A Large Scale Grid Data Analysis Platform for DSOs

Abstract

Abstract:The number of fluctuating distributed energy resources (DER) in electricity grids is continuously rising. Due to the lack of operational information on low-voltage (LV) networks, conservative assumptions are necessary to assess the connection of generators to the grid. This paper introduces the hosting capability (HC) as a measure to assess the amount of DER that can be integrated in LV-feeders. The HC of a feeder is the minimum amount of DER that can be hosted in a feeder without reinforcement needs for a given DER-scenario and for a given admissible voltage rise. The hosting capability assessment was performed on the entire LV-grid data of two Austrian Distribution System Operators (DSOs) with more than 36,000 LV-feeders. In total, 40 HC-scenarios were calculated with varying admissible voltage rise levels, DER-scenarios and reactive power control strategies. It turned out that only few feeder parameters such as the resistance at the end node and the lowest ampacity value of feeders show a high correlation with the calculated HC. Further, the impact of the DER-scenario on the share of voltage and loading constrained feeders is rather limited. The gathered results are suitable to validate equivalent LV-feeders models to perform integrated power flow studies on the transmission and distribution grids. Besides the results obtained for the network data of the two DSOs, a performant, modular and parallelizable tool has been developed to automatically analyze large LV network sets.

Data Analytics-Based Multi-Objective Particle Swarm Optimization for Determination of Congestion Thresholds in LV Networks

“…Distribution networks have their end users spread, distributed in any form, independently if regarding a LV feeder and its phases, or among different feeders connected to the same power transformer in a secondary substation [4,34,35]. Moreover, each end user, whether consumer, prosumer, or generator, has its own arbitrary pattern, even more diverse with the proliferation of DER technologies.…”

“…Conventional development of the distribution networks has been driven by a worst-case fit-and-forget principle to tackle most of medium-voltage (MV) and low-voltage (LV) congestion challenges due to the long lifetime of their assets and planning horizon [1][2][3][4]. This led to limited distributed energy resources (DER) capacity allowed to be connected and costly network reinforcements, although peak loads generally occur only for a few hours in a year [1].…”

On the Classification of Low Voltage Feeders for Network Planning and Hosting Capacity Studies

“…The calculation of these hosting capacity-related parameters is more complex than the first type of parameters. For this, an adapted Newton-Raphson algorithm has been implemented in the simulation environment [26]. This tool increases the generation connected to the considered feeder until one of the planning limits (voltage or current constraint) is reached.…”

“…The scenario "uniform" assumes a uniform distribution of the generation long the feeder, the scenario "weighted" assumes that the size of the generator is proportional to the annual electricity consumption of the LV-customers (motivated by e.g., self-consumption) and the scenario "end of feeder" assumes that the generation is located at the end of the feeder, which represents a worst-case. For all these scenarios, load an generation are assumed to be balanced (equally distributed over the three phases) Further scenarios have been defined and analyzed in [23]-a complete overview of these scenarios is provided in [26].…”

“…In this second step, all the feeder parameters which have been previously defined (see Section 2.1) are computed through automated scripts [26]. These variables can be divided into two families of variables:…”