Simple Strategies to Improve Power Reliability

By Steve Eckles, El Paso Electric Co.

On average, 92 percent of all customer outages result from problems on the power distribution system, and these power interruptions are costly-both for customers and utilities.

An unscheduled power outage costs the average large industrial customer approximately $40,000, increasing to $75,000 for a four-hour outage. The Electric Power Research Institute estimates annual outage costs to society amount to $119 billion. In some states, reliability also has a direct economic impact on utilities through monetary incentives (e.g. performance-based rates) or punishments (e.g. fines imposed for poor reliability performance).

Electric utilities need more bang for their buck as regulators and customers focus on reducing power outages. This article describes short circuit (also known as faults) causes and types and explains a five-step strategy to improve distribution reliability while accounting for fault statistics.

Data Gathering

The first step toward cost effectively improving power reliability is compiling basic interruption data, including date and time of the outage, the feeder affected, type of device that operated, outage cause, number of customers affected, and outage duration. Many utilities track this data using an outage management system (OMS) or a simple database program or spreadsheet.

Analyzing the Data

When it comes to the war on interruptions, utilities must “know their enemy.” Periodically, outage data should be evaluated and standard reliability indices should be calculated.

Indices are used to measure, report, and track sustained outages, and prioritize system improvements. The most common reliability indices are SAIFI, SAIDI, CAIDI and MAIFI. Sustained interruptions are defined as outages five minutes or greater in duration (although some utilities use a different definition).

SAIFI (System Average Interruption Frequency Index) is the total number (cumulative sum) of interrupted customers divided by the total number of customers served.

SAIDI (System Average Interruption Duration Index) is the sum of customer interruption durations (cumulative sum of the product of the number of customers interrupted and outage duration) divided by the total number of customers served.

CAIDI (Customer Average Interruption Duration Index) is the (cumulative) sum of customer interruption duration divided by the total number of customer interruptions. This is also equal to SAIDI divided by SAIFI.

SAIFI and SAIDI are the most commonly reported indices. They can vary greatly among utilities depending on climate (commonality of snow, ice and/or wind storms), terrain (mountainous, desert or coastal), load density (urban or rural) and system design (radial, looped or 3-wire). Typical U.S. utility SAIFI is 1.4 interruptions per year, SAIDI is 110 minutes each year, and CAIDI is 79 minutes per year.

Two lesser-used indices are ASAI and CEMI. ASAI (Average service availability index) is the percentage of time service was available to customers. It is calculated by using 8,760 hours in a year (8,784 hours in a leap year) and SAIDI. Utilities with a SAIDI of 110 minutes (1.83 hours) during the year have an ASAI of 99.98 percent.

ASAI = [(8760 – SAIDI)/8760] x 100%

CEMI (customers experiencing multiple interruptions) measures the percent of overall customers that have experienced more than a specific number of interruptions. Decreasing this index has the propensity to decrease the number of customer complaints. Three interruptions during a six-month period typically doubles the number of complaints that a utility would receive for three outages during a full year. Complaints increase even more dramatically when customers experience more than three service interruptions within six months. Outage interval is also important; however, rural customers seem to accept outages better than urban ones.

Electronic equipment tends to misoperate or reset following momentary interruptions, such as those occurring when transient faults are cleared. Some regulators now require reporting of MAIFI index (Momentary Average Interruption Frequency Index), which is calculated by the total number of customers (cumulative sum) affected by an interruption of less than five minutes divided by the total number of customers served.

Most unknown outages can usually be attributed to wildlife.
Click here to enlarge image

Outages from major events, having a disproportionate impact, are usually not included in reported index numbers. These include ice storms, hurricanes, earthquakes, flooding and rioting. Major events are usually predetermined by establishing wind speed or precipitation thresholds that are monitored and reported by the National Weather Service. Another common threshold is if 10 percent of customers in a certain area are affected. Sorting the data by interruption cause will indicate where resources should be applied for optimal results. It makes little sense to significantly invest in more lightning arresters if animals caused three times more outages than lightning. Often an “unknown cause” will have a higher percentage of entries. Detailed investigations show that most unknown outages can usually be attributed to wildlife.

Outage Prevention

By definition, both SAIFI (outage frequency) and SAIDI (outage duration) can be reduced by preventing sustained outages. Approximately 75 percent of overhead faults have a temporary cause such as lightning, animals, trees or debris in the line, or vehicles hitting poles causing conductors to slap together. Reclosing-the practice of de-energizing (tripping) the faulted circuit momentarily then re-energizing (reclosing) it-may cause some customers’ electronic appliances to stop or reset, but will probably prevent a sustained outage. For multiple reclosing, the probability of successful re-energization decreases progressively with each attempt. Most utilities never exceed three reclose attempts.

