TCFD Compliance¶

Guide for climate-related financial disclosures under TCFD recommendations.

What is TCFD?¶

The Task Force on Climate-related Financial Disclosures provides recommendations for consistent climate-related financial risk disclosures. Many jurisdictions now require TCFD-aligned reporting.

TCFD Pillars¶

TCFD organizes disclosures into four pillars:

1. Governance — Board and management oversight
2. Strategy — Climate risks and opportunities
3. Risk Management — Identification and management processes
4. Metrics and Targets — Emissions data and reduction targets

The Green module primarily supports Metrics and Targets.

Metrics Requirements¶

TCFD recommends disclosing:

Cross-Industry Metrics¶

Metric	Green Module Support
Scope 1 emissions	N/A (direct emissions)
Scope 2 emissions	N/A (purchased energy)
Scope 3 emissions	✅ Category 1 (AI inference)
Climate-related risks	⚠️ Manual assessment needed
Climate-related opportunities	⚠️ Manual assessment needed

Industry-Specific Metrics¶

For technology/software companies:

Metric	Green Module Support
Energy consumption	✅ Derived from factors
Carbon intensity per unit	✅ Per token, per call
Emissions from cloud services	✅ AI inference portion

Export Format¶

Generate TCFD-compliant export:

openclaw green export --format tcfd --period 2025 --baseline 2024

Output Structure:

{
  "absoluteEmissions": {
    "scope3Cat1_tCO2eq": 0.01245,
    "reportingPeriod": "2025",
    "comparisonToBaseline": {
      "baseYear": 2024,
      "changePercent": 24.5
    }
  },
  "carbonIntensity": {
    "perMillionTokens_gCO2eq": 142.29,
    "perApiCall_gCO2eq": 6.74
  },
  "targets": [
    {
      "target": {
        "targetId": "target-123",
        "name": "Net Zero 2030",
        "baseYear": 2025,
        "targetYear": 2030,
        "targetReductionPercent": 50,
        "pathway": "1.5C"
      },
      "currentYearEmissionsGrams": 12450,
      "progressPercent": 75.1,
      "onTrack": true,
      "projectedEndYear": 2028
    }
  ],
  "historicalTrend": [
    { "period": "2025-Q1", "emissions_tCO2eq": 0.0028 },
    { "period": "2025-Q2", "emissions_tCO2eq": 0.0031 },
    { "period": "2025-Q3", "emissions_tCO2eq": 0.0033 },
    { "period": "2025-Q4", "emissions_tCO2eq": 0.0033 }
  ]
}

Intensity Metrics¶

TCFD emphasizes intensity metrics for comparability:

Per Million Tokens¶

Intensity = Total CO₂ (g) / Total Tokens (M)

Enables comparison across: - Different time periods - Different organizations - Industry benchmarks

Per API Call¶

Intensity = Total CO₂ (g) / Total Calls

Alternative metric tied to business activity.

Uncertainty Disclosure¶

TCFD expects disclosure of data limitations:

openclaw green intensity

Shows uncertainty range (e.g., ±30%) based on data quality.

Target Disclosure¶

TCFD recommends disclosing:

1. Baseline year and emissions
2. Target year and reduction goal
3. Progress toward target
4. Methodology for target setting

Set targets using SBTi framework:

openclaw green targets:add \
  --name "50% Reduction by 2030" \
  --base-year 2025 \
  --target-year 2030 \
  --reduction 50 \
  --pathway 1.5C

Historical Trend¶

TCFD values trend data for understanding trajectory:

"historicalTrend": [
  { "period": "2025-Q1", "emissions_tCO2eq": 0.0028 },
  { "period": "2025-Q2", "emissions_tCO2eq": 0.0031 },
  { "period": "2025-Q3", "emissions_tCO2eq": 0.0033 },
  { "period": "2025-Q4", "emissions_tCO2eq": 0.0033 }
]

Shows quarterly progression for the reporting period.

Example Disclosure¶

Metrics and Targets Section¶

## Climate Metrics

### Greenhouse Gas Emissions

| Scope | 2024 | 2025 | Change |
|-------|------|------|--------|
| Scope 3 Cat 1 (AI) | 10.0 kg | 12.4 kg | +24% |

### Carbon Intensity

| Metric | 2024 | 2025 | Change |
|--------|------|------|--------|
| g CO₂e per M tokens | 140.2 | 142.3 | +1.5% |
| g CO₂e per API call | 6.67 | 6.74 | +1.0% |

Intensity metrics remained stable despite 24% growth in absolute
emissions, reflecting efficiency improvements from model selection
and caching strategies.

### Emission Reduction Targets

| Target | Base Year | Target Year | Reduction | Progress |
|--------|-----------|-------------|-----------|----------|
| Net Zero 2030 | 2025 | 2030 | 50% | On track |

We have set a science-based target aligned with the 1.5°C pathway
to reduce AI inference emissions 50% by 2030 from a 2025 baseline.

### Methodology and Limitations

Emissions calculated using per-token factors from academic research.
Data quality: Average secondary data (DQS 3).
Uncertainty: ±30%.

Limitations include lack of supplier-specific data and unknown
data center locations. We are engaging providers to improve data quality.

Scenario Analysis¶

TCFD recommends scenario analysis for climate risks. For AI emissions:

Physical Risks¶

• Data center cooling costs may increase
• Extreme weather may disrupt cloud services
• Water scarcity may constrain operations

Transition Risks¶

• Carbon pricing may increase API costs
• Regulation may require provider transparency
• Customer demand for low-carbon AI may shift market

Opportunities¶

• Efficient models reduce costs and emissions
• Carbon-aware scheduling optimizes timing
• Provider selection based on renewable energy

Related Standards¶

• GHG Protocol — Emission calculation methodology
• CDP — Aligned disclosure framework
• SBTi — Target-setting methodology