A transparent technique for mass balancing and data adjustment of complex metallurgical circuits

Abstract

To operate a metallurgical/chemical plant efficiently, metallurgists need to rely on consistent data sets that describe the process with sufficient accuracy. Such data are obtained from measurements with inherent errors generally through the use of commercially available mass balancing packages that involves some non-linear optimisation procedure for the minimisation of measurement errors and data adjustment. This practice somewhat limits the ability to incorporate some known constraints pertaining to the given operation and/or to see how the data adjustment procedure actually works. In this paper, principles of mass balancing and data adjustment procedures of some well-known methods are briefly discussed and a simple method of developing one's own spreadsheet for the above purpose, using Excel with its standard routines such as Solver, is demonstrated. Such spreadsheet calculations have the advantages of being transparent and indicate where most of the large errors occur, their contribution to the overall error and also shed light on which measurements are less reliable. It also provides the opportunity to modify the objective function to suit any specific needs and incorporate plant specific constraints.