumer: food choices
• Specific objective: Produce estimates of GHG emissions for:• major dairy products and related services: milk, cheese, butter,
cream, milk powder, manure, and traction;• predominant dairy production systems (e.g. grass-based, mixed crop-
livestock); • main world regions and agro-ecological zone; and• major activity steps along the dairy chains.
• Approach:• Draw from national inventories and a growing body of literature• Methodological issues and preliminary results discussed with a group
of experts (WUR, INRA, SIK, ILRI, Danone, ITE, Agroscope, JRC)• Coupled with economic modeling – cost effectiveness analysis, poverty
and food security implications.
The choice of LCA
• Widely accepted in industry and agriculture as a method to evaluate environmental impacts of production, and identify the emission-intensive processes within a product’s life cycle.
• Ability to provide a holistic assessment of production processes in terms of resource use and other environmental impact, as well as to consider multiple parameters.
• Provides a framework to identify the most effectiveways to reduce environmental burdens.
• Capacity to evaluate how changes within a production process may affect the overall life-cycle balance, and therefore prevent shifting environmental problems from one phase of the life cycle to another.
A food-chain perspective of GHG emissions• Emissions from feed production
• chemical fertilizer fabrication• chemical fertilizer application• on-farm fossil fuel use• livestock-related deforestation• C release from ag. soils
• Emissions from livestock rearing• Methane from enteric fermentation• Methane and Nitrous Oxide from manure
• Post harvest emissions• slaughtering and processing• international transportation
ForestryEnergy
Transport and energy
Industry and energy
Industry and energy
Agriculture
Agriculture
Agriculture / livestock
IPCC attribution
Functional units
• Dairy-cattle production systems produce :• Edible products: meat and milk• Non-edible products and services: draught power, leather,
manure and capital.
• Functional Units: kg of FPC milk and bone- free meat at the farm gate.
• GHG emissions expressed in CO2eq.
System boundaries
System boundary
GHG emissions sources includedFrom cradle to farm gate• Processes for producing grass, feed crops, crop residues, by-
products, and concentrates (production and application of fertilizer and pesticides; application/deposition of manure; management of crops residues; energy used; transport of feed; changes in carbon stocks associated with land use change and related nitrogen losses.
• Enteric fermentation by ruminants.• Emissions from manure storage.• Energy consumption in animal production units for heating, cooling,
milking etc.
From farm gate to grave• Transport of milk and animals to dairies and slaughterhouses.• Processing of raw milk into commodities such as cooled milk,
yoghurt, cheese, butter, milk powder and bone-free meat.• Packaging and waste handling.• Refrigeration.• Transport of processed products to retailer.
GHG emissions calculations
• Follows IPPC Tier 2
• Input data• animal numbers• feed composition (digestibility, N content)• feed production parameters (land use change, fertilization,
mechanization)• manure management
• Land use change• 3 types of soybean considered • no emissions under constant management practices, 20 years time
frame
• Post farm gate• 7 commodities• statistics and literature review
Worldwide average of milk processing chains
Raw milk100 kg
Cheese production
Whey1
38.2 kg
Milk powder production
Fresh milk27.9 kg
Fresh and fermented milk production
Cream2
5.5 kg
1: whey is sold as feed and as whey powder, 2: cream is sold as such or processed into butter
Fermented milk
5.2 kg
Cheese4.5 kg
Milk powder2.2 kg
Water 16.5kg
45 % 20 % 35 %
GHG emissions calculations
• Emissions related to goods and services other than meat and milk calculated separately and deducted from overall emissions before attribution to meat and milk.
• Allocation rules• beef versus dairy: based on relative protein content.• manure: fertilizer value attributed to crops, remainder to
livestock; manure burnt exits the system after deposition.• financial and insurance services: no emissions allocated.• draught power: physical allocation based on extra longevity of
animals.
Allocation - Comparison of 6 techniques
Case study in France, Dollé and Bertrand, 2009
Farming systems
• Starting point: Livestock Production Systems (Seré & Steinfeld, 1996), adapted to the needs of the analysis
• Specific requirements:• Disaggregate calculation: capture main determinants of
emissions• Realistic / data availability• Support data management
Overview of system classification
Specie Techn. 1 Climate Country
Cattle
Dairy
Pure beef
Dual purpose
Techn. 2
Mixed
Grazing
Mixed
Grazing
Mixed
Grazing
AB...
aridhumidtemperatearidhumidtemperatearidhumidtemperatearidhumidtemperatearidhumidtemperatearidhumidtemperate
AB...
AB...
AB...
AB...
AB...
Data management and calculation
• Entirely based on GIS• Finest resolution: 0.0089 arc degrees, or ca. 10km x 10
km• Calculation implemented in GIS software• Results re-aggregated at various levels (e.g. species,
farming system, region, climatic zone)
Key parameters - Animals
Animal
Products & services:
milkmeateggsdraught manuresavings...
