Sustainable management of mining operations with accidents: A mean-variance optimization model

Tsan Ming Choi

Research output: Journal article publicationJournal articleAcademic researchpeer-review

19 Citations (Scopus)


Sustainable management of mining operations requires careful considerations of environmental sustainability, economic sustainability and corporate social responsibility (CSR) related issues. Motivated by prior studies on accidents associated with mining operations, this paper builds a formal optimization model to address the above three sustainability related issues for a mining operation with the optimal decision on mining quantity. To be specific, we model the number of accidents as a Poisson distribution with a quantity dependent distribution parameter. We formulate the objective function via the mean-variance approach, and incorporate analytical constraints which relate to environmental sustainability and CSR into the model. We analytically derive the algorithm which can find the globally optimal solution for the optimization problem. After that, we further analyse when the mining company should consider implementing (i) the pollutant reduction technology, and (ii) the accident reduction technology. It is interesting to find that the mining company's degree of risk aversion affects the choice of pollutant reduction technology, but not the choice of accident reduction technology. Several other important insights are also analytically derived.
Original languageEnglish
Pages (from-to)116-122
Number of pages7
JournalResources Policy
Publication statusPublished - 1 Dec 2015


  • Analytical studies
  • Corporate social responsibility
  • Economics sustainability
  • Environmental sustainability
  • Mean-variance approach
  • Mining operations

ASJC Scopus subject areas

  • Sociology and Political Science
  • Economics and Econometrics
  • Management, Monitoring, Policy and Law
  • Law


Dive into the research topics of 'Sustainable management of mining operations with accidents: A mean-variance optimization model'. Together they form a unique fingerprint.

Cite this