loading...

INESC Coimbra

UC | INESC | IPL

Multiobjective Models in Energy Efficiency Evaluation Problems

Title: Multiobjective Models in Energy Efficiency Evaluation Problems

Coordinator: Carlos Henggeler Antunes

Dates start/end: Outubro 2007 / Setembro 2010

Entity: FCT (PTDC/ENR/64971/2006)

Synopsis: The promotion of energy efficiency, namely those actions aimed at changing the shape and level of the demand, has been implemented by the utilities with operational or economic goals, or supporting energy and environmental policies. These usually conflicting objectives maintain their attractiveness even in a restructured electricity sector. In this scenario, several entities exist in each branch of activity meaning that the goals when implementing energy efficiency actions may also vary from entity to entity, thus making the assessment of such actions harder. In this context, the main objective of this project is the development of multiobjective models able to be used in the assessment of energy efficiency actions at different demand aggregation levels and according to the objectives of different entities that may be interested in the promotion of such actions.

Among all the energy efficiency promoting actions those aimed at changing the shape and the level of the demand were the most often implemented. However, a previous assessment of these actions must be done in a detailed and careful way in order to avoid eventual negative effects, such as increase in maximum demand, strong reduction of revenues or discomfort caused to consumers. Such evaluation requires suitable load models that allow the simulation of both the demand and energy efficiency actions at different demand aggregation levels. These models are based on physically-based end-use models. In order to be able do deal with the multiple objectives and combinatorial characteristics of the problem, adequate multiobjective mathematical models are used, explicitly addressing the multiple, conflicting and incommensurate aspects of evaluation (of technical, economical and quality of service nature). The suitable assessment of energy efficiency actions impacts at different demand aggregation levels and according to different entities existing in the several branches of activities will be possible in this way. In the context of multiobjective programming models the search for non-dominated solutions will be based on genetic algorithms (GAs). The choice of GAs as the most suited methodological approach is based on experience of the research team. These problems are characterized by difficult search spaces, in which both the morphology and the location of solutions are unknown. GAs, while search and optimization tools, are well suited for complex and combinatorial multiobjective problems since they work with a set of potential solutions in each generation allowing a broader characterization of the search space. In multicriteria problems (in which the alternatives are known a-priori) the goal is to develop decision support systems that enable the classification of the energy efficiency actions according to pre-defined ordered categories. The members of the research team have a fruitful experience on the development, use and implementation of methodological tools devoted to these problems (such as the IRIS method).

The main objective of this project is the development and implementation of models and methodologies for energy efficiency assessment in restructured power systems. Four main tasks will be carried out: development of multiobjective models able to deal with all the aspects under evaluation and usable by different entities; development of a computational platform capable of simulating both the demand and the energy efficiency actions at different demand aggregation levels; implementation of a genetic algorithm suited for the identification and search for non-dominated solutions; development of a decision support system (DSS) to be used in the assignment of energy efficiency alternatives to pre-defined ordered categories.

Site: http://www2.deec.uc.pt/~ch/ProjMMEE.html