Scholarlink

<p>Lithium-rich (Li-rich) sedimentary brine has emerged as a valuable unconventional resource, driven by the blooming global market, advancements in direct extraction technologies, and a lower environmental impact compared to traditional mining methods. However, resource delineation and estimation remain challenging due to inefficient field sampling and unreliable correlations between Li&#xa0;concentration ([Li]) and environment-sensitive geochemical indicators. Supported by public data and newly acquired measurements of water chemistry for Alberta Devonian brines, we developed a cutoff-based data-driven approach to extract Li-rich environmental characteristics in the probability domain to predict [Li] levels at locations with water chemistry data but without [Li] measurements. The approach relies solely on commonly available geospatial (coordinates, stratigraphic position) and geochemical features, including contents of total dissolved solids&#xa0;(TDS) and cations of Na, K, Mg, and Ca. Validated against about one hundred Li-labeled samples measured after May 2022, the approach achieved a minimum precision and accuracy of 97% and 84%, respectively, for predicting three [Li] cutoff levels (i.e., &gt; 35&#xa0;mg/L, &gt; 50&#xa0;mg/L, and &gt; 75&#xa0;mg/L). It was subsequently applied to predict [Li] levels of formation water from 897 different locations with legacy water chemistry data. The results align spatially with observed trends of Li-rich brines in Alberta Devonian formations and expand resource delineation and estimation capabilities to areas and formations with limited [Li] data availability.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Data-Driven Approach for Exploring Unconventional Lithium Resources in Devonian Sedimentary Brines, Alberta, Canada

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年4月18日

查看全文
  • Xiaolong Peng,
  • Zhuoheng Chen,
  • Chunqing Jiang,
  • Wanju Yuan,
  • Jiangyuan Yao

摘要

Lithium-rich (Li-rich) sedimentary brine has emerged as a valuable unconventional resource, driven by the blooming global market, advancements in direct extraction technologies, and a lower environmental impact compared to traditional mining methods. However, resource delineation and estimation remain challenging due to inefficient field sampling and unreliable correlations between Li concentration ([Li]) and environment-sensitive geochemical indicators. Supported by public data and newly acquired measurements of water chemistry for Alberta Devonian brines, we developed a cutoff-based data-driven approach to extract Li-rich environmental characteristics in the probability domain to predict [Li] levels at locations with water chemistry data but without [Li] measurements. The approach relies solely on commonly available geospatial (coordinates, stratigraphic position) and geochemical features, including contents of total dissolved solids (TDS) and cations of Na, K, Mg, and Ca. Validated against about one hundred Li-labeled samples measured after May 2022, the approach achieved a minimum precision and accuracy of 97% and 84%, respectively, for predicting three [Li] cutoff levels (i.e., > 35 mg/L, > 50 mg/L, and > 75 mg/L). It was subsequently applied to predict [Li] levels of formation water from 897 different locations with legacy water chemistry data. The results align spatially with observed trends of Li-rich brines in Alberta Devonian formations and expand resource delineation and estimation capabilities to areas and formations with limited [Li] data availability.