In this article we will discuss about:- 1. Sources of Voltage Sags 2. Area of Vulnerability 3. Equipment Sensitivity 4. Interruptions 5. Performance Evaluation 6. Protection.

Voltage sag is a short duration (typically 0.5 to 30 cycles) reduction in rms voltage caused by faults on the power system and by the starting of large loads, such as motors. Momentary interruptions (typically no more than 2 to 5 s) cause a complete loss of voltage and are a common result of the actions taken by utilities to clear transient faults on their systems. Sustained interruptions of longer than 1 min are generally due to permanent faults.

Utilities have been faced with rising numbers of complaints about the quality of power due to sags and interruptions. There are a number of reasons for this, with the most important being that customers in all sectors (residential, commercial, and industrial) have more sensitive loads. The influx of digital computers and other types of electronic controls is at the heart of the problem.

Computer controls tend to lose their memory, and the processes that are being controlled also tend to be more complex and, therefore, take much more time to restart. Industries are relying more on automated equipment to achieve maximum productivity to remain competitive. Thus, an interruption has considerable economic impact.

Sources of Voltage Sags:


Voltage sags and interruptions are generally caused by faults (short circuits) on the utility system. Consider a customer that is supplied from the feeder supplied by circuit breaker 1 on the diagram shown in Fig. 2.1. If there is a fault on the same feeder, the customer will experience a voltage sag during the fault followed by an interruption when the breaker opens to clear the fault. If the fault is temporary in nature, a reclosing operation on the breaker should be successful and the interruption will only be temporary.

It will usually require about 5 or 6 cycles for the breaker to operate, during which time a voltage sag occurs. The breaker will remain open for typically a minimum of 12 cycles up to 5 s depending on utility reclosing practices. Sensitive equipment will almost surely trip during this interruption. A much more common event would be a fault on one of the other feeders from the substation, i.e., a fault on a parallel feeder, or a fault somewhere on the transmission system (see the fault locations shown in Fig. 2.1).

In either of these cases, the customer will experience a voltage sag during the period that the fault is actually on the system. As soon as breakers open to clear the fault, normal voltage will be restored at the customer. Note that to clear the fault shown on the transmission system, both breakers A and B must operate. Transmission breakers will typically clear a fault in 5 or 6 cycles. In this case there are two lines supplying the distribution substation and only one has a fault. Therefore, customers supplied from the substation should expect to see only a sag and not an interruption. The distribution fault on feeder 4 may be cleared either by the lateral fuse or the breaker, depending on the utility’s fuse-saving practice.


Any of these fault locations can cause equipment to misoperate in customer facilities. The relative importance of faults on the transmission system and the distribution system will depend on the specific characteristics of the systems (underground versus overhead distribution, lightning flash densities, overhead exposure, etc.) and the sensitivity of the equipment to voltage sags.

Figure 2.2 shows an example of the breakdown of the events that caused equipment misoperation for one industrial customer. Note that faults on the customer feeder only accounted for 23 percent of the events that resulted in equipment misoperation.

This illustrates the importance of understanding the voltage sag performance of the system and the equipment sensitivity to these events. The system was restored after the second operation.


There are few things to note about this typical event:

1. The voltage did not go to zero during the fault as is often assumed in examples. There are few examples of the case in real life.

2. The line recloser detected the fault and operated very quickly. There is a common misconception that fault interruption is slower on the distribution system than on the transmission system. While it can be slower, it can also be faster.

3. Since the voltage did not collapse to zero during the fault, induction machines will continue to have excitation and continue to feed the fault. This can be an especially important consideration for distributed generation.


Motor Starting Sags:

Motors have the undesirable effect of drawing several times their full load current while starting. This large current will, by flowing through system impedances, cause a voltage sag which may dim lights, cause contactors to dropout, and disrupt sensitive equipment. The situation is made worse by an extremely poor starting displacement factor usually in the range of 15 to 30 percent.

The time required for the motor to accelerate to rated speed increases with the magnitude of the sag, and an excessive sag may prevent the motor from starting successfully. Motor starting sags can persist for many seconds, as shown in Fig. 2.3.

