85-6 An Exploration Vector for Associated Carbon Dioxide Mineralization and Nickel Resources using Sparse Prediction of Olivine Abundance

Session:

Presenting Author:

Ashton Wiens

Authors:

Wiens, Ashton¹, Jones, Matthew Madden², Steup, Kadie³, Blondes, Madalyn S.⁴, Lahiri, Nabajit⁵, Stanfield, C. Heath⁶, Thakurta, Joyashish⁷, Miller, Quin⁸, Schaef, H. Todd⁹

(1) U.S. Geological Survey, Wichita, KS, USA, (2) U.S. Geological Survey, Reston, VA, USA, (3) Pacific Northwest National Laboratory, Richland, WA, USA; U.S. Geological Survey, Reston, VA, USA, (4) U.S. Geological Survey, Reston, VA, USA; Carbfix, Reykjavik, Iceland, (5) Pacific Northwest National Laboratory, Richland, WA, USA, (6) Pacific Northwest National Laboratory, Richland, WA, USA, (7) Natural Resources Research Institute, Minerals and metallurgy, Duluth, MN, USA, (8) Pacific Northwest National Laboratory, Richland, WA, USA, (9) Pacific Northwest National Laboratory, Richland, WA, USA,

Abstract:

Demand for nickel, cobalt, and other critical minerals is projected to increase over the coming years due in large part to rapid upscaling of battery production. However, the supply of Ni from newly developed sulfide and laterite deposits is not sufficient to meet demand, and production from these traditional deposit types has been associated with adverse surface environmental effects. Mining lower-grade resources requires moving larger volumes of rock and consuming more energy, whereas alternative methods of enhanced recovery could ameliorate the environmental burden of mining. Deposits of the mineral olivine (Mg,Fe)₂SiO₄ are well known both for having elevated Ni content and for being a highly reactive target for CO₂ storage via mineralization; hence, methods are currently being developed to recover critical minerals like Ni while mineralizing CO₂ as stable carbonate minerals. To better understand the geologic potential for coupled Ni recovery and CO₂ mineralization, we present an exploration vector (EV) based on a dataset of geochemical, mineralogical, and microanalytical analyses of rock samples from spatially extensive and geologically representative mafic and ultramafic units in the United States. The EV ranks Ni-CO₂ mineralization resource prospectivity with a knowledge-driven fuzzy logic approach composed of relevant factors for CO₂ reactivity and leachable Ni content, such as divalent cations and modal abundance of olivine. To make future ranking more accessible, we emulate this ranking scheme using only elemental concentrations that are commonly measured in routine geochemical methods. To accomplish this, we predict olivine abundance from geochemical data using regularized regression and log-ratio transformations, producing a data-driven alternative to normative mineralogy that can better account for serpentinization and carbonation. Model selection via cross-validation provides high predictive accuracy for olivine abundance for the mafic and ultramafic samples. Predictive models provide a means to rank potential Ni-CO₂  mineralization resources based on more commonly measured data without a need to incorporate modal mineralogy data. These developments provide statistical tools needed to analyze complex high-dimensional datasets for novel technologies such as critical mineral recovery coupled with subsurface injection of CO₂ in mafic and ultramafic rock formations.

Geological Society of America Abstracts with Programs. Vol. 57, No. 6, 2025

doi: 10.1130/abs/2025AM-9965

© Copyright 2025 The Geological Society of America (GSA), all rights reserved.