Sitemap Contact Events New Home

Database of research projects FEDRA

Presentation

Research actions

Persons

Search

Research and applications > Projectendatabanken > Database of research projects FEDRA

Multiscale Modelling of Electrical Energy Systems (M2E2S)

Research project P7/02 (Research action P7)


Persons :

  • Prof. dr.  DRIESEN Johan - Katholieke Universiteit Leuven (K.U.Leuven)
    Coordinator of the project
    Financed belgian partner
    Duration: 1/4/2012-30/9/2017
  • Dr.  VANDEVELDE Lieven - Universiteit Gent (RUG)
    Financed belgian partner
    Duration: 1/4/2012-30/9/2017
  • Dr.  SLODICKA Marian - Universiteit Gent (RUG)
    Financed belgian partner
    Duration: 1/4/2012-30/9/2017
  • Dr.  GEUZAINE Christophe - Université de Liège (ULG)
    Financed belgian partner
    Duration: 1/4/2012-30/9/2017
  • Dr.  CLENET Stéphane - Arts et Métiers ParisTech (AMP)
    Financed foreign partner
    Duration: 1/4/2012-30/9/2017
  • Dr.  HADJSAID Nouredine - Grenoble Institute of Technology (Gren_INP)
    Financed foreign partner
    Duration: 1/4/2012-30/9/2017
  • Dr.  NORDSTROM Lars - Royal Institute of Technology Sweden (KTH)
    Financed foreign partner
    Duration: 1/4/2012-30/9/2017

Description :

During the previous phases of the IAP programme the members of the consortium have successfully developed increasingly sophisticated numerical methods for analysing and simulating a variety of coupled electromagnetic problems, with focus on electrical energy transducers and systems. In the last IAP phase, these numerical methods
have allowed to optimise applications, to analyse and interpret measurements by solving inverse problems and to identify model parameters. In this new IAP phase, the enlarged consortium wants to focus on the natural « next step »: modelling and managing complete electrical energy systems (multiple electrical energy devices linked through a power grid), based on the measurements and the solution of inverse problems.

This problem is inherently multiscale (in space and time), as the interactions in the fines may significantly influence the solution of the largest scale problem. Understanding this multiscale behaviour is becoming increasingly important as the power grid evolves into a complex system of interconnected distributed generation and electronically controlled loads and contains less and less natural, stabilizing inertia.

The present project concerns fundamental research in this domain. Three main parts can be distinguished:
● Mathematical modelling of electrical energy systems;
● Applications in electrical energy systems;
● Validation environment.

Clearly, the appropriate modelling of the extremely low in inertia electrical energy system requires a fundamentally new framework as the classical methodologies are all based on the presence of inertia and linked energy storage. Such modelling methodology is a prerequisite for the massive deployment of renewable energy resources as a part of the current energy policy.

Part 1. Mathematical modelling of electrical energy systems
In a first part original numerical models of electrical energy systems are investigated and developed. Furthermore, the partners investigate model-order reduction techniques to replace time-consuming numerical models of generation and loads by simplified models suitable for inclusion in grid simulators. Another aim is to construct, analyse and solve well designed mathematical models, describing the fast dynamics of low inertia power systems, taking into account uncertainties due to stochastic effects at generation, load, and grid (transmission and distribution) side. A last challenge to tackle is devising suitable algorithms for managing the instantaneous balance in a power grid. This is part of the very general setting of optimization and inverse problems in which the consortium has built up a large expertise. The fundamental mathematical tools to develop must satisfy a number of additional requirements in terms of speed and robustness, and have to be extensively tested both under normal and extreme conditions. Finally, the resulting control methods themselves should allow for distributed deployment and extended safe autonomy, so as to avoid introducing a central point-of-failure into the system.
Main novelty aspects with respect to the state of the art:
● Development of fast (“real-time”) simplified models of generators and loads;
● Development of original comprehensive mathematical models of low inertia power grids;
● Convergence and error analysis of the numerical approximation schemes, reliability of the obtained solutions;
● Development of new paradigms for the incorporation of stochastic uncertainties in (open and closed-loop) dynamically changing systems, together with robustness and stability assessment of the systems with respect to these uncertainties ;
● Sensitivity analysis with respect to uncertain grid and resource parameters;
● Optimal choice of a regularization parameter in ill-posed inverse problems.

Part 2. Applications in electrical energy systems
Herein the models investigated and developed in Part 1 are tailored to applications in all domains of electrical energy systems: generation, transmission, distribution and loads and the interconnection between them. In a first work package, an open object-oriented system for the multi-domain modelling and simulation of complex electrical energy systems with focus on the interconnection between generation and load is developed. The main effort lies in the combination of the first fundamental theoretical part of the project, with special attention to the balancing between the generation, the renewable resources, and load of the electrical energy system. The second work package deals with the investigation of the different operating conditions of the transmission system with extremely low inertia. The third work package focuses on the design methodology for control strategies for microgrids, taking into account the power electronic converters, connecting loads, generation units and storage devices. Herein both islanded and grid-connected mode, fundamentally different operating modes, are considered. Finally, multiscale models of the loads (both in time and space), mainly electric drives, are investigated.
Main novel aspects are:
● Object-oriented descriptions of energy system components to describe flexible and dynamic electricity systems;
● Introduction of stochastic variables in the electrical energy system models;
● Development and analysis of control strategies (on different time scales) for active distribution networks (microgrids), accounting for the lack of inertia and the distribution network characteristics;
● Design of new grid architectures able to accommodate Distributed Generation;
● Development of observability tools for active distribution grid;
● Development of advanced tools for optimizing grid operation;
● Development of the concept of self-healing in active distribution grids;
● Models of inertialess systems (or with very little inertia);
● New optimized control methodologies for the joint operation of power electronic converters in HVDC installations and renewable energy resources.


Part 3. Validation environment
The developed models and methods will be validated by means of real-life applications, namely: 1) microgrids and Virtual Power Plants (VPP); 2) High Voltage Direct Current (HVDC) and 3) power electronic drives. The first work package deals thus with small-scale electrical energy systems with a maximum degree of techno-economic autonomy. They can be stand-alone (microgrid) or grid-connected (virtual power plant, VPP). The aim is to develop and validate an objective and neutral approach to assess the maximally feasible degree of autonomy for a given set of technical resources or to optimally determine the degrees of freedom of device (e.g. size) or control setting within a microgrid or VPP. In the second work package, it is essential to investigate control schemes for both the renewable energy resources and the HVDC transmission in order to optimise the use of the available energy in the future inertialess systems while avoiding any detrimental effects.
In the last work package, the developed reduced order models are used in time-simulation packages to study the dynamic behaviour of controlled power converters with electric machines as load. These reduced order models will help building hybrid simulators with a precise description of the loads.
The main added values comprise:
● Multiobjective optimisation of microgrids and virtual power plants designs;
● Validation of microgrid/VPP designs and control strategies by simulation and (lab) experiment;
● Validation of self-healing strategies with advanced protections for active distribution grids;
● Validation of advanced DMS functions for active distribution grids;
● Validation of the impact of ICT on grid performance (e.g. fault location and restoration);
● Validation of the operational strategies for renewable generation sites which are decoupled from the main grid by a DC grid.


About this website

Privacy

© 2017 Science Policy PPS