Selection optimale des quantites d'explosifs basee sur des contraintes de production et un resultat d'analyse de regression multiple

CIM Bulletin, Vol. 83, No. 937, 1990

M. Laflamme et J.L. Collins, Departement de mines et metallurgie, Universite Laval, Quebec

A study, underway since 1987, tries to evaluate factors affecting mines' productivity. Information has been collected from seven underground mines in Quebec and this paper shows preliminary results from one. In particular, it shows how a combination of multiple regression analysis, mathematical programming and a set of realistic mine constraints can be used for a particular application of mine planning. More specifically, the paper shows a parallel between maximizing production of tons of broken ore, minimizing explosives costs or minimizing the costs of explosives per ton of broken ore by an appropriate selection of ANFO and TOVEX quantity of explosives. Finally, the paper shows that minimizing the costs of explosives per ton of broken ore is the best alternative to follow providing a special optimization program can be set up.

Mots Clés: Underground mining, Explosives, Regression analyses.