Outage-cause analysis should influence feeder design and construction. If lightning is a major outage culprit, shield-wire construction or additional lighting arresters should be installed, and feeder Basic Insulation Level (BIL) should be evaluated. Raising BIL is often accomplished in overhead circuits by increasing clearances from phase conductors to uninsulated guy wires and other grounded equipment.

Anti-perch devices, such as this triangle, can mitigate bird-caused outages.
Click here to enlarge image

Birds cause nearly 25 percent of all overhead distribution interruptions in the United States. To mitigate bird-caused outages, clearances can be increased between energized conductors by using longer cross-arms, covered jumpers, bird guards/covers and perch prevention devices. Many birds of prey (raptors) have wide wing spans. Increasing conductor spacing to 60 inches or covering conductors with less clearance is generally adequate to prevent raptor electrocution. Increasing three-phase clearance may require a 12-foot (or greater) crossarm or pole-top insulator and conductor construction-a 10-foot crossarm may be used if installed far enough below the top to provide 60 inches of space between pole-top and crossarm conductors.

Three-phase sectionalizer with bird guards.
Click here to enlarge image

Some utilities try to provide a safe nesting area or perch for every anti-perch device they install. Since state and federal regulations may prevent removing certain bird nests, supplying birds with alternative nesting areas can be beneficial.

Some utilities provide birds with alternate perching areas.
Click here to enlarge image

Proactive maintenance programs and patrols that concentrate on replacing failing and overloaded equipment will reduce outages. Infrared inspections can help identify failing cable elbows and connections as well as overloaded transformers. Tree trimming and other vegetation control also belong in this category. Some utilities use vertical construction not only due to tight right-of-way constraints but to also reduce tree contact. For U.S. utilities, data shows that trees cause five times more outages by falling into distribution lines than by growing into them.

Underground cable is immune to many overhead outage causes, but the higher installation cost prevents more extensive underground cable use. Cable has limited life, therefore, directly burying cable should be avoided in favor of installing cable in conduit. Some proactive utilities test the condition of aged cable to determine whether it should be replaced or injected with silicone-fluid under pressure (usually economical for direct-buried cable only) to reconstitute its dielectric strength and delay failure.

The topic of outage prevention design would be incomplete without touching on ungrounded (and high-impedance grounded) distribution systems that are used in much of the world. The advantage of these systems is that the most predominate type of fault-phase-to-ground-may produce insufficient amounts of fault current to initiate overcurrent device operation. This leaves customers energized for transient phase-to-ground faults. If the fault persists, ground detection alarms alert the utility, and workers are dispatched to find and resolve the condition, hopefully before substantial damage occurs. The disadvantages to ungrounded systems are the added cost of insulating utility equipment for phase-to-phase voltage, difficulty locating faults and public hazard. Furthermore, as ungrounded systems grow, large capacitive currents may prevent faults from self-clearing.

Infrared inspections can help identify failing cable elbows and connections.
Click here to enlarge image

Critical loads such as medical, commercial and small manufacturing centers may justify spot networks or automatic throw-over (ATO) switching. Spot networks generally have tied underground secondary systems fed from transformers from different primary circuits. Network protectors detect and open for reverse power flow preventing one primary circuit from feeding into another primary circuit’s fault. ATO switches between two feeders are less complex but involve a brief outage before re-energizing customers.

Segmenting and Sectionalizing

It is impossible to build a fault-free distribution system. Even with good anti-outage design and construction, equipment degrades and eventually fails. A reasonable number of faults are expected. For these faults, the primary goal is to sectionalize feeders to reduce customer outages. Sectionalizing also shortens restoration time by narrowing the troubleshooting range.

First and foremost, all lateral taps and transformers should be fused. The completely self-protected (CSP) transformer is a misnomer. They offer little protection against animals faulting across the primary bushing to the grounded transformer tank. Installing primary fuses on CSP transformers prevents lateral fuses from blowing for faults across distribution transformer primary bushings reducing corresponding outages to one transformer. Lateral taps with just one transformer (or transformer bank) may be more economically protected by placing the transformer fuse(s) at the tap.

In addition to being low cost, installing fuses affords reliable and predictable operation with a low probability of fuse holders faulting. They are best employed when they “coordinate” with other fuses in series. Coordination occurs when downstream fuses are small enough-with respect to upstream fuses-that they clear before upstream fuses start to melt for the entire range of potential fault current through both fuses. For significantly long feeders and lateral taps, multiple fuses in series may prove beneficial. Fuses have the drawback of not automatically resetting, thereby failing to take advantage of the high percentage of temporary fault causes.

