Viés hídrico da homogeneização na caracterização de resíduos sólidos urbanos.

Abstract

Gravimetric characterization of municipal solid waste (MSW) is essential for sizing sorting, composting, and recycling systems, and depends on sampling protocols that prescribe mechanical homogenization of the sample before mass reduction. This thesis investigated whether this step, whose neutrality with respect to material properties is presumed by ABNT NBR 10.007:2004 and ASTM D5231 without experimental verification, alters the distribution of moisture among fractions and biases the composition obtained. The results are organized in four articulated chapters. An integrative review compiled 123 studies with verified metadata and established that homogenization is applied consistently, but that the sampling theory legitimizing municipal solid waste sampling does not treat moisture as a redistributable property among fractions. A comparative study conducted at the Integrated Complex of Basic, Professional and Technological Education (CIEB) in Eunápolis, Bahia, applied three protocols, one per campaign, on separate days: the General Method (integral sorting, 72.65 kg), ABNT NBR 10.007:2004 (homogenization with quartering, 21.53 kg), and ASTM D5231 (homogenization without quartering, 104.80 kg). The protocol that sorted the least mass (ABNT, 21.53 kg) registered only six of the ten categories, compared with ten detected under ASTM and nine under integral sorting, and aggregation into functional macrogroups reduced compositional distances between protocols by 1.4 to 2.6 times, concealing internal divergence rather than eliminating it. Because protocol and collection day remain confounded in that design, the mechanism was isolated through a paired experiment, over four campaigns in which the same waste lot was simultaneously subjected to the three methods. Homogenization transferred water from the biodegradable organic fraction to the dry fractions of plastic and paper/cardboard; since the total water mass of the sample did not vary, this is internal redistribution, not material gain. This absorbed water inflated the apparent mass of dry recyclables by up to one quarter of the reported mass, with correction factors of 0.72 to 0.77 for plastics and paper/cardboard, validated by regression. The thesis consolidated these findings into a proposition to adjust the sampling standards: it recognizes MSW as a hygroscopic material, recommends sorting without homogenization when feasible, requires reporting of composition on a dry basis per fraction and recording of moisture conditions, and provides for recording the mass effectively sorted, without instrumentation beyond that already specified in the characterization routine. The results evidence a class of preparation error not modeled by classical sampling theory, with implications for the sizing of treatment systems and for solid waste policies.