1

Voltage Flicker Primer

Submitted to the IEEE SCC 21 P1547 Working GroupJune 17, 1999

Jeff W. SmithElectrotek Concepts, Inc.

Knoxville, TNjeffsmith@electrotek.com

IntroductionFlicker is a relatively old subject that has recently gained considerable attention due to theincreased awareness of issues concerning power quality. Power engineers first dealt with flickerin the 1880’s when the decision of using ac over dc was of concern [1]. Low frequency acvoltages resulted in a “flickering” of the lights. To avoid this problem, a higher 60 Hz frequencywas chosen as the standard frequency in North America.

What is Flicker?The term flicker is sometimes considered synonymous with voltage fluctuations, voltage flicker,light flicker, or lamp flicker. The phenomena being referred to can be defined as a fluctuation insystem voltage that can result in observable changes (flickering) in light output. Because voltageflicker is mostly a problem when the human eye observes it, usually it is considered to be aproblem of perception. There are, however, rare cases where voltage flicker can affectequipment operation such as electric drives and UPS systems [1].

Voltage flicker can be separated into two types: cyclic and noncyclic. Cyclic flicker is a result ofperiodic voltage fluctuations in the system voltage, with noncyclic referring to occasional voltagefluctuations. An example of sinusoidal-cyclic flicker is shown in Figure 1.

-200

-150

-100

-50

0

50

100

150

200

0.0

00

0.0

58

0.1

17

0.1

75

0.2

33

0.2

92

0.3

50

0.4

08

0.4

67

0.5

25

0.5

83

0.6

42

0.7

00

0.7

58

0.8

17

0.8

75

0.9

33

Time (s)

Vol

tag

e (V

)

Figure 1. Example Graph of Voltage Flicker

As shown in Figure 1, flicker is simply amplitude modulation where the main signal (60Hz forNorth America) is the carrier signal and flicker is the modulating signal. The usual method forexpressing voltage flicker is similar to that of amplitude modulation. Voltage flicker is usuallyexpressed as a percent of the total change in voltage with respect to the average voltage (∆V/V)over a specified time interval.

In the early 1900’s, many studies were done on humans to determine the levels of voltage flickerthat were both observable and objectionable. Many curves, such as the one shown in Figure 2,were developed by various companies to determine the severity of voltage flicker. The flickercurve shown in Figure 2 was developed by C.P. Xenis and W. Perine in 1937, and was basedupon data obtained from 21 groups of observers. In order to account for the nature of flicker, theobservers were exposed to various waveshape voltage variations, levels of illumination, andtypes of lighting [2].

Vo

ltag

e C

han

ge

(in

Vo

lts)

on

120

V S

yste

m

Figure 2. General Flicker Curve

The frequency content of voltage flicker is extremely important in determining whether or notflicker levels are observable (or objectionable). The typical frequency range of observable flickeris from 0.5Hz to 30.0 Hz, with observable magnitudes starting at less than 1.0%. As shown inFigure 2, the most sensitive frequency range for voltage flicker is approximately 5-10 Hz. Inessence, this means that the human eye is more susceptible to voltage fluctuations in this 5-10 Hzrange. As the frequency of flicker increases or decreases away from this range, the human eyegenerally becomes more tolerant to luminance fluctuations.

One issue that was not considered in the development of Figure 2 is that of multiple flickersignals. Generally, most flicker producing loads contain multiple voltage flicker signals (ofvarious magnitudes and frequencies), thus making it more difficult to use flicker curves inevaluating the levels of voltage flicker.

SourcesTypically speaking, voltage flicker occurs on relatively weak systems with a low short-circuitratio. This, in combination with considerable variations in line current over a short period oftime, results in voltage flicker. As the load increases, the current in the line increases, thusincreasing the voltage drop across line. This phenomenon results in a sudden reduction in busvoltage. Depending upon the change in magnitude of voltage and frequency of occurrence, thiscould result in observable amounts of voltage flicker. If a lighting load were connected to thesystem in relatively close proximity to the fluctuating load, observers would see this as adimming (or flickering) of the lights. A common situation which could result in voltage flickerwould be a large plant located at the end of a weak distribution feeder.

Whether the resulting voltage fluctuations cause observable flicker is dependent upon thefollowing parameters:

! Size of flicker producing source! System impedance (stiffness of utility)! Frequency of resulting voltage fluctuations

A common load which often causes voltage flicker would be an electric arc furnace (EAF) orwelder. EAFs are non-linear, time-varying loads, which often cause large voltage fluctuationsand harmonic distortion. Large induction machines undergoing startup or widely varying torquechanges are also known to produce voltage fluctuations on systems. As shown in Figure 3, as aninduction machine is started up, most of the power drawn by the motor is reactive. This resultsin a large voltage drop across distribution lines. Figure 4 demonstrates the affect on voltage ofinduction motor starting and torque variations.