Motor Starting Methods:


Energizing the motor in a single step (full voltage starting) provides low cost and allows the most rapid acceleration. It is the preferred method unless the resulting voltage sag or mechanical stress is excessive.

i. Autotransformer starters have two autotransformers connected in open delta. Taps provide a motor voltage of 80.65 to 50 percent of system voltage during start-up. Line current and starting torque vary with the square of the voltage applied to the motor, so the 50 percent tap will deliver only 25 percent of the full voltage starting current and torque. The lowest tap which will supply the required starting torque is selected.

ii. Resistance and reactance starters initially insert an impedance in series with the motor. After a time delay, this impedance is shorted out. Starting resistors may be shorted out over several steps; starting reactors are shorted out in a single step. Line current and starting torque vary directly with the voltage applied to the motor, so for a given starting voltage, these starters draw more current from the line than with autotransformer starters, but provide higher starting torque. Reactors are typically provided with 50, 45 and 37.5 percent taps.

iii. Part-winding starters are attractive for use with dual-rated motors (220/440 V) or 230/460 V). The stator of a dual-rated motor consists of two windings connected in parallel at the lower voltage rating, or in series at the higher voltage rating. When operated with a part-winding starter at the lower voltage rating, only one winding is energized initially, limiting starting current and starting torque to 50 percent of the values seen when both windings are energized simultaneously.

iv. Delta-wye starters connect the stator in wye for starting and then, after a time delay, reconnect the windings in delta. The wye connection reduces the starting voltage to 57 percent of the system line-line voltage; starting current and starting torque are reduced to 33 percent of their values for full- voltage start.

Calculation of Sag Severity during Full Voltage Starting:

As shown in Fig. 2.3, starting an induction motor results in a steep dip in voltage, followed by a gradual recovery. If full voltage starting is used, the sag voltage, in per unit of nominal system voltage, is

where V(pu) = actual system voltage, in per unit of nominal

kVALR = motor locked rotor kVA

kVASC = system short-circuit kVA at motor

Voltage Sag due to Electric Arc Furnace:

The voltage across an electric arc, which is relatively independent of current magnitude, consists of three components: anode drop, cathode drop and arc column component, which amounts to about 12 volts/cm of arc length. Typical values of arc voltages are in the range of 150-500 volts. Since the arc is extinguished at current zero, the power factor plays an important role on arc reignition. Fig. 2.4 shows how arc voltage, power factor, input power; arc power and reactive power vary with arc current for a particular tap setting on the furnace transformer. The furnace is normally operated near maximum arc power, which corresponds to a power factor of 70%.

The three basic changes in operating states of an electric arc furnace, which can produce distinguishable voltage disturbances on power system, are open circuit condition, short circuit condition and the normal operation. The measurable data of interest for an electric arc furnace load include the three-phase quantities- supply voltage, real and reactive power, flicker, frequency and total harmonic distortion in respective phases.

Because of the non-linear resistance, an arc furnace acts as a source of current harmonics of the second to seventh order, especially during the meltdown period. Voltage fluctuations are produced in this way through impedance on the value of harmonic currents supplied and the effective impedances at the harmonic frequencies.

The harmonic current of the arc furnace forms a parallel tuned circuit consisting of capacitor C with reactive power and mains inductance, resulting from the mains short circuit power. When this tuned circuit resonates at a harmonic frequency, its reactance is high and a harmonic voltage arises, which is damped by the resistance of the resistive component of the supply system consumers’ equipment.

The Q factor of this tuned circuit is low at times of full load, and no resonant peaks occur. But in slack periods with combinations of low load with high resistance and Q factor values, harmonic voltages are expected at levels sufficient to cause appreciable interference.

Fault Clearing Issues:

Fault clearing practices have a great influence on the voltage sag and interruption performance at a distribution-connected load.

Utilities have two basic options to continue to reduce the number and severity of faults on their system:

1. Prevent faults.

2. Modify fault-clearing practices.

Fault prevention activities include tree trimming, adding line arresters, insulator washing, and adding animal guards. Insulation on utility lines cannot be expected to withstand all lightning strokes. However, any line that shows a high susceptibility to lightning-induced faults should be investigated. On transmission lines, shielding can be analyzed for its effectiveness in reducing direct lightning strokes. Tower footing resistance is an important factor in backflashowers from static wire to a phase wire.

If the tower footing resistance is high, the surge energy from a lightning stroke will not be absorbed by the ground as quickly. On distribution feeders, shielding may also be an option as is placing arresters along the line frequently. Of course, one of the main problems with overhead distribution feeders is that storms blow tree limbs into the lines. In areas where the vegetation grows quickly, it is a formidable task to keep trees properly trimmed.

Improved fault-clearing practices may include adding line re-closers, eliminating fast tripping, adding loop schemes, and modifying feeder design. These practices may reduce the number and/or duration of momentary interruptions and voltage sags, but utility system faults can never be eliminated completely.

Therefore, the detection of faults and the clearing of the fault current must be done with the maximum possible speed without resulting in false operations for normal transient events.

The two greatest concerns for damage are typically:

i. Arcing damage to conductors and bushings.

ii. Through fault damage to substation transformers, where the windings become displaced by excessive forces, resulting in a major failure.

