Processing GFM

The Processing Gap Filling Module calculates greenhouse gas emissions from food processing operations. It covers a wide range of processes including freezing, drying, baking, juice production, and many more. Processing can significantly affect the environmental footprint of food products, making this module essential for accurate life cycle assessments.

Quick Reference

Property	Description
Runs on	`ModeledActivityNode` with `FoodProductFlowNode` parent, or `FoodProcessingActivityNode` for combined products
Dependencies	`OriginGapFillingWorker`, `AttachFoodTagsGapFillingWorker`, `ConservationGapFillingWorker`, `MatchProductNameGapFillingWorker`, `AddClientNodesGapFillingWorker`
Key Input	Processing tags (glossary terms), product origin, nutritional composition
Output	Processing activity nodes with electricity consumption, raw material flows
Trigger	Product matched to specific glossary term combinations

When It Runs

The module triggers when:

A product has a ModeledActivityNode with a parent FoodProductFlowNode containing a product name
The product has matching glossary terms that trigger a processing model
All dependency GFMs have completed (origin, conservation, food tags)
The product is not matched to a non-food term

Key Output

The module adds processing activity nodes to the calculation graph:

Electricity consumption: Country-specific electricity market flows
Raw material flows: Input materials required for processing
Processing activity: Complete life cycle inventory for the process

Scientific Methodology

The Processing GFM uses two fundamentally different approaches based on the type of processing:

Simple Processing Models

Simple processing models add processing activities on top of existing base products. These processes:

Have a 1:1 input-to-output ratio (1 kg input = 1 kg output)
Scale linearly with product weight
Do not fundamentally change product properties

Examples: Cutting, shredding, mixing, cooling, grinding

Complex Processing Models

Complex processing models create new life cycle inventories for transformed products. These processes:

May have non-unit input-to-output ratios (for example, more raw material needed than output)
Depend on product characteristics (nutritional composition)
Can create fundamentally different products

Examples: Drying, freezing, juice production, jam production, yogurt production

Processing Model Categories

The Processing GFM implements several model types organized in a class hierarchy:

AbstractProcessingModel
    |
    +-- ProcessWithElectricityNode (electricity-only models)
    |       |
    |       +-- ProcessWithFixedElectricityAmount
    |       +-- FreezingProcessingModel
    |
    +-- ProcessWithBWNode (Brightway/Ecoinvent models)
            |
            +-- ProcessWithUnitRawMaterial
            +-- ProcessWithNonUnitRawMaterial
            +-- DryingProcessingModel

ProcessWithFixedElectricityAmount

Simple processes that only require electricity consumption, scaled by product weight.

Formula:

Electricity [kWh] = Fixed electricity amount [kWh/kg] * Production amount [kg]

Processes using this model:

Grinding
Cutting
Shredding
Fermenting
Puffing
Freeze-drying
Cooling
Mixing

ProcessWithUnitRawMaterial

Processes that use Brightway/Ecoinvent activity nodes with a 1:1 raw material ratio.

Formula:

Raw material [kg] = Production amount [kg]
Processing flow = Brightway activity * Production amount

Processes using this model:

Dried Fruit Grinding
Carbonating
Smoking
Roasting
Chopping
Heating
Transportation

ProcessWithNonUnitRawMaterial

Processes that use Brightway/Ecoinvent activity nodes where raw material input differs from output.

Formula:

Raw material [kg] = Raw material ratio * Production amount [kg]
Processing flow = Brightway activity * Production amount

Processes using this model:

Baking
Jam production
Juice production
Yogurt production

Implemented Processes

The following table lists all implemented processing types with their model class:

