VOLTAGE DIPS AND SENSITIVITY OF CONSUMERS IN LOW
VOLTAGE NETWORKS
Günther Brauner
Vienna University of Technology
Summary
In distribution systems power quality is of
increasing importance as the low voltage
consumers use microelectronic components for
control and operation, which are sensitive to
voltage dips and power interruptions. In an
investigation
of
VEÖ
(Verband
der
Elektrizitäswerke Österreichs, Union of Austrian
Utilities) the functional sensitivity of low voltage
consumers on voltage dips investigated. Short
circuit faults and large load variations may cause
voltage dips. After system faults not only voltage
dips but also interruptions or low voltage situations
of longer duration are possible in electrical
distribution systems. These phenomena are purely
random in nature and their amplitude and duration
varies in a significant range.
Point of malfunction
∆U/U
Area of Malfunction
time
Figure 1: Functional area of low voltage
consumers
Measurements of the immunity of low voltage
devices to voltage depressions of different
amplitude and duration have shown, that the area of
malfunction can be described in many cases by a
single point, which represents the minimum voltage
needed for continuous operation and the maximum
permissible duration of a voltage dip (fig. 1). If the
voltage falls below the minimum for steady state
operation or exceeds the permissible duration of a
voltage dip, a malfunction will occur.
It was found, that the allowed duration of a short
interruption for personal computers is between
80 ms to 450 ms with an accumulation around a
value of about 200 ms. The respective values for
video recorders are between 50 ms and 200 ms. Gas
discharge lamps are able to withstand short
interruptions with durations of 10 ms to 100 ms.
For an easy comparison with the above immunity
values, the typical duration and amplitude of
voltage depressions in low voltage supply networks
Christian Hennerbichler
Vienna University of Technology
now: ABB Automation Ltd, Baden, CH
can be found in fig. 1. Faults in low voltage
networks have a clearing time between
approximately 5 ms if fuses are used and up to
150 ms for circuit breakers. Short interruptions by
auto-reclosing lead to zero voltage in the range of
300 ms to 500 ms in medium voltage networks. In
high voltage systems the duration of supply
interruption by auto-reclosing is up to 750 ms. But
as these networks have a redundant structure, in
most cases no interruption of energy flow can be
observed in underlying low voltage networks but
only a voltage dip.
To improve the immunity of devices against
voltage dips in principle two measures are possible:
at the consumer side and at the network side.
At the consumer side the power electronic can be
redesigned to withstand voltage dips and voltage
deviations.
An improvement of the ride through capabilities of
consumers under voltage dips can be achieved, if
the smoothing capacitor is large enough for a ride
through of minimum about 300 ms. If the switched
mode converter following the rectifier and the
smoothing capacitor are is designed for low steady
state voltage situations e.g. for 50% of nominal
voltage, in many cases malfunctions are excludes.
Such measures only lead to small additional costs
on the equipment side but helps to reduce the
number of malfunctions significantly.
On the network side the costs for redundant
structures can be very high, especially in rural areas
with low consumer density. In the liberalized
electricity market the utilities are constrained to
save costs. So the tendency to rebuild a distribution
network to reduce the number of voltage dips will
in future decrease.
VOLTAGE DIPS AND SENSITIVITY OF CONSUMERS IN LOW VOLTAGE
NETWORKS
Christian HENNERBICHLER
Vienna University of Technology
now ABB Power Automation Ltd, Baden
Günther BRAUNER
Vienna University of Technology
Abstract – In distribution systems power quality is of
increasing importance as the low voltage consumers
use microelectronic components for control and
operation, which are sensitive to voltage dips and
power interruptions. In an investigation of VEÖ
(Verband der Elektrizitäswerke Österreichs, Union
of Austrian Utilities) the functional sensitivity of low
voltage consumers on voltage dips investigated [1].
Short circuit faults and large load variations may
cause voltage dips. After system faults not only
voltage dips but also interruptions of longer duration
are possible in electrical distribution systems. These
phenomena are purely random in nature and their
amplitude and duration varies in a significant range.
•
The deviation from the rated voltage caused by
a short circuit depends on the fault distance and
the short-circuit power of the network. The
duration is given by the fault clearing time of
the protection and the circuit-breaker.
•
Large load changes, e.g. by motor starting, may
result in voltage dips with a duration of up to
some seconds.
In this paper the behaviour of different low voltage
consumers during the occurrence of voltage
reductions is considered. For the investigated devices
the combinations of amplitude and duration of
voltage depressions which result in a malfunction are
evaluated.
