Frac Disclosures

**Overview** Accurate chemical disclosure of hydraulic fracturing fluids requires extensive domain knowledge, coupled with numerous rules to address misspellings and diverse naming conventions, and machine learning to combat obfuscation. Frac Disclosure from IHS Markit is a comprehensive and detailed frac ingredient database for North America, backed by IHS Markit’s industry-leading well and production data. This interpretation-ready dataset of 150,000 North American wells brings together chemical disclosure information from all available registries and reconciles wells to proprietary IHS Markit 14-digit UWIs. This rigorous process provides transparent, un-obscured access to the most current library of supplier mappings to ingredients and purposes, and corrects errors commonly found in other industry datasets. Dataset Overview: - Frac Data is augmented with IHS Markit E&P data to reconcile from 10- to 14- digit well UWIs - Proppant detail includes proppant type, resin coating, sand and ceramic quality and mesh size – augmented with IHS Markit information - Ingredient mappings aid in identification of frac type, crosslinking type, surfactant mixtures, in addition to local and regional preference trend information - Targeted data selection easily removes bias of dominant ingredients and enables comparison of frac type, wellbore integrity protection ingredients, and other individual or groups of ingredients for competitive benchmarking - Includes all US and Canadian FracFocus data, and all US state-reported data not in FracFocus - Ingredients are mapped to main purpose and sub-purpose to differentiate key active ingredients from solvents/carriers and other components - Machine learning processes identify patterns in ingredients, supplier, tradename, CAS#, purposes and percentages to predict missing data - Development process includes consultation with industry experts in frac chemistry, completions techniques, E&P operations and machine learning techniques **Use cases** - Economic Impact Analysis - Analyze supply and demand balance - Supply Chain - Save on procurement costs - Competitive Analysis - Competitor well benchmarking**概述** 精准披露水力压裂液（hydraulic fracturing fluids）的化学成分，需具备深厚的领域专业知识，辅以大量规则以处理拼写误差与多样的命名规范，并结合机器学习技术对抗信息混淆。 IHS Markit旗下的Frac Disclosure压裂披露数据集是一套面向北美地区的全面且详尽的压裂成分数据库，依托IHS Markit行业领先的油井与生产数据作为支撑。该预适配解读分析的数据集涵盖15万个北美油井的相关信息，整合了所有可获取登记处的化学成分披露数据，并将油井匹配至IHS Markit专属的14位唯一井标识符（Unique Well Identifier，UWI）。这套严谨的处理流程可实现透明、无遮蔽的访问权限，获取最新的供应商-成分-用途映射库，并修正其他行业数据集常见的各类误差。 **数据集概览** - 压裂数据已结合IHS Markit勘探与生产（Exploration and Production，E&P）数据进行增强，可将10位唯一井标识符转换为14位标准格式 - 支撑剂详情涵盖支撑剂类型、树脂涂层、石英砂与陶瓷材质及目数，并辅以IHS Markit的相关信息进行补充增强 - 成分映射可辅助识别压裂类型、交联类型、表面活性剂混合物，同时提供本地及区域的偏好趋势信息 - 通过针对性的数据筛选，可有效消除主流成分带来的偏差，支持开展压裂类型、井眼完整性防护成分及其他单一或成组成分的对比分析，以进行竞争性基准测试 - 涵盖所有美国与加拿大的FracFocus数据集，以及所有未纳入FracFocus的美国各州上报数据 - 成分已映射至主要用途与子用途，以区分关键活性成分与溶剂/载体及其他组分 - 机器学习流程可识别成分、供应商、商品名、CAS编号（Chemical Abstracts Service Number，CAS#）、用途及占比的模式，从而预测缺失数据 - 数据集开发过程征询了压裂化学、完井技术、勘探与生产作业及机器学习技术领域的行业专家意见 **应用场景** - 经济影响分析：开展供需平衡分析 - 供应链管理：降低采购成本 - 竞争性分析：开展竞品油井基准测试

搜集汇总

背景与挑战

背景概述

Frac Disclosure是一个涵盖北美15万口井的水力压裂成分数据库，整合多源披露数据并通过机器学习进行纠错和补全。该数据集提供成分用途分类、供应商映射等详细信息，支持供应链优化、竞争分析等工业应用。