All utilities maintain statistics on the reliability of the electricity supply to consumers. In recent years, utilities have begun to track other power quality characteristics of the distribution system that may impact consumers. Carolina Power and Light Co. (CP&L), located in Raleigh, N.C., has one of the largest distribution power quality monitoring systems in the world. The utility currently monitors its entire distribution system. In this article, you'll read about this ambitious project and the resulting benefits, including a benchmarking indice for evaluating feeder performance.

At CP&L, the monitoring system includes power quality monitors on 285 substations and 970 distribution feeders, operational alarms in 255 substations, and DSCADA in 75 substations. The monitoring system also controls 285 substation capacitor banks and approximately 1700 distribution-feeder capacitor banks.

As you might expect, voltage sag performance for distribution systems can vary dramatically from system to system, depending on the design (i.e., overhead vs. underground), line exposure, the lightning flash density in the area, voltage levels, trees, animals, nearness to the ocean, and many other factors. That's why it's difficult to measure the effectiveness of distribution system maintenance by the number of voltage sags.

A power system's voltage sag performance is heavily impacted by system faults. Both distribution-circuit faults (i.e., faults on all of a substation's feeder circuits) and transmission-system faults contribute to a system's overall sag performance.

In the case of transmission systems, utilities track the fault performance by the number of faults per 100 miles per year. These numbers are compiled for each class of voltage on the system, and they can be used to evaluate individual line performances against the system averages.

At CP&L, the monitoring system monitors every substation and feeder circuit on the system. This allows engineers to track the fault performance of distribution circuits in a manner similar to transmission lines. Such statistics allow performance evaluations at the individual feeder level, and they can be used to target feeders for improvements and maintenance.

The part of CP&L's overall monitoring system that captures power quality information for substation and feeder performance is called the feeder monitoring system (FMS). CP&L also uses FMSs in some of the SCADA-controlled substations and in all of the substations used for feeder capacitor control.

The FMS consists of a remote terminal unit in the substation, with transducers on each feeder and a set of transducers on the substation bus. The sample rate is 16 samples per cycle. The sample rate is more than adequate for characterizing faults on the feeder circuits. Besides calculating feeder fault performance, the system also locates faults on the feeder circuits, which helps reduce system repair times.

The FMS can capture any voltage or current event that exceeds set triggers down to 1 cycle in duration. Waveforms and rms-vs.-time records characterize each event. Besides triggered events, the system collects periodic measurements of voltage, current, real and reactive power, power factor, and harmonic distortion. All of the data is gathered three times a day and stored for future use.

How Faults Affect Customers

The typical operation of a distribution feeder breaker and the effect of a fault on an adjacent feeder are shown in Fig. 1, on page 8. The figure indicates the typical recloser settings used by CP&L.

On a faulted feeder, customers experience a voltage sag followed by an interruption (see Fig. 2). This may repeat if recloser operations do not clear the fault successfully.

On adjacent feeders, customers experience voltage sags during the fault, but the voltage is restored when the recloser opens to clear the fault.

At CP&L, reclosers try to restore power three times before lockout occurs. With the first operation, a fault is cleared 68% of the time. With the subsequent two operations, the fault is cleared 27% of the time. Finally, 5% of the time, the breaker will lock out and require a serviceman to clear the fault and reset the breaker. The benefits of using reclosers are obvious — 95% of all faults are successfully cleared without causing a long outage for any of the customers on the feeder.

Distribution-system faults can have a variety of different characteristics. The type of fault, the distance from the substation, and the fault impedance can all impact voltage sags experienced by customers.

A few examples of fault characteristics are illustrated in Fig. 3, on page 10, and Fig. 4. These monitoring results illustrate the performance of the protection equipment during a fault. Engineers can use the monitored results to quickly identify problems with recloser operation or coordination between the reclosers and downline protective devices (e.g., fuses).

Feeder Performance Indice

During a four-year period, CP&L's monitors recorded more than 720,000 data events. Engineers used these results to develop performance benchmarking data for individual feeders and overall distribution systems.

The concept of using faults as a measure of feeder performance is a logical one. Every fault causes a power quality event (sag), and many of the faults lead to a reliability event (when a customer experiences an outage lasting more than 5 minutes). Fault events are counted using 1-minute aggregation. This prevents multiple counts for the same fault event due to recloser operations.

For the benchmarking effort, engineers calculated the fault counts on individual feeder circuits. These performance indices are normalized by the lengths of the feeder circuits — a method similar to the one used by engineers to normalize transmission line performance (i.e., faults/100 miles/year). The index for distribution feeders is faults/feeder mile (FFM).

After engineers calculated the base indices, they performed a data correction to eliminate faults during major storm events. Such events are considered unusual, and faults during these events cannot be prevented by normal maintenance and system-design practices. Therefore, indices designed to help prioritize system maintenance and investment should not include these events.

The table summarizes the feeder performance statistics on the CP&L system over the benchmarking period. These are annualized indices. It shows that the average fault performance for feeders on the CP&L system is about 1 FFM. Engineers can use this figure as a convenient gauge when evaluating the performance of individual feeder circuits. Note that other factors, including the number of customers impacted by the faults, are important for prioritizing expenditures to improve fault performance.

Conclusion

The fact that the substation monitoring system at CP&L tracks the performance of individual feeder circuits on the system is only one of the benefits realized by utility engineers. Other benefits discovered by the engineers include automatic location of faults on the feeder circuits; identification of equipment problems, such as blown fuses in capacitor banks; identification of incipient faults that may indicate arrester failures or cable splice problems; and identification of unusual fault conditions, such as those caused by galloping conductors.

CP&L staff members are continuing to develop advanced applications for the monitoring data. More information on the types of advanced applications under development can be found at www.powermonitoring.com/substation.htm.

Mark McGranaghan directs power quality projects and product development at Electrotek Concepts. You can reach him at mark@electrotek.com.

Scott Peele is a power quality project engineer at Carolina Power and Light Co. in Raleigh, N.C. You can reach him at scott.peele@pgnmail.com.