The Bitcoin mining hosted in the blockchain network consumes enormous amounts of energy and generates electronic waste at an alarming rate. The paper aims to model and predict the future values of these two hazardous variables linked to conventional Bitcoin mining. We develop two predictive models using Facebook's Prophet algorithm and deep neural networks to identify and explain energy consumption and electronic waste generation patterns. The models rely on several explanatory features linked to the blockchain microstructure and the Bitcoin marketplace. We assess the predictive performance of the two models based on daily data of energy consumption and electronic waste generation and eleven key input features. We use local interpretable model-agnostic explanation (LIME) and Shapley additive explanation (SHAP) for explaining how these inputs can predict and control energy consumption and electronic waste generation. The findings assist in accurately estimating the future figures of energy discharge and electronic waste accumulation in the present Bitcoin mining setup. The study also reveals the block size to be the major driver.