Distributed Energy Storage Simulations in Real-Time Environment (DESiRE)

 This document provides detailed documentation of the lab access project titled, "Distributed
 Energy Storage Simulations in Real-Time Environment" with the acronym DESiRE. The project
 deals with real-time scheduling of Battery Energy Storage devices, namely BESS, which can
 support a medium-voltage distribution network with high PV penetration. The project makes use
 of the planning of a distribution network and delves further into the scheduling problem. Real
time BESS scheduling is important, as it covers the forecasting errors, and thus the undesired
 effects of perturbations not considered in the planning phase can be minimized.
 The project considers an IEEE-33 bus distribution network for the optimal placement and sizing
 of a predetermined number of BESS.Aplanningalgorithm is developed based on the objectives
 namely voltage deviation, line loss minimization and the minimization of line loading. Apart
 from the predetermined BESS number constraints, the constraints are other energy balance
 and power balance equality constraints and limit-based inequality constraints. The planning
 scenario offers the size and the locations of the storage units in the network, and the scheduling
 algorithm works on a limited search space and thus can run an optimization algorithm in a short
 time. However, the scheduling algorithm conventionally does not support deviations in the
 standard load profiles, and therefore, a beta distribution function is used to generate a small
 amount of forecasting errors in the profiles. Modification in the load profile input helps take into
 account, the changes in the load in real-time, thus rescheduling the BESS profiles.
 The scheduling method is made more robust in the DESiRE project as a collaboration between
 the University of Malta and RWTH Aachen. The entire process of scheduling is now translated
 to real-time simulations, thus making it an online optimization approach. This is pursued with
 the help of VILLASNode, which is a flexible gateway that enables interaction between MATLAB
 and RTDS, even for different IP. The VILLASNode framework actuates a two way communica
tion between MATLAB and RTDS, and thus MATLAB is emulated as a Controlled-in-the-loop.
 In order to simulate the IEEE-33 bus network on two-cores of RTDS available in the home in
stitution, i.e., the University of Malta, a holistic reduction of the network was performed while
 maintaining the sanctity of the full network. Therefore the MATLAB based optimization was
 performed on the full network to generate the power references for the BESS, and the same
 references was used by the reduced network in the RTDS. A feedback loop was designed
 to reduce the error between the standard load profiles and the forecasted load profiles. The
 proposed feedback loop helps in mitigating this effect.