A Monte Carlo Simulation-based algorithm for updating Combined Variance cost function in optimally locating additional drillholes


Soltani-Mohammadi S., Sohrabian B., Safa M., TERCAN A. E.

INTERNATIONAL JOURNAL OF MINING RECLAMATION AND ENVIRONMENT, vol.36, pp.305-322, 2022 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 36
  • Publication Date: 2022
  • Doi Number: 10.1080/17480930.2022.2036558
  • Journal Name: INTERNATIONAL JOURNAL OF MINING RECLAMATION AND ENVIRONMENT
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, Agricultural & Environmental Science Database, Applied Science & Technology Source, Compendex, Computer & Applied Sciences, Environment Index, Geobase, Greenfile, INSPEC, Pollution Abstracts
  • Page Numbers: pp.305-322
  • Keywords: Particle swarm optimisation, uncertainty, boreholes, combined variance, cokriging, PARTICLE SWARM OPTIMIZATION, GENETIC ALGORITHM
  • Hacettepe University Affiliated: Yes

Abstract

Optimally locating additional boreholes in the boundaries of ore deposits is an important problem in mining projects. To solve this problem, combined variance has been proposed as a cost function to be minimized. An issue about combined variance is its dependence on the value of variable at unknown location. This value is achieved through a round-based algorithm that estimates the probability of occurrence of ore and rounds it to the nearest integer without considering all scenarios. This study aims to consider all scenarios using a Monte Carlo simulation-based algorithm. This approach uses ordinary cokriging to estimate the probability of occurrence of ore at unknown locations. Then, the estimated probabilities are entered into a Monte Carlo simulation procedure to generate various realizations. The method was applied in the Chadormalu ore deposit to propose ten additional boreholes. The round-based algorithm proposed additional drill holes in the middle of ore-intersected and non-intersected initial drill holes. However, the Monte Carlo-based algorithm is sensitive to the thickness such that in boundaries with thicker ore, the additional boreholes are suggested farther away from the ore-intersected drill holes and closer to the non-intersected ones and vice versa.