Short T.A. Electric Power Distribution Handbook

Подождите немного. Документ загружается.

Voltage Sags and Momentary Interruptions 537

All I know is that when shit hits the fan, management calls us we don’t call them.

In response to: Do linemen feel they are respected by management and

coworkers for the jobs they are doing, do management and coworkers

understand what you do?

www.powerlineman.com

1791_book.fm Page 537 Monday, August 4, 2003 3:20 PM

Other Power Quality Issues

While voltage sags and momentary interruptions cause the most widespread

power quality problems, several other power quality disturbances can dam-

age equipment, overheat equipment, disrupt processes, cause data loss, and

annoy and upset customers. In this chapter, we explore several of these,

including transients, harmonics, voltage ﬂicker, and unbalance.

11.1 Overvoltages and Customer Equipment Failures

Often, customers complain of equipment failures, especially following

power interruptions. Is it lightning? Voltage swells during faults? Some sort

of switching transient? Sometimes explanations are obvious, sometimes not.

Several events can fail equipment during a fault/interruption, either from

the disturbance that caused a fault, the voltage sag during the fault, a voltage

swell during the fault, or the inrush while the system is recovering. Some

possibilities are

•

Overvoltages

— Lightning and other system primary-side overvolt-

ages can enter the facility and damage equipment.

•

Grounding —

Poor facility grounding practices can introduce over-

voltages at equipment from fault current.

•

Capacitive coupling —

Reclose operations and other switching tran-

sients can create fast-rising voltage on the primary that capacitively

couples through the transformer, causing a short pulse on the

secondary.

•

Inrush current —

While recovering from a voltage sag or momentary

interruption, the inrush current into some electronic equipment can

blow fuses or fail semiconductor devices.

•

Unbalanced sags —

Three-phase electronic equipment like adjustable-

speed drives can draw excessive current during a single-phase sag

1791_book.fm Page 539 Monday, August 4, 2003 3:20 PM

540

Electric Power Distribution Handbook

or other unbalanced sag. The current can blow fuses or fail the front-

end power electronics.

•

Equipment aging

— Some equipment is prone to failure during turn

on, even without a voltage transient. The most obvious example is

an incandescent light bulb. Over time, the ﬁlament weakens, and

the bulb eventually fails, usually when turned on. At turn on, the

rapid temperature rise and mechanical stress from the inrush can

break the ﬁlament.

Lightning can cause severe overvoltages, both on the primary and on the

secondary. Damaging surges can enter from strikes to the primary, strikes

to the secondary, strikes to the facility, strikes to plumbing, and strikes to

cable-television or telephone wires. Poor grounding practices can make light-

ning-caused failures more likely.

Another source of severe overvoltages is primary or secondary conductors

contacting higher voltage lines. Other overvoltages are possible; normally

these are not severe enough to damage most equipment, except for sensitive

electronics:

•Voltage swells — Peaks at about 1.3 per unit on most distribution

circuits.

• Switching surges — Normally peaks at less than 2 per unit and

decays quickly.

• Ferroresonance — Normally peaks at less than 2 per unit.

Just as arresters on distribution lines are sensitive to overvoltages, arresters

inside of electronic equipment often are the ﬁrst thing to fail. The power

supply on most computers and other electronics contains small surge arrest-

ers (surge suppressors) that can fail quickly while trying to clamp down on

overvoltages, especially longer-duration overvoltages. These small suppres-

sors have limited energy absorption capability.

In addition to proper grounding, surge arresters are the primary defense

against lightning and other transients. For best protection, use surge protec-

tion at the service entrance and surge protection at each sensitive load.

Surge arresters work well against short-duration overvoltages — lightning

and switching transients. But arresters have trouble conducting temporary

power-frequency overvoltages; they absorb considerable energy trying to

clamp the overvoltage and can fail. Small arresters often are the ﬁrst com-

ponent to fail in equipment. Using a higher voltage rating helps give more

protection to the surge arrester during temporary overvoltages (for example,

end users should not use arresters with a maximum continuous operating

voltage below 150 V). Surge arresters should be coordinated; the large surge

arrester at the service entrance should have the lowest protective level of

all of the arresters within the facility. Because arresters are so nonlinear, the

unit with the lowest protective level will conduct almost all of the current.