•
Voltage variations or voltage fluctuations
Voltage variations are within limits of +10%
around the nominal voltage. According to the
specification of power quality they should not
cause malfunctions of typical electrical devices.
Voltage variations will therefore not be regarded
further in this paper.
•
Voltage dips
A voltage dip is a sudden reduction of the supply
voltage to a value between the range of 90% to 1%
of nominal voltage and with a short duration from
10 ms up to one minute.
•
Short term interruptions
These are interruptions of supply up to a duration
of 3 minutes. Interruptions are defined as system
state with a voltage at the point of common
coupling (PCC) below 1% of the agreed voltage.
•
Long term interruptions
Here the duration is above 3 minutes and the
voltage at the PCC is below 1%.
1. INTRODUCTION
Voltage variations, voltage dips and short interruptions
may cause functional problems, e.g. breakdown of a
computer. To evaluate the compatibility level of modern
single-phase A.C. equipment to such kind of voltage
depression, a case study [1] was performed. In this
investigation the combination of duration and voltage
deviation from rated voltage of voltage depressions that
lead to malfunction have been evaluated for several
types of low voltage A.C. equipment
3.
Low voltage consumers will under the condition of
short term and long term interruptions normally have a
malfunction, so these types of voltage depression are
not regarded in this paper.
VOLTAGE DIPS AND INTERRUPTIONS IN
DISTRIBUTION NETWORKS
The main reasons for malfunctions of electrical low
voltage devices are voltage dips and short interruptions.
These are caused during short circuits and subsequent
fault clearing by protection equipment or by a sudden
change of load such as motor starting. Most of these
voltage depressions have a duration of less than 500 ms.
According to table 1 after short circuit the duration
depends on the kind of protection system.
2. DEFINITION OF VOLTAGE DEPRESSIONS
According to EN 50160 [2] different types of voltage
depression are defined in terms of their duration and
deviation from rated voltage:
Protection Scheme
duration of voltage dip
Fuse
5 to 10 ms
Breaker
80 to 150 ms
Auto-Reclosing
300 to 700 ms
Back up protection
300 ms to some seconds
of all relevant power quality criteria, e.g. voltage dips,
voltage variations, short term interruptions, harmonics,
voltage fluctuations for flicker investigation and other
voltage forms.
The simulated voltage depressions are characterised by:
PC
function
generatorboard
hardware
Table 1 Influence of protection on dips
This is shown in many investigations e.g. in 1994 (see
EN 61000-4-11 [3]), where the voltage dips in
distribution networks have been measured. Nearly all of
them showed a duration less than one second. In fig. 1
the results of this study are shown. The voltage
depressions are put in order according to their duration
and deviation from nominal voltage. Main reason for
voltage dips are short circuit, switching operations and
motor starting.
software
measurement
software
v(t)
device under test:
PC, video recorder,
etc.
amplifier
Figure 2:
Power quality simulator for
testing of low voltage consumer types
70
•
60
50
•
•
number of
voltage 40
depressions
per year 30
Amplitude of voltage depression related to the
nominal voltage Vn:
∆V/Vn = 1 – V/Vn
Duration of the voltage depression: ∆t
Phase angle of the mains voltage related to zero
voltage crossings at positive voltage slope: ∆ϕ
20
10 bis <30
10
30 bis< 60
60 bis <100
0
10 bis
<100
100 bis
<500
duration [ms]
100
500 bis
<1000
1000 bis
<3000
deviation from
nominal voltage
[%]
Fig. 3 shows an example for a voltage depression with a
sudden rectangular voltage change of 57 ms duration
and voltage reduction of 50 %, starting at a phase angle
of 90°.
Figure 1:
Number of voltage depressions per
year in public distribution networks [3]
4.
POWER QUALITY SIMULATOR
Type tests of the immunity level of electrical devices
against voltage depressions can be performed according
to EN 61000-4-11 [3], with testing levels according to
relevant product standards.
For the investigation of low voltage devices a power
quality simulator was developed, allowing to reproduce
voltage depressions with characteristics according to
EN 61000-4-11 or of any other characteristic. In fig. 2
this system is shown. Its main component is a power
amplifier of 4,5 kW (alternatively 3x1,5 kW) output
power, that is controlled by a personal computer with a
function generator board. At the output terminals of the
amplifier the electrical device can be tested with
predefined functions. This system allows the simulation
Figure 3:
Voltage dip of ∆V/Vn = 50%, with
57 ms duration and phase angle ∆ϕ = 900
Measuring Procedure
•
•
V/V n [%]
For the actual measurements, mainly voltage
depressions according to [3] with ∆ϕ = 00 were used.
