Statistical model for heating value of municipal solid waste in Brazil based on gravimetric composition

•Statistical models of LHV based on Brazilian wastes physical characteristics.•Empirical measures of moisture and heating value from samples of MSW in Brazil.•Samples were collected based on statistically representative sampling methodology.•Two models for accurate prediction of experimental data for LHV was built. Municipal solid waste (MSW) management is a serious problem for public administrations, especially in terms of treatment and final disposal. These wastes have a high energy content and one of the possibilities of treatment is the recovery of energy through thermochemical processes. The parameter used to measure the amount of useful energy available in collected waste when submitted to thermochemical processes is called the lower heating value (LHV), which is usually determined in the laboratory or through empirical models from the literature. To this end, this paper aims to present two models for prediction of the LHV in the municipal solid waste of the municipality of Santo André. Samples were collected from 36 garbage trucks in the above-mentioned city, from September 2015 to January 2016. The models were developed based on the results of the gravimetric composition and laboratory analysis. The technique used to develop the models was the multiple linear regression by least squares method. As a result, the models obtained mean absolute percentage error (MAPE) indexes of 5.09% and 5.52%, considered excellent according to the literature classification. In addition, the calculated LHV of the Santo André municipal waste was 7.03 MJ/kg, which indicates a great potential for energy recovery using thermochemical processes. These are the first LHV prediction models developed in Brazil, which has been a significant accomplishment in Brazil. The proposed models were developed using empirical measurements of moisture in the solid waste and the LHV on samples collected with a statistically representative sampling method.