1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0

Slip

Rea

l Pow

er

Rea

ctiv

e P

ower

Q

P

Figure 3. Real and Reactive Power During Induction Machine Startup

Figure 4. Voltage Fluctuations Caused by an Induction Machine

Although starting large induction machines across-the-line is not considered to be arecommended practice, it does however occasionally occur. Generally, motors are brought up tospeed using reduced voltage starters or variable speed drives.

InterharmonicsIn certain circumstances, superimposed interharmonics in the supply voltage can lead tooscillating luminous flux and cause light flicker. Voltage interharmonics are components in theharmonic spectrum that are non-integer multiples of the fundamental frequency. Thisphenomenon can be observed with incandescent lamps as well as with fluorescent lamps.

Sources of interharmonics include static frequency converters, cycloconverters, sub-synchronousconverter cascades, induction machines, and arc furnaces [3].

Present Evaluation and Measurement TechniquesRMS Strip ChartsHistorically, voltage flicker has been measured using rms meters, load duty cycle, and a flickercurve. If sudden rms voltage deviations occurred with specified frequencies exceeding valuesfound in flicker curves, such as the one shown in Figure 2, the system was said to haveexperienced voltage flicker. An example graph of rms voltage variations is shown in Figure 5,where large voltage deviations up 9.0V rms (∆V/V = ± 8.0% on 120V base) are found. Uponcomparing this to the flicker curve in Figure 2, the feeder would be experiencing voltage flicker,regardless of the duty cycle of the load producing the voltage flicker because any sudden totalchange in voltage greater than 7.0V rms results in objectionable flicker, regardless of thefrequency. The advantage to such a method is that it is quite simple in nature and the rms datarequired is rather easy to acquire. The most obvious disadvantage to such a method would be thelack of accuracy and inability to obtain the exact frequency content of the voltage flicker.

Figure 5. RMS Voltage Variations

Flicker MetersThe most accurate method for measuring flicker is to use what are known as flicker meters.Generally, these meters can be divided up into three sections. In the first section the inputwaveform is demodulated, thus removing the carrier signal. As a result of the demodulator, thedc offset and higher frequency terms (sidebands) are produced. The second section removesthese unwanted terms using filters, thus leaving only the modulating (flicker) signal remaining.The second section also consists of filters that weight the modulating signal according to theparticular meters specifications. The last section usually consists of a statistical analysis of themeasured flicker.

US Flicker StandardsCurrently, a standard method for measuring voltage flicker does not exist in the United States. InEurope, however, the International Electrotechnical Commission (IEC) has developed standardsfor measuring voltage flicker. The IEEE Task Force on Light Flicker is presently consideringmodifications necessary to the IEC standards in order for it to be used in the United States.

Although the US does not currently have a standard for flicker measurement, there are IEEEstandards that address flicker. IEEE Stds 141-1993 and 519-1992 both contain flicker curves thathave been used as guides for utilities to evaluate the severity of flicker within their system. Bothflicker curves, Stds 141 [4] and 519 [5] are shown in Figure 6.

Figure 6. IEEE Stds 141 and 519 Flicker Curves

The IEEE Task Force on Flicker (P1453) is currently working on a Recommended Practice formeasuring voltage flicker and in February of 1998, the Task Force voted unanimously to embraceand enhance the IEC Flickermeter measurement approach (discussed in next section). However,in order for the IEC flicker meter to be adopted by IEEE, various modifications must be madesuch as adjusting the weighting filters for 120V/60Hz lamps (instead of 230V/50Hz).

IEC Flicker StandardsThe IEC method for flicker measurement is defined in the IEC Std. 1000-4-15 (formerly the IEC868 [6]). The IEC flickermeter is the standard for measuring voltage flicker in Europe and othercountries currently adopting IEC standards. The IEC flicker meter consists of five blocks whichare shown in Figure 7.