Their duration was increased in steps of 5 ms or 10 ms,
respectively at a constant ∆V/Vn until a malfunction of
the low voltage device was observed. For the next data
point the amplitude ∆V/Vn was changed to a new
constant value and the duration of the voltage
depression was altered again until another malfunction
occurred.
5. RESULTS
The measuring procedure described above results in a
sequence of amplitude/duration combinations of voltage
depressions that are leading to malfunction of a device.
In fig. 4 the measured amplitude/duration combinations
leading to malfunction of a high-pressure mercuryvapour lamp are shown as an example.
For the mercury lamp under investigation, a short
interruption with a duration less than 5 ms does not
result in malfunction. Increasing this duration to 10 ms
results in extinguishment of the arc and therefore the
luminous flux. If the voltage depression does not fall
short of 79 % of nominal voltage, then certainly the
luminous flux is reduced but there is no interruption of
the illumination. All combinations of amplitude and
duration of voltage depressions lying above the curve in
fig. 4 do not result in malfunction.
A similar behaviour as shown in fig. 4 is found for the
class of electrical low voltage devices, that is equipped
with a switched-mode power supply. As shown in fig. 5,
which is valid for personal computers, the area of
malfunction is formed by a rectangular region. The
permissible duration of voltage interruptions depends on
the size of the smoothing capacitor that is used in the
switched-mode power supply. The minimum voltage
amplitude for long duration voltage dips is determined
by the switch mode power electronic following the
capacitive smoothing. A similar behaviour can be
observed for all low voltage devices using a rectifier
with capacitive smoothing.
In accordance with fig. 5, the area of malfunction can
simply be described by a single point, characterising the
left upper corner of this region, which is approximately
rectangular. This holds for a PC like the one in fig. 5
and approximately for the high-pressure mercuryvapour discharge lamp in fig. 4 as well as for the other
investigated devices.
In fig. 6 these left upper points of the malfunction areas
are shown for different devices.
Low voltage devices under investigation
The following types of devices have been investigated:
Fluorescent lamps with inductive, capacitive and
electronic ballast elements
High-pressure mercury-vapour and sodium-vapour
90
80
70
60
50
40
30
20
10
0
0
200
400
600
800
1000
∆ t [ms]
Figure 4:
•
•
Amplitude/duration combinations for
malfunction of a high-pressure
mercury-vapour lamp
lamps
Personal Computers
Video recorders
Fluorescent lamps
are not very sensitive to voltage dips. After
extinguishing of the arc and luminous flux a reignition
will immediately occur, thus only resulting in short light
flicker.
Compact fluorescent lamps
have a similar behaviour than fluorescent lamps and an
even faster ignition capability.
High-pressure gas-discharge lamps
are sensitive even to very short voltage interruptions.
They will e.g. already extinguish at voltage dips with
V/Vn = 70 % and a duration of about 5 to 10 ms. For
reignition a time delay of up to 15 minutes is necessary.
Personal Computers
Personal Computers have a power electronic with an
rectifier with capacitive smoothing, as shown in fig. 7.
Behind the capacitor is switched mode converter
operated at some kHz with a small transformer. Two
influencing factors are relevant for this equipment.
In fig.°6 the results for different personal computer is
shown. In the worst case a computer will have a
malfunction at a voltage dip of 80 ms duration, in the
best case after 450 ms. The steady state voltage drop
can be between 65% and in the best case about 30%.
70
area of operation
V/Vn [%]
60
50
40
30
R
area of malfunction
20
C
10
0
electronic
0
500
1000
1500
2000
Figure 7:
∆t [ms]
Figure 5:
•
•
Area of malfunction of a personal
computer
The permissible duration of interruption, also called
“ride through capability” depends mainly on the
size of the smoothing capacitor
The minimum steady state amplitude of supply
voltage. Here the minimum voltage for the
switched mode components for stable work is
relevant.
70
U/U n in [%]
60
Switched mode power supply
Video recorders
Disturbances of the recording process were selected as
criterion to state malfunctions of video recorders.
During standby oper-ation a loss of preprogrammed TV
channels for later recording could not be observed.
Thus, voltage depressions especially affect the
recording and playback function of a video recorder.
