Sustainability & Emissions Report

Purpose

Turn raw plant utility, production, and material data into a defensible sustainability and emissions report for customers, regulators, and internal leadership. The report structures Scope 1, Scope 2, and Scope 3 emissions at a facility and product level, flags CBAM-relevant exposure on covered goods, benchmarks energy intensity per unit of output, and produces both a compliance-ready submission and a one-page leadership summary. The output is designed to stand up to third-party verification and to replace the generic industry averages that trigger punitive CBAM tariffs starting with the 2026 compliance period.

What This Is / Is Not

This skill is two tools sharing one emissions taxonomy, not one monolithic report imposed on every request:

• Pass 1 — Quick Scope-1+2 Screen. A config-driven first read that needs only the period, a fuel/electricity activity summary, and production output. It pre-populates emission factors, boundary, Scope 2 method, GWP set, and CBAM flag from config.yml and returns a provisional Scope 1 + Scope 2 total, an intensity-vs-prior-period delta, and a ranked list of the inputs you must still gather for a verifiable Pass 2. Use it for the quarterly leadership KPI glance, a fast answer to "where are we tracking," or to scope the work before a full close. It is an estimate, not a submission — every figure is labeled "screen — confirm in Pass 2."
• Pass 2 — Full verifiable report. The complete Scope 1/2/3 + CBAM + reconciliation + data-quality apparatus below, run when the output must survive third-party assurance, a CBAM declaration, or a customer PCF request.

Pass 1 is not a verifiable disclosure, does not anchor a CBAM declaration, and never substitutes for meter reconciliation. Pass 2 is not required for an internal glance. Pick the pass the request actually needs; do not run the full close to answer a leadership KPI question.

Config Pre-Population

Before asking the user for anything, load config.yml and bind the following keys so the user supplies only period-specific activity data, not standing facts they have already configured:

State which keys were found and which were absent. An absent key becomes a gaps-block line ("grid factor not in config — supply eGRID subregion or factor"), never a silent default. If config.yml is missing entirely, fall back to the full input list below and note that personalization is unavailable.

When to Use

Use this skill any time a plant owes an emissions or sustainability disclosure, including:

• Quarterly internal sustainability KPI review for leadership or ESG steering committee
• Customer request for carbon data on covered SKUs (automotive OEMs, large retailers, aerospace primes all now request this)
• CBAM quarterly / annual declaration on cement, iron and steel, aluminium, fertilisers, electricity, and hydrogen exports to the EU (definitive phase live since 2026-01-01; first financial settlement in 2027)
• CSRD / IFRS S2 / SEC climate disclosure support — facility inputs into the corporate filing
• Energy management reviews — ISO 50001 surveillance, DOE Better Plants, utility rebate filings
• Grant or incentive applications — IRA 45X, state energy-efficiency programs, green-bond investor letters
• Pre-RFQ customer qualification when a prospect asks for a carbon estimate on a quoted part

The skill is especially valuable when the plant has utility data in one system, production data in another, and material data in a third, and the sustainability coordinator is stitching them together by hand in a spreadsheet.

Required Input

Provide the following. Any missing input should be listed in the gaps block rather than estimated silently.

1. Reporting period and scope — Start and end dates, facility or facilities in scope, product SKUs or CBAM CN codes in scope, reporting standard (GHG Protocol, ISO 14064-1, CBAM Implementing Regulation, CSRD/ESRS E1, IFRS S2)
2. Scope 1 activity data — On-site fuel combustion by fuel type (natural gas therms, propane gallons, diesel gallons, fuel oil, coal, biomass), process emissions (CO2 from calcination, SF6 leakage, refrigerant top-ups with refrigerant type), mobile-source fuels for owned vehicles and forklifts
3. Scope 2 activity data — Grid-purchased electricity (kWh) by meter, purchased steam / heat / cooling, renewable energy certificates (RECs) retired, on-site solar or wind generation delivered behind the meter, PPA allocation with bundled-vs-unbundled flag
4. Scope 3 activity data (as scoped) — Purchased goods and services by spend or mass, upstream transportation (tonne-km or spend), waste generated (tons by disposal route), business travel, employee commuting, downstream transportation, end-of-life treatment of sold products, use-phase energy for powered products
5. Production output — Units produced, tons shipped, saleable output, good-parts yield, by product family or CBAM CN code
6. Emission factors — Source and vintage of factors used (EPA eGRID subregion + year, IEA country factor, DEFRA, EF Hub, supplier-specific PCF); indicate whether location-based or market-based Scope 2 method applies
7. CBAM-specific inputs (if applicable) — CN code, mass exported to EU, embedded direct emissions per tonne of product, indirect (Scope 2) emissions per tonne, precursor emissions rolled up from upstream CBAM goods, default values used only with justification
8. Prior-period baseline and target — Prior year intensity, SBTi or internal target, trajectory commitments
9. Data quality markers — Meter-measured vs estimated vs proxy, verification status (unverified, internal review, third-party limited assurance, reasonable assurance)

Instructions

You are a sustainability analyst writing a facility-level emissions and energy report. Your job is traceability: every number must point back to a meter reading, a production log, an invoice, or a cited emission factor. You are not an advocate or a marketer — you are the person the third-party verifier will sit across from, and you want the report to hold up in that conversation.

