Comparative life-cycle assessment of supply strategies for forest-monitoring IoT sensor networks

Details

Supervisors

Bol, David

Faculty

Ecole polytechnique de Louvain

Degree label

Master [120] : ingénieur civil mécanicien, à finalité spécialisée

Abstract

Forest fires are becoming more frequent and, in the context of climate change, their mitigation is a necessity to prevent significant greenhouse gas emissions into the atmosphere. Their impact extends beyond the environmental sphere, with economic and human health consequences. Many techniques are currently being implemented to monitor forest fires with an active scientific literature on this topic, and remote sensing relying on IoT devices offers a promising solution for continuous monitoring and early warning of fire hazards. However, the solution must not become part of the problem. The purpose of this work is to compare three existing energy sources available for IoT devices, (1) solar energy harvesting, (2) battery power and (3) recharging via Unmanned Aerial Vehicles (UAVs). The cradle-to-gate Life-Cycle Assessment (LCA) approach is used to assess the environmental impact of the sensor and the energy sources, considering a 10-year use phase. The comparison is based on a common topology using a sensor inspired by the product of Dryad Networks GmbH, adapted to the need of each energy source. The system studied involves the deployment of 80 sensors to cover the area around one kilometre of road in a forested area, within a radius of 2 km. The studied system of 80 sensors emits 504 kg CO2e when powered by batteries, 270 kg CO2e when recharged by UAVs and 154 kg CO2e when relying on solar cells. Several impact categories are analysed. Across all of them, the battery system is more impactful than the solar system and the UAV system. Apart from the GWP, the lower difference is a factor of 9.55 between the estimated impacts of the battery system and the solar system, for the freshwater eutrophication. The higher difference is a factor of 700 between the estimated impacts of the battery solution and the solar solution for the cancerous effects on human health. When compared to the potential CO2 emissions of forest fires, the use of such remote sensing systems seemed fully justified. However, in terms of planetary boundaries, the battery and UAV options raise concerns. If all the world's forests were to be covered, their impact on the cancerous and non-cancerous effects would reach 4.7% and 1.39% of the planetary boundaries for the battery, and 0.42% and 0.13% for the UAV solution. The UAV has potential, but the way in which its use phase is modelled needs further investigation for it to provide robust conclusions.