Método para identificación de periodos en análisis de descomposición LMDI de la intensidad de carbono del sector eléctrico
The Logarithmic Mean Divisia Index (LMDI) method allows decomposing a time series associated with a dependent variable of interest, for example, carbon intensity into explanatory variables of the phenomenon that can be structural (for instance, economic dispatch) or intensity (for instance, energy e...
- Autores:
-
Rojas Lozano, Daniela Del Pilar
- Tipo de recurso:
- Fecha de publicación:
- 2020
- Institución:
- Universidad de los Andes
- Repositorio:
- Séneca: repositorio Uniandes
- Idioma:
- spa
- OAI Identifier:
- oai:repositorio.uniandes.edu.co:1992/50984
- Acceso en línea:
- http://hdl.handle.net/1992/50984
- Palabra clave:
- Dióxido de carbono atmosférico
Generación de energía
Mitigación de dióxido de carbono
Ingeniería
- Rights
- openAccess
- License
- https://repositorio.uniandes.edu.co/static/pdf/aceptacion_uso_es.pdf
| Summary: | The Logarithmic Mean Divisia Index (LMDI) method allows decomposing a time series associated with a dependent variable of interest, for example, carbon intensity into explanatory variables of the phenomenon that can be structural (for instance, economic dispatch) or intensity (for instance, energy efficiency). This method can be carried out on a multi-time basis, considering several consecutive periods of a certain duration. The selection of the duration of each period has been made according to arbitrary criteria that do not consider changes in the trend in the variable of interest. Usually, the selection of each period is carried out in accordance with the national quadrennial or five-year plans. As the LMDI method is applied point to point, the selection of periods of arbitrary duration prevents the capture of the explanatory variables associated with a change in trend that is not properly considered in the selection of the periods. Consequently, it is necessary to develop a methodology that allows identifying the relevant periods in the decomposition analysis by LMDI. This work seeks to meet this need. The proposed method consists of dividing the time series into subgroups that share the same trend with a minimum mean square deviation. Four algorithms are proposed for this purpose. The partitioning of the subgroups is done in such a way that there is no loss of information when performing the segmentation. The method has been applied to the time series for Agreggated Carbon Intensity (ACI) of the electricity sector in Colombia and in Latin America and the Caribbean. The advantages of the method are discussed by comparing the results obtained with the solutions based on arbitrary periods reported in the literature. |
|---|