Deep Technical Guide: GHG Protocol Scope 1, 2, and 3 Calculation Logic for Multinational Corporations

1) Core Accounting Architecture

1.1 Organizational Boundary (Who is included)

Choose one consolidation approach and apply consistently:

• Equity share: account for emissions in proportion to equity ownership.
• Financial control: account for 100% where financial control exists.
• Operational control: account for 100% where operational control exists (most common for MNC inventories).

Rule: Boundary choice affects all scopes and all geographies. Maintain legal-entity-to-site mapping and ownership/control metadata by reporting period.

1.2 Operational Boundary (What is included)

• Scope 1: direct emissions from owned/controlled sources.
• Scope 2: indirect emissions from purchased energy (electricity, steam, heat, cooling).
• Scope 3: all other indirect value-chain emissions (15 categories).

For MNCs, operational boundary must be linked to:

• ERP chart of accounts,


• procurement/supplier master,


• travel and logistics systems,


• fixed asset registry,


• utility meters/contracts.

1.3 General Calculation Equation

For any emission source \(i\):
\[
E_i = AD_i \times EF_i \times (1 - ER_i) \times GWP_g
\]
Where:

• \(AD\): activity data (fuel, kWh, ton-km, spend, etc.)


• \(EF\): emission factor per activity unit (often by gas or CO2e)


• \(ER\): oxidation/carbon capture/removal efficiency adjustment when applicable


• \(GWP\): global warming potential for gas \(g\), per chosen assessment report and reporting requirement

If EF is gas-resolved:
\[
E_{CO2e} = \sum_g (AD \times EF_g \times GWP_g)
\]

1.4 Data Hierarchy (best to worst)

1. Primary measured activity (metered fuel/energy/production data)
2. Supplier-specific cradle-to-gate factors / product carbon footprints
3. Physical-model or engineering estimates
4. Spend-based proxy factors
5. Industry-average assumptions

Track data quality score per line item.

1.5 Temporal and Currency Normalization

• Convert all activity to reporting period (monthly close preferred).
• For spend methods: convert local currency to reporting currency with documented FX policy (transaction-date or period-average), then apply factor currency basis consistently.
• Handle leap year/partial-period acquisitions explicitly.

1.6 Biogenic Carbon and Land-use

• Report biogenic CO2 separately from fossil CO2e totals.


• CH4/N2O from biomass combustion are still included in CO2e totals.


• Land-use and removals follow separate accounting frameworks; avoid netting inside gross inventory unless standard explicitly allows.

2) Scope 1: Direct Emissions Calculation Logic

Typical sub-sources for MNCs:

1. Stationary combustion


2. Mobile combustion (fleet)


3. Process emissions


4. Fugitive emissions (refrigerants, SF6, methane leaks)

2.1 Stationary Combustion

\[
E = Fuel\_Quantity \times NCV \times EF_{fuel,gas}
\]
Or direct EF per unit fuel.

Technical points:

• Prefer fuel purchase + stock reconciliation or meter data.


• Distinguish HHV vs LHV/NCV basis and align with EF basis.


• Apply oxidation factor if protocol/factor requires.


• Country/site-specific EFs where available.

2.2 Mobile Combustion

Two approaches:

• Fuel-based (preferred): liters/gallons by fuel type.


• Distance-based (fallback): km by vehicle class × fuel economy assumptions × EF.

Include:

• owned and controlled vehicles only (Scope 1),


• refrigerant leaks from transport cooling units if controlled.

2.3 Process Emissions

Use stoichiometric or mass-balance models:
\[
E_{CO2} = \sum_j (Material_j \times Carbon\ Content_j \times Conversion\ Factor_j)
\]
Examples: clinker production, lime, ammonia, metals.

2.4 Fugitive Emissions

Refrigerants:

\[ E = (Charge_{start} + Purchases - Recoveries - Charge_{end}) \times GWP \] Alternative screening: \[ E = Installed\ Charge \times Leak\ Rate \times GWP \] for missing records.

SF6 / CH4 leakage:

Use equipment-level leakage rates or measured top-ups.

3) Scope 2: Purchased Energy Calculation Logic

Report both:

1. Location-based (grid-average factors)


2. Market-based (contractual instruments + supplier-specific data)

3.1 Location-Based Method

\[
E_{LB} = \sum_s (kWh_s \times EF_{grid,location,s})
\]

• Use subnational grid EF where possible (state/province balancing area).


• For steam/heat/cooling: supplier/region thermal EF.

