Modelos para estimativa da geração de resíduos sólidos urbanos em municípios brasileiros

Abstract

In order to know and understand the relationship between Municipal Solid Wastes (MSW) generation and demographic, economic and cultural aspects of the population, for example, mathematical models can be employed. In this context, the objective of this study was to develop mathematical models to predict the amount of MSW generated, based on the database from Sistema Nacional de Informações sobre Saneamento – SNIS, and socioeconomic variables. Multiple regression was used. Because of the high diversity of municipalities in Brazil, the sample was grouped into some class of population size and income. It was formulated models for each of the dependent variables: mass of MSW, per capita generation rate 1 (calculated based on the municipal urban population) and per capita generation rate 2 (calculated based on the municipal total population). About 25% of database was saved for models validation. Results showed that grouping the sample based on the municipal population resulted in valid models for almost all the class and that for municipalities with population up to 3,000,000 inhabitants the generation of USW increased with increasing population. When grouping the sample by population classes and income subclasses it was observed that, for municipalities with population between 30,001 and 100,000 inhabitants, which had not presented any valid model for the condition adopted, presented one valid model for income rate above R$ 29,000.00 per capita. For population up to 250,000 inhabitants, it was observed a direct correlation between MSW generation intensity and income level. When grouping the sample into income classes and population subclasses, the analysis of per capita generation rate showed that an increase on the municipal population is associated with an increase on the MSW generation intensity. The valid models formulates presents mean relative error of 23%, and can be employed at municipalities that corresponds to 60.1% of Brazilian population (relative to the year 2015). Given such results, it can be concluded that the generation of MSW was related to the income level and size of the municipal population. Moreover, results showed that the per capita generation rate calculated based on the total municipal population presented a more uniform behaviour than that of the per capita generation rate calculated based on the total urban population, possibly because the degree of uncertainty associated with the prediction of the urban population of the municipality. It was possible develop mathematical models for mass MSW and to rates 1 and 2 of MSW.