This title appears in the Scientific Report :
2016
On the Estimation of Regional and Sectoral Electricity Load Profiles: A Generalized Cross-Entropy Approach
On the Estimation of Regional and Sectoral Electricity Load Profiles: A Generalized Cross-Entropy Approach
This paper concerns the estimation of region and sector specific electricity load profiles on the bases of partial and possibly conflicting information by means of a generalized cross-entropy model. Usually, region and sector specific load profiles are derived by means of disaggregating a national o...
Saved in:
Personal Name(s): | Schröder, Thomas (Corresponding author) |
---|---|
Többen, Johannes | |
Contributing Institute: |
Systemforschung und Technologische Entwicklung; IEK-STE |
Imprint: |
2016
|
Conference: | North Sea Conference, &th International Syposium on energy Challenges and Mechanics, Inverness (Great Britan), 2016-08-14 - 2016-08-18 |
Document Type: |
Conference Presentation |
Research Program: |
Helmholtz Interdisciplinary Doctoral Training in Energy and Climate Research (HITEC) Assessment of Energy Systems – Addressing Issues of Energy Efficiency and Energy Security |
Publikationsportal JuSER |
This paper concerns the estimation of region and sector specific electricity load profiles on the bases of partial and possibly conflicting information by means of a generalized cross-entropy model. Usually, region and sector specific load profiles are derived by means of disaggregating a national one by assuming that electricity demand is proportional to indicators as GDP or population density. Although such assumptions are to some extend plausible, they bear the danger of delivering poor results if regional or sectoral structures strongly deviate from the national average. Therefore, a favourable solution would be the integration of as much sector or region specific information as possible in order to improve the quality of estimates. However, such attempts are often hampered by information conflicts between data from different sources. In this paper we propose a nonlinear programming model based on the principle of minimal cross-entropy for harmonising. The starting point of our approach is an initial estimate derived by regionalizing sectoral electricity demands at the national level on the bases of regional shares in national gross output by sector. In the second step, our model determines the target values by minimizing the information distance from the initial estimate, subject to constraints imposed by the additional data to be integrated. Potential conflicts between data constraints are resolved in term of finding compromise values between data points. These compromise values are found by including the minimization of the information distance from data points in their objective function. To our knowledge this paper is the first that proposes a flexible model for systematically disaggregating load profiles.In a case study for Germany we estimate the hourly electricity demands for 16 federal states, 35 sectors (NACE-classification) and nine different typical day categories including a distinction between seasons (summer, winter, spring/autumn) and weekdays (working day, Saturday and Sunday). This proceeding enables us to analyse the economic consequences of various power outage scenarios varying different impact factors. Furthermore the specific sectoral differentiation makes an application to input-output-tables possible. |