HemeraScope · EventsDemo
Pawston Annual 2026 · 2026-03-19

Methodology

The summary version of the methodology PDF (event-v1.2.0).

1. Boundary

The boundary follows ISO 20121 event-management practice. We report what the organiser controls or significantly influences: attendee travel, venue energy allocated by booked-hours share, catering, accommodation, materials and waste. Out of scope: attendee leisure travel after the event, third-party private hospitality, and travel between attendees' own offices.

  • · Geographic: UK + inbound international attendees.
  • · Temporal: 2026-03-19 (single day, plus travel days +/-1).
  • · Consolidation: Operational control — Pawston is the organiser.

2. Categories — included and excluded

CategoryGHG ProtocolTiertCO₂e
Attendee travelScope 3 · Cat 6T2 · Survey41.2
Venue energy (allocated)Scope 2 (location-based)T1 · Primary3.2
CateringScope 3 · Cat 1T2 · Survey1.8
AccommodationScope 3 · Cat 6T3 · Spend-based0.7
Materials, signage & wasteScope 3 · Cat 1 + 5T3 · Spend-based0.4
Excluded with rationale: Scope 1 mobile combustion (no organiser-owned vehicles); pre-event marketing energy (de minimis under 0.5% screen); attendee leisure travel (out-of-scope per ISO 20121).

3. Calculation methods, per category

Attendee travel
Scope 3 · Cat 6
Data: Survey + statistical extrapolation. Factor: DEFRA 2024 passenger transport.

Per-attendee responses for 312 of 450 tickets; remaining 138 projected from respondent mode-mix banded by distance.

Venue energy (allocated)
Scope 2 (location-based)
Data: Half-hourly meter data. Factor: DEFRA 2024 UK grid (location-based).

Sub-metered hall, allocated by booked hours / total available hours.

Catering
Scope 3 · Cat 1
Data: Caterer menu lines × meals served. Factor: DEFRA 2024 (pinned); Poore & Nemecek 2018 as benchmark default for low-data menu lines.

Three menu options across 460 covers, plus 220 morning pastries and break-time snacks.

Accommodation
Scope 3 · Cat 6
Data: Survey-reported room-nights. Factor: Cornell CHSB 2023 benchmark default (UK average, carried at wide uncertainty).

48 room-nights reported across respondents; non-responders assumed local (postcode-banded).

Materials, signage & waste
Scope 3 · Cat 1 + 5
Data: Spend-based, EE-IO. Factor: DEFRA 2024 (waste) + Exiobase v3.8.2 (signage spend).

Lanyards, signage, printed handouts; waste streams from venue weighbridge.

4. Emission factor sources (pinned at publish)

SourceVersionUsed forRole
DEFRA UK government GHG conversion factors2024 v1.1Travel, energy, wastePinned calculation source
Cornell Hotel Sustainability Benchmarking Index (CHSB)2023AccommodationBenchmark default for low-data estimates, carried at wide uncertainty
Poore & Nemecek (food)2018CateringBenchmark default for low-data estimates, carried at wide uncertainty
Exiobasev3.8.2Spend-based residualsPinned calculation source
DEFRA 2024 and Exiobase v3.8.2 are the pinned calculation sources. Cornell CHSB and Poore & Nemecek are benchmark defaults used only for low-data estimates, carried at wide uncertainty; they are not production calculation sources.

5. Allocation decisions

Venue energy: allocated by a recorded allocation decision — method booked_hours here (booked hours / total available hours); headcount_share where the venue hosts a concurrent activity. The share is stored as numerator / denominator and disclosed in the methodology PDF. Where no decision is recorded, the whole meter is attributed to the event and a share of 1.0 is disclosed. Sub-metered hall + half-hourly grid factor.

Shared transport: coach and rail allocated per-pkm (passenger-kilometre), so capacity changes do not transfer between attendees.

Catering: meal-level allocation by menu line × portions served. Surplus produced but not served is allocated to waste rather than catering.

6. Uncertainty analysis

Monte Carlo iters
5,000
CI level
95% two-tailed
σ source
Structural σ engine
Headline width
±4.9 tCO₂e

Records are correlated in the MC roll where they share a factor family — short-haul flight σ moves together across attendees, not independently — so the headline band does not understate the true uncertainty. Non-ecoinvent σ values are taken from Schulte (2024), IPCC AR6 (2019), the PEF guide and Meinrenken (2022).

7. Double-count guard

Venue energy is reported under the event but flagged so it does not also appear under the venue operator's annual report. Two-way attestation handled at publish.

Staff travel by Pawston employees attending is reported here (Cat 6 business travel) and excluded from Pawston's commuting line for the same dates.

Bundled catering covered by the venue contract is allocated to the event, not the venue, to avoid splitting one meal across two reports.

8. Limitations and caveats

The 31% non-responder share is the largest single source of uncertainty. Their travel is projected from respondent mode-mix banded by distance — not assumed to be local. Extrapolated buckets carry an additional log-space σ of 0.50 × (1 − response rate), composed in quadrature with the bucket's structural σ.

Hotel room emissions are estimated from a benchmark default (Cornell Hotel Sustainability Benchmarking Index 2023, UK average), carried at wide uncertainty — not a pinned calculation source. Finer detail requires per-property disclosures that most hotels do not publish.

The footprint is illustrative of what HemeraScope can produce. Live numbers depend on the organiser's data quality at intake.

Action-first reading

Once the headline is on the page, the next decision is the invitation list and venue catchment — that's where the lever sits, not in the calculation.

What we claim — and what we don’t

What we claim
  • · Every line names DEFRA factor + version + vintage in plain text on the page.
  • · Every number is σ-banded — central value + 95% CI on the page, not buried in an appendix.
  • · Every line names its approver with ISEP credential. Override decisions auditable to a named carbon accountant.
  • · Inputs hash + MC seed published with the headline. A second run on the same inputs returns the same number, bit-identical.
  • · Non-responder treatment documented on the page, including which cohort the imputation drew from and what σ that adds.
  • · Double-count guard runs across every line — three checks (cross-event, cross-scope, cross-org-allocation). Passes shown explicitly.
  • · Methodology PDF mirrors the page content — same numbers, same factor versions, generated from the same publication snapshot.
  • · Replay is one command — the inputs hash anchors a deterministic recompute.
What we don’t claim
  • · No verified third-party assurance — ISO 14064-3 assurance requires an external assuror. Hemera produces the evidence pack; the assuror signs.
  • · No sector-specific add-on protocols wired in v1.
  • · Factor lineage is to DEFRA + CIBSE only in v1; ecoinvent integration is roadmap.
  • · Demo numbers are illustrative for a fictional Pawston event, not a real footprint.

Roadmap

The v1 surface is deliberately narrow. The following items are on the path, not in the shipped engine today:

  • · Sector-specific add-on protocols — ICT-event, hospitality-event, broadcast-event modules.
  • · ecoinvent factor lineage — DEFRA + CIBSE today; ecoinvent integration with full Ciroth pedigree on roadmap.
  • · Third-party assurance integration — evidence pack export hand-off to ISO 14064-3 assurors with stable record IDs.
  • · Mixture-of-candidates σ (v2) — structural σ today; v2 widens the band where one activity maps to multiple plausible factors.
  • · Year-on-year longitudinal integrity — change-log + restatement protocol across publication snapshots.

Methodology PDF

Pawston Annual 2026 — Methodology Report

The audit-grade PDF mirrors this page plus three appendices: full activity records, attendee survey responses (anonymised to outward postcode), and the change log against the prior publication.

Methodology version event-v1.2.0 · published 2026-04-04
Demo · PDF not generated