Power Quality in European Electricity Supply Networks - 1 Edition Network of Experts For Standardisation
Power Quality in European Electricity Supply Networks - 1 Edition Network of Experts For Standardisation
Power Quality in European Electricity Supply Networks - 1 Edition Network of Experts For Standardisation
Ref: 2002-2700-0005
.......................................................................................................
Power Quality
in European Electricity Supply Networks
- 1st edition
.......................................................................................................
Its mission is to contribute to the development and competitiveness of the Electricity Industry and to
promote the role of electricity in the advancement of society.
As a centre of strategic expertise, the Union of the Electricity Industry - EURELECTRIC will identify
and represent the common interests of its members and assist them in formulating common solutions
to be implemented and in coordinating and carrying out the necessary actions. To that end it will also
act in liaison with other international associations and organisations, respecting the specific missions
and responsibilities of these organisations.
The Union of the Electricity Industry - EURELECTRIC is also the association of the Electricity
Industry within the European Union representing it in public affairs, in particular in relation to the
institutions of the EU and other international organisations, in order to promote the interests of its
members at a political level and to create awareness of its policies.
The reports published by EURELECTRIC are the result of the work of its structure of expertise: they
represent one of the most direct methods of circulating knowledge and information throughout the
sector, on subjects of common interest.
They are intended for wide circulation both within the electricity supply industry and outside it.
Ä Please do not hesitate to ask for the latest available printed EURELECTRIC publications
catalogue (with summaries of EURELECTRIC reports) from:
Ä You can also use the EURELECTRIC Internet Web site, which provides the following
information:
- EURELECTRIC general information
- EURELECTRIC positions and statements
- Events & Conferences
- Publications Catalogue
http://www.eurelectric.org
1
International Union of Producers and Distributors of Electrical Energy
2
European Grouping of Electricity Undertakings
Power Quality
in European Electricity Supply Networks
............................................................................................
The measurement results presented in this report show that continuous care for EMC in the supply
networks is needed. Although the standardised compatibility levels for the different phenomena within
the field of power quality are still not exceeded in general – what correlates with the results from other
surveys – this issue is to be considered in a differentiated way. The development of the past years
may be summarised in the following recognitions:
o During the past ten years a steady increase of the harmonic levels can be recognised, causing
increasing costs which may be counted in hundreds of B$ worldwide. When considering appro-
priate measures to maintain EMC, the cumulative effect of equipment and systems is to be con-
sidered.
o The mean values of flicker are generally quite low compared with the compatibility level; beside the
excession of the compatibility levels by maximum flicker levels, some cases of excession have
been identified in some networks.
o Voltage dips get an increasing importance for the functionality of electrical equipment, facing an
increasing sensitivity against that phenomenon. Appropriate measures on the equipment side allow
cost-effective mitigation.
o If considering changes concerning the recent building of standardisation concerning EMC, related
to procedures as well as to limits and requirements, careful consideration of the consequences of
such changes are indispensable.
INTRODUCTION
Electricity supply represents one of the most essential basic services for the support of an industrial
society. From the point of view of electricity consumers this basic service is required
§ to be available all the time (i.e. a high level of continuity) and also
§ to enable all the consumers’ electrical equipment to work safely and satisfactorily (i.e. a high level
of power quality).
Electricity -- nowadays referred to as a product -- is rather a unique product because of its intangible
and transient nature. Strictly, it is a product which exists only for an instant at a given point of delivery,
and comes into existence at the same instant at which it is being used. It is immediately replaced by a
new product with rather different characteristics, and its characteristics are different at each separate
point of delivery. Moreover, it is a product whose quality depends not only on the elements which go
into its production, but also on the way in which it is being used at any instant by the equipment of
multiple users.
The most important characteristics of electricity are described in EN 50160 [1]. This European
standard gives information on several phenomena which have an effect on the power quality and
describes the likely maximum levels of those phenomena under normal operating conditions.
