Integrid - Bridging the Gap

31 Table 4.2 - UC2 KPIs KPI Result Description KPI 1.1: Voltage Violation Frequency Reduction (VFR) 92,31% Number of voltage violations registered in the baseline scenario (n_baseline) with the number of voltage violations registered with the LVC tool application (n_LVC). KPI 1.2: Voltage Violation Magnitude Reduction (VMR) 98,12% Magnitude of voltage violations registered in the baseline scenario (d_baseline) with the magnitude of voltage violations registered with the LVC tool application (d_LVC). KPI 2: Flexibility Activation 28,92% Percentage of the total flexibility available from private consumers that is used to manage operation without technical constraints. KPI 3: Voltage constraint management success rate 92,31% Percentage of the total voltage violations managed by the Low Voltage Control tool recurring to the available flexibilities in the LV network. KPI 4: Operator awareness 100% This is a qualitative indicator on a scale of 1 – 100 for the operator awareness with respect to the state of the system. An automatic approach to the activation was followed, enabling a 100% result for the KPI. KPI 5.1: Mean absolute error of voltage magnitudes (MAE_D) 0,63% Dispersion of the algorithm to perform state estimations. Performance is evaluated for voltage magnitude. KPI 5.2: Mean absolute error of Active power (MAE_P) 0,39kW Dispersion of the algorithm to perform state estimations. Performance is evaluated for active power. KPI 6.1: Maximum absolute deviation of voltage magnitudes and active power (MAD_V) 2,82% Determines the maximum dispersion of the algorithm to perform state estimations. Performance is evaluated for voltage magnitude. KPI 6.2: Maximum absolute deviation of active power (MAD_P) 6,67kW Determines the maximum dispersion of the algorithm to perform state estimations. Performance is evaluated for active power magnitude. Results from HLUC02 The sequence diagram shown in Figure 4.9 explains how the flexibility made available by customers through the InteGrid app and gm-hub (step 1) is considered by DSO tools, when computing the necessary flexibility to allow more efficient grid operation (step2). In a third step, the flexibility activation is then taken by the DSO and, in a fourth step, it is sent to customers through the InteGrid ICT infrastructure. KPI calculation Table 4.2 below outlines the KPI results. A brief description of each KPI is also provided. Several trials were performed in this second test wave at the LV demonstration sites of Valverde and Alcochete. Despite the tighter voltage violation gaps considered (undervoltage U ≤ 98% U_N and overvoltage U ≥ 105% U_N), it was proven that low voltage flexibility provided by domestic customers is the best alternative to actively manage LV networks. These types of resources are very effective at solving voltage violations (due to their distributed nature), while having limited impact on the remaining grid. The low amount of flexibility activated (from the available pool) is also a good indicator, demonstrating that the DSO would have more alternatives to solve severe grid problems. Three main aspects are critical for the success of a scalable solution: the first is the integration of all the architecture; the second is active customer participation; and the third, and most important for this UC, is the quality of the algorithms that generate the forecasts and the flexibility activation setpoints. Results from HLUC03 – Predictive Maintenance With regard to the predictive maintenance results in the Portuguese demonstrator, there is some evidence which indicates that dry transformers (MV/LV), which are subjected to greater load, fail sooner. To reach this conclusion, the HLUC03 team analysed 25 dry transformers and the corresponding sensor data and historical failure records. The former refers to transformers with failures, whereas the latter relates to the measured load from 2015-2019. Moreover, sensor data provides the load that each dry transformer was subject to in 15 second intervals during this period.