Analysis of the Environmental Impact of the Use of Electric Generators in the 2024 Energy Crisis in the Canton of Cuenca

The 2024 energy crisis in Ecuador, driven by severe drought conditions, significantly impacted hydroelectric power generation, leading to prolonged blackouts in urban areas such as Cuenca. In response, the widespread use of electric generators increased considerably, raising concerns about their environmental and health impacts. This study presents a novel methodology to estimate pollutant emissions (specifically CO2, CO, HC, and NO_x) based solely on environmental noise levels. A stratified sample of 248 generators was measured across different urban zones using a sound level meter and a gas analyzer, following Ecuadorian regulations. Three machine learning models (linear regression, decision tree, and neural network) were trained to predict emissions from noise data. The decision tree model showed the best predictive capacity, with a maximum R² of 0.66 for CO₂, a value that, while moderate, is considered acceptable in environmental estimation involving a single predictor variable. The trained model was then applied to a broader dataset of 710 generators to estimate total emissions during 691 hours of power outages recorded over 89 days. The results revealed the emission of 428 tons of CO₂, 14.91 tons of C}, 0.72 tons of HC, and 7.6 tons of NO_x. In parallel, noise levels exceeded local regulatory limits, particularly in central and residential areas, indicating significant public health implications. The study supports the implementation of regulatory strategies, improvements in generator technology, and a shift towards cleaner energy sources in emergency contexts.