3.2 Market-Based Method

\[
E_{MB} = \sum_s (kWh_s \times EF_{contractual,s})
\]
Factor hierarchy typically:

1. Supplier-specific emission rate


2. Energy Attribute Certificates (EACs: RECs, GOs, I-RECs), PPAs matched to load


3. Residual mix


4. Grid average (if above unavailable, per guidance)

Quality criteria controls:

• Vintage matching (same reporting year)


• Geographic market boundary consistency


• Exclusive claim (no double counting of attributes)


• Correct certificate retirement evidence

3.3 Scope 2 Data Model for MNCs

Per site-month:

• meter kWh,


• utility supplier,


• contract type,


• EAC quantity/vintage/region,


• residual mix EF source.

Then calculate LB and MB in parallel; prevent cross-netting between sites unless certificate allocation rules allow.

4) Scope 3: Value-Chain Calculation Logic (15 Categories)

Scope 3 requires category-by-category method selection. Use hybrid logic: supplier-specific where material, activity-based where available, spend-based for tail spend.

\[
E_{cat} = \sum_{line} AD_{line} \times EF_{line,method}
\]

4.1 Upstream Categories (1–8)

Category 1: Purchased goods and services

Methods:

• Supplier-specific PCF (preferred): quantity × supplier EF
• Activity-based: mass/units × LCA factor
• Spend-based: spend × EEIO factor
• Hybrid: top suppliers primary data + spend model for remainder

Controls:

• map SKUs/material groups to emission factor taxonomy,


• avoid counting capital goods here (send to Cat 2),


• ensure cradle-to-gate boundary alignment.

Category 2: Capital goods

CapEx-based life-cycle factors for machinery/buildings/IT.
\[
E = \sum (CapEx_{asset} \times EF_{capital\ class})
\]
or quantity/material BOM-based LCAs for major projects.

Category 3: Fuel- and energy-related activities (not in Scope 1/2)

Includes:

• upstream extraction/production/transport of purchased fuels,
• T&D losses of purchased electricity,
• WTT emissions for electricity/steam.

\[ E = Fuel/Energy\ Activity \times EF_{upstream/T\&D} \]

Category 4: Upstream transportation and distribution

\[ E = \sum (Mass \times Distance \times EF_{mode,load,region}) \] or spend/logistics-provider data. Include third-party warehousing energy allocated by floor area, pallet-days, or throughput.

Category 5: Waste generated in operations

\[ E = \sum (Waste\ by\ type \times Treatment\ route\ EF) \] Route-specific EFs: landfill, incineration, recycling, composting, wastewater treatment.

Category 6: Business travel

Hierarchy:

1. carrier-specific flight/train data with radiative forcing policy stated,
2. distance-class factors,
3. spend proxies.

For hotels: room-night × country/hotel-class EF.

Category 7: Employee commuting

\[ E = \sum (Employees \times Commute\ distance \times Mode\ split \times Workdays \times EF) \] Use survey-based mode split; include remote work if policy requires.

Category 8: Upstream leased assets

If not in Scope 1/2 due to boundary approach: \[ E = Energy/Fuel_{leased} \times EF \] Need lease metadata by IFRS/GAAP and control approach.

4.2 Downstream Categories (9–15)

Category 9: Downstream transportation and distribution

Same logic as Cat 4 but after point of sale. Use distributor/carrier data where possible.

Category 10: Processing of sold products

\[ E = \sum (Sold\ intermediate\ product\ quantity \times Processing\ EF_{customer\ stage}) \] Requires assumptions on customer process routes and yields.

Category 11: Use of sold products

Most material for appliances, vehicles, electronics, fuels. \[ E = Units\ sold \times Lifetime\ energy\ use \times EF_{use\ phase\ energy} \] Key assumptions:

• average lifetime,
• usage intensity profiles by region,
• grid decarbonization trajectory choice (static vs dynamic, disclose method).

Category 12: End-of-life treatment of sold products

\[
E = \sum (Material\ mass \times EoL\ route\ share \times EF_{route})
\]
Use region-specific waste route mixes.

Category 13: Downstream leased assets

Energy/fuel consumed by leased-out assets during lease term.

Category 14: Franchises

Franchisee operational emissions not in Scopes 1/2.

Category 15: Investments

Financed emissions methodology (e.g., attribution factor): \[ E_{financed} = \sum (EVIC/loan\ share\ attribution \times Investee\ emissions) \] Data quality strongly depends on investee disclosures and model estimates.

5) Method Selection Logic for Multinationals

5.1 Materiality-driven tiering

• Rank suppliers/categories by expected emissions and spend.
• Apply primary data programs to top contributors.
• Use modeled factors for long tail.

