Design of heliostat fields using a multi-objective evolutionary algorithm

The paper discusses the interest of the multi-objective optimization approaches for the design of complex energy systems and the basics of the original evolutionary algorithm used. The specific problems linked to the design of the concentrators of solar tower power plants are then introduced to illustrate the objective of the work presented. The first analysis deals with the optimization of the x-y positioning of each of the 500 rectangular reflectors of an existing predesigned field. In spite of the large number of variables (1000) the algorithm successfully calculated an optimum positioning of each concentrators and the main induced variations are shown. The second part of the paper describes a new program, which was made and coupled with the same algorithm to optimize the whole design in a two-objective approach (specific energy cost versus investment cost). The main hypothesis made, to keep the number of variables within a reasonable domain, was to distribute identical reflectors along a set of concentric ellipses around the tower. The interest of the method is illustrated for one given period with the Pareto curve showing all the optimum solutions (lowest specific energy cost at any given investment). The investment cost includes the cost of the heliostats themselves, the cost of land, the cost of the tower, etc. Variables include the x-y positioning, the height of the support and the dimensions of each rectangular reflectors as well as the height of the tower and the number of reflectors. These decision variables being given for each solution, the extracted energy is calculated by a program of CIEMAT [Sanchez, Romero, 2003]. Details of the design parameters are illustrated for selected optima along the Pareto curve References M. Sanchez, M. Romero Optimisation of heliostat field layout in central receiver systems based on yearly normalized energy surfaces. ISES, 2003, Goteborg

Data reduction for mixed integer linear programming in complex energy systems

Luc Girardin, Julia Granacher, Natalija Ljubic, François Maréchal, Luise Middelhauve

The task of optimally designing and scheduling energy systems with a high share of renewable energies is complex and computationally demanding. A widespread method for tackling this task is to apply mixed integer linear programming (MILP). Even though the branch-and-bound algorithms used for solving these programs have seen significant improvements in the last years, many problems cannot be solved without further time series aggregation (TSA) methods. State of the art approaches tackle TSA by using well known machine learning techniques to cluster yearly input data to typical periods. However, latter algorithms are usually evaluated by indicators on the performance of the algorithms themselves rather than the MILP optimization model. Furthermore, the selection of the optimal number of typical periods is commonly a subjective imposition of thresholds on these performance indicators. The issue of computational effort is eased by this generally accepted algorithm, but is still limited by the size of the problem, especially the number of integer decisions. This paper aims at proposing a algorithm for systematically reducing the input data for MILP optimization models and choosing the appropriate size. Contrary to most existing studies, the focus is on the impact on the objective function as well as the integer decision rather than on the quality of the clustering algorithm. The subject is addressed by exploiting the two-stage character of optimal design and scheduling of the system by sequentially performing k-medoids clustering. The demonstration of the algorithm on two case studies shows that a few typical periods are sufficient to achieve near optimal decisions. Multi objective optimization (MOO) is performed to assess the quality of the data reduction. The proposed approach is outperforming state of the art algorithms for TSA by reducing CPU time of more than 40%. The case study furthermore reveals that the runtime of the MOO can be reduced by approximately 90% while diverting less than 2 % on Pareto optimal solutions.

2021

Decentralized production in urban systems, the vision of the Institute of Energy Sciences of EPFL

François Maréchal

In 1998, the Board of the Swiss Federal Institute of Technology established the challenging vision of the 2000 W society. The goal is to reach, for Switzerland by the year 2050, an energy intensity target of 2000 Wyear/year/cap. Referring to a present intensity of 5000 W year/year/cap, this means an expected factor 4 reduction of the intensity. For the Swiss Federal Institute of Technology, the challenge is to contribute to this target by its Research and Development activities. In the Ecole Polytechnique Fédérale de Lausanne (EPFL), the Institute of Energy Sciences (ISE) of the School of Engineering (STI) is a group of research laboratories that has committed to contribute to this target in the field of efficient conversion and distribution of energy. Decentralised production of electricity in urban areas is one of the multidisciplinary headlight projects that has been selected to reach the 2000 W target. Decentralised production is considered as a way of producing the energy services required by a community in an integrated approach allowing for polygeneration of heating, cooling, refrigeration and electricity services. This new systemic approach requires a new vision where several networks that have to be operated simultaneously. Electricity has to be distributed through microgrids connected with the major grids. In this field, the laboratories from electrical engineering are developing new tools for the design and the operation of such grids as well as new equipments that guarantee the safe operation of the system and its power quality. The design of district heating networks and their integration with the energy services requirements is a complex task since the demands vary with the seasons, the hours of the day and are by nature stochastic due to the behavior of the inhabitants and the weather conditions. Furthermore, the profitability and the conversion efficiency of such systems mainly rely on the appropriate system design, i.e. the size of the piping and the equipments, as well as on its optimal management. Applying techniques such as thermo-economic modeling, process integration , exergy analysis and multi-objective optimisation, the Laboratory for Industrial Energy Systems (LENI) develops computer aided design methods for helping the different stakeholders (energy services companies, technology developers, final users, policy makers) in defining the most appropriate investments both from the side of the infrastructure and the energy conversion equipments and also from the optimal operation management side in order to place the technologies in the energy market context. The Laboratories of the Energy Sciences Institute also work on the development of new equipments for advanced energy conversion, such as high speed turbocompressors for heat pumping, solid oxide fuel cells (SOFC) and cogeneration concepts combining SOFC and gas turbines. Within the Institute, this advanced R&D effort is performed in a multidisciplinary approach combining experimental work with multi-scale multi-physics computer aided design tools like CFD and optimisation.

2005

SCCER-FURIES - A decision support tool for the geo-localised design and assessment of the integration of multi-carrier/multi-services/multi-grids innovative energy conversion systems in urban systems and industrial clusters - Deliverable 1.5.4

Luc Girardin, François Maréchal, Paul Michael Stadler

The computing platform (OSMOSELua) has been developed in the Industrial Process and Energy Systems Engineering group IPESE at EPFL as a flexible and robust tool for the design of complex integrated energy systems. The decision tool has been developed using the script language Lua, to help decision making for the design and operation of various systems (energy, supply chains, industrial processes etc.). A research tool in essence, it also aims at capitalizing the knowledge acquired and methods developed by the research group. OsmoseLua can handle GIS data based on a generic object‐oriented model of each urban energy system component type (buildings, utilities and networks). Multi‐carrier/multi‐services/multigrids layers are treated using energy flows such as fuel or electricity associated to unit process. A multi‐period description allows to characterize the main stages of the urban architectural morphing bringing the long term vision needed to perform the coordinated planning of multienergy networks.