Voltage sags and what to do about them

Voltage sags are the most common events that affect power quality. They are also the most costly. Equipment used in modern industrial plants, such as process controllers, PLCs, adjustable speed drives, and robots, becomes more sensitive to voltage sags as the complexity of the equipment increases. Relays and contactors in motor starters are sensitive to voltage sags, resulting in downtime when ...

By Jack Smith, Senior Editor, Plant Engineering Magazine August 8, 2002

Key Concepts

  • Consider the entire plant when addressing power quality problems.
  • Power quality problems occur on both sides of the electric meter.
  • At full load, a constant voltage transformer can be more than 90% efficient.

Voltage sags are the most common events that affect power quality. They are also the most costly.

Equipment used in modern industrial plants, such as process controllers, PLCs, adjustable speed drives, and robots, becomes more sensitive to voltage sags as the complexity of the equipment increases. Relays and contactors in motor starters are sensitive to voltage sags, resulting in downtime when they drop out.

What are voltage sags?

Voltage sag, as defined by IEEE, is a reduction in voltage for a short time. The voltage reduction magnitude is between 10% and 90% of the normal root mean square (RMS) voltage at 60 Hz. The duration of a voltage sag event, by definition, is less than 1 min and more than 8 msec, or a half cycle of 60-Hz electrical power.

RMS voltage variations include interruption, swell, and sag (Fig. 1.). An interruption is a complete loss of voltage, or a drop to less than 10% of nominal voltage in one or more phases. Interruptions can be momentary (8 msec to 3 sec), temporary (3 sec to 1 min), or long duration (greater than 1 min).

Fig. 1. RMS voltage variations include interruption, swell, and sag. An interruption is a complete loss of voltage; a swell is voltage in excess of 110% of nominal; and sag is a momentary reduction in voltage.

When a voltage swell occurs, the RMS voltage exceeds 110% of nominal for less than 1 min. RMS voltage exceeding 110% of nominal for longer than 1 min is called a long-duration overvoltage.

Other power quality issues involving voltage amplitude and duration include undervoltage (similar to sags, but lasting more than 1 min), transients, voltage unbalance among phases, voltage fluctuations (flicker), harmonics, and electrical noise (see sidebar, “Power quality problems and solutions”). Although they are important power quality issues, only voltage sags are addressed in this article. However, it should be noted that power quality issues are interrelated. Address power quality problems from an entire plant approach. Sometimes fixing one problem makes another worse. Looking at the big picture enables you to doctor the cause and not just the symptom.

What causes voltage sags?

When voltage sags occur, the local utility usually is blamed. However, many voltage sag events are caused by equipment within the plant. Power quality problems occur on both sides of the electric meter.

It is sometimes difficult to determine the source of voltage sags or other power quality problems. Even experienced power quality experts do not always agree.

Voltage sags that occur on the utility side of the power meter can either be human created or natural events. The most common human-created events are switching operations. The most common natural events include:

  • Lightning
  • Wind
  • Trees falling onto power lines
  • Construction workers digging into buried cables
  • Squirrels and rodents
  • Equipment failures
  • Traffic accidents.

A voltage sag on the power grid can affect users in greater than a 100-mile radius from the causing event.

There is potential for a difference in individual phase voltages and an associated phase-angle shift during a voltage sag. There is also a nonsinusoidal characteristic to the voltage waveform during the sag. Many sags are caused by a single-line-to-ground fault (SLGF). Double and three-phase symmetrical faults are infrequent, occurring less than 20% of the time. Asymmetrical faults cause voltage unbalance and phase shift.

A large motor starting within the plant can also cause voltage sags. However, these usually have a balanced effect on all three phases. The electrical arc furnace is another culprit of voltage sags internal to those industrial plants that process metal. Sags produced by arc furnaces tend to be unbalanced. Workers within the plant also cause internal sags and unbalances when they shut off certain legs.

When a sag occurs, the power supply inside electronic devices uses some of its stored energy to compensate for the loss of input voltage. If enough energy is lost due to the sag, then the power supply may lose its ability to maintain an exact DC voltage to all the active components, such as integrated circuits, inside the device — even for a few milliseconds. This is long enough to corrupt data in microprocessor-based electronics and to cause malfunctions of digital equipment.

Typically, ac adjustable frequency drives are touted to have good voltage sag ride-through capability. However, closer examination may indicate the contrary. The phase shift associated with voltage-sag-induced unbalance has a direct effect on ac drive response because the drive’s diode bridge responds to the maximum difference between any two of the line voltages. Whether voltage magnitude or phase angle, the momentary unbalance in supply voltage during sags may lead to faulty operation of the drive or its protective devices because of excessive current unbalance in the line side of the drive.