Electrical equipment has become increasingly complex in terms of the functions it fulfils and the way in
which it interacts with other electrical equipment. Frequently, that interaction takes place through the
medium of the electricity network, which is the common energy source for all the equipment. It arises
because the network, intended to be a common energy source, also provides a conducting path inter-
linking all equipment.
As a result of the increasing complexity, some types of electrical equipment today are more sensitive
to deviations from the sinusoidal supply voltage. At the same time there is an increasing tendency for
the same or other equipment to cause modifications to the characteristics of the supply voltage. An
1. GENERAL
Quality related terms
The quality of the electricity supply (in some French documents the term is translated as “Qualité
de l’alimentation”) is a function of its suitability as an energy source for the electrical equipment
designed to be connected to the supply network. The two primary components of supply quality are:
– Continuity (freedom from interruption): the degree to which the user can rely on its availability
at all times
– Voltage level: the degree to which the voltage is maintained at all times within a specified
range.
Electrical energy is delivered by means of voltage derived from single-phase or three-phase sinusoidal
systems and having the following main parameters: magnitude, frequency, waveform and unbalance.
– Harmonics and other departures from the intended frequency of the alternating supply voltage
– Voltage fluctuations, especially those causing flicker
– Voltage dips and short interruptions
– Unbalanced voltages on three-phase systems
– Transient overvoltages, having some of the characteristics of high-frequency phenomena.
Power quality can be defined as the degree of any deviation from the nominal values of the above-
mentioned characteristics. It can be also defined as the degree to which both the utilisation and
delivery of electric power affects the performance of electrical equipment. Draft IEC 61000-4-30 [9]
defines power quality as “The characteristics of the electricity at a given point on an electrical system,
evaluated against a set of reference technical parameters” (with a Note saying that “these parameters
might, in some cases, relate to the compatibility between electricity supplied on a network and the
loads connected to that network”).
Security of supply: Technically safe and long term sufficient energy supply in terms of quantity and
quality at any time, based on the sufficient availability of energy resources and appropriate low risks in
infrastructure, regarding the conditions of international competition and environmental compatibility.
– To ensure that equipment has a sufficient level of immunity to maintain adequate performance
in the presence of the electromagnetic disturbances to which it is subjected.
These requirements concern both high frequency disturbances, affecting especially equipment having
telecommunication or information processing functions, and low frequency disturbances, affecting
equipment utilising energy drawn from electricity networks in general.
There is a great deal of ground that is common to both the EMC and PQ fields. The phenomena
mentioned already in connection with power quality are also electromagnetic disturbances, the
emission of which needs to be limited and immunity from which is necessary for the proper functioning
of equipment.
Since many of these disturbances are generated as a result of the particular way in which some utili-
sation equipment makes use of the supplied energy, limitation of emissions from users’ equipment is
as vital for PQ as it is for EMC. In fact, it has always been a condition of the contract for electricity
supply that the electricity user must avoid disturbing the network and other users. Thus, meeting of the
Equally, given the technical as well as the economic reality that power quality can never be “perfect”, it
is important that the EMC immunity standards be implemented, so that the equipment utilising the
electricity supply can continue to operate satisfactorily in the real conditions of the electromagnetic
environment, of which the supply networks are a significant part.
Therefore, the effective implementation of the EMC Directive, the essential requirements of which are
focussed on enabling the consumer to be provided with equipment which neither generates distur-
bance excessively nor is excessively susceptible to disturbance, is very important for power quality.
Indeed, the need to protect the electricity networks from disturbance is recognised quite explicitly in
the recitals of that Directive.
In relation to power quality, the important point is that the compatibility level is intended to represent a
measure for the cumulative disturbance level in a given environment, which is expected to be
exceeded quite rarely. That is why the measured values mentioned in this report are compared with
the compatibility levels for the relevant environment.
From a power quality perspective, any tendency for the actual cumulative disturbance level to exceed
the compatibility level would be that electricity users’ equipment would be jeopardised to get degraded
or hindered in its intended operation by disturbances conducted to it by the electricity networks, and
that the networks themselves would be hindered in their function of delivering energy to the utilisation
equipment in the form of a voltage with the intended characteristics.
