Title:
Optimization of Waste Management Scenarios by Principal Component Analysis: a Case Study in Reunion Island
Author(s):
Hatik, C., Gatina, J.C.
Document(s):
Paper
Poster
Abstract:
Knowledge of waste composition and production is a requirement to build an efficient waste management scenario. Moreover, analysis of this data at a detailed level of observation (regional or communal) is useful to create adapted local scenarios, thus optimizing the overall waste management. However, working at a detailed level of observation multiplies the number of scenarios to build. In this article, we make use of Principal Components Analysis (PCA) to identify similarities between local administrative areas. This simple method is directly based on the quantitative and qualitative evaluation of wastes at the desired level of observation. The PCA provides an optimal two-dimensional or three-dimensional graph on which all the administrative areas are classified. The axes are based on the waste characteristics that are the most discrimining in the dataset. So, two administrative areas close the one from another in the graph have similar waste production. By grouping administrative areas based on their waste production, this analysis is an efficient way to drastically reduce the number of local waste management scenarios to define and it also favorizes cooperation among administrative areas of the same group. To illustrate our methodology, we focus on the specific case of biowaste of Reunion Island. Reunion Island is a French oversea territory of approximately 2500 square miles, located in the Indian Ocean, and composed of 24 municipalities. Waste are managed by the Regional Council (Conseil Régional) at the island-scale. With landfills having reached their maximal capacity and an ever-growing population, the waste management scenario must be revised. An optimal global scenario is particularly difficult to build due to the specificities of biowaste composition and production of each municipality. Indeed, coastal municipalities are urbanized and developed whereas the one in the highlands are more rural, specialized on crop and breeding. These differences motivate the definition of biowaste management scenarios at the municipality scale, thus facing the creation of 24 local scenarios. The PCA analysis enabled to identify 5 groups of municipalities, thus reducing the number of required scenarios to build drastically. Through the example of Reunion Island, we demonstrate effectiveness of PCA analysis to identify biowaste composition and production patterns among different administrative areas. One of the assets of PCA is the fact that it features graphical representation of the dataset, enabling anyone to visualize correlation existing between the different administrative areas. Principal Components Analysis of biowastes production can then be used for different purposes: biowaste management scenarios pooling, cooperation among administrative authorities, location and optimization of waste treatment plants, and more generally waste management decision-making.
Keywords:
decision making, principal component analysis, town's typology and territories analysis, waste management scenarios
Topic:
Biomass Resources
Subtopic:
Municipal and Industrial Wastes
Event:
25th European Biomass Conference and Exhibition
Session:
1DV.1.86
Pages:
234 - 236
ISBN:
978-88-89407-17-2
Paper DOI:
10.5071/25thEUBCE2017-1DV.1.86
Price:
FREE
