Integrid - Bridging the Gap

30 During the demonstration trials, it was observed that on each occasion that voltage constraints and/or overloading occurrences were foreseen, MPOPF could successfully provide a solution to avoid these technical problems. Furthermore, when the focus of the optimization study was set to reduce active power losses, it was possible to achieve effective cost reductions. For this specific KPI, knowledge on active power loss cost reduction is also meaningful in terms of %. Although the optimization process was run for the next 24 hours, flexibility resources were only available during six of these hours, which resulted in a decrease of 6.68%. Regarding the time execution, all the predictivemode studies, which run for a 24 hour period, were computed in less than oneminute, while real-time mode studies, which run for a 1 hour period, took an average of 1.3 seconds. This metric is of utmost importance as it shows that the developed optimization tool not only provides the expected output, but also complies with non-functional requirements, such as computational time. The results of MVLA, which intend to provide a reliable network snapshot toMPOPF, also proved their accuracy. MVLA properly corrected MV/LV substation forecasts in order to make them coherent with the forecasts available for the corresponding upstream feeder. KPI calculation The following metrics were then calculated as a result of the HLUC01 demonstration. With the exception of the KPI associated to the active power losses, the remaining results show an average of all the outcomes gathered during the three demonstration trials. This exception is explained by the fact that active power loss reduction was focused on in only one of these trials. The differences between each of the trials are based on the lessons learned. In other words, the experience gained in the first trials was used to enhance the algorithms and include new functionalities. Table 4.1 – UC1 KPIs and description KPI Result Description Fulfilment of Voltage Limits(%) 100 % of reductions in voltage violations. Focused on the assessment of voltage levels in accordance with Portuguese legislation which follows the standard EN 50160, stating that the voltage must remain in +/- 10% of its nominal value considering an integrating period of 10 minutes. Although not indicated in the name of the KPI, it also accessed the % of reductions in overloading occurrences. Total cost of active power losses reduction (€) 36.18 Comparison between the total cost for the active power losses obtained for the MV network operating scenarios without optimization and the cost of the losses where optimization was considered Total execution time (m) 0.61 (predictive) 0.022 (real-time) Measures the function total execution time for evaluating the scalability of the tool. The execution time is used for comparison between operating scenarios with different dimensions and of variable complexity Mean absolute estimation error (MW/Mvar) 0.053MW 0.072Mvar Indicates the average absolute deviation between estimated and forecast values for different times Maximum absolute estimation error (MW/Mvar) 0.104MW 0.145 Mvar Indicates the maximum absolute deviation between estimated and forecaste values for different times Figure 4.9 – Sequence Diagram for UC2 LV Consumer flexibility is available and is managed by the HEMS acting in smart appliances DSO data (hist.+Rn as input will feed forecast algorithms and technical constraints shown in ADMS Activation platform will consider both DSO and consumer assets to pre-book flexibility and will communicate such flexibility through gm-hub to customers Flexibility will be broken down for customers by real-time Flexibility Plan + Activation solving voltage constraints in the grid