The EMC problem usually involves a number of parties with some kinds of interrelationships. There-
fore, to properly resolve the problem, cooperating consideration of related problems, in a coordinated
way, is of importance.
The costs associated with solving EMC problems, such as the purchase of a conditioning system,
must be taken into account.
Thus, when a consumer enquires about the connection of a major load to a low-voltage or medium-
voltage public supply network, upon request he shall give the information required to assess the toler-
able disturbance emissions caused by the load.
Equally, the electricity supplier will make available relevant information concerning the quality of
supply parameters at the point of common coupling when required.
The network operator is responsible for the overall control of disturbance levels under normal
operating conditions in accordance with standard requirements.
The consumer is therefore responsible for maintaining his emissions at the PCC (point of common
coupling) below the limits specified by the utility.
In this report, EURELECTRIC presents an overview of the present PQ situation in European electricity
supply networks, relating to the main PQ phenomena.
Chapter 2 gives a hint on the position of standardisation in serving customers, manufacturers as well
as electricity suppliers at the economic use of electricity and a short overview of the general concept
for EMC standardisation having been defined by the International Electrotechnical Commission (IEC)
and the European Committee for Electrotechnical Standardisation (CENELEC).
Chapter 7 gives a short overview of the causes and ranges of values of transient overvoltages.
Chapter 8 summarises the conclusions that may be drawn from the measurement results.
With a comprehensive Annex beside others additional examples of results from measurement cam-
paigns performed in different countries are given as well as overviews of the measurement methods
and the standards per phenomenon; also the network impedance, that one not representing a compo-
nent of power quality but an important value in relation to the disturbing effect of some devices that
electricity users connect to the network, are dealt with there.
- the definition of “compatibility levels” per supply network level (e. g. LV): for a given
phenomenon a reference level is defined, which should not be exceeded by 95 % of time and
location
- the definition of “emission limits” for certain kinds of electrical equipment or systems:
aiming at maintaining the sum of disturbances, being generated by electrical equipment, below
a level ensuring that customers’ equipment may be operated with a sufficient reliability, limits
for the conducted disturbances caused by the single equipment are defined per phenomenon.
- the definition of the “immunity level” for certain kinds of electrical equipment or systems:
aiming at ensuring the reliable function of customers’ equipment or systems, considering the
compatibility level as well as an appropriate functional reliability regarding an economic
design, immunity requirements are defined per phenomenon. The design of the margin
between immunity level and compatibility level depends on the environment in which the
related equipment will be normally operated and on the normally required functional reliability.
C o m p a tib ility
le v e l
E q u ip m e n t
S y s te m im m u n ity
d is tu rb a n c e le v e l
le v e l
Probability density
P ro b . = 5 %
a p p ro x .
D is tu rb a n c e le v e l
Manufacturers as well as electricity suppliers are cooperating in this field of standardisation, whose
success is the basis of economic manufacturing of equipment on the one side and provision of an
economical electricity supply with appropriate power quality.
In Annex D an overview of the standards related to the issue of this Report is given.
3. HARMONICS
3.1 INTRODUCTION
Both electricity distribution networks and consumers’ equipment are affected by harmonic distortion of
the voltage waveform. Harmonic distortion levels have increased rapidly in electric power systems in
recent years due primarily to the increasingly widespread application of non-linear semi-conductor
devices, which produce the majority of harmonic distortion.
In particular, switched-mode power supplies have over the last twenty years increasingly replaced
th
transformer/rectifier power supplies in electronic equipment. These cause large 5 harmonic
emissions, relative to the rated power consumption of the equipment, which are not randomly orien-
tated in-phase and so are strongly additive. Consequently, in Europe, electricity supply systems are
experiencing increasing background levels of harmonic voltage distortion at all voltage levels up to
400 kV. As harmonic voltage distortion increases, the risk grows of widespread problems arising from
premature ageing of equipment and overloading of neutral conductors.
In response to that, many network operators are implementing monitoring systems to obtain a better
overview of the present status of harmonic loads and of the related development.
