Performance analysis of biochar and W. Robusta palm waste reinforced green mortar using response surface methodology and machine learning methods

Abstract

The pressing issue of growing carbon dioxide (CO2) levels and global warming has prompted efforts to create new building materials with the capacity to absorb and store CO2. Using cement substitutes helps to preserve the environment by slowing the spread of carbon dioxide. As a result of the abundance of Washingtonia robusta waste (WRW) produced during maintenance operations on Washingtonia robusta (WR) and sand from the dunes of Algeria's Sahara desert, as well as the production of biochar made from a pyrolysis at 500°C of these residues (WRWB) on the flexural, compressive properties and porosity of cementitious mortars are explored. The approach is based on gradually replacing cement with WFRB biochar and WRW fibre waste at varying rates: from 0 % to 2 % with a step of 0.5 % and treated for different periods of time (24, 72, and 168 hours) at different NaHCO3 concentrations (4, 8 and 16 %). According to the response surface method (RSM) and artificial neural networks (ANN), the optimal cement substitution of rates were 1,8 % of WRWB and 1,3 % of WRW treated with 4 % of CaCO3 concentration for a 23,6 hour. Furthermore, it is appropriate to note that predictive accuracy using ANN models is higher than that of RSM models, as demonstrated by their good correlation with developed models' experimental data. These techniques have increased the use of green mortars and their acceptance as construction materials.