1791_book.fm Page 540 Monday, August 4, 2003 3:20 PM

Other Power Quality Issues

541

So, we want the arrester with the most energy capability to absorb most of

the energy.

11.1.1 Secondary/Facility Grounding

Grounding problems within a facility can lead to equipment failures or

malfunctions. A common problem is ground loops. If the secondary neutral

has multiple connections within the facility, external ground currents from

lightning or from faults can induce voltages along the neutral. Figure 11.1

shows an example. The neutral voltage shifts can impose overvoltages on

sensitive equipment. For more information, see IEEE Std. 1100-1999 or IEEE

Std. 142-1991. Single-point grounding of the neutral breaks the loops and

reduces noise and possible overvoltages.

Grounding loops between multiple “ports” can create damaging scenarios.

Any signiﬁcant wired connection to equipment counts as a port — power,

telephone, cable TV, printer cables, and networking cables. Figure 11.2 shows

an example of a television where the cable enters via a different route and

has a different ground than its electric service. A fault or lightning strike on

the utility side can elevate the potential of the electric-supply ground relative

to the television cable ground. Even if both the electric and cable television

cables have independent surge protection, surges still create voltage differ-

ences between components. To avoid multiple-port difﬁculties, arrange to

have all power and communications enter a building at one location, provide

a common ground, and apply surge protection to all ports. To overcome this

problem, users can install a “surge reference equalizer” rather than indepen-

dent surge protection on each port (EPRI PEAC Solution No. 1, 1993). Both

the electric supply and the other port (telephone, cable TV, ethernet, etc.)

plug into the surge reference equalizer. The equalizer provides surge pro-

tection for both incoming ports and a single common grounding connection.

FIGURE 11.1

Ground loops within a facility.

Neutral

Sensitive load

120-V phase

Service

entrance

Grounding wire

Grounding loop Grounding loop

1791_book.fm Page 541 Monday, August 4, 2003 3:20 PM

542

Electric Power Distribution Handbook

For equipment, overvoltages between the phases and neutral are most

important. While it is commonly believed that neutral-to-ground overvolt-

ages cause problems for sensitive electronics, EPRI PEAC found that com-

puters are quite tolerant of neutral-to-ground overvoltages. In tests, neither

a continuous 50-V neutral-to-ground voltage nor a 3-kV, 100-kHz ring wave

upset the operation of a computer (EPRI PEAC Brief No. 21, 1994).

Pole-mounted controllers for regulators, reclosers, capacitors, and switches

face particularly severe environments. Because of their location, they have

signiﬁcant exposure to transients from lightning strikes to nearby poles.

Additionally, they are right in the path of fault current returning from faults

downstream of the controller. Some utilities have had reliability problems

with controllers; some problems may stem from poor powering and ground-

ing arrangements. Figure 11.3 shows how signiﬁcant voltages can develop

on the low-voltage supply when the power is supplied at a remote pole and

a fault occurs downstream. Lightning current following this same path can

create very severe voltages. Additionally, two-port vulnerabilities can arise

FIGURE 11.2

Example of a two-port voltage stress at a television between its power source and the cable TV

cable.

FIGURE 11.3

Pole-mounted controller exposure to faults along the circuit.

Cable TV

Fault or

lightning

current

Voltage rise

along the neutral

Neutral

Controller

120-V phase

Fault

1791_book.fm Page 542 Monday, August 4, 2003 3:20 PM

Other Power Quality Issues

543

between the controller power and potential transformers (PTs). Other sources

of two-port problems are current transformers and communication lines.

Consider the following strategies to better protect these controllers:

•

Surge protection

— On the power-supply inputs, apply extra low-

voltage surge protection, especially in high lightning areas.

