“20th PSCC 2018 papers submission and review platform

Full Program »

Modelling Active Distribution Networks under Uncertainty: Extracting Parameter Sets from Randomized Dynamic Responses

Distribution grids are expected to host an increasing amount of Dispersed Generation Units (DGUs). These units will be more and more active in response to disturbances. Therefore, there is an increasing need to build dynamic models of such Active Distribution Networks (ADNs).

One major issue, however, lies in the uncertainty affecting the behaviour of DGUs and even more of loads. A traditional approach is to perform Monte Carlo simulations involving random variations of the parameters of the model. Thus, for given disturbances and operating points, randomized time-domain responses are generated.

The next challenge is to compress this huge amount of
information into a small number m of models representative of the system behaviour. Each model is an instance of the same analytical model but with different parameter values. The objective is thus to identify m parameter sets among those drawn at random yielding representative dynamic responses.

The results deal with a 75-bus 11-kV distribution grid, feeding 75 loads and hosting 75 DGUs. The load model has a static and a dynamic equivalent motor part. Some DGUs have fault-ride-through and reactive current injection
capabilities to support the grid voltage during faults. The load and DGU parameters are randomized over the Monte
Carlo simulations, as well as from one bus to another.

Author(s):

Gilles Chaspierre    
University of Liège
Belgium

Patrick Panciatici    
RTE - R&D Dept.
France

Thierry Van Cutsem    
FNRS and University of Liège
Belgium

 

Powered by OpenConf®
Copyright ©2002-2014 Zakon Group LLC