As the penetration of renewable sources in the distribution networks increases, distribution network operators are concerned with possible voltage fluctuation problems in the conventional tree-configured distribution network. The voltage fluctuation problem can be managed at the nodes by reactive power control devices like SVC or DSTATCOM [1-3]. BTB inverter type controller was introduced to regulate voltage by controlling power flow . Energy storage system is an effective way to prevent reverse power flow. But, these solutions are too costly to be applied to the wide range of distribution networks.
Here, we introduce the define feeder loop line as the line with switching devices placed between feeders. The distribution networks can have loop configuration by switching on the loop lines.
The bus voltage at a distribution network can be controlled by controlling network configuration, i.e., by controlling on or off status of the loop lines. The distribution network can be resumed to the original tree configuration when loop line is switched off. The DGs are assumed to be connected to the distribution network of which the voltage level is 22.9 kV or less.
2. The Reverse Power Flow Problem in Distribution Network
The reverse power caused by the DG installed at the distribution network flows into the feeder as shown in Fig.1. The voltage at the node to which DGs are connected increases as the reverse power flow increases.
Fig. 1.The flow of the reverse power flow generated by the DG
Thus the node voltage may excurse over the allowed operating voltage ranges. The operating voltage ranges and voltage profile with reverse power flow from dispersed DGs are shown in Fig. 2.
Fig. 2.Overvoltage by reverse power flow
3. Feeder Loop Line Control
The feeder loop line control is defined as a method to control the status of the feeder loop line by switching on or off the loop line placed between the feeders. The surplus power of feeder shifted to the other feeder through the loop line. As a result, the overvoltage at the node is dropped to the normal voltage range.
3.1 The advantage of the feeder loop line
Fig. 3 shows the power loss depending on the DG output power level with the feeder loop line switched on and off, respectively.
Fig. 3.The power loss index with the feeder loop line switched on and off
Though appropriate DG in the distribution network reduces the power loss, an excessive DG output increases the power loss by the reverse power flow. The looped distribution network with the feeder loop line switched on can reduce the power loss as shown in Fig. 3. Fig. 4 shows the voltage evaluation index [A] defined in Eq. (2). The values of the voltage evaluation index depend on the DG’s output power levels and distribution network configurations. The node overvoltage can be prevented by the looped distribution network.
Fig. 4.The voltage variation index with the feeder loop line switched on and off
3.2 The place of the loop line at the distribution network
The loop lines are placed between two feeders with the same nominal voltages, i.e. the same region. The node at feeder 1 in region i is connected to the node at feeder 2 in region i through the feeder loop line. This strategy allows the region with the feeder loop line to have the same nominal voltage level. The loop lines are not placed to connect two nodes in different regions as shown in Fig. 5.
Fig. 5.Loop lines installed between the nodes with the same nominal voltages
4. Feeder Loop Line Status Control
The feeder loop lines are placed between the feeders to add power flow paths to control the voltages at the nodes to which DGs are connected. However, the feeder loop lines are not always needed to be switched on depending on the distribution network conditions. The control variables are the status of the feeder loop lines, on and/or off. If n loop lines are placed between two feeders, 2n loop line status can be constituted. The most reasonable configuration of the networks is selected considering the power loss and the node voltage.
For the simplicity of this study, two-feeder system was chosen as in Fig. 6.
Fig. 6.Node range allocated to loop lines
The system can be divided by regions with different nominal voltage and voltage ranges, Vi,nom ± ΔVi,nom . Where, is the nominal voltage at region i , and is voltage deviation allowed at region i.
We assume that one loop line has been prepared at each region for looping the feeders and the region voltage is represented by the node voltage.
4.1 The control objective for the loop line control
4.1.1 The voltage evaluation index
The voltage evaluation index [A] is defined by Eq. (1).
where, sk : the status of kth loop line, 1 or 0
: the voltage of region i, measured at each feeder
The voltage evaluation index is zero if the node voltage is equal to the nominal voltage and higher than 1 if the node voltage is out of the allowed operating range.
4.1.2 The power loss index
The power loss at a line of impedance Z between two nodes is given by Eq. (4).
Then, the power loss at a network is approximately given by Eq. (4).
4.1.3 The cost function
The feeder loop line control is to determine the status of sk so that node voltages and power losses are to be minimized.
The cost function for the control is given by Eq. (6).
The optimal distribution network is one of the 2K possible configurations of the distribution networks that can be constituted by K loop lines. The simplest way to calculate the cost function is to solve power flows2K times.
6. Simulation Results
The simulation study has been carried out by using PSCAD/EMTDC software.
We assumed that the power loss at loop lines, Pk,ll, are all the same by 5 kw and that the nominal voltage of center of the feeder is 22.9kV. The operating voltage range is +/- 2% of the nominal voltage. The parameters of line constants are shown in Table 1.
Table 1.Parameters of loop line
The loads are assumed to be distributed equally to each node throughout the feeder. In the simulation, two load cases were considered, one is heavy load case with 10 MVA with power factor of 0.9 and the other is light load case with 7 MVA with power factor of 0.9.
Fig. 7.Voltage ranges of a feeder
As shown in Fig. 8, the voltage profiles of a feeder with heavy load is lower than those of a feeder with light load.
Fig. 8.Voltage profiles before DG applied.
5.1 Case studies
The simulation studies for two cases were conducted. One is the case when the network consists of three heavy loaded feeders as in Fig. 9, the other is the case when the network consists of two heavy loaded feeders and one light loaded feeder as in Fig. 10.
Fig. 9.Configuration of case study 1
Fig. 10.Configuration of case study 2
A generation source is placed in the middle of a feeder in any cases. The influence of various output level to voltage profiles was investigated. The output levels we considered are 0%, 7%, 14%, 20%, 30% of the total load.
The weighting factor α and β were given 0.5 respectively.
The simulation results were summarized in Table 2. ‘o’ represents that the loop line is on status, on the other hand ‘x’ represents the off status.
Table 2.Result of simulation
When the loop line control is not applied, the number of nodes with overvoltage in case 2 are larger than that of case 1. That is why reverse power under light load is greater than the heavy load case. The voltage profiles of case 2 are shown in Figs. 12 and Fig. 13. Overvoltage over the feeder caused by reverse power flow from distributed generator has been suppressed into the specified voltage range by the feeder loop line control.
Fig. 11Simulation results (case 2)
Fig. 12.Voltage profiles of case2 without control
Fig. 13.Voltage profiles of case 2 with control
When distributed generators are included in the distribution networks, the feeder voltages tends to rise over the specified range proportionally to the output levels of the generators. The loop line control can be a solution to the overvoltage problem and reduce power losses of distribution networks. Further studies are needed to develop optimization methods to get the solutions considering taps of transformations.