Application of GridEye for Grid Analytics

Transcription

1 Application of GridEye for Grid Analytics This document provides a use case for the application of GridEye for the monitoring of low voltage grids. GridEye modules primarily measure the electrical quantities and process the measurements using the distributed intelligence on every module. Thus, only useful data for the system operators and/or endcustomers are communicated and stored. In this way, the high communication costs and big data issues are avoided. GridEye monitoring is an efficient and scalable approach for providing observability in low voltage grids and ensuring its secure operation. This functionality of GridEye is not only useful for the grids with PV production, but it is also a valuable tool for understanding the current and voltage levels, and balanced operation of three phases systems. Various analysis using the temporal profiles and statistics are accessible through the GridEye interfaces. The transmitted data is used for monitoring and visualization purposes through a dedicated interface. Author: Omid Alizadeh-Mousavi Date: March 8th, 2017 DEPsys SA 1

2 1. INTRODUCTION OF THE USE CASE AND GRIDEYE INSTALLATION The grid topology of the use case and the installations of GridEye modules are shown in Figure 1. The electricity consumptions in the grid are mainly residential and agricultural. Moreover, there are PV installations at three locations shown by G1, G2 and G3. In this use case, the GridEye modules are used for the monitoring of electric quantities. It is worth noting that the monitoring request can be from system operators and/or from end-customers. In this use case, the local distribution system operator is interested in monitoring and optimal control of the shown grid as well as the monitoring and optimization of the energy consumptions and productions of a private end-customer. For this private customer in node 104, shown within the dashed line, an additional GridEye module is installed. The installations of GridEye modules at the MV/LV transformer as well as on distribution cabinets are demonstrated in Figure ANALYSIS OF MONITORING DATA The GridEye measurements are used to analyze the voltages, currents, power flows, and loadings in the transformer and the low voltage grid. The temporal profiles and statistics of the measurements are used for the detailed analysis of the devices Transformer daily loading The profiles and distributions of the transformer loading for a winter and a summer day are shown in Figure 3. The distribution of transformer loading Figure 3-a and Figure 3-b) illustrate the number of hours that the transformer is operated at each loading level. The temporal profile of the active, reactive, and apparent powers of the transformer are illustrated in Figure 3-c and Figure 3-d). During the winter day, the transformer is loaded between 30% and 110% of its nominal capacity (i.e. 250 kva). The observed overloading is for a short period of time and it is due to consumption peak around 19:00 as illustrated in Figure 3-c. Therefore, the transformer overloading can be an issue for the system operator, specifically if additional consumption is planned to be added to the transformer. Figure 1. The use case grid and the installations of GridEye modules. During the summer day, the transformer is loaded between 0% and 80% of its nominal capacity which is lower than the winter loading. The highest loading level in the summer day is due to PV production as a) Distribution cabinet c) MV/LV transformer b) Rogowski coils for the current measurements at transformer LV d) PV power plant Figure 2. Installation of GridEye modules 2

3 Winter day Summer day a) distribution of transformer loading b) distribution of transformer loading c) temporal profile of transformer powers d) temporal profile of transformer powers Figure 3. Distribution and temporal profile of transformer loading for a winter day (a and c) and a summer day (b and d). Sunny day Cloudy day a) temporal profile of voltage temporal profile of voltage b) c) temporal profile of current d) temporal profile of current Figure 4. Temporal profile of transformer s voltage and current for a sunny day (a and c) and a cloudy day (b and d). 3

4 indicated in Figure 3-d. Therefore, the production of the existing PV installations does not cause transformer overloading. However, the impact of additional PV installations on the transformer loading should be taken into consideration Transformer daily voltage and current variations The temporal profile of the transformer s voltage and current for a sunny and a cloudy day are shown in Figure 4 and the dynamics of voltage and current variations are analyzed. The period with negative currents, indicated in Figure 4-c and Figure 4-d, corresponds to the instances where production is higher than consumption. It is worth noting that for these two days, the maximum current production does not deviate from the nominal current of the transformer (i.e. 362 A). During the days with PV production, in general, the voltage profile follows the current profile. In other words, the increase of PV power injection (i.e. larger negative current) results in the increase of the voltage level. During the sunny day, the highest voltage level (i.e. 246V) is observed around 13:30 that corresponds to the instance with the highest PV production. Furthermore, the observed voltage steps at 6:00 and 9:00, specified in Figure 4-a, are due to the change of a tap-changer in MV grid, since there are no current variations at these instances. During the cloudy day, the large voltage variations are observed because of the large variations of the PVs productions. The current variation during the day reaches 470 A (from 130 A consumption to 340 A production) and the largest current step is approximately 300 A within a 2- minutes period. Similarly, the voltage variation during the day reaches 10 V and the largest voltage step is 4 V over 2- minutes. Here, the utilized measurement data has 2- minutes time interval. In this case, measurements with lower time intervals are needed to capture the voltage variations using shorter time steps. It also emphasizes the importance of power quality measurements October 2016) is shown in Figure 5, Figure 6, and Figure 7, respectively. In these figures, every data point represents the daily average of the electrical quantities. The transformer s loading, shown in Figure 5, is lower during the summer months and the higher loading levels occur during the winter months. The higher PV productions during the summer months results in the negative power flows from LV to MV grid for almost 4-5 months. The current profiles, shown in Figure 6, follow the active power profile and have negative values during the summer months. It is worth noting that the neutral current does not vary in different months that the level of balanced operation of the grid does not change over the seasons. The voltage profiles, shown in Figure 7, follow the PV production and have higher values during the summer months. Figure 5. Daily average profile of transformer active, reactive, and apparent powers for 18 months. Figure 6. Daily average profile of transformer currents for 18 months Yearly statistics Transformer power, current, and voltage average daily dynamics The temporal profile of transformer s powers, currents, and voltages for 18 months of measurements (April Figure 7. Daily average profile of transformer voltages for 18 months. 4