•

Power source

— To help avoid two-port problems, try to power the

controller from a PT or small transformer on the same structure as

the controller.

11.1.2 Reclose Transients

One source of possibly damaging surges is from capacitive coupling through

transformers. The very short-duration pulses come from electrostatic cou-

pling through the transformer. Much analysis of this has been done for larger

power transformers at generating stations (Abetti et al., 1952). At high fre-

quencies, the transformer acts like a capacitor. A steep surge can pass from

the primary to the secondary. How much of the surge gets from the primary

to the secondary depends on the capacitances of the transformer and the

secondary load — not the transformer turns ratio. A transformer has capac-

itance from the high-voltage winding to the low-voltage winding (

) and

capacitance from the secondary winding to ground (

). As a ﬁrst approxi-

mation, the voltage on the secondary as a function of the voltage on the

primary is a capacitive voltage divider (Greenwood, 1991):

Line energization and capacitor switching and other transients can create

surges that pass through the transformer. It does not even have to be an

overvoltage on the primary, just a fast rise to the nominal peak — a normal

reclose operation — can create a surge on the secondary. When a line is

energized near its peak voltage, right when the switch engages, a traveling

wave with a very steep front rushes down the line. At a transformer, this front

can be steep enough to couple capacitively through to the secondary at volt-

ages much higher than the turns ratio of the transformer. Figure 11.4 shows

a 1-

sec wide transient to just over 2000 V following a reclose operation.

These capacitively coupled surges are worse with

• Higher voltage transformations from the primary to the secondary

(34.5 kV to 480 V is worse than 12.5 kV to 480 V).

• Loads close to the substation or recloser (the wave front ﬂattens out

with distance).

• Minimal resistive load on the secondary.

1791_book.fm Page 543 Monday, August 4, 2003 3:20 PM

544

Electric Power Distribution Handbook

We can protect against these surges with:

• Line-to-ground capacitor banks in the facility (either power-factor

correction or surge capacitors)

• Surge arresters

11.2 Switching Surges

Transients are triggered from capacitor switching, from line energization,

and from faults. Capacitor switching transients normally cause the highest

peak magnitudes. If a capacitor is switched just when the system voltage is

FIGURE 11.4

Measurements of a transient captured during a circuit reclose. (Recordings courtesy of Francisco

Ferrandis Mauriz, Iberdrola Distribucion Electrica.)

Millisecond-scale view of the transient:

Microsecond-scale view:

1791_book.fm Page 544 Monday, August 4, 2003 3:20 PM

Other Power Quality Issues

545

near its peak, the capacitor pulls the system voltage down (as current rushes

into the capacitor to charge it up). The system rebounds; the voltage over-

shoots and oscillates about the fundamental-frequency waveform (theoreti-

cally rising to two per unit, but normally less than that). This transient

normally decays quickly. The oscillation occurs at the natural resonance

frequency between the capacitor bank and the system, usually in the hun-

dreds of hertz:

where

= frequency, Hz

= capacitance, F

= system inductance up to the capacitor bank at nominal frequency,

= line-to-ground impedance of one phase of the capacitor bank at

nominal frequency

= system impedance up to the capacitor bank at nominal frequency

MVA

= three-phase short-circuit MVA at the point where the capacitor is

applied

Mvar

= three-phase Mvar rating of the capacitor

Capacitor switching transients tend to be more severe when the capacitor

is at a weaker point on the system. Other capacitors on the circuit normally

help, but if two capacitors are very close together, switching in the second

capacitor can create a higher voltage transient. Capacitors with signiﬁcant

separation can magnify the switching surges.

Normally, switching surges are not particularly severe on distribution

systems. The voltages decay quickly, and magnitudes are normally not

severe enough to fail line equipment (Figure 11.5 shows a typical example).

Switching transients can be large enough to affect sensitive end-use loads,

particularly adjustable-speed drives and uninterruptible-power supplies.

The oscillation frequency is in the hundreds of hertz, low enough for the

transient to pass right through distribution transformers and into custom-

ers’ facilities.

