Admissibility-First Energy Measurement (EA-1): Closure, Integrity, and Refusal-Coded Claims for Auditable Energy-System Results

EA-1 (Admissibility-First Energy Measurement) is a measurement-governance instrument for evaluating energy-system performance claims (energy, power, efficiency/COP, heat rate, and related statements) under adversarial or high-dispute conditions. It requires the claimant to declare a complete ClaimTuple (measurement boundary, channels, time window, pipeline, thresholds, uncertainty floor, and disclosures) and then applies a deterministic gate stack that prioritizes closure, identifiability, and integrity before any interpretive conclusion is permitted. EA-1 produces a replayable, citable decision artifact rather than a narrative: a VerdictObject with outcomes PASS, REFUSE, or DOWNSCOPE, plus a ReceiptManifest and content-addressed hashes binding the verdict to a specific run. The standard distinguishes “a run occurred” from “a publishable claim exists,” routes common integrity failures (e.g., time skew, drift, clipping, aliasing, provenance breaks) into explicit refusal states, and enforces monotonic downscoping so reduced claims cannot become stronger than the original. It also specifies permitted/prohibited reporting language to prevent overstated conclusions and defines a required citation block that ties public statements to verifiable hashes, boundaries, and verdict status. In short, EA-1 turns contested energy measurements into audit-ready, refusal-coded, hash-bound objects designed to reduce interpretive drift and make admissibility mechanically checkable. Keywords: energy measurement, measurement & verification (M&V), metrology governance, admissibility-first evaluation, ClaimTuple specification, boundary conditions, measurement boundary closure, channel integrity, data provenance, reproducible measurement, deterministic gate stack, refusal semantics, structured failure modes, PASS / REFUSE / DOWNSCOPE, monotonic downscoping, integrity screening (time skew / drift / clipping / aliasing), uncertainty floor, identifiability constraints, audit-ready receipts, content-addressed hashes, canonicalization, replay-verifiable verdicts, receipt manifest, citation block standard, permitted/prohibited claim language, anti-narrative inflation, adversary-aware measurement protocol.