Area of Vulnerability:

The concept of an area of vulnerability has been developed to help evaluate the likelihood of sensitive equipment being subjected to voltage lower than its minimum voltage sag ride-through capability. The latter term is defined as the minimum voltage magnitude a piece of equipment can withstand or tolerate without misoperation or failure. This is also known as the equipment voltage sag immunity or susceptibility limit.

An area of vulnerability is determined by the total circuit miles of exposure to faults that can cause voltage magnitudes at an end-user facility to drop below the equipment minimum voltage sag ride-through capability. Figure 2.5 shows an example of an area of vulnerability diagram for motor contactor and adjustable- speed-drive loads at an end-user facility served from the distribution system.

The loads will be subject to faults on both the transmission system and the distribution system. The actual number of voltage sags that a facility can expect is determined by combining the area of vulnerability with the expected fault performance for this portion of the power system. The expected fault performance is usually determined from historical data.

Equipment Sensitivity to Voltage Sags:

Equipment within an end-user facility may have different sensitivity to voltage sags. Equipment sensitivity to voltage sags is very dependent on the specific load type, control settings, and applications. Consequently, it is often difficult to identify which characteristics of a given voltage sag are most likely to cause equipment to misoperate.

The most commonly used characteristics are the duration and magnitude of the sag. Other less commonly used characteristics include phase shift and unbalance, missing voltage, three-phase voltage unbalance during the sag event, and the point-in-the-wave at which the sag initiates and terminates. Generally, equipment sensitivity to voltage sags can be divided into three categories.

Equipment Sensitive to only the Magnitude of a Voltage Sag:

This group includes devices such as under-voltage relays, process controls, motor drive controls, 6 and many types of automated machines (e.g., semiconductor manufacturing equipment). Devices in this group are sensitive to the minimum (or maximum) voltage magnitude experienced during a sag (or swell). The duration of the disturbance is usually of secondary importance for these devices.

Equipment Sensitive to both the Magnitude and Duration of a Voltage Sag:

This group includes virtually all equipment that uses electronic power supplies. Such equipment misoperates or fails when the power supply output voltage drops below specified values. Thus, the important characteristic for this type of equipment is the duration that the rms voltage is below a specified threshold at which the equipment trips.

Equipment Sensitive to Characteristics other than Magnitude and Duration:

Some devices are affected by other sag characteristics such as the phase unbalance during the sag event, the point in the wave at which the sag is initiated, or any transient oscillations occurring during the disturbance. These characteristics are more subtle than magnitude and duration, and their impacts are much more difficult to generalize. As a result, the rms variation performance indices defined here are focused on the more common magnitude and duration characteristics. For end users with sensitive processes, the voltage sag ride-through capability is usually the most important characteristic to consider.

Voltage Sags and Interruptions:

These loads can generally be impacted by very short duration events, and virtually all voltage sag conditions last at least 4 or 5 cycles (unless the fault is cleared by a current- limiting fuse). Thus, one of the most common methods to quantify equipment susceptibility to voltage sags is using a magnitude-duration plot as shown in Fig. 2.6.

It shows the voltage sag magnitude that will cause equipment to misoperate as a function of the sag duration. The curve labeled CBEMA represents typical equipment sensitivity characteristics. The curve was developed by the CBEMA and was adopted in IEEE 446. Since the association reorganized in 1994 and was subsequently renamed the Information Technology Industry Council (ITI), the CBEMA curve was also updated and renamed the ITI curve.

Typical loads will likely trip off when the voltage is below the CBEMA, or ITI, curve. The curve labelled ASD represents an example ASD voltage sag ride-through capability for a device that is very sensitive to voltage sags. It trips for sags below 0.9 pu that last for only 4 cycles. The contactor curve represents typical contactor sag ride-through characteristics. It trips for voltage sags below 0.5 pu that last for more than 1 cycle.

The area of vulnerability for motor contactors shown in Fig. 2.3 indicates that faults within this area will cause the end-user voltage to drop below 0.5 pu. Motor contactors having a minimum voltage sag ride-through capability of 0.5 pu would have tripped out when a fault causing a voltage sag with duration of more than 1 cycle occurs within the area of vulnerability.

However, faults outside this area will not cause the voltage to drop below 0.5 pu. The same discussion applies to the area of vulnerability for ASD loads. The less sensitive the equipment, the smaller the area of vulnerability will be (and the fewer times sags will cause the equipment to misoperate).

Transmission System Sag Performance Evaluation:

The voltage sag performance for a given customer facility will depend on whether the customer is supplied from the transmission system or from the distribution system. For a customer supplied from the transmission system, the voltage sag performance will depend on only the transmission system fault performance. On the other hand, for a customer supplied from the distribution system, the voltage sag performance will depend on the fault performance on both the transmission and distribution systems.