EPRI’s distribution power quality (DPQ) study found regular but mild

transient overvoltages, most presumably due to switching operations. Tran-

sients measured on the distribution primary above 1.6 per unit are rare,

averaging less than two per year (where per unit is relative to the peak of

the nominal sinusoidal voltage wave). Figure 11.6 shows the average distri-

bution of transients measured during the study (EPRI TR-106294-V2, 1996).

This graph shows occurrences of transients with a peak magnitude between

1.05 and 1.9 per unit with a principal oscillation frequency between 240 and

3000 Hz (excluding transients associated with faults).

MVA

Mvar

===

60 60

1791_book.fm Page 545 Monday, August 4, 2003 3:20 PM

546

Electric Power Distribution Handbook

During EPRI’s DPQ study, most of the transients measured occurred dur-

ing the morning hours, from 5 a.m. to 10 a.m., when most switched capacitors

are coming on line. The wide majority of oscillating transients measured in

the study had dominant frequencies between 250 and 800 Hz.

AEP staged various switching tests on a 12.47-kV circuit and found that

most capacitor switching and line energization transients were less than two

per unit on the primary (Kolcio et al., 1992). In their tests, switching a 450-

kvar capacitor bank produced the highest transient, just under 2.3 per unit

on closing — vacuum switches produced slightly higher transients than oil

switches. Line energizations caused peak primary voltage of just under 1.9

per unit. The transients decayed in less than 5 msec.

Switching transients are normally more problematic on higher voltage

distribution circuits, such as 34.5 kV. Several instances of equipment failures

FIGURE 11.5

V2.

An Assessment of Distribution System Power Quality: Volume 2: Statistical Summary Report.

Reprinted with permission.)

FIGURE 11.6

Average magnitudes of voltage transients measured on the distribution primary in EPRI’s DPQ

study. (Data from [EPRI TR-106294-V2, 1996; Sabin et al., 1999b].)

1.2 1.4 1.6 1.8

Peak voltage transient, per unit

Average number of events per year

exceeding the x-axis value

1791_book.fm Page 546 Monday, August 4, 2003 3:20 PM

Other Power Quality Issues

547

have been reported on these higher voltage systems, usually from switching

a large (10 plus Mvar) capacitor bank (Shankland et al., 1990; Troedsson et

al., 1983). Capacitors tend to be larger, and sources are weaker. Plus, at higher

voltages, primary level insulation capability is lower on a relative basis

compared to lower voltage systems.

In capacitive circuits, switch

restrike

prestrike

generates more severe

transients. A restrike can occur when switching a capacitor off; the switch

opens at a current zero, trapping the peak voltage on the capacitor (see Figure

11.7). As the system voltage wave decreases from the peak, the voltage across

the switch rapidly increases. By the time the system voltage reaches the

opposite system peak, the voltage across the switch is 2 per unit (this extra

voltage is the main reason that switches have a different rating for breaking

capacitive circuits). If the switch restrikes because of this overvoltage, the

voltage swings about the new voltage with a peak-to-peak voltage of 4 per

unit, forcing the line-to-ground voltage to a peak of 3 per unit (theoretically).

The interrupter can clear at another current zero and trap even more voltage

on the capacitor, which can cause another more severe restrike; such multiple

restrikes can escalate the voltage. Fortunately, such voltage escalation is rare.

Restrikes are most likely if the interrupting contacts have not fully separated

when the current ﬁrst interrupts. Restrikes can fail switches and cause higher

voltages on the system. Most switching technologies, including vacuum, can

restrike under some conditions.

Prestriking occurs when a switch closes into a capacitor. If the gap between

a switch’s contacts breaks down before the contacts have fully closed, the

system voltage charges up the capacitor. Because the contacts are not closed,

the switch can clear at a current zero, leaving the capacitor charged. We are

FIGURE 11.7

Restrike of a capacitor bank.

Switch opens

Switch restrikes

Vo ltage trapped

on the capacitor

1791_book.fm Page 547 Monday, August 4, 2003 3:20 PM