Process Name	Model Type	Description
Grinding	ProcessWithFixedElectricityAmount	Mechanical size reduction
Dried Fruit Grinding	ProcessWithUnitRawMaterial	Grinding of dried fruits
Carbonating	ProcessWithUnitRawMaterial	Adding carbon dioxide
Cutting	ProcessWithFixedElectricityAmount	Mechanical cutting
Shredding	ProcessWithFixedElectricityAmount	Mechanical shredding
Fermenting	ProcessWithFixedElectricityAmount	Fermentation process
Smoking	ProcessWithUnitRawMaterial	Smoking preservation
Puffing	ProcessWithFixedElectricityAmount	Expansion by heat/pressure
Freeze-drying	ProcessWithFixedElectricityAmount	Lyophilization
Cooling	ProcessWithFixedElectricityAmount	Refrigeration
Baking	ProcessWithNonUnitRawMaterial	Thermal processing
Jam production	ProcessWithNonUnitRawMaterial	Fruit preserve making
Juice production	ProcessWithNonUnitRawMaterial	Liquid extraction
Yoghurt production	ProcessWithNonUnitRawMaterial	Dairy fermentation
Roasting	ProcessWithUnitRawMaterial	Dry heat cooking
Chopping	ProcessWithUnitRawMaterial	Mechanical cutting
Heating	ProcessWithUnitRawMaterial	Thermal treatment
Drying	DryingProcessingModel	Moisture removal
Freezing	FreezingProcessingModel	Freezing preservation
Mixing	ProcessWithFixedElectricityAmount	Combining ingredients
Transportation	ProcessWithUnitRawMaterial	Transport to processing

Freezing Processing Model

The freezing model calculates energy requirements based on product composition using thermodynamic principles.

Energy Calculation Formula

The total freezing energy is calculated as:

Q_total = Q_heat_load + Q_refrigerator

Where:

Q_heat_load: Energy to cool and freeze the product
Q_refrigerator: Energy for refrigeration system operation

Product Heat Load

Q_product = W_pr / t_pr * (c_u * (T_i - T_if) + L * 1000 + c_fr * (T_if - T_out))

Where:

W_pr: Product mass (kg)
t_pr: Residence time (s)
c_u: Unfrozen specific heat (J/kg K)
T_i: Initial temperature (20 degrees Celsius)
T_if: Initial freezing temperature (-1.7 degrees Celsius)
L: Latent heat of freezing (kJ/kg)
c_fr: Frozen specific heat (J/kg K)
T_out: Outlet temperature (-18 degrees Celsius)

Nutritional Composition Impact

The model uses water and fat content to calculate thermophysical properties:

# Solid fraction
x_s = 1 - x_w - x_f

# Bound water fraction (limited by total water)
x_b = min(0.25 * x_s, x_w)

# Ice fraction
x_i = (x_w - x_b) * (1 - T_if / T)

# Latent heat of freezing
L = 334 * x_i  # kJ/kg

Heat Load Components

For continuous air blast freezing:

Component	Percentage
Product	60%
Fans	20%
Pull-down	0%
Defrost	15%
Other	5%

Physical Constants

Parameter	Value	Unit
Frozen food temperature (T)	-23	degrees Celsius
Center temperature (T_c)	-15	degrees Celsius
Cooling medium temperature (T_f)	-32	degrees Celsius
Coefficient of Performance (COP)	2.75	-
Water density (p_w)	1000	kg/m3
Fat density (p_f)	930	kg/m3
Solids density (p_s)	1450	kg/m3
Water specific heat (c_w)	4180	J/kg K
Ice specific heat (c_i)	2110	J/kg K
Latent heat of water (L1)	334	kJ/kg

Drying Processing Model

The drying model calculates water evaporation based on the nutrient upscale ratio from the Nutrient Subdivision GFM.

Water Loss Calculation

Water to evaporate [kg] = (Upscale ratio - 1.0) * Production amount [kg]

The upscale ratio represents how much fresh product is needed to produce 1 kg of dried product.

Evaporation Formula

The percentage of water loss is calculated as:

Water loss % = 100 * {1 - [(Final water % / 100) * (Total weight - Water weight)] /
                         [(1 - Final water % / 100) * Water weight]}

For dried products:

Final water content: 0-20% (dry weight content 80-100%)
Based on Afolabi 2014 research

Integration with Nutrient Subdivision GFM

The drying model works in conjunction with the Nutrient Subdivision GFM:

Nutrient Subdivision GFM: Calculates the upscale ratio based on water content difference
Drying GFM: Adds the evaporation energy based on the upscale ratio
Processing Activity: Attaches to the existing upscaling node

