Optimization Of A Chemical Coal Cleaning Process Via Simple Genetic Algorithm

Mehta, R. K. ; Dieudonne, V. ; Han, K. N. ; Queneau, P. B. ; Yoon, R. H.
Organization: Society for Mining, Metallurgy & Exploration
Pages: 15
Publication Date: Jan 1, 1993
A study based on a statistically designed sets of experiments (central composite design) was conducted to determine the optimum conditions for leaching a subbituminous coal (7.0 % ash) from the middle Wyodak seam in mild sulfuric acid solutions. The statistical analysis was conducted with the ash content and ash removal as the dependent variables and with the coal loading, acid concentration, ferric ion concentration, and temperature as the independent variables. Treatment of the experimental data via a simple genetic algorithm (GA) revealed that acid concentration and temperature are the two most critical variables. Unlike the case with bituminous coal, the ferric ions addition shows no beneficial effect. The optimum "macro response" predicted by GA was verified experimentally; the ash content was reduced to 2.7%, which is close to the value predicted by GA. The ash content was further reduced to 1.7% in a two-stage leaching process devised by incorporating the optimization results with that of leaching kinetic studies.
Full Article Download:
(437 kb)