Mikel Armendariz

Multiagent-Based Distribution Automation Solution for Self-Healing Grids

A multiagent-based distribution automation solution is proposed to be used in the distribution of self-healing grids to solve the service restoration part of the Fault Location, Isolation and Service Restoration (FLISR) task. The solution reduces the grid topology to an undirected weighted graph and executes a distributed implementation of Prims minimum spanning tree algorithm to solve the problem. The solution is compliant with state-of-the-art standards within smart grids, including but not limited to IEC61850. To test the performance of the algorithm, a testbed is assembled consisting of a physical dc grid model and several Arduino microcontrollers and Raspberry Pi computers. The test results show that the proposed algorithm can handle complex FLISR scenarios.

A co-simulation platform for medium/low voltage monitoring and control applications

The increasing demand for electricity and the penetration of renewable energy resources into the electrical grid are triggering the electric power system to undergo changes and complex modifications at high, medium and low voltage levels. There are large and complex projects like Smart Grid Gotland that are developing solutions to deal with these challenges. However, there are several scenarios that could be more practical to be first studied at the simulation level. Therefore, the purpose of this study is to propose and develop a co-simulation platform that allows performing real-time monitoring and control tests and simulations for MV/LV grids. This platform combines both electrical and information and communication technology (ICT) infrastructure simulations. It consists of a real-time power system simulator (Opal-RT) connected to a communication network emulator (OPNET) through an automatic meter reading interface that is based on commercial off-the-shelf (COTS) and low cost devices like Raspberry Pi and Arduino boards. The platform allows the analysis of realistic MV/LV grids and communication networks and will facilitate the simulation and comparison of low voltage control strategies for real-time test scenarios that could be more complicated to test in the reality due to cost and complexity aspects (i.e. high power losses on the low voltage grid).

A method to identify exposed nodes in low voltage distribution grids with High PV penetration

The impact of introducing distributed energy resources at the low voltage side of the distribution grid is currently raising new challenges for utilities. In particular, the high penetration of photovoltaic panels (PVs) in radial grids is increasing the active power losses in the branches and the voltage level at some of the nodes. Principally nodes next to PV array installations. This paper presents a methodology based on design of experiments (DOE) to detect such exposed nodes and branches, together with the identification of the main scenarios that cause such problems, characterized by: season, type of day, solar radiation and outdoor temperature levels. The methodology is simulated on a LV network based on the Cigre benchmark Grid with real utility data. The exposed nodes are classified for each feeder from most to least problematic and showed (as expected) sensitivity to seasonality (summertime), characterized by high solar radiation and outdoor temperatures.

A method to place meters in active low voltage distribution networks using BPSO algorithm

This paper proposes a method to be used by a Distribution System Operator (DSO) to optimally place sensors at medium voltage/ low voltage (MV/LV) substation and some low voltage cable distribution cabinets. This method aims to improve the estimation of the grid states at low voltage distribution networks. This method formulates a multi-objective optimization problem to determine the optimal meter placement configuration. This formulation minimizes the low voltage state estimation error and the cost associated to a particular meter deployment configuration. The method uses Binary Particle Swarm Optimization (BPSO) to solve the optimization problem and it has been tested on a network based on the Cigré LV benchmark grid. The simulation results show that the method can be applied to both situations where smart meter measurements are available and situations where they are not. In the latter situation the measurements are replaced by pseudo-measurements, which represent meter readings by using smart meter historical data and prediction models.

Two-stage network processor for an independent HVDC grid supervisory control

This paper proposes a two-stage network processor for the supervisory control of multi-terminal HVDC grid that connects different AC areas. The proposed network processor processes the DC substation topology locally in the first stage and analyzes the HVDC grid connectivity at the supervisory controller. The processor uses the k-means clustering method to detect the islands in HVDC grid and prepares required information to carry out the converter control mode assignation. The performance of the method has been tested for an islading scenario in a 7-terminal HVDC grid using a real-time co-simulation platform. The output of network processor helps the control application at the supervisory level to make decisions based on the control modes of the available converters in the islands. This two-stage architecture has been compared with the centralized architecture in terms of computational complexity.

Facilitating distribution grid network simulation through automated common information model data conversion

In R&D projects as well as in development work, simulations of grid configurations and control methods are a common tool for assessing different technical options. At the same time, these simulations are often limited by difficulties in data exchange, simply due to the fact that systems are not compatible and data format in operational systems is not aligned with data in simulation and analysis tools. To this extent a tool has been developed that allows large scale import of real-world grid data using CIM XML to state of the art simulation tools based on MATLAB. The tool is used in a large FP7 project involving leading European DSOs.