The Levenberg–Marquardt method was used to solve the nonlinear least squares curve-fitting problem, applying an integration step of 10−3 with precision of 10−8. The Levenberg–Marquardt method was programmed in Fortran language, combining the Gauss and the Steepest Descent methods (Press et al., 1996). The sum of quadratic differences (the residual) between the experimental flux data and the flux data computed with the Field model was minimized, obtaining the best adjusted parameter (kn). This procedure was carried out for each n value in the Field equation. Thus, the best equation to describe the fouling occurrence will be that with the lowest residual value, since this will be the best equation to describe the experimental data.