Feeder J1


Fig. 1. Map of Feeder J1

Feeder J1, located in the northeastern US, was selected for analysis because 1.7 MW of customer-owned PV systems already exists on this feeder and customers have complained of overvoltage that is suspected to be caused by PV. The map of this feeder (Feeder J1) is shown in Fig. 1 with the feeder conductors drawn as purple lines (3-phase solid lines, single-phase dotted lines).

With a peak load of approximately 6 MW, this 12-kV feeder serves approximately 1300 residential, commercial, and light industrial customers via 58 total miles of primary line. The substation is located at the southern-most edge of the feeder, as shown in Fig. 1 (green triangle), while the 1.7-MW PV systems are located near the center of the feeder, approximately 5 circuit miles from the distribution substation.

While not a heavily loaded feeder, it does require the use of a load tap changer (LTC) as well as multiple feeder regulators and switched capacitor banks to provide voltage regulation. As shown in Fig. 1, there are four feeder regulators utilized throughout, along with five capacitor banks, three of which are voltage controlled.

Fig. 2. Feeder J1 satellite view showing 1.7-MW PV systems (customer owned) located near a small town.

Fig. 2 shows a satellite view showing the 1.7-MW PV. The small town, located at the crossroads west of the PV plants, has a population less than 1,000. Feeder J1 serves the town as well as surrounding rural areas and farms. At the current penetration level of PV, the utility has encountered concerns or complaints of high voltage on this feeder, likely caused by the 1.7 MW of solar generation. A common misnomer is that feeder regulators should be used to mitigate voltage issues caused by solar PV. While they can in fact compensate for the fluctuations caused by PV, the intentional time delays programmed within the regulators prevent them from operating faster than 45-90 seconds. As EPRI has shown in previous analysis, solar PV can ramp fast enough to cause voltage violations prior to regulator operation. This could perhaps be the source of the overvoltage violations.

Per request by the local utility, the site owner has agreed to lower the power factor on selected PV inverters to conduct a trial of increased reactive power flow and the potential effects on voltage control. The expectation is that the voltage issues will be resolved at a slightly inductive power factor. In addition to installing the pole-mount systems along the feeder, EPRI’s monitoring system is installed at the solar PV sites, collecting both active and reactive power as well as interconnect voltage. The monitoring systems allow EPRI to capture the feeder voltage response due to the solar PV, both before and after the power factor adjustment is made. This can be used to confirm the overvoltage conditions, as well as identify the effectiveness of the power factor change.

Sample Results

Along with the measurement data, the feeder model developed in the OpenDSS can be used to simulate both conditions. It can further be used to evaluate the overvoltage conditions, simulate the feeder regulator responses due to the PV, different power factor setpoints, volt/var functions, and increasing PV levels.

Fig. 3 shows the feeder voltage profile as a function of distance. Each color represents a separate phase, while the solid lines represent primary voltages and the hashed lines represent secondary voltages. As expected, due to customer load along the feeder, the voltage is reduced as the distance from the substation increases (x-axis is distance from the distribution substation). However, the step increases in voltage at 5 km, 8 km, and 13 km are points along the feeder where regulators increase the voltage in order to prevent unacceptable undervoltages from occurring. Without the feeder regulators, the end-line voltages would fall well below the acceptable threshold of 0.95 pu (note that the ANSI C84.1 acceptable range of 0.95 - 1.05 pu is shown as red horizontal lines).

Fig. 3. Feeder Voltage Profile without PV (Solid: Primary, Hashed: Secondary)

One of the most common tests for determining whether or not variable generation, either distribution or transmission connected, could potentially result in unacceptable voltage variations on the grid is to perform what is known as a “voltage change” test. This test essentially involves solving the circuit without generation, then locking all regulation equipment controls such that they do not operate, adding the PV to the model, and resolving. By performing the analysis this way, the resulting voltages are those experienced on the grid before regulation equipment operates. This approach assumes the generation can ramp fast enough before regulation equipment has time to compensate for the change in output. Given that the potential ramp rates associated with variable generation are less than 1 minute, this is a reasonable test.

The resulting voltage profile is shown in Fig. 4. The PV site, located approximately 6.5 km from the substation, increases the Phase A and B voltages that are downstream from the first feeder regulator well above the 1.05 pu limit. Because of the increased voltages along the feeder, downstream voltages on Phase B are also found to increase above the allowable threshold.

Fig. 4. Feeder Voltage Profile with PV before Regulation Equipment Operation (Solid: Primary, Hashed: Secondary)

While this only illustrates an initial analysis of the voltage impacts on the feeder, it is further verified with feeder monitoring and detailed analysis.