The computational burden and the time required to train a deep reinforcement learning (DRL) can be appreciable, especially for the particular case of a DRL control used for frequency control of multi-electrical energy storage (MEESS). This paper presents an assessment of four training configurations of the actor and critic network to determine the configuration training that produces the lower computational time, considering the specific case of frequency control of MEESS. The training configuration cases are defined considering two processing units: CPU and GPU and are evaluated considering serial and parallel computing using MATLAB® 2020b Parallel Computing Toolbox. The agent used for this assessment is the Deep Deterministic Policy Gradient (DDPG) agent. The environment represents the dynamic model to provide enhanced frequency response to the power system by controlling the state of charge of energy storage systems. Simulation results demonstrated that the best configuration to reduce the computational time is training both actor and critic network on CPU using parallel computing.@en

This paper presents a descriptive statistical analysis (DSA) of time-series of electro-mechanical quantities related to the frequency control (e.g. kinetic energy (KE), electrical frequency and power demand) of the Nordic power system (NPS). The idea of the DSA is to identify main observables features and patterns between these variables. Historical data publicly available has been used in this research paper; pre-processing included evaluating and identify missing data, and it filled by using the linear interpolation. The DSA uses descriptive statistical indicators to obtaining observable features. The dispersion analysis is used to observes how affects the KE to the electrical frequency. The data is grouped by weeks, days and hours, and its correlation coefficient was calculated. A correlation analysis between the KE and the power demand was computed, and the linear regression was used to construct a prediction model.@en

This paper presents a dynamic data-driven state of charge (SoC) management strategy for a battery energy storage system (BESS) providing fast frequency response (FFR) services. The SoC limits are continuously updated based on an expected frequency profile which is derived from a statistical analysis of real frequency data for three different power systems, namely the Continental European (CE), Great Britain (GB) and Dominican Republic (DR) power systems, spanning the period 2014-2017. Simulations presented indicate that the control scheme developed can successfully maintain the SoC of the BESS inside the desired limits while crucially providing FFR services to the grid.

@en

Looking to the future, there are several challenges that electricity networks will face: prosperity, global sustainable growth and security. The electricity industry situation is complex because resources worldwide are becoming scarce and the need for sustainable growth is important. Dealing with this challenge led to the foundation of micro-grids. The classical 'micro-grid' defines a mix of distributed energy resources (DER) capable of providing sufficient and continuous energy to a significant portion of the internal demand. Transmission system friendly micro-grids is a novel concept to encourage the power electronic converters to provide grid services and functionalities in line with new standards such as IEEE 1547- 2018. This paper presents a method for reducing losses by injecting reactive power from the grid-connected DER. The methodology is developed in Python and the system is modelled in DIgSILENT PowerFactory. The paper includes numerical results of the proposed methodology to show the suitability of the proposed concept.

@en

The increased adoption of renewable energy generation is reducing the inertial response of the Great Britain (GB) power system, which translates into larger frequency variations in both transient and pseudo-steady-state operation. To help mitigate this, National Grid, the transmission system operator in GB, has designed a control scheme called enhanced frequency response (EFR) specifically aimed at energy storage systems (ESSs). This study proposes a control system that enables the provision of EFR services from a multi-electrical ESS and at the same time allows the management of the state of charge (SOC) of each ESS. The proposed control system uses a Fuzzy Logic Controller to maintain the SOC as near as possible to the desired SOC of each ESS while providing EFR. The performance of the proposed controller is validated in transient and steadystate domains. Simulation results highlight the benefits of managing the SOC of the energy storage assets with the proposed controller. These benefits include a reduced rate of change of frequency and frequency nadir following a loss of generation as well as an increase in the service performance measure which renders into increased economic benefits for the service provider.

@en

The growth of vehicle electrification is driven by the desire to reduce environmental pollution, and it is fueled by advancements in battery technology. If left unmanaged, electric vehicle (EV) charging will increase peak demand and put a strain on the electricity networks. However, if properly managed, EVs can provide useful services to the power system operator such as fast active-power injection which serves to improve the system frequency response (SFR) after a disturbance. The objective of this paper is to assess the impact that clusters of EVs, connected to frequency-responsive charging stations, have on the provision of SFR after a loss of generation event. The assessment considers EV charging demand in Great Britain (GB) for the year 2025 considering three different daily charging patterns. A generic model for the EV clusters is developed which includes the effects of measurement delays and control charger time response. The model and scenarios are integrated into a single-Area model representative of the GB power system and the minimum expected values for the system's inertia in the year 2025 are used. The results obtained highlight the benefits on the SFR of utilizing EVs as a dynamic energy storage system for different types of charging and the impact of the measurement delay on the dynamics of the response.

@en