Feeder reclosers may also be installed to clear temporary faults and for feeder segmenting. They detect overcurrent and trip accordingly to de-energize the line downstream, allowing the fault to clear before re-energizing the circuit. If the fault persists, the recloser will typically stay open (lockout) after the fourth time the fault is detected. Some utilities use single-phase reclosers to keep two-thirds of the customers in service for more common phase-to-ground faults.

Feeder reclosers are installed to clear temporary faults and for feeder segmenting.
Click here to enlarge image

Reclosers can be employed that trip quickly upon detecting the first or second fault before lateral fuses reach their melting point. This technique, known as “fuse saving,” is an effective means to keep all customers in service for transient faults. For persistent faults, reclosers will delay subsequent tripping to allow a downstream fuse to blow and isolate the faulted circuit. The disadvantages to fuse saving are increases in momentary outages, more stress on substation transformers and higher probability of fault damage.

When coordination is difficult with reclosers and fuses, sectionalizers can be used to segment feeders and major lateral taps.
Click here to enlarge image

When coordination is difficult with reclosers and fuses, sectionalizers can be used to segment feeders and major lateral taps. Sectionalizers can be three- or single-phase and detect typical fault current magnitudes. However, sectionalizers require upstream feeder breakers or reclosers to interrupt fault current. They open during line de-energization to “sectionalize” the fault before the upstream device re-energizes the line to restore service to customers upstream of the open sectionalizer. They are preset to open after the first, second or third time they detect fault current.

Motorized switches are also used to segment faulted feeders.
Click here to enlarge image

Other methods to segment faulted feeders are remote controlled or automated motorized switches. Advanced systems can be programmed and linked via radio to detect fault location and change state to isolate the fault while reconfiguring the circuit to restore service quickly to unisolated customers. Another segmenting principle is installing more, but shorter, feeders that reduce average feeder exposure.

Quicker Outage Restoration

Quicker outage restoration will improve SAIDI and CAIDI, but not SAIFI, so it may not be the best category on which to spend substantial resources. Assuming repair time is relatively constant, fast outage troubleshooting and switching are the keys to restoring customers in less time. As referenced above, troubleshooting time is shorter if fuses, reclosers or sectionalizers are installed to segment faults. Nevertheless, if one of these devices fails to operate correctly, fault locating time may lengthen dramatically.

To assist in fault locating, faulted circuit indicators (FCIs) can be installed permanently or temporarily on underground or overhead phase conductors, typically on long unsegmented runs and feeder circuit branches. FCIs monitor conductor current and indicate faults with a flag, target and/or flashing light. Today’s designs discriminate between inrush current and capacitive current backfeeding into faults. Some FCIs offer several hours of reset delay after circuit re-energization to allow utilities to find temporary, recurring faults. Technology allows some FCIs to transmit fault information back to utility system operators. Care must be taken to install FCIs per manufacturer’s guidelines to reduce misoperation. Investment will be wasted if utility personnel are misdirected or lose faith in their accuracy.

Voltage regulators with bird guards and covered jumpers.
Click here to enlarge image

Anticipating possible permanent fault locations in advance and documenting possible backfeeding scenarios will assist system operators in backfeeding circuits quicker during the heat-of-the-battle. Distribution power flow software is becoming more powerful. With accurate data, it can quickly determine the best switching configuration to restore the most customers during repairs.


Cost effectively improving electrical distribution reliability starts with good feeder design and outage data collection. Analyzing outage data and calculating reliability indices provide a good way to track improvements and determine where to direct more resources. Preventing sustained outages will lower overall system interruption frequency and duration. The next step is to reduce the number of customers interrupted by employing sectionalizing overcurrent protective devices. This will also reduce restoration time by narrowing the troubleshooting and patrolling scope. FCIs, automatic, or predetermined switching may reduce outage time also. In any event, utility efforts to keep the lights on have far-reaching societal cost and safety benefits.

Steve Eckles has been a distribution engineer for 13 years and is a licensed PE in New Mexico and Texas. He obtained a BSEE from San Diego State University and a MSEE from going through New Mexico State University’s Electrical Utility Management Program (EUMP). He has previously authored technical papers in electrochemistry, photovoltaics, electrical utility distribution and power quality.

Previous articlePOWERGRID_INTERNATIONAL Volume 11 Issue 4
Next articleSuntech Power receives UL cert for photovoltaic modules

No posts to display