Detecting voltage sags

A power quality monitor can detect voltage sags, swells, interruptions, and other power quality anomalies. It measures power as it enters a facility and compares it to currently accepted standards, such as ITIC (formerly CBEMA).

Fig. 2. Web-enabled power monitoring systems provide system health status by indicating total harmonic distortion (THD), voltage, power factor, current demand, voltage unbalance, voltage sag, voltage swell, and alarms for multiple locations. (Courtesy of SquareD/Schneider Electric)

Some currently available power quality monitors or power analyzers are web-enabled, allowing plant engineers to monitor multiple plant sites. This capability is especially valuable if there is great distance between sites. For example, a company located in central Tennessee can monitor its facilities in Mexico, Canada, and multiple U.S. locations from one browser.

A web-based monitor can log electrical parameters, such as voltage, current, power factor, harmonics, and sag (Fig. 2.). They provide alarms when these and other parameters exceed predetermined values. Functionality within the software allows plant engineers to drill down to the offending site, or even the offending equipment, depending on system configuration. It is also possible to capture waveforms of the offending event (Fig. 3.).

Fig. 3. Web-enabled power monitoring systems can also provide waveform captures of offending events. The voltage trace indicates a phase B to neutral voltage sag. ( Courtesy of SquareD/Schneider Electric)

Simulation devices

When faults that produce sag, undervoltage, and interruption events occur outside the plant, usually the utility bears the responsibility. Contacting the utility regarding power quality problems is in order. However, when these faults occur within the plant, it’s up to plant personnel to determine the culprit.

When offending equipment, such as large motors, affect most of the plant, the situation must be addressed. Providing more power, lowering voltage drop, and employing soft-starting techniques may help alleviate this problem.

When an isolated machine or two trips out or locks up randomly, it could be that this piece of equipment is a little more sensitive to sags than companion equipment. How do you find the cause when this is the problem?

Fig. 4. Voltage sag generators are used to simulate voltage sags. They feature precise timing and phase angle control, fine amplitude resolution, and point-on-wave selection throughout 360 deg of the waveform in 1-deg increments. (Courtesy of EPRI-PEAC)

Simulating voltage sags, interruptions, or undervoltage can be accomplished using a form of sag generator. A voltage sag generator is used to simulate voltage sags (Fig. 4.). There is a little more to it than just lowering the supply voltage until a machine fails. Typically, voltage sag generation devices feature precise timing and phase angle control, enabling the simulation of sags with durations from 0.25 cycles to 3 sec. They also offer fine amplitude resolution and point-on-wave selection throughout 360 deg of the waveform in 1-deg increments. The sag generation device is recommended for quickly and efficiently assessing the sensitivity of a wide variety of industrial devices, including PLCs, motion controllers, sensors, relays, emergency shutdown systems, motor drives, and other single-phase loads, both linear and nonlinear, up to 200 A.

Voltage sag reduction

It is important to isolate individual pieces of equipment that seem to be more sensitive to voltage sags and to determine where to deploy sag correction devices.

Location can play a role in a machine’s voltage sag sensitivity. Also, wiring is sometimes to blame, and increasing wire size can reduce voltage drop. Sag correction devices can be applied at various locations, including the control panel, machine level, bus level, or even at the plant service entrance. The percent of sensitive loads within the plant, ease of installation, downtime cost, solution cost, and return on investment are all factors in determining the best fit. Three phase solutions are often necessary to address larger loads.

Power line conditioning technologies include the uninterruptible power supply (UPS), the constant voltage transformer (CVT), and solid-state sag correction devices.

A UPS protects equipment from voltage sags, momentary power loss, and extended power outages for up to several minutes. When the UPS circuitry senses a voltage sag, it transfers the protected load to a battery-based inverter. The UPS supplies power as long as the battery or batteries have stored energy, which can range typically from 3 to 20 minutes.

UPSs are usually designed for constant loads, such as computers and other electronic equipment. Their performance with dynamic loads, such as large motor starting, is limited. Some UPS designs have an output that approaches a square wave, which is rich with odd-order harmonics and may not be appropriate for factory automation equipment.

Fig. 5. The CVT uses ferroresonant technology to reduce the effect of voltage sags to individual pieces of equipment. The tank circuit provides a constant, clean output voltage that shuns most types of power disturbances. (Courtesy of Sola/Hevi-Duty)

The CVT uses ferroresonant technology to provide reliable power protection in a single unit (Fig. 5). The technology is not new, but it is still widely used. All CVTs incorporate the same basic design — a saturating transformer preloaded with a resonant tank circuit (Fig. 6).

