Metodología para la estimación de incertidumbres asociadas a concentraciones de sólidos suspendidos totales mediante métodos de generación aleatoria

Abstract

In this work a methodology for the estimation of uncertainties associated to Total Suspended Solids Concentrations (TSS) concentrations obtained during laboratory analysis, that takes into account not only the precision of the laboratory instruments, but also subsampling, sample, and equipment manipulation, is proposed. The proposed methodology includes the computation of replicates uncertainties with analytical methods, the comparison between replicates through t-tests, and computation of uncertainties applying a Monte Carlo method. As an example, this methodology was applied to TSS concentration data obtained in laboratory. The data set used corresponds to real samples with low, medium, and high TSS concentrations and with different subsampling variabilities. Subsequently, comparisons between uncertainties obtained from the proposed methodology and those obtained when only equipment precision is taken into account were undertaken. From these comparisons it was observed that some important differences can exist between concentrations uncertainties as well as samples mean concentrations.