Real-time adaptive optimization engine algorithm for integrated Volt/VAr optimization and conservation voltage reduction of smart microgrids

Farhangi, Hassan (Hassan Farhangi (hassan_farhangi)) (author)
Manbachi, Moein (author)
Palizban, Ali (author)
Arzanpour, Siamak (author)
(editor)
(translator)

© 2012, M. Manbachi, H. Farhangi, A. Palizban, S. Arzanpour.

Proceedings from CIGRÉ Canada Conference, Montreal, Sept. 2012. In recent decade, smart microgrids have raised the feasibility and affordability of adaptive and real-time Volt/VAr optimization (VVO) and Conservation Voltage Reduction (CVR) implementations by their exclusive features such as using smart metering technologies and various types of dispersed generations. Smart distribution networks are presently capable of achieving higher degrees of efficiency and reliability through employing a new integrated Volt/VAr optimization system. For VVO application, two well-known approaches are recommended by different utilities and/or companies: Centralized VVO and Decentralized VVO. In centralized VVO, the processing system is placed in a central controller unit such as DMS in the so called “Utility Back Office”. The DMS uses relevant measurements taken from termination points (i.e. utility subscribers) supplied to it from either field collectors or directly from MDMS, to determine the best possible settings for field-bound VVO/CVR assets to achieve the desired optimization and conservation targets. These settings are then off-loaded to such assets through existing downstream pipes, such as SCADA network In contrast, decentralized VVO utilizes VVO/CVR engines which are located in the field and in close-proximity to the relevant assets to conserve voltage and energy according to local attributes of the distribution network. In this case, local measurements do not need to travel from the field to the back-office, and the new settings for VVO/CVR assets are determined locally, rather than from a centralized controller. Without having any preference between above mentioned VVO techniques, this paper studies an adaptive optimization engine for real-time VVO/CVR in smart microgrids based on Intelligent Agent technology. The optimization algorithm provides the best optimal solution for VVO/CVR problem at each real-time stage through minimizing system loss cost and improves system energy efficiency as well as voltage profile of the relevant distribution system. The algorithm may employ distributed generation sources to address the Volt/VAr optimization problem in real-time. Coordinated VVO/CVR requires real-time data analysis every 15 minutes. It utilizes a distributed command and control architecture to supply the VVO Engine (VVOE) with the required data, and secures real-time configuration from the VVO engine for the VVO control devices such as On-Load Tap Changers (OLTCs), Voltage Regulators (VRs) and Capacitor Banks (CBs). It also has the option of employing distributed generation (DG) as well as modelling load effects in VVO/CVR application. The algorithm minimizes the distribution network power loss cost at each time stage, checks the voltage deviation of distribution buses and distributed generation sources considering different types of constraints such as system power flow, distribution network power factor, system active and reactive power constraints and switching limitations of Volt/VAr control devices. The algorithm receives required real-time data from an intelligent agent. Then, it starts to solve the real-time VVO/CVR problem in order to find the best optimal configuration of the network in real-time. The paper uses British Columbia Institute of Technology (BCIT) distribution network as its case study in order to explore the effectiveness and the accuracy of the optimization engine. Moreover, the VVO/CVR optimization algorithm is implemented in different configurations; a) VVO/CVR confined to the substation and b) VVO/CVR optimization algorithm within the substation and along distribution feeders. The algorithm also checks the availability of DGs to assist VVO/CVR control functions and assesses the impact of new distributed sources such as: Flywheel Energy Storage System (FESS) on real-time VVO/CVR. For this reason, the algorithm classified DGs in a microgrid based on their impacts and instantiates them based on their application feasibility for real-time VVO/CVR.

https://www.researchgate.net/publication/256374804_Real-Time_Adaptive_Op...