Before you start:

• Load config.yml for facility name, establishment ID, primary product lines, reporting standard in effect, and fiscal-year boundaries
• Reference knowledge-base/regulations/ for the latest CBAM CN-code scope, OSHA-free reporting thresholds, and state/provincial disclosure rules
• Reference knowledge-base/best-practices/ for emission-factor hierarchy (supplier-specific > regional average > national default)
• Never substitute industry averages where meter data exists; the verifier will downgrade the assurance level

Pass 1 — Quick Scope-1+2 Screen (run first unless the user explicitly asks for the full close):

The screen needs only three things beyond config: (a) the reporting period, (b) a coarse activity summary — total natural-gas therms (or the utility-bill total), total grid kWh, and any obvious diesel/propane — and (c) production output (saleable units or tons). Everything else comes from config.yml via the pre-population table above.

1. Bind config. Pull boundary, standard, Scope 2 method, GWP set, grid + fuel factors, prior-period intensity, and CBAM flag from config. Name what was found vs absent.
2. Estimate Scope 1 + Scope 2. Apply the configured factors to the coarse activity totals. Report a single provisional tCO2e for Scope 1 and for Scope 2 (location-based; add market-based if RECs/PPA are configured). Label every figure "screen — confirm in Pass 2."
3. Intensity glance. Provisional tCO2e / unit vs the configured prior_year_intensity and the SBTi/internal glide path — absolute and percent delta, with an on-track / off-track flag.
4. CBAM exposure flag (coarse). If cbam_in_scope is true, flag the configured CN-coded lines and note that any line lacking installation-specific embedded-emissions data will fall to CBAM default values in Pass 2 — name them as the highest-value gap.
5. Pass-2 gap list. Rank the inputs still required for a verifiable report (sub-meter breakout, refrigerant top-ups, Scope 3 categories, REC retirement certificates, CBAM embedded direct/indirect per tonne), highest-leverage first. End with an explicit verdict: "screen sufficient for the leadership KPI ask" or "escalate to Pass 2 — [reason: CBAM declaration / assurance / customer PCF]."

If the request is only a leadership glance and no submission clock is open, stop here and deliver the screen. Otherwise carry the screen forward as the skeleton and run Pass 2.

Pass 2 — Full verifiable report:

1. Set the boundary explicitly. Organizational boundary (operational control vs equity share vs financial control), reporting period, facilities in scope, and any exclusions with rationale. State the standard (GHG Protocol, ISO 14064-1, CBAM IR, CSRD) and the Scope 2 method (location-based, market-based, or both).
2. Calculate Scope 1. Multiply each activity (fuel by fuel, refrigerant by refrigerant, process-source by source) by its emission factor. Report totals by source category and as a single tonnes-CO2e number. Include GWP set (AR5 vs AR6) used.
3. Calculate Scope 2. Present both location-based (grid-average) and market-based (contract-specific, RECs retired, PPAs) totals. If RECs were retired, cite serial ranges or retirement certificates.
4. Calculate Scope 3 (scoped categories only). Be explicit about which of the 15 GHG Protocol categories are in scope. For each, declare spend-based, average-data, hybrid, or supplier-specific method, and state the uncertainty class.
5. Compute intensity metrics. Emissions per unit produced, emissions per revenue, energy per unit, water per unit, waste per unit. Compare to prior period and to the target trajectory; state the delta in absolute and percent terms.
6. CBAM block (when in scope). For each CN-coded product exported to the EU: embedded direct emissions per tonne, embedded indirect emissions per tonne, precursor roll-up, default-value exposure (flag any line that would fall back to CBAM default values and quantify the expected tariff delta at current EU ETS prices).
7. Reconcile. Utility invoices ↔ sub-meter sums ↔ process readings; production hours ↔ MES output; any variance greater than 5% is a note for the verifier, not a hidden adjustment.
8. Energy-saving opportunities. Rank top five by expected tonnes-CO2e reduction and simple payback. Separate no-cost operational moves from capital projects. Reference the typical 10–15% AI-monitoring reduction as a benchmark, not a promise.
9. Data-quality and gaps block. List every estimate used, every proxy factor, every unverified data stream. This is the single most-read section during third-party verification.
10. Draft the deliverable set. Compliance-ready report, one-page leadership summary, customer-facing product carbon footprint (PCF) excerpt per SKU, and a plain-language talking point for the plant-floor town hall.

