中国湖北省のセメント工場5社におけるクリンカー生産量オンライン監視のデュアルトラック比較・品質評価データセット（2026年）
A dataset of dual-track comparison and quality assessment for clinker output online monitoring of five cement plants in Hubei, China, 2026 (原題)
Hu Hongbo, Meng Lingdong, Ye Meng, Wang Xingpeng, Liang Defeng
🤖 gxceed AI 要約
日本語
中国湖北省のセメント工場5社・7生産ラインを対象に、2026年3〜6月のクリンカー生産量オンライン監視データを公開したデータセット。原料-クリンカー転換係数法とバケットエレベーター電力法のデュアルトラック比較に、在庫検証や計測器トレーサビリティ点検を組み合わせ、116ライン日分を収録する。両手法の相関はr=0.99905、平均絶対相対偏差1.46%、拡張不確かさU=3.08%（k=2）と±5%基準を大きく上回る品質を示した。炭素会計のデータ品質評価やオンライン監視の受入手法に実証的基盤を与える。
English
An open dataset of daily clinker output monitoring from 7 production lines at 5 cement plants in Hubei, China (March–June 2026). It cross-validates a raw-meal conversion method against a bucket-elevator power method, plus inventory and instrument traceability checks, covering 116 valid line-days. The two tracks show r=0.99905, mean absolute relative deviation 1.46%, 99.1% pass rate at ±5%, and expanded uncertainty U=3.08% (k=2). It offers empirical grounding for carbon accounting data quality and online monitoring acceptance.
Unofficial AI-generated summary based on the public title and abstract. Not an official translation.
📝 gxceed 編集解説 — Why this matters
日本のGX文脈において
セメントはSSBJ・Scope1排出の主要源であり、生産量計測の品質保証は排出量報告の信頼性に直結する。中国の炭素計量レビュー実務は、日本の排出量検証・原単位管理の高度化にも示唆を与える。
In the global GX context
Cement is a hard-to-abate, high-emission sector central to Scope 1 accounting under ISSB/CSRD. This dataset demonstrates a replicable dual-track quality-assurance protocol for activity data, directly relevant to verification rigor and carbon metrology review globally.
👥 読者別の含意
🔬研究者:活動量データの不確かさ評価・デュアルトラック検証手法の実証例として参照できる。
🏢実務担当者:セメント・素材企業の排出量報告における生産量計測の品質保証設計に活用できる。
🏛政策担当者:炭素計量レビューやオンライン監視システムの受入基準策定の根拠として参考になる。
📄 Abstract（原文）
Online monitoring data of cement clinker output are the core metrological data for carbon emission accounting and carbon metrology review in the cement industry, and their quality directly affects the accuracy and traceability of enterprise carbon emission reports. However, daily on-site monitoring data cross-validated by multiple methods are still absent from public datasets worldwide. This dataset originates from on-site data quality traceability assessments of 7 clinker production lines in 5 cement plants in Hubei Province, China, conducted from March to June 2026. The daily output was compared using a dual-track approach, i.e., the raw meal-to-clinker conversion coefficient method (primary method) versus the bucket elevator power/current method (reference method), supplemented by stock inventory verification, metrological instrument traceability inspection, and transmission quality monitoring. The dataset covers 116 valid line-days from March 26 to June 21, 2026, accompanied by instrument traceability ledgers (49 instruments), device-level daily packet loss records (1351 entries), monthly material balances (36 entries), hourly- and minute-level raw acquisition samples, and a log of 14 data quality issues, totaling 15 data files and 3701 records. Quality control results show that the Pearson correlation coefficient between the two tracks is r = 0.99905; the mean absolute relative deviation |delta| = 1.46%; the pass rate against the ±5% acceptance threshold is 99.1%; Bland-Altman analysis yields a relative bias of 0.12% with 95.7% of samples within the 95% limits of agreement; and the relative expanded uncertainty of daily output is U = 3.08% (k = 2), significantly better than the ±5% acceptance requirement. The dataset underwent unified-caliber recalculation, record-level verification, and enterprise anonymization, with field-level traceability to the original assessment reports and supporting materials. It provides empirical data support for carbon accounting data quality evaluation, acceptance methodology of online monitoring systems, and carbon metrology review in the cement industry.
🔗 Provenance — このレコードを発見したソース
- scidb https://doi.org/10.57760/sciencedb.j00001.001mmfirst seen 2026-10-21 23:36:12 · last seen 2026-11-06 06:08:44
🔔 こうした論文の新着を逃したくない方は キーワードアラート に登録（無料・3キーワードまで）。
gxceed は公開メタデータに基づく研究支援データセットです。要約・翻訳・解説は AI 支援で生成されています。 最終的な解釈・検証は利用者が原典資料に基づいて行うことを前提とします。