Example tiering:

• Tier A (top 70–80% emissions): supplier-specific/activity-based


• Tier B (next 15–20%): hybrid


• Tier C (tail): spend-based

5.2 Decision Tree (practical)

1. Is primary activity data available and auditable? → use activity-based.


2. Is supplier cradle-to-gate EF/PCF available with boundary metadata? → use supplier-specific.


3. Is physical proxy available (mass, ton-km, kWh)? → use activity proxy.


4. Else use spend-based EF with conservative assumptions.

6) Emission Factors: Governance and Versioning

Maintain centralized EF library with:

• source (IPCC, IEA, DEFRA, EPA, ecoinvent, national inventories),


• geography, year, sector coverage,


• unit basis and calorific basis,


• gas breakdown and GWP set,


• validity period and version ID.

Never overwrite historical factor versions; recalculate only under formal base-year restatement policy.

7) Allocation, Avoiding Double Counting, and Consolidation

7.1 Internal double counting

Prevent overlap:

• Scope 1 fuel combustion not repeated in Scope 3 Cat 3 combustion portion.
• Capital goods excluded from Cat 1.
• Intercompany transactions eliminated in consolidated reporting where required.

7.2 Value-chain double counting

Cross-company double counting is expected in Scope 3 and not an error; disclose this clearly.

7.3 Allocation rules

Use physically causal allocators where possible:

• mass, energy content, machine hours, floor area, revenue (last resort).

Document allocator per process.

8) Uncertainty Quantification and Data Quality

For each emissions line:

• activity uncertainty (%),


• EF uncertainty (%),


• model uncertainty (%).

Propagate (independent approximation):
\[
U_{total} \approx \sqrt{U_{AD}^2 + U_{EF}^2 + U_{model}^2}
\]

Portfolio uncertainty via Monte Carlo recommended for large Scope 3 categories.

Track data quality dimensions:

• technological representativeness,


• temporal,


• geographic,


• completeness,


• reliability.

9) Base Year, Recalculation, and M&A Handling

Recalculate base year when structural changes are significant:

• acquisitions/divestments,


• outsourcing/insourcing,


• methodological changes,


• major data error correction.

For MNC M&A:

• define inclusion rule by close date,


• pro-rate partial year where policy requires,


• maintain pre/post-acquisition audit trail.

10) Implementation Blueprint (System Level)

10.1 Data pipeline

1. Ingest: ERP, AP, utility, fuel cards, TMS, HR, travel, supplier portal.
2. Normalize: units, currency, calendar.
3. Classify: scope/category mapping rules engine.
4. Factor match: geography-year-method-aware lookup.
5. Calculate: line-level CO2e (gas-level where possible).
6. QA/QC: outlier checks, variance to prior year, intensity sanity checks.
7. Consolidate: legal entity → country → region → group.
8. Report: Scope 1, Scope 2 LB/MB, Scope 3 by category, uncertainty, method mix.

10.2 Pseudocode (simplified)

```text
for line in activity_data:
boundary = map_org_boundary(line.entity, reporting_policy)
if not boundary.included: continue

scope_cat = classify_scope_category(line)
method = select_method(line, data_quality_rules, materiality_rules)

ef = fetch_emission_factor(
scope_cat, method, geography=line.country,
year=reporting_year, unit=line.unit, contract=line.contract_type
)

emissions = convert_units(line.activity, ef.unit_basis) * ef.value

if ef.gas_breakdown:
emissions = sum(gas_amount * gwp[gas] for gas_amount in emissions.by_gas)

store(line.id, scope_cat, method, emissions, ef.version, dq_score(line))
```

11) High-Risk Technical Pitfalls

• Mixing HHV/LHV fuel bases.
• Using grid-average factors for market-based Scope 2 claims with EACs.
• Currency-year mismatch in spend-based Scope 3.
• Applying supplier PCFs with inconsistent boundaries (cradle-to-gate vs gate-to-gate).
• Missing refrigerant bank reconciliation.
• Not separating biogenic CO2.
• Inconsistent treatment of leased assets with boundary approach.
• No residual mix usage where required for unbundled claims.

12) Minimum Disclosure Set for Defensible Inventories

• Organizational boundary method and changes.


• Scope 1 breakdown by source type and gases.


• Scope 2 LB and MB with instrument details.


• Scope 3 categories, included/excluded, and estimation methods share (% primary vs secondary).


• EF sources, versions, GWPs used.


• Base year and recalculation triggers.


• Uncertainty approach and key assumptions (lifetimes, usage profiles, allocation keys).

Bottom line

For multinationals, high-quality GHG accounting is a data engineering + methodological governance problem: line-level activity data, strict boundary logic, dual Scope 2 reporting, hybrid Scope 3 methods, versioned factors, and auditable uncertainty/disclosure controls.