Drying Process Variants

Process	Glossary Term	Description
General drying	J0116	Default drying process
Dairy drying	J0116 + dairy terms	Milk powder production
Grain drying	J0116 + grain terms	Cereal drying
Legume drying	J0116 + legume terms	Bean/pea drying
Nut drying	J0116 + nut terms	Nut dehydration
Oilseed drying	J0116 + oilseed terms	Seed drying
Spice drying	J0116 + spice terms	Herb/spice drying
Fruit drying	J0116 + fruit terms	Fruit dehydration
Vegetable drying	J0116 + vegetable terms	Vegetable drying

Implementation Details

Trigger Tag System

Processing is triggered by specific combinations of FoodEx2 glossary terms. The configuration is stored in a Google Sheet with:

Columns B-D: Glossary term combinations
Column E: Brightway ID or electricity amount
Column F: GFM involved
Column G: Comments (location: GLO, RER, etc.)

Tag Priority

When multiple tag combinations match, the most specific one is selected:

# The processing with a higher number of tags is prioritized
# as they are more specific for this particular product
if len(processing_xid_and_tags.tags) > len(viable_processing[0].tags):
    viable_processing = [processing_xid_and_tags]

Country-Specific Electricity

The module uses country-specific electricity markets:

# Low voltage electricity for processing
electricity_node = electricity_node_by_country.low_voltage.get(country_code)

# Fallback to global if country not available
if electricity_node is None:
    electricity_node = electricity_node_by_country.low_voltage.get("GLO")

Transportation Addition

Transportation to processing facilities is added when:

The glossary term Z0001 (Adjunct Characteristics of Food Not Known) is present
The process creates a new LCI (not just adding processing on top)
Excludes: Cooling and Freezing (to avoid double-counting)

# Add transport term for new inventory creation
if "Z0001" in filtered_tag_term_xids:
    filtered_tag_term_xids.add("EOS_Transportation")

Calculation Graph Integration

Node Structure

The Processing GFM creates the following graph structure:

FoodProductFlowNode (parent)
    |
    +-- FoodProcessingActivityNode (processing activity)
            |
            +-- FoodProductFlowNode (raw material flow)
            |       |
            |       +-- ModeledActivityNode (original Brightway node)
            |
            +-- PracticeFlowNode (processing flow)
                    |
                    +-- ModeledActivityNode (electricity or Brightway process)

Mutation Sequence

Remove edge between parent flow and original Brightway node
Create FoodProcessingActivityNode
Add edge from parent flow to processing activity
Duplicate parent flow as raw material flow
Update amounts on raw material flow
Add edge from raw material flow to original Brightway node
Create PracticeFlowNode for processing
Connect processing flow to electricity/Brightway process

Calculation Example

Scenario: 1 kg of frozen peas, produced in Germany

Step 1: Determine Nutritional Composition

From nutrient data:

Water content (x_w): 78.9%
Fat content (x_f): 0.4%

Step 2: Calculate Thermophysical Properties

# Solid fraction
x_s = 1 - 0.789 - 0.004 = 0.207

# Bound water fraction
x_b = min(0.25 * 0.207, 0.789) = 0.052

# Ice fraction at -23 degrees Celsius
x_i = (0.789 - 0.052) * (1 - (-1.7) / (-23)) = 0.682

# Latent heat
L = 334 * 0.682 = 227.8 kJ/kg

Step 3: Calculate Energy Requirements

# Unfrozen specific heat
c_u = 4180 * 0.789 + 1600 * 0.207 + 1900 * 0.004 = 3638 J/kg K

# Frozen specific heat (simplified)
c_fr = 2650 J/kg K

# Product heat load
Q_product = 1 * (3638 * (20 - (-1.7)) + 227800 + 2650 * ((-1.7) - (-18)))
          = 79,005 + 227,800 + 43,185 = 349,990 J

# Total including freezer components (60% product load)
Q_heat_total = 349,990 / 0.6 = 583,317 J

# Refrigerator energy (COP = 2.75)
Q_refrigerator = 583,317 / 2.75 * 1.175 = 249,425 J

# Total energy
Q_total = 583,317 + 249,425 = 832,742 J = 0.231 kWh/kg

Step 4: Apply Country-Specific Electricity

Emissions = 0.231 kWh * German electricity mix [kg CO2eq/kWh]

Final Output

The module adds:

Processing activity node for freezing
Electricity flow (0.231 kWh) connected to German low-voltage electricity market

Known Limitations

Model Coverage

Not all processing types have dedicated models
Some processes use simplified electricity-only approaches
Complex multi-step processes may not be fully captured

Data Gaps

Default electricity amounts may not reflect actual industrial practice
Some product-specific processing parameters are unavailable
Regional variations in processing efficiency not captured

Double-Counting Risk

Products with existing processing in the base inventory may have processing counted twice. A potential solution involves tagging Brightway activities that already include processing.

