• Aucun résultat trouvé

Modelling the integrated management of organic waste at a territory scale

N/A
N/A
Protected

Academic year: 2021

Partager "Modelling the integrated management of organic waste at a territory scale"

Copied!
1
0
0

Texte intégral

(1)

Modelling the integrated management of organic waste

at a territory scale

J.-C. Souliéa and T. Wassenaara a CIRAD – PERSYST Department

Research Unit Recycling & Risk - TA B-78/01 Avenue Agropolis Lavalette

34398 Montpellier Cedex 5 Email: jean-christophe.soulie@cirad.fr

Abstract: Modelling integrated management of organic waste at a farm level is still a key challenge, but one can find several works in the literature. However, the shift to a larger (e.g area, district or county) scale is even more complex. Consequently, associated modelling work is still considerably under-developed. Here we use the UPUTUC model (french acronym of Production Unit, Transformation Unit, Consumption Unit) to examine fluxes of organic matter residues within the Territoire de la Côte Ouest (TCO), on the west of Réunion Island.

UPUTUC is an agent-based model and is developed using the AnyLogic platform. This platform is a multi-method modeling environment and allows for the combination of discrete event models, agent-based models and system dynamics models. It also provides spatial features in order to use information coming from GIS. UPUTUC uses three agents. These agents are: (i) production units that produces organic matter or waste (for example: pork, poultry breeding and sludge produced by a wastewater treatment plant), (ii) transformation

units that produces compost or fertilizer with organic matter coming from the production unit, and (iii) consumption units that uses transformed organic matter from the transformation unit or organic matter coming

directly from the production unit (for example: a sugar cane field, a market gardening and a grassland). State variables of production units are a production rate, a stock, and all times and distances to the transformation

units and consumption units in the model. State variables of transformation units are a production rate,

equations the transform organic matter into compost or fertilizer, a transformation time, a stock, capacity of trucks that delivers products to consumption units, and finally, all times and distances to consumption units in the model. State variables of consumption units are an area, compatible fertilizers with the unit, fertilizer requirements, and periods during which fertilization is possible. In addition, the agents are spatially situated based on GIS information. The French Land Parcel Identification System (LPIS) and the Farmers’ block from the European Common Agricultural Policy provide this information. Moreover, as Réunion Island is a volcanic island, the roads are very rough and consequently the logistic dimension is crucial in the model. Therefore, as the dynamic of logistic characteristics of the associated road network is not simulated (for example, the time for a vehicle to transport organic matter from a production unit to a transformation unit) these variables are pre-computed off line in a previous step and are considered as parameters. In UPUTUC, there are 51 production

unit agents, 2272 consumption unit agents and, finally, only 4 transformation unit agents. Several scenarios

are tested. First, the current situation in order to check how the model behave. Second, the introduction into the first scenario of a new transformation unit (called UT1) that takes as inputs only litter, manure and crushed green wastes and combines them into a specific compost used by all the consumption units. Third, the introduction into the second scenario of two new transformation units: the first one (called UT2) takes as inputs only poultry dropping, vinasse and crushed green wastes and combines them into a specific compost, and the second one takes as input compost from UT1 and UT2 and produces granulates with them.

Keywords: Agent based model, GIS, AnyLogic, organic waste management, Réunion Island

B2. Advances in Agent-Based Modelling of complex biological, ecological and agricultural systems

EXTENDED ABSTRACT ONLY

Références

Documents relatifs

This paragraph is devoted to a quick review of some of the principal facts concerning Fourier and Fourier-Stieltjes series, measures, and convolutions, which

For specific requests about statistical resources and databases, you can address to the documentalist responsible for the document collection, which is in an

Copyright and disclaimer: The project EASIT has received funding from the European Commission under the Erasmus+ Strategic Partnerships for Higher Education programme, grant

This reading list presents a selection of references related to the X elements of Unit X in the EASIT materials.. Please note that trainers and trainees using this reading list

Subject: Write the type of document (e.g. Reading list, Additional material).. Author: Write the name of the partners who are in charge of

2. Remember to go through the Enter the number of the question here. • Questionnaires will be based only on video lectures. • Each questionnaire should have around 15 questions. •

Once you find the document you need and it belongs to the Statistics Unit (location: C. Socials-Estadística), you should write down the call number (an alphanumeric code that

Representations of linear functionals In this section we prove the following main result of the paper.. Theorem