Type:
cattle: dairy/beefbuffalosheep&goatspigspoultry (chicken)
Herd composition
Live weightsbirthadult“market”
Rates reproductionreplacement diseasesdeath growth
Performance
Key parameters - feed basket
Feed basket Feed components:• grass• feed crops/crop residues• wastes• external feed
Virtual feedmill:• defining the animals ration • defining total energy and protein• calculating emissions related to the real
feed mill
Key parameters - Manure
Manure Manure management SystemPartitioning manure:• application in LPS• use out LPS, in arable crops or fuel
Estimated distribution of farming systems
Source: FAO-ILRI
Input data example (i)
Age at first calving (year)
Source: various
Input data example (ii)
Death rate of calves (%)
Source: various
Input data example (iii)
Estimated cattle distribution
Source: FAO
Input data example (iv)Estimated net primary productivity in areas dominated by
pasture (g. of C per m2 per year)
Source: Prince and Goward, 1995
Input data example (v) Estimated maize, wheat and barley production for animal feed
Source: FAO-IFPRI
The dairy herd model - overview
• Main input parameters • total number of cattle• number of milked cows• death rates, age at first calving fertility and prolificacy• cows per bull• milk production and composition, weight at slaughter
• Outputs• replacement young female• adult male and replacement young male• meat young female and male by difference: beef herd
Results of the herd model – the Netherlands
Herd constitution matters
Total cattle 3 730 000
Dairy herd Beef herd
Milked cows 1 450 000 139 500
Replacement female 1 025 032 43 860
Male for reproduction 14 500 5 580
Replacement male 15 117 5 695
Meat female 233 398 54 687
Meat male 682 998 59 613
Dairy related herd 3 421 045 308 955
Preliminary results - overview
• The contribution of the global milk production, processing and transportation to total anthropogenic emissions is estimated to range between 2.0 and 3.5 percent.
• The global average of emissions from milk production, processing and transport is estimated to range between 1.8 and 3.0 kg Co2eq. per kg of milk at farm gate.
• The overall emissions attributed to the dairy herd plus milk processing and transport activities are estimated to contribute between 3.0 and 5.1 percent of total anthropogenic emissions. This includes the production of milk and milk products, the production of meat from dairy related animals (old stock and young fattened stock), as well as the provision of draught power.
Preliminary results – regional variations
0
1
2
3
4
5
6
7
8
Nor
th A
mer
ica
Cen
tral &
Sou
thA
mer
ica
Wes
tern
Eur
ope
Eas
tern
Eur
ope
Rus
sian
Fed
erat
ion
Wes
t Asi
a &
Nor
ther
n A
frica
Sub
Sah
aran
Afri
ca
Sou
th A
sia
Eas
t Asi
a
Oce
ania
Wor
ld
GH
G e
mis
sion
s (k
g C
O2 e
quiv
alen
ts/k
g FP
CM
)
GHG deforestationGHG milk processingGHG milk production
GHG emissions per kg of FPCM, averaged by main regions and for the world.
Preliminary results – compared performances of dairy and beef herds
0.02
0.02
0.04
Beef herd
0.06
0.07
0.32
Dairy herd
Biological efficiency (kg of protein per /kg live weight)
Emissions to proteins (kg CO2eq./kg protein)Country
455128Brazil
543160India
17646The Netherlands
Beef herdDairy herd
Preliminary results - allocation
Milk
Live weight
0.000
5.000
10.000
15.000
20.000
25.000
65 70 75 80 85 90 95 100
% of emissions allocated to milk
kg C
O2e
q. p
er k
g m
ilk/m
eat
Preliminary results - sensitivity analysis
• Three hundred model runs performed for the Swedish situation
• Random fluctuation:• feed digestibility by -/+ 10%, • conversion factor for enteric
fermentation by -/+ 15%• emission factors for manure
and N application by -/+ 50% • energy use for feed
production by -/+ 25 % 0
2
4
6
8
10
12
14
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 More
GHG emissions (CO2-equivalents/kg milk)
Freq
uenc
y (%
)
Preliminary conclusions
• Emissions per unit of animal protein is substantially lower in the dairy system (including meat) than in the beef system (46 kgCO2eq./ kg protein vs 176 kg CO2eq./ kg protein in the Netherlands).• emissions from young fattened stock are similar• most difference come from cows and replacement females
• Emissions from milk and beef production need to be addressed in an integrated approach
• Milk emissions expressed per unit of output are marginally sensitive to allocation rule, contrary to emissions from meat production (total volume of output is greater for milk – factor 5 in the Netherlands).
• Feed digestibility and milk yield are key factors influencing emission level (C fluxes related to land use and land use changenot yet included)
Next steps
• Adapt the model for buffalo, small ruminants, pig and poultry;
• Complete data collection;• Improve sensitivity analysis;• Estimate C fluxes associated with land use (pastures) and
land use change (deforestation) ?
• Completion estimated by late 2010