Nutritional Dependency

For drying and freezing:

Models require nutritional composition data
If nutrients unavailable, default values are used
Default assumes 90% dry weight content for drying

Process String Matching

The following processing strings are automatically matched to their respective processes:

# Drying variants
"drying", "dairy drying", "grain drying", "legume drying",
"nut drying", "oilseed drying", "spice drying",
"fruit drying", "vegetable drying"

# Production processes
"fruit jam production", "fruit yoghurt production (cow milk)",
"fruit yoghurt production (soy milk)", "fruit juice production",
"fruit juice concentrate production"

# Mechanical processes
"grinding", "dried fruit grinding", "cutting", "shredding",
"tree nut chopping", "chopping", "mixing"

# Thermal processes
"roasting", "nut roasting", "baking", "heating",
"cooking with fat or oil", "uht pasteurization"

# Preservation processes
"freezing", "cooling", "smoking", "fermenting", "freeze-drying"

# Other processes
"artificial carbonation", "puffing",
"transportation for processing", "cooled transportation for processing"

References

Afolabi, I.S. (2014). Moisture Migration and Bulk Nutrients Interaction in a Drying Food Systems. Food and Nutrition Sciences, 5(8), 692-714. http://dx.doi.org/10.4236/fns.2014.58080
ASHRAE Handbook - Refrigeration (2018). Chapter 19: Thermal Properties of Foods.
Camenzind, R. (2016). Food Processing LCA Data. ZHAW Report for Eaternity.
Ecoinvent Database v3.6. Swiss Centre for Life Cycle Inventories.
FoodEx2 Classification System. European Food Safety Authority (EFSA).

Quick Reference​

When It Runs​

Key Output​

Scientific Methodology​

Simple Processing Models​

Complex Processing Models​

Processing Model Categories​

ProcessWithFixedElectricityAmount​

ProcessWithUnitRawMaterial​

ProcessWithNonUnitRawMaterial​

Implemented Processes​

Freezing Processing Model​

Energy Calculation Formula​

Product Heat Load​

Nutritional Composition Impact​

Heat Load Components​

Physical Constants​

Drying Processing Model​

Water Loss Calculation​

Evaporation Formula​

Integration with Nutrient Subdivision GFM​

Drying Process Variants​

Implementation Details​

Trigger Tag System​

Tag Priority​

Country-Specific Electricity​

Transportation Addition​

Calculation Graph Integration​

Node Structure​

Mutation Sequence​

Calculation Example​

Step 1: Determine Nutritional Composition​

Step 2: Calculate Thermophysical Properties​

Step 3: Calculate Energy Requirements​

Step 4: Apply Country-Specific Electricity​

Final Output​

Known Limitations​

Model Coverage​

Data Gaps​

Double-Counting Risk​

Nutritional Dependency​

Process String Matching​

References​

Quick Reference

When It Runs

Key Output

Scientific Methodology

Simple Processing Models

Complex Processing Models

Processing Model Categories

ProcessWithFixedElectricityAmount

ProcessWithUnitRawMaterial

ProcessWithNonUnitRawMaterial

Implemented Processes

Freezing Processing Model

Energy Calculation Formula

Product Heat Load

Nutritional Composition Impact

Heat Load Components

Physical Constants

Drying Processing Model

Water Loss Calculation

Evaporation Formula

Integration with Nutrient Subdivision GFM

Drying Process Variants

Implementation Details

Trigger Tag System

Tag Priority

Country-Specific Electricity

Transportation Addition

Calculation Graph Integration

Node Structure

Mutation Sequence

Calculation Example

Step 1: Determine Nutritional Composition

Step 2: Calculate Thermophysical Properties

Step 3: Calculate Energy Requirements

Step 4: Apply Country-Specific Electricity

Final Output

Known Limitations

Model Coverage

Data Gaps

Double-Counting Risk

Nutritional Dependency

Process String Matching

References