OPENDSS SIMULATIONS ON KAUPINRINNE LV-NETWORK

Transcription

1 - 1 - OPENDSS SIMULATIONS ON KAUPINRINNE LV-NETWORK Table of Contents Introduction... 1 The Model... 2 General description of the modelled network... 2 Model composition... 3 Simulations... 4 Power flow with symmetrical loadings... 4 Power flows with asymmetrical loadings... 6 Power flow with asymmetrical generation... 9 Daily simulations Introduction This document contains demonstration simulations with OpenDSS on a single LV-network model. The goal of these simulations is to show what kind of simulations can be easily run on OpenDSS when the system asymmetry is considered. The network itself is modelled as symmetrical in respect to impedance values of lines and the asymmetry in the system is considered to be due to asymmetry in network loads and generation. The three key points examined here are: 1. Voltage asymmetry on LV network side 2. Overloads due to asymmetric loading currents 3. Asymmetry in power drawn from MV-network The model is composed so that it can be easily attached to the previously created MV-network model if enough LV-network data is supplied to model other LV-networks.

2 - 2 - The Model General description of the modelled network The modelled network is Kaupinrinne LV-network which feeds the residential area of Kaupinrinne near the centre of Orivesi city. The network has 101 customers on 51 connection points, most of them being detached houses. LV network is fed by 500 kva Dyn11 distribution transformer and the network lines are mostly underground cables. The residential area is circled red on the map excerpt on figure 1. The customer connection points are circled green on the circuit plot on figure 2, the blue circle is the distribution transformer. Letter O next to connection point stands for detached house, the letter K stands for apartment house and the letter A stands for auxiliary connection (here this is streetlight connection). Figure 1: Kaupinrinne residential area

3 - 3 - Figure 2: Circuit plot of the LV-network Model composition For these simulations the MV network feeding the system is modelled with 20 kv voltage source and series impedance (Thevenin equivalent). The source feeds LV-network through distribution transformer modelled with OpenDSS s Transformer object. The general composition of the model is depicted in figure 3.

4 - 4 - Figure 3: Model composition On the LV side each customer connection point is modelled with a symmetrical 3-phase load and 3 single phase loads for asymmetrical simulations. Also, every customer connection point has an attached voltage recording element. Simulations Power flow with symmetrical loadings First simulations are done with only symmetrical loads computed from the measured customer loads. Two simulations are run, first with mean loading and secondly with maximum loading. The mean load is computed as average of all measured hourly loads per customer and the maximum loading is computed by selecting the highest hourly load from the measured period. The simulation results are shown on figure 4 and the buses on circuit plot are depicted in figure 5 (bus 0 is the distribution transformer primary).

5 Voltage Profiles - Kaupinrinne LV grid VOLTAGE (MEANLOAD) pu VOLTAGE (MAXLOAD) pu VOLTAGE BUS Figure 4: LV-network voltages with symmetrical loads

6 - 6 - Figure 5: Buses on circuit plot Power flows with asymmetrical loadings In these simulations additional 3.3kW 1-phase load to symmetrical average load is attached to each customer connection point. In first simulation all 1-phase loads are attached to 1 st system phase and in the second simulation the system phase is selected at random. The simulation results for first case are shown on figure 6.

7 - 7 - VOLTAGE P Voltage profile w\1-phase loads on 1st phase Phase1 Phase2 Phase BUS Figure 6: 3-phase voltages of LV-network with 1-phase loads on 1st phase Connecting all the 1-phase loads to same phase causes an overload in the circuit. In table 1 an excerpt of the OpenDSS s overload report is shown. Table 1: Excerpt from OpenDSS's overload report Element Terminal I1 %Normal %Emergency I2 %I2/I1 I0 %I0/I1 Line For these simulations the normal ampacity of a line is determined by maximum current allowed by line fuses. The emergency ampacity is the maximum current of a conductor type. After inspecting the phase currents of the overloaded line (with OpenDSS command: Visualize Currents Line ) it s apparent that the overload is caused by current (252.5 A) in phase 1. The maximum normal current is 250 A. Overloaded line is show on circuit plot on figure 7.

8 - 8 - Figure 7: Overloaded line on circuit plot In the next simulation same loads as in previous simulations are attached to the system. Yet in this simulation the connection phase of an individual asymmetrical load is selected at random. The system voltages after simulation are depicted on figure 8

9 - 9 - VOLTAGE P Voltage profile w\1-phase loads on random phases Phase1 Phase2 Phase BUS Figure 8: 3-phase voltages of LV-network with 1-phase loads on random phase Power flow with asymmetrical generation In this simulation 3.7 kw single phase generation is attached to each connection point with detached house load ( see fig. 2 ). Mean loads are used for symmetrical base load. Each generation is attached to the 1 st system phase. System voltages are depicted on figure 9.

10 VOLTAGE 3P Voltage Profiles w\ single phase generation Phase1 Phase2 Phase BUS Figure 9: 3-phase voltages of LV-network with 1-phase generation on 1st phase Daily simulations In these simulations a 24 h period is chosen during which the power drawn from the MV-network is recorded and examined. First a simulation with just the symmetrical load is run. Time period for the simulation is from :00 to :00. Simulation results are shown on the figure 10.

11 Apparent power drawn from MV-network Power (kva) S1 (kva) S2 (kva) S3 (kva) Hour Figure 10: Power drawn from the MV-network with base loading Next, single phase EV charging loads are attached to the network. In this simulation the charging load is 3.3 kw with charging time between 6-8 h. The charging is determined to begin between 16 pm to 19 pm. So, the load profile for EV charging beginning at 17 pm and lasting for 7 hours would be as depicted on figure 11. When all EV loads are summed the total additional load would be as on figure 12. The charging time, the beginning hour, and the connection phase for each individual load is selected at random (with criterions above). The power drawn from MV-network with these loads is shown on figure 13.

12 Figure 11: An example of an EV charging load profile

13 Figure 2: Sum of EV charging loads Apparent power drawn from MV-network Power (kva) S1 (kva) S2 (kva) S3 (kva) Hour Figure 3: Power drawn from the MV-network with base loading and additional EV loads In the last simulation a 1-phase solar generation is attached to each detached house. The size of the generation unit is same for each connection point and the connection phase is selected at random. The simulation period is changed to :00 to :00. On figure 14 the power drawn without 1-phase generation is depicted.

14 Apparent power drawn from MV-network Power (kva) S1 (kva) S2 (kva) S3 (kva) Hour Figure 4: Power drawn from the MV-network with base loading Every generation unit follows the generation curve depicted on figure 15 during the simulation. The power drawn from MV-network is shown on figure 16. Generation profile Power (kw) Hour P_gen Figure 5: the generation profile of a single generation unit

15 Apparent power drawn from MV-network Power (kva) S1 (kva) S2 (kva) S3 (kva) Hour Figure 6: Power drawn from the MV-network with base loading with additional generation