Predicting the hydraulic conductivity of compacted soil barriers in landfills using machine learning techniques
Soil; Waste Disposal Facilities
Machine learning models (MLMs) were developed to predict saturated hydraulic conductivity of compacted soil barriers and help to identify appropriate soils for the construction of landfill liners and covers. Data from hydraulic conductivity tests on compacted soil barriers were collected from the literature and compiled into a database for MLM construction. The database contains 329 records of hydraulic conductivity tests associated with 12 selected impact factors covering physical properties, compaction efforts, and hydration and mineralogy behaviors of compacted soil barriers. Three machine learning algorithms (random forest, gradient boosting decision tree, and neural network) were used to develop MLMs, and a statistical technique (multiple linear regression) was used to compare the precision of predictions with the MLMs. Results from this study showed that the random forest model provided the best prediction of the hydraulic conductivity of compacted soil barriers, with 100% of predicted hydraulic conductivity within 100-time differences to measured hydraulic conductivity and 93% within 10-time differences. Feature importance analysis showed that void ratio after compaction, fines content, specific gravity, degree of saturation after compaction, and plasticity index of soils are the top-five factors (in descending order) that influence the hydraulic conductivity of compacted soil barriers and are recommended for a precise prediction. Three predictive MLMs were created for industries as simple tools to screen the soils prior to the construction of compacted soil barriers in landfill liners and covers.
Waste management (New York, N.Y.)
Digital Object Identifier (DOI)
Tan, Yu; Zhang, Poyu; Chen, Jiannan; Shamet, Ryan; Hyun Nam, Boo; and Pu, Hefu, "Predicting the hydraulic conductivity of compacted soil barriers in landfills using machine learning techniques" (2023). UNF Faculty Research and Scholarship. 3212.