In this section the harmonic voltage distortion in LV, MV and HV networks measured over the
past twenty years is analysed and trends are identified, which reflect the steadily increasing
use of mass marketed equipment, in particular IT equipment (ITE) and TV receivers which
have created the harmonic distortion.
To obtain an understanding of the generation of harmonic distortion, actual data have been classified
by countries and voltage levels, taking into account different distribution systems structures.
5th harmonic
th
Fig. 3.1 – Fundamental, 5 harmonic
An electric power supply system is inherently very passive and the generator outputs and voltages
measured on the system would be nearly sinusoidal at the declared frequency if no load was con-
nected to the system.
For a pure sinusoidal voltage wave, with 50 Hz frequency and 230 V amplitude, the spectrum is equal
to zero for all frequencies except 50 Hz, for which the value is 230 V.
Signal Signal
500 500
0 0
50 100 150 Hz
-500
t -500
For a distorted voltage wave, the spectrum contains harmonic frequencies which are characteristic of
the nature of distortion.
The fundamental is the component of the spectrum at which the network is designed to work. It is
normally the first and greatest component of the spectrum. The term Total Harmonic Distortion (THD)
is used to describe the r.m.s. sum of the voltages of all harmonic frequencies that are present relative
to the fundamental.
Harmonic currents are mainly generated by industrial and residential loads with non-linear
characteristics. Sources may produce harmonics at a constant or varying level, depending on the
method of operation.
Industrial loads which are a source of significant levels of harmonic distortion include power converters
(rectifiers), variable speed drives, induction furnaces, arc furnaces, etc.
Studies published during the past 10 years show the origin of harmonic distortion as being given by:
- harmonic currents injected by industrial and commercial installations. In general, the harmonic
currents produced by industrial installations are high, and very often, for mitigation purposes, a filter
- harmonic currents injected by residential equipment connected to LV public network as for example
TVs, PCs, self-ballast lamps, air conditioners. In general, the single equipment has a low rated
power and produces only a low harmonic current, but because they are very often applied at the
same time by customers and because there is no self-cancelling effect, the cumulative effect of all
these small contributions of harmonic emission is not insignificant (► see also sub-clause 3.4).
- Consumers and distribution systems are sometimes forced to derate their transformers because of
the heating effects of harmonic currents. Transformer manufacturers recommend derating by 10 %
when a transformer supplies more than 30 % of its nominal capacity to non-linear load, and the
same recommendation applies to generating plant.
- Neutral conductors of supply systems and installations have the same cross-sectional area as
phase conductors. There is already evidence of the use of neutral conductors of larger cross-
section in newer commercial installations to take account of the increased third harmonic currents.
The retrospective installation of such larger neutral conductors in existing networks would result in
astronomical costs, including significant increase in demand for copper and aluminium, which
cannot be realistically contemplated.
- The flow of harmonic currents in power supply systems, caused by non-linear loads of customers,
may affect telephone communication.
- Poor power factors associated with non-linear loads are responsible for a substantial increase in
the currents flowing in power supply systems and consumer installations, and hence in the cost of
losses.
The following indicates how problems with consumers’ equipment and network components [8]
increase as Total Harmonic Voltage Distortion, THD, measured on public networks increases:
Harmonic voltages in excess of the compatibility levels, would cause distributors to replace their
transformers, switchgear and lines at prohibitive cost. The resulting networks would be inefficient as
harmonic distortion represents reactive power flow.
As mentioned above (► sub-clause 3.3) – the cumulative effect of harmonic distortion can be of great
significance in relation to conducted low frequency disturbances when considering mass produced
equipment. In specifying emission limits, the need is to ensure that the limit applicable to
equipment groups is in proportion to their contribution to the cumulative disturbance level.
Harmonised standards in the framework of the EMC Directive [2], which specify limits and require-
ments to meet the requirements of Article 4 of the Directive, need therefore take into account the
characteristics, use and the expected number and dispersion of equipment to ensure a fair balance
between appropriately manufactured equipment and appropriately designed electricity supply
systems.
