Current Journal of Applied Science and Technology

Growing demand for electricity, particularly in developing countries, is putting pressure on fragile grids and exacerbating energy insecurity among low-income households, which spend a significant portion of their budget on electricity. Traditional energy management solutions, such as smart meters or differentiated tariffs, prove to be limited without direct control mechanisms, leading to unpredictable bills. This study proposes the implementation of an intelligent system that defines an energy consumption threshold and adjusts behaviors (ISECTB), based on an ESP32 microcontroller and current sensors (SCT-013) to monitor consumption in real time. An algorithm sets a monthly threshold of 150 kWh. The system uses an IoT architecture with sensors, a monitoring module (ESP32/Arduino), a management algorithm and actuators (relays) to adjust loads.

Tested for three months in five households, the system reduced consumption by 10.3% to 13.9%, with an average of 11.8%. It stabilised consumption and limited peaks, while maintaining essential comfort. The system reduced peak collective power demand by 14.3% during nighttime peak consumption periods (6 PM - 10 PM). This demand stabilization demonstrates its usefulness in securing the local electricity grid. Thus, the system simultaneously benefits households and the stability of the distribution network.

The ISECTB (Implementation of an Intelligent System which sets an Energy Consumption Threshold and adjusts Behaviors), proves to be an effective and socially acceptable alternative to prepaid meters, enabling a significant reduction in consumption. Prospects include the integration of renewable energies, the use of machine learning for predictive management, and larger-scale deployment with socio-economic assessment.