Transmission line faults and the subsequent opening of the protective devices rarely cause an interruption for any customer because of the interconnected nature of most modern-day transmission networks. These faults do, however, cause voltage sags. Depending on the equipment sensitivity, the unit may trip off, resulting in substantial monetary losses. The ability to estimate the expected voltage sags at an end-user location is therefore very important.

The magnitude of the lowest secondary voltage depends on how the equipment is connected:

i. Equipment connected line-to-line would experience a minimum voltage of 33 percent.

ii. Equipment connected line-to-neutral would experience a minimum voltage of 58 percent.

This illustrates the importance of both transformer connections and the equipment connections in determining the actual voltage that equipment will experience during a fault on the supply system. Math Bollen 16 developed the concept of voltage sag “types” to describe the different voltage sag characteristics that can be experienced at the end-user level for different fault conditions and system configurations. The five types that can commonly be experienced are illustrated in Fig. 2.7.

These fault types can be used to conveniently summarize the expected performance at a customer location for different types of faults on the supply system.

The actual expected performance is then determined by combining the area of vulnerability with the expected number of faults within this area of vulnerability. The fault performance is usually described in terms of faults per 100 miles/year (mi/yr).

Most utilities maintain statistics of fault performance at all the different transmission voltages. These system-wide statistics can be used along with the area of vulnerability to estimate the actual expected voltage sag performance.

Figure 2.8 gives an example of this type of analysis. The figure shows the expected number of voltage sags per year at the customer equipment due to transmission system faults. The performance is broken down into the different sag types because the equipment sensitivity may be different for sags that affect all three-phases versus sags that only affect one or two phases.

Utility Distribution System Sag Performance Evaluation:

Customers that are supplied at distribution voltage levels are impacted by faults on both the transmission system and the distribution system. The analysis at the distribution level must also include momentary interruptions caused by the operation of protective devices to clear the faults. These interruptions will most likely trip out sensitive equipment. The example presented here illustrates data requirements and computation procedures for evaluating the expected voltage sag and momentary interruption performance. The overall voltage sag performance at an end-user facility is the total of the expected voltage sag performance from the transmission and distribution systems.

Figure 2.9 shows a typical distribution system with multiple feeders and fused branches and protective devices. The utility protection scheme plays an important role in the voltage sag and momentary interruption performance.

The critical information needed to compute voltage sag performance can be summarized as follows:

i. Number of feeders supplied from the substation.

ii. Average feeder length.

iii. Average feeder reactance.

iv. Short circuit equivalent reactance at the substation.

Protection from Voltage Sags:

Several things can be done by the utility, end-user and equipment manufacturer to reduce the number and severity of voltage sags and to reduce the sensitivity of equipment to voltage sags.

Figure 2.11 illustrates voltage sag solution alternatives and their relative costs. As this chart indicates, it is generally less costly to tackle the problem at its lowest level, close to the load. The best answer is to incorporate ride through capability into the equipment specifications themselves. This essentially means keeping problem equipment out of the plant, or at least identifying ahead of time power conditioning requirements.

Several ideas, outlined here, could easily be incorporated into any company’s equipment procurement specifications to help alleviate problems associated with voltage sags:

1. Equipment manufacturers should have voltage sag ride-through capability curves available to their customers so that an initial evaluation of the equipment can be performed. Customers should begin to demand that these types of curves be made available so that they can properly evaluate equipment.

2. The company procuring new equipment should establish a procedure that rates the importance of the equipment. If the equipment is critical in nature, the company must make sure that adequate ride-through capability is included when the equipment is purchased. If the equipment is not important or does not cause major disruptions in manufacturing or jeopardize plant and personnel safety, voltage sag protection may not be justified.

3. Equipment should at least be able to ride through voltage sags with a minimum voltage of 70 percent (ITI curve). The relative probability of experiencing a voltage sag to 70 percent or less of nominal is much less than experiencing a sag to 90 percent or less of nominal. A more ideal ride- through capability for short-duration voltage sags would be 50 percent, as specified by the semiconductor industry in Standard SEMI F-47.17. As we entertain solutions at higher levels of available power, the solutions generally become more costly.

If the required ride-through cannot be obtained at the specification stage, it may be possible to apply an uninterruptible power supply (UPS) system or some other type of power conditioning to the machine control. This is applicable when the machines themselves can withstand the sag or interruption, but the controls would automatically shut them down. At level 3 in Fig. 2.11, some sort of backup power supply with the capability to support the load for a brief period is required. Level 4 represents alterations made to the utility power system to significantly reduce the number of sags and interruptions.