This implies that some limits might need only to be applied to certain categories of products. Therefore
equipment groups could be defined and indications given in a matrix which limits would apply to which
equipment group.
There are several reasons why harmonic distortion has to be limited. Firstly in the industrial plant itself
in order to prevent interference, malfunctions, overloading etc of the equipment in the installation itself.
The second is to limit the emission of harmonic currents out to the supplying network and fulfil the
regulation on harmonic distortion applied by the electricity company.
The timeframe is important when considering the different methods to reduce harmonic distortion. At
the planning stage of a new plant or when new equipment shall be installed certain tools are available.
If harmonic distortion in an existing plant has to be reduced, the tools are partly different.
- reduction of the generation of harmonic currents by increasing the pulse number of converters.
If several sources are involved, connecting them through transformers with different phase
shifts such that the vectorial summation of harmonic currents will have favourable diversity
- installation of a filter at the bus with the disturbing source. A similar method is to install reactors
in series with existing shunt capacitors so as to make the impedance inductive for the
characteristic harmonics. Care has to be taken with parallel (anti-) resonances which may occur
below the tuning frequency or between tuning frequencies if several filters are installed.
Passive mitigation consists of a low impedance filter for one or several harmonic orders. They are
based on series resonant filters. Passive filters at the point where the loads are connected to the
public distribution systems are technically difficult to be applied due to continuous change of system
conditions. In fact, it is technically impossible to install a passive filter at each point of connection for
each residential installation because connection of a large number of identical filters on the same
network would create other serious problems.
Active mitigation may be applied at equipment, site or system level. It consists of filters that -- by
electronic techniques -- may continuously adapt to changing conditions in the network. Active
mitigation is more expensive but technical improvements made these filters more cost effective now.
Mitigation measures at equipment level are less costly to be implemented in general.
Some previous analysis have been performed by utilities regarding the extra cost in power systems
due to harmonic pollution. Evaluation methods are based on the calculation of harmonic levels,
increase of cost growing with the square of the growth of the harmonic voltage, calculation of ageing
according to operation duty cycle of equipment and other factors as above-mentioned.
The main hypothesis is based on a constant increase of harmonic voltage level in networks, extra-
polating from the development during the past twenty years.
As a first conclusion, based on a five-year global cost estimation and adopting different hypotheses for
calculation of these extra costs, the following estimations can be shown:
1) France: Basic data: Gross Domestic Product (GDP) = 1400 B€; overall consumption [GWh} =
450;. Population = 60 M. EDF has performed an analysis based on the hypothesis of incorpo-
rating active filters in HV/MV substations and reinforcing the neutral wire in LV networks A total
extra cost of roughly 25 B€ has been estimated.
2) Quebec (Canada): GDP = 150 B€; GWh = 120; Population = 7,5 M. Hydro-Quebec has
performed a study based on the calculation of Joule losses and ageing of equipment. A total
extra-cost of 650 M$ has been estimated.
3) Germany: Basic data: GDP = 2300 B€; GWh = 500; Population = 85 M. An approximate estima-
tion based on reinforcement of all network installations throughout the country would lead to an
extra cost of 20 – 25 B€.
4) Spain: Basic data: GDP = 560 B€; GWh = 180; Population = 40 M. An approximate calculation
based on the application of a combination of the above several hypothesis would lead to an
extra cost of 6 – 7 B€.
A gross extrapolation to the entire world would lead to quite high numbers.
Anyone with a basic technical background in electricity and an understanding of rational trouble-
shooting techniques can monitor conducted disturbances, but often the interpretation of the infor-
mation about the power quality can seriously be undertaken only by experts.
It is important to note that the measurement criteria necessarily refer to the compatibility levels.
However, the compatibility levels are only reference values for coordination purposes, which enable
definition of, with appropriate methods and margins, emission limits and immunity tests levels for
equipment connected to the supply. They are supposed to cover 95 % of the cases, allowing for the
3.7.1 Harmonic levels in European networks. General survey over the past 20 years.
Trends of the 5th harmonic in the past 20 years
In recent years, measurements in different countries showed increasing harmonic voltage distortions
in public networks.
th
Fig.3.3 - Trends of the 5 harmonic in different countries
th
A general survey within UNIPEDE (Fig.3.3) shows a mainly significant increase of the 5
harmonic, with a constant slope, during the last twenty years.
th
In fact, the 5 harmonic is the most significant one and, consequently, directly affects the Total har-
monic distortion factor (TDH) in the same manner.