Fig. 6. Constant-voltage transformers incorporate a saturating transformer preloaded with a resonant tank circuit such as the one shown in this schematic. (Courtesy of Sola/Hevi-Duty)

The power reserve comes from the tank circuit created by the inductive reactance of the transformer and the internal capacitor. The result is constant, clean output voltage that shuns most types of power disturbances. Since the transformer is always in saturation, fluctuations of input voltage have little effect on the output voltage. For light loads, the input can drop below 60% of nominal voltage continuously with less than a 10% fluctuation in output (Fig. 7).

Fig. 7. For light loads, the input to a constant-voltage transformer can drop below 60% of nominal voltage continuously with less than a 10% fluctuation in its output. At full load, it can be more than 90% efficient. (Courtesy of Sola/Hevi-Duty)

The down sides are efficiency (heat), size, weight, and availability in limited size ranges. In addition CVTs have difficulty in handling dynamic and harmonic rich loads often requiring significant oversizing. Oversizing provides better performance and sag correction, but with a penalty of less efficiency, size, weight, and cost.

Apply CVTs directly between the supply power and each piece of equipment that is determined to be most sensitive to voltage sags — probably the PLCs, PC-based controllers, and dedicated controls that make wide use of microprocessor or digital technology. CVTs that provide step-down voltage are also available. This eliminates the need to buy an extra step-down transformer.

Solid-state sag correction devices use microprocessor-based technology to monitor incoming power quality, which allows them to be activated within1/ 8 th of a cycle. Some devices use capacitors to provide power, and other devices draw additional current from the utility during the sag and convert this extra current to the missing voltage. The result is very deep sag correction, and even the ability to cover short outages in the 12 cycle range often resulting from utility breaker reclosure operations.

Voltage sags are prevalent power quality problems. They can be detected and corrected in a number of ways. Regardless of the technologies employed, providing appropriate power quality solutions requires a full understanding of the problems.

PLANT ENGINEERING magazine extends its appreciation to Eaton Corp., Eaton | Cutler-Hammer, EPRI PEAC, SoftSwitching Technologies, Sola/Hevi-Duty, and Square D/Schneider Electric for the use of their materials in the preparation of this article.

Power quality: A relationship-based challenge

Discussions about power quality (PQ) often focus exclusively on technology. While understanding technology is important, understanding the relationships behind PQ is a useful step in assuaging its impact on operating facilities, and for minimizing the cost of solving problems.

Relationship #1: Plant equipment and power both define power quality

The fundamental definition of power quality is the ability of the electric power supply to meet the needs of the load. Interestingly, most PQ problems aren’t caused by a change in the power coming from the utility, but by the addition of a new, more-sensitive device to an existing plant. Power that was “just fine” before, is now suddenly of inadequate quality. Power quality problems are brought about by both power and the load. The optimum solution will emerge from balancing both.

Relationship #2: Electric power is a raw material

Manufacturers are intimately familiar with the relationship between the quality of raw materials and that of the finished product—if a higher-quality finished product is desired, then increasing the quality of essential raw materials is usually a good first step. To improve productivity and product quality, many facilities have employed sophisticated electronic controls in their processes, causing power quality problems to surface. The relationship between electric power quality and product quality is no different, and improving its quality can become as essential a step as with any other raw input.

Relationship #3: Knowledge, not technology, solves PQ problems

Keeping a person healthy can’t be done reliably with simple quick-fix measures, such as merely taking a pill. Good health usually requires an ongoing process of maintenance, measurement, correction and, if necessary, dramatic intervention. Solving power quality problems requires a similar understanding of the broad relationship between electric power and a facility’s devices and processes. Many PQ problems that have broad impact on a facility (for example, shutting down an entire manufacturing line) are caused by misoperation of a single, small component, such as a relay, sensor, or programmable controller. Simply protecting or replacing the sensitive device may be all that is required. Regardless, the solution comes from intimate and fundamental knowledge of how the plant’s systems work.

Relationship #4: Solving PQ problems requires a relationship of trust

A common complaint among PQ mitigation vendors is that facilities won’t pay for PQ solutions. Why would a plant that is experiencing, say, losses of $100,000 per year in scrap, misdirected labor, and lost production not be willing to spend even half that amount to prevent these losses? When PQ problems strike, facility managers first turn to resources with which they have had long, stable relationships, including local electricians or trusted familiar vendors. They tend to shy away from unfamiliar providers even if they can offer well-designed solutions. Exploring PQ options and resource providers, such as the local utility and specialty consultants before problems strike can help build such relationships of trust.

Contributed by Bill Howe, PE, Director of Technology Information Businesses, EPRI PEAC Corp.