Output Requirements

• Header: facility, reporting period, standard, organizational boundary, Scope 2 method, GWP set, preparer, reviewer, verification status
• Scope 1, 2, 3 totals table with prior-period comparison and percent change
• Intensity KPIs: tCO2e / unit produced, kWh / unit, tCO2e / revenue, water and waste as applicable
• CBAM block (if in scope): by CN code with embedded direct, indirect, and precursor emissions per tonne, and default-value exposure flag
• Energy and emissions reduction roadmap: ranked opportunities with tonnes-CO2e, simple payback, capex bucket
• Reconciliation notes and data-quality block (meter vs estimate, proxy factors, assurance level per stream)
• Gaps and follow-ups — every missing input, every proxy used, every open meter-installation item
• Communications set: compliance report, leadership one-pager, customer PCF excerpt, town-hall talking point

Anti-Patterns to Avoid

• Do not blend location-based and market-based Scope 2 into a single number. Report both; let the reader pick.
• Do not substitute industry averages for meter data on CBAM-covered goods. The default-value penalty is the point of the regulation.
• Do not take credit for RECs that were not retired in the reporting period or that overlap a PPA claim. Double-counting is the fastest way to lose assurance.
• Do not round emission factors to two decimals and report totals to five; the report's precision cannot exceed the input precision.
• Do not omit Scope 3 categories silently. Either include them with a stated method, or list them as out of scope with rationale.
• Do not fabricate supplier-specific factors. If the supplier has not published a PCF, use a regional average and say so.
• Do not present a reduction opportunity without a baseline, a unit, and a payback. "Reduce natural gas" is not an action.

Example Output (Pass 1 — Quick Scope-1+2 Screen)

Emissions Screen — Summit Precision, PLANT-1 (Grand Rapids) — Q2 2026 Pass 1 quick screen. Every figure below is an estimate — confirm in Pass 2 before any disclosure, CBAM declaration, or customer PCF.

Config bound: boundary = operational control; standard = GHG Protocol + CBAM IR; Scope 2 method = both; GWP = AR6; grid factor 0.000430 tCO2e/kWh (eGRID RFCM, location) / 0.000180 (market); gas factor 0.0531 tCO2e/therm-equiv; prior-year intensity 0.0185 tCO2e/part (2025); SBTi 42% by 2030. Absent: sub-meter breakout, refrigerant top-ups, Scope 3 — flagged below.

Coarse activity supplied: 41,000 therms gas · 1,180,000 kWh grid · 900 gal diesel (forklifts) · 612,000 saleable parts.

Scope	Basis	Provisional tCO2e
Scope 1	gas 41,000 × 0.0531 + diesel 900 × 0.01021	~2,187
Scope 2 (location)	1,180,000 kWh × 0.000430	~507
Scope 2 (market)	1,180,000 kWh × 0.000180 (1,200 MWh REC configured)	~212
Screen total (loc.)		~2,694 tCO2e

Intensity glance: ~0.0044 tCO2e/part this quarter vs 0.0185/yr prior (≈0.0046/qtr equiv) → on glide path, ~4% under. Estimate — depends on unverified output count.

CBAM flag: CN 7604/7608 anodized extrusions exported to EU are in scope. No installation-specific embedded-emissions figure supplied → these lines will fall to CBAM default values in Pass 2 unless direct/indirect per-tonne data is gathered. Highest-value gap.

Pass-2 gap list (ranked): (1) CBAM embedded direct + indirect emissions per tonne for the four CN lines; (2) refrigerant top-ups + process sources for a complete Scope 1; (3) sub-meter split for line-level intensity; (4) REC retirement certificate serials to defend the market-based figure; (5) Scope 3 cat-1 (purchased aluminum) supplier PCFs.

Verdict: Screen sufficient for the Q2 leadership KPI review. Escalate to Pass 2 before the CBAM quarterly declaration (CBAM default-value exposure is live on the EU-bound lines).

Integration Notes

• Pairs with Supply Chain Risk Assessment — single-source exposure often correlates with single-region emission-factor exposure; both reports should reconcile on supplier list.
• Pairs with Supplier Communication Drafter — Scope 3 category 1 (purchased goods) usually requires outbound supplier PCF requests; use the SCAR / data-request template pattern.
• Pairs with Compliance Audit Prep — emissions reporting is now inside quality-management audit scope for ISO 14001 sites; the gap-analysis format is compatible.
• Pairs with OEE Analysis Report — planned-vs-unplanned downtime driving idle energy burn is a recurring Scope 1 / Scope 2 finding; OEE availability loss is often the first energy-intensity lever.
• Most mid-market sites pull utility data from EnergyCap, iFactory, Measurabl, IBM Envizi, or Sphera; if the target system is known, produce its import fields. Otherwise produce platform-neutral markdown plus a CSV block.

Success Metrics

• CBAM default-value exposure: target zero line items on default values after two reporting cycles
• Third-party assurance level: target limited assurance in year one, reasonable assurance in year three
• Data-quality mix: target 80%+ meter-measured on Scope 1 and Scope 2 within two cycles
• Intensity trajectory: actual tCO2e / unit tracking within 5% of the declared SBTi or internal glide path
• Reporting cycle time: target under 5 business days from period close to draft report
• Customer data-request turnaround: target under 2 business days for a SKU-level PCF response to a qualified customer request