Modelling waste generated during construction of buildings using regression analysis.

Area of Science:

• Construction Management
• Environmental Engineering
• Waste Management

Background:

• The construction industry generates significant waste, necessitating effective management strategies.
• Existing theoretical models for construction waste may not fully capture local practices.
• Accurate measurement and modeling are crucial for developing targeted waste reduction plans.

Purpose of the Study:

• To measure and analyze construction waste generation in Porto Alegre, Brazil.
• To develop predictive models for total waste and waste based on construction stage and time.
• To assess the statistical performance of developed waste generation models.

Main Methods:

• Analysis of waste data from 18 building construction sites.
• Regression analysis to develop two distinct waste generation models.
• Model 1: Predicting total waste based on building attributes.
• Model 2: Predicting waste based on construction stage and time schedules.

Main Results:

• An average construction waste generation rate of 0.151 m³ m⁻² was observed.
• Model 1 (total waste) achieved an R² of 0.81, indicating good predictive performance.
• Model 2 (time-dependent waste) achieved an R² of 0.91, showing an S-shaped relationship with time.

Conclusions:

• The developed regression models demonstrate satisfactory statistical validity for predicting construction waste.
• These models can significantly aid in creating more effective waste management plans during the preconstruction phase.
• Understanding waste generation dynamics based on building attributes and project timelines is key to sustainable construction practices.