COMPANY PROFILE
"Providing innovative solutions across the project life cycle for the geosciences"
Summary
Life Cycle Geo (LCG) provides innovative solutions for its clients across the mine project life cycle. LCG, in business since 2019, is 100% remote with more than a dozen geoscientists and data scientists distributed across North America. Our scientists have highly specialized and diversified skillsets that are grounded in geochemistry domain expertise and augmented by strong mechanistic and data-driven predictive modeling capabilities. LCG is currently engaged in innovative work across the globe in various sectors including mining, power, and water resources and all project stages including exploration, permitting, feasibility/design, operations, closure, and post-closure.
Main Leadership
- Tom Meuzelaar, Founder
- Jake Waples, Principal Consultant
Our Services
Geochemical modeling and data sciences
Project Highlight
Client: Equinox Gold
Services Provided:
- Innovative data collection and analysis techniques used to classify fine-grained, clay rich ore and waste
- Collection of mineralogical data using infrared (IR)-based spectral data
- Supervised machine-learning based predictions of mill ore, leach ore and waste ore grade classes
- Cloud-hosted workflow for data ingestion and analysisThe Cactus Project is a 100%-owned brownfield porphyry copper project located on private land, in a tier 1 mining jurisdiction. The Project is situated at the convergence of three major copper trends in Arizona, near the city of Casa Grande, Arizona, USA. Access to onsite and nearby infrastructure, including water, power, highways, rail and access to skilled labour, is reflected within the Project’s first quartile capital intensity, while an advanced permitting schedule lowers its associated risks. The city of Phoenix and the Sky Harbor International Airport are located 45 miles to the north and Tucson is approximately 75 miles to the southeast. The Project itself covers approximately 5,700 acres.
Project Overview:
The Castle Mountain Gold project, located in the historic Hart Mining District at the southern end of the Castle Mountains (~110 south of Las Vegas, NV) aims to recover 3 Moz of gold over a 16-year mine life by mill and heap leach. Epithermal style gold mineralization occurs in a Miocene-age volcanic rocks. Field classification of mill ore, leach ore and waste is very difficult given the fine-grained and clay-rich nature of the ore-hosting volcanic rocks.
Life Cycle Geo (LCG) developed a proxy for grade control using infrared (IR)-based core scan mineralogy data in order to separate high-grade mill ore (>1.3 g/t), leach ore (0.17 to 1.3 g/t) and waste (<0.17 g/t). Machine learning algorithms (deep learning approach using neural networks) were employed to predict the three material classes. Initial trained models indicate an 80% success rate in classifying waste material and 70% success rate in classifying ore. The trained model currently indicated low accuracy in segregating of mill ore from leach ore given that mill ore occurs in micro-quartz veinlets not detected by the hyperspectral approach. Addition of RGB core images to the model would like resolve this issue. Scope of Services:
LCG was engaged to support Equinox in the following activities:
- Conduct unsupervised exploratory data analysis on short-wave infrared (SWIR) and long-wave infrared (LWIR) hyperspectral mineral spectra and core face imagery to identify ore grade and waste classes that are extremely difficult to identify by hand
- Develop grade domains and a classification conceptual model by linking IR data to assay-based grade classes for composited drillhole intervals
- Train a supervised deep learning (neural network-based) machine learning model to identify grade and waste material classes based on core face imagery and spectral data
- Construct a cloud-based software workflow to ingest data, train and test the model and present results
Notable Accomplishments:
This innovative work indicated an 80% success rate in identifying low-grade waste (<0.17 g/t) and a 70% success rate in identifying combined mill and leach ore. The model was less successful in separating mill and leach ore- this is likely due to high grade gold residing in micro quartz veinlets that are not picked up by IR spectra. Future work could aim to combine high-resolution RGB-based core photography with IR-spectra in an ensemble deep learning approach to better separate ore grade classes.
This work was presented at the Society for Economic Geologists meeting in Denver in 2018.
Project Highlight
Client: KSM Mining ULC
Services Provided:
- Geochemical characterization: waste, borrow and ore
- Support development of project geometallurgical and environmental conceptual models
- Use of unsupervised material classification and domaining techniques
- Supervised regression/classification-based material property prediction to support block model development, and future blast assay programs and operational segregation
Project Overview:
The KSM project is a proposed gold-copper, silver and molybdenum project located approximately 950 km northwest of Vancouver in British Columbia, Canada. The Project is the world’s largest undeveloped gold project, and the world third largest undeveloped copper project, comprises five separate ore bodies, with estimates of 47.3 million ounces of gold and 7.3 billion pounds of copper.