5 Transformer current and voltage weekly dynamics The temporal profile of transformer s currents and voltages for 18 months of measurements (April 2015 October 2016) is shown in Figure 8 and Figure 9, respectively. In these figures, average, minimum, and maximum values of the currents and voltages for every week is represented by a data point. In Figure 8, it is observed that the current variation is higher during the summer weeks due to higher PV production. In the consumption side, the maximum current exceeds the nominal current in a winter week. In the production side, the maximum current reaches the values close to the nominal current. It is also observed that the neutral current and the imbalance between three phases is low. In Figure 9, it is observed that the voltages in winter are lower than in summer because of the relatively weak transformer (i.e. 250 kva) Distribution of transformer loading The distribution of the transformer apparent power for one year is shown in Figure 10-top. It demonstrates the number of hours that the transformer is operated at every loading level. Although the transformer is mostly operated in average loading levels, the loadings between 90% and 110% are observed for almost 5 hours during the year. Therefore, the risk of transformer overloading should be considered in its operation and specifically in planning studies. The distribution of the transformer active power for the same period of time is shown in Figure 10-middle. The active power distribution contains the direction of power flow in addition to the loading level. It is observed that the transformer is operated more on the consumption side (i.e. positive power flow) and that a 100% loading level is reached for less than an hour. Thus, the impact of the increase of the LV grid consumption on the transformer loading level should be considered. Although the transformer is operated for less number of hours on the production side (i.e. negative power flow), it is interesting to mention that 90% loading on the production side is experienced for almost two hours. This observation emphasizes that additional PV installations in the LV grid can result in overloading on the production side. The distribution of the transformer reactive power for the same period of time is shown in Figure 10-bottom. It is observed that the transformer s reactive power is less than 20% of its nominal capacity for more than 98% of the time period. The low reactive power loading of the transformer indicates that in the LV grid the active power losses is low and the power factor is high. This information is important for the system operator and the end-customer specifically if the endconsumers are charged for the reactive power. Moreover, the transformer is operated for 99% and 1% of the time in the inductive and capacitive loadings that respectively correspond to the positive and negative reactive power flows. In the inductive side, the transformer is loaded between 20% and 50% of the nominal capacity for 140 hours. Figure 8. Weekly mean, minimum, and maximum profiles of transformer currents for 18 months. This observation justifies the need for the optimal voltage control in this use case. From a planning point of view, further PV installations at these nodes, without voltage control, will definitely result in overvoltage issues. Figure 9. Weekly mean, minimum, and maximumprofiles of transformer voltages for 18 months. 5

6 Figure 10. Distribution of transformer loading for one year, apparent, active, and reactive powers Distribution of voltages and currents throughout the grid The distribution of the three phase voltages at the transformer and several grid nodes are shown in Figure 11. The voltage distributions demonstrate the number of hours that each node is operated at every voltage level. The first observation is that the voltages across all nodes are within the acceptable limits (230 ± 10%) during the whole year. Nevertheless, the voltages at node 101 and 102 reach to the maximum allowed voltage level (i.e. 253V) which is due to the PV installations at these two nodes. The high voltage levels at these nodes require paying particular attention to ensuring the secure operation of the grid. Figure 11.Three phase voltage distributions at several nodes of the gridover one year. 6

7 loading reaches to 20%. The cable is mostly used on the consumption side. There is small PV production at node 102 that results in a small negative loading. Although from the loading perspective, this cable has a large margin and additional PV installations can be considered at node 102, the additional PV installations can cause overvoltage issues. Therefore, a holistic planning analysis should simultaneously consider both the voltage and current aspects. Figure 12.Three phase and neutral current distributions at the transformer and several cables of the grid over one year. The distribution of three phases and neutral currents for the transformer and several cables over one year are shown in Figure 12. The current distributions show the number of hours that the transformer or the cables are operated at every loading level while the direction of current flow is preserved (i.e. positive for consumption and negative for production). The loading level of the transformer and the cables are determined by dividing the value of currents by their nominal current values. The transformer is mostly used on the consumption side and the maximum consumption current goes above 100% of its nominal current. Although the transformer is operated for less period on the production side, its maximum current on the production side reaches up to 90% because of PV productions. The overloading of the transformer should be studied in the case of additional consumption and PV installations. The cable is often used on the consumption side. However, the highest loading level (i.e. -52%) is reached on the production side which is because of the PV installations at node 101 and 102. The cable is mostly used on the consumption side. The maximum loading levels on the production and consumption sides are observed at -25% and 48%, respectively. The production part is due to the PV installations at node 104 and the consumption part includes the loads at node 104, 105, and 106. The cable and the cable are only operated on the consumption side because there are no PV productions at nodes 103 and 106. These two cables are only slightly loaded as their maximum Furthermore, the voltage and current distributions provide important information about the balanced operation of the three phases. This information can be used for improving the operation and planning of the distribution grids. In the case of unbalanced loading of three phases, the distribution of the voltages and currents of the phases are shifted to left/right with respect to each other and the distribution of the neutral current shows higher values. In this use case, it is observed that the voltages and currents of three phases have similar distributions and the neutral currents are not significant. These observations denote a balanced operation of the three phases. For more information please contact: DEPsys SA Route du Verney 20B 1070 Puidoux, Switzerland Phone : Omid Alizadeh-Mousavi R&D Director omid.mousavi@depsys.ch Antony Pinto Electrical Engineer antony.pinto@depsys.ch Joël Jaton Chief Technology Officer joel.jaton@depsys.ch 7