The left upper points, characterising the malfunction
rectangles are presented in fig. 8 for video recorders and
compared to personal computers and gas discharge
lamps. There is a broad range of variation in permissible
amplitude and duration of voltage dips especially for the
class of PCs and video recorders. As an example it was
found, that the best PC that still worked after a short
interruption with a duration of 460 ms. If the voltage
recovery after this fault reaches only 33%, the PC will
50
PCs
40
video recorders
gas discharge lamps
100
30
20
0
0
100
200
300
400
500
∆ t in [ms]
V/V n [%]
80
10
60
40
20
Figure 6:
Left upper points of malfunction of
personal computers
0
0
Typical functional behaviour can bee seen form fig. 5.
The area of malfunction is limited on the left side by the
permissible duration of interruption and on the upper
side by the steady state voltage reduction. This
behaviour is typical for all low voltage consumers
having a rectifier with capacitive smoothing.
A malfunction of personal computers due to voltage
dips and short interruptions was estimated, if there was
no response to key strokes at the keyboard or if the
computer was performing a new start of the operating
system. Both cases can result in loss of data.
100
200
300
400
500
∆ t [ms]
Figure 8:
Left upper points of the malfunction
areas of different electrical devices
operate fine, too, independently from the duration. On
the other hand the worst PC showed a malfunction after
a short interruption of just 80 ms and needed a
minimum voltage of 64 % of nominal voltage after
recovery.
Acknoledgement
6. SUMMARY
Measurements of the immunity of low voltage devices
to voltage depressions of different amplitude and
duration have shown, that the area of malfunction can
be described in many cases by a single point, which
represents the minimum voltage needed for continuous
operation and the maximum permissible duration of a
voltage dip. If the voltage falls below the minimum for
steady state operation or exceeds the permissible
duration of a voltage dip, a malfunction will occur.
According to fig. 6 the allowed duration of a short
interruption for personal computers is between 80 ms to
450 ms with an accumulation around a value of about
200 ms. The respective values for video recorders are
between 50 ms and 200 ms. Gas discharge lamps are
able to withstand short interruptions with durations of
10 ms to 100 ms.
For an easy comparison with the above immunity
values, the typical duration and amplitude of voltage
depressions in low voltage supply networks can be
found in fig. 1. Faults in low voltage networks have a
clearing time between approximately 5 ms if fuses are
used and up to 150 ms for circuit breakers. Short
interruptions by auto-reclosing lead to zero voltage in
the range of 300 ms to 500 ms in medium voltage
networks. In high voltage systems the duration of
supply interruption by auto-reclosing is up to 750 ms.
But as these networks have a redundant structure, in
most cases no interruption of energy flow can be
observed in underlying low voltage networks but only a
voltage dip.
To improve the immunity of devices against voltage
dips in principle two measures are possible: at the
consumer side and at the network side.
At the consumer side the power electronic can be
redesigned to withstand voltage dips and voltage
deviations.
An improvement of the ride through capabilities of
consumers under voltage dips can be achieved, if the
smoothing capacitor is large enough for a ride through
of minimum about 300 ms. If the switched mode
converter following the rectifier and the smoothing
capacitor are is designed for low steady state voltage
situations e.g. for 50% of nominal voltage, in many
cases malfunctions are excludes. Such measures only
lead to small additional costs on the equipment side but
helps to reduce the number of malfunctions
significantly.
On the network side the costs for redundant structures
can be very high, especially in rural areas with low
consumer density. In the liberalized electricity market
the utilities are constrained to save costs. So the
tendency to rebuild a distribution network to reduce the
number of voltage dips will in future decrease.
The Autors like
to thank VEÖ (Verband der
Elektrizitätswerke Österreichs / Union of Austrian
Utilites) for supporting of this work.
7. REFERENCES
1. Verband der Elektrizitätswerke Österreichs:
Empfindlichkeit von elektrischen Geräten gegenüber
Spannungseinbrüchen und Kurzzeitunterbrechungen.
Wien, 2001.
2. EN 50160: Merkmale der Spannung in öffentlichen
Elektrizitätsversorgungsnetzen.
3. IEC 61000-4-11: Electromagnetic compatibility Part 11: Voltage dips, short interruptions and voltage
immunity tests.
4. Stade, D.; Schau, H.; Hünermund, J.: A
measurement program for analysing and estimating
of voltage dips in the electric power system.
Proceedings of the 5th International Conference on
Electrical Power Quality and Utilisation. Cracow
1999.