Life Cycle Geo (LCG), in partnership with Geosyntec and WSP, was engaged from 2018 to 2024 to support characterization, classification and mine materials management support from operations to post-closure for ore, borrow and waste. LCG, Geosyntec and WSP developed innovative data-driven, machine learning-based workflows to assist KSM in optimizing materials management through the project life cycle. Scope of Services:
LCG, Geosyntec and WSP were engaged to support KSM in the following activities:
- Conduct multiple, advanced, long-term materials characterization tests to support development of project conceptual models for ore, borrow and waste material behavior
- Identify material domains (or classes) by integrating all lab and field characterization data using standard multivariate unsupervised machine learning approaches
- Develop regression and classification-based supervised machine learning workflows to extend ore, borrow and waste material classes to the exploration assay database
- Engage in multi-stakeholder discussions to evaluate long-term project materials and water management options during the operational and post-closure stages of the life cycle
Notable Accomplishments:
The innovative data-driven methods developed and employed by LCG and Geosyntec represent a step change improvement in material classification and volume estimation accuracy over traditional block modelling approaches which are typically binary (e.g. PAG/NPAG) and use single parameter assay as proxy. These improvements will result in considerable long-term project risk reduction and cost savings.
LCG and Geosyntec are presenting portions of this work at the International Conference for Acid Rock Drainage in Halifax, Nova Scotia in September 2024, and have previously presented at the BC MEND ARD/ML workshop.
Project Highlight
Client: Ivanhoe Electric Inc.
Services Provided:
- Geochemical characterization of all mine material types
- Historic and baseline water quality fingerprinting
- Predictive water quality models for underground workings and above ground waste storage facilities
- Environmental block model development support
- Feasibility, design and permitting support
Project Overview:
The Santa Cruz Copper Project includes a cluster of porphyry deposits buried under deep cover, part of the larger Arizona porphyry copper belt. Indicated mineral resources total 227 million tonnes grading 1.2% total copper with estimated copper production at 1.6 million tonnes over a 20-year mine life. The Initial Assessment base case assumes that onsite renewable energy infrastructure will generate 70% of the electric power required for the Santa Cruz Project to minimize CO2 equivalent emissions.
Ivanhoe Electric has been receiving valuable support from Life Cycle Geo (LCG) since 2022. LCG has been instrumental in providing materials characterization, water quality assessment, predictive water quality modeling, environmental block model estimates, and overall feasibility, design, and permitting support. Due to Ivanhoe Electric’s innovative and forward-thinking approach, LCG frequently employs advanced unsupervised and supervised machine learning workflows to provide cutting-edge water and materials management support. Scope of Services:
LCG has been engaged to support the Santa Cruz Project in the following activities:
- Conduct materials characterization testing for overburden and bedrock waste, borrow materials, all ore types, and paste-cemented neutralized processing residues towards assessing long-term acid rock drainage (ARD) and metals leaching potential
- Identify unique water quality fingerprints associated with baseline overburden and mineralized bedrock, fractured wallrock in underground excavations, and paste backfill using unsupervised analysis methods
- Develop predictive water quality models associated with long-term operational and post-closure underground and surface water management
- Support the development of volume estimates of materials with various environmental characteristics for mine planning, using an array of unsupervised and supervised machine learning methods
- Assist Ivanhoe Electric with technical studies, engineering design, and long-term permitting support under the Arizona Aquifer Protection Permit (APP) program using Best Available Demonstrated Control Technology (BADCT)
Notable Accomplishments:
The innovative data-driven methods developed by LCG are a step change improvement in mine materials and water management during the entire mine project life cycle, contributing to long-term CAPEX and OPEX reduction.
Ivanhoe Electric is being supported by LCG to meet all permitting objectives and deadlines, and LCG is collaborating with Ivanhoe Electric as they ensure proactive and transparent engagement with all project stakeholders.LCG and Geosyntec are presenting portions of this work at the International Conference for Acid Rock Drainage in Halifax, Nova Scotia in September 2024, and have previously presented at the BC MEND ARD/ML workshop.