Block 1

Detector andgain control

Signal generatorfor calibration

checking

Input voltageadaptor

Block 2

Demodulatorwith

squaringmultiplier

Block 4

Squaringmultiplier

1st

orderslidingmeanfilter

Squaring andsmoothing

Block 5

A/Dconverter

Samplingrate

≥ 50 Hz

64 Levelclassifier

Outputinterfaces

Programming of short and longobservation periods

Statistical evaluation of flicker level

InputTransformer

R.M.S.meter

Output 1half cycle r.m.s. voltageindication

Block 3

.05 35 Hz

dB

-31

0 8.8 Hz

Weighting filters

Rangeselector

%∆V

V

0.51.02.05.010.020.0

Output 2weightedvoltagefluctuation

Square rooter

∫

1 i i t t

Output 3rangeselection

Output 4short timeintegration

Output 5recording

Outputand datadisplayandrecording

Figure 7. Diagram of the IEC FlickerMeter

Block 1 is an input voltage adapter that scales the input half-cycle rms value to an internalreference level. This allows flicker measurements to be made based upon a percent ratio ratherthan dependent upon the input carrier voltage level.

Block 2 is simply a squaring demodulator which squares the input to separate the voltagefluctuation (modulating signal) from the main voltage signal (carrier signal), thus simulating thebehavior of the incandescent lamp.

Block 3 consists of multiple filters that serve to filter out unwanted frequencies produced fromthe demodulator and also to weight the input signal according to the incandescent lamp eye-brainresponse. The basic transfer function for the weighting filter is shown in Eq. 1.

( )( )43

221

21

/1/1

/1

2)(

ωωω

ωλω

ss

s

ss

sksH

+++⋅

++= (1)

Block 4 consists of a squaring multiplier and sliding mean filter. The voltage signal is squared tosimulate the non-linear eye-brain response, while the sliding mean filter averages the signal tosimulate the short-term storage effect of the brain. The output of this block is considered to bethe instantaneous flicker level. A level of one on the output of this block corresponds toperceptible flicker.

Block 5 consists of a statistical analysis of the instantaneous flicker level. The output of block 4is divided into suitable classes, thus creating a histogram. A probability density function (PDF)is created based upon each class and from this a cumulative distribution function (CDF) can beformed.

Flicker level evaluation can be divided into two categories, short-term and long-term. Short-termevaluation of flicker severity, PST, is based upon an observation period of 10 minutes. Thisperiod is based upon assessing disturbances with a short duty-cycle or those that producecontinuous fluctuations. PST can be found using Eq. 2.

ssssST PPPPPP 5010311.0 08.028.00657.00525.00314.0 ++++= (2)

where the percentages P0.1, P1s, P3s, P10s, P50s are the flicker levels that are exceeded 0.1, 1.0, 3.0,10.0, and 50.0 percent of the time. These values are taken from the cumulative distribution curvediscussed previously. A PST of 1.0 on the output of block 5 represents the objectionable (orirritable) limit of flicker.

For cases where the duty cycle is long or variable, such as in arc furnaces, or disturbances on thesystem that are caused by multiple loads operating simultaneously, the need for the long-termassessment of flicker severity arises. Therefore, the long-term flicker severity, PLT, is derivedfrom PST using Eq. 3.

31

3

N

PP

N

iSTi

LT

∑== (3)

where N is the number of PST readings. The number of PST readings (N) is determined by theduty cycle of the flicker producing load. The purpose is to capture one duty cycle of thefluctuating load. If the duty cycle is unknown, the recommended number of PST readings is 12(two hours of measuring).

ConclusionThis primer has served as an introduction to flicker in general, with an emphasis on flickermeasurement. As evident in this text, flicker can be an extremely difficult power quality problemto quantify and previous techniques for evaluating and measuring flicker have proven toinadequate. Therefore, it is recommended that the SCC 21 P1547 Working Group follow asimilar approach to that of the IEEE Task Force on Flicker (P1453) by considering the use of therecently adopted IEC method for flicker measurement.

References Cited

[1] Owen, Edward L., “Power Disturbance and Power Quality – Light Flicker VoltageRequirements”, IEEE IAS Annual Meeting, pp. 2303-2309, Denver, Colorado, October1994.

[2] Xenis, C. P., Perine, W., “Slide Rule Yields Lamp Flicker Data.” Electrical World, Oct.23, 1937.

[3] Interharmonics in Power Systems, IEEE Interharmonic Task Force, Cigre 36.05/CIRED 2CC02 Voltage Quality Working Group.

[4] IEEE Standard 141-1993: Recommended Practice for Power Distribution in IndustrialPlants, IEEE, 1993.

[5] IEEE Standard 519-1992: Recommended Practices and Requirements for HarmonicControl in Electrical Power Systems, IEEE, 1993.

[6] IEC Publication 868, "Flickermeter: Functional and Design Specifications," CEI, 1986.