In Annex A.1.2, examples from several surveys, conducted in several countries, can be found.
General surveys made throughout different European countries, show results based on statistical
treatment of measurement in multitude test sites, and so are considered reasonably representative.
th
The following figure shows the results in the three voltage levels, concerning the 5 harmonic, of a
general survey made in Austria in June 2000 and January 2001 by about 15 utilities, elaborated within
the Austrian Association of Electricity Companies.
6
95-%-Percentile of the 5th Harmonic
0
0 1 10 100 1 .0 0 0 1 0 .0 0 0
S s c in M V A
Evaluation of the results concerning the 95-%-value shows a similar behaviour, with somehow lower
values compared with the maximum levels, by nature.
A correlation between the measurement results in LV, MV and HV networks can be recognised.
Further examples from several surveys, conducted in several countries, can be found in Annex A.1.3.
Weekly and daily diagrams show the cumulative effect produced by harmonic generating equipment,
like TVs.
As an example, the following describes higher harmonic levels as having been measured on French
LV supply systems. Related measurement results have been obtained during a survey carried out for
th
the period of about 12 days in a winter sports resort. The 95-%-value of the 5 harmonic voltage was
equal to 11,2 %.
In Annex A.1.4, further examples from several surveys, conducted in several countries, can
be found.
3.8 CONCLUSIONS
In this report, most surveys show a constant increase of the THD voltage as well as the predominance
th
of the 5 harmonic in LV, MV and HV networks. Only in a few countries, there has been a quite stable
situation during the last years.
3) The rising number of ITE equipment, particularly PCs and TVs, results in a considerable total non-
linear load in residential areas that causes a continued considerable rise in the THD and harmonic
th
levels, in particular of 5 order. In consequence, compatibility levels can be expected to be widely
exceeded leading to more frequent damages to existing equipment.
5) Further voltage components with frequencies up to some tens of kHz are likely to occur in
distribution networks, due to the increasing application of, for example, switched mode power
supplies and adjustable speed drives.
6) The use of non-linear loads over a day may be observed by the distribution of harmonic values
along the 24-hours profile, showing peak hours.
7) In MV, the impact of low order harmonics is also significant. In several large MV systems, the
harmonics exceed the compatibility levels in around 5 % of the cases. Taking into consideration
that harmonics can transfer from low to higher voltage levels, it may be expected that harmonic
pollution in MV and even in HV networks could dangerously increase.
th
8) In general, in MV networks, the annual increase of the THD and 5 harmonic is similar to that in
LV networks.
The connection of a wide range of non-linear devices to supply networks needs to be considered in
relation to the cumulative effect of disturbances emitted into these networks.
The number of problems caused by harmonics is not yet a significant widespread problem, but the
potential for such problems is much greater today than it was 10 years ago.
Monitoring of harmonic pollution of public electricity networks has shown a steady increase in voltage
waveform distortion over the past twenty years. Particular harmonics that are characteristic for the
power supplies used in televisions, personal computers and similar equipment have increased by a
factor of 3 and have been shown to follow patterns of usage of such equipment.
The principal harmonics are now found at levels of several percent on distribution networks, and of
about 2 % on many transmission networks. Some decades ago the related levels were only a fraction
of these present levels. The difference mainly stems from electronic products, as e.g. TV sets, the
peak harmonic levels closely corresponding to times of peak television viewing.
The voltage of an electrical network varies all the time under the influence of various switching on and
off operations of electrical equipment connected to the supply network.
The voltage variation can be slow or fast, depending on whether it is a progressive variation of the
total load supplied by the grid, or it is an abrupt variation of a large load.
The level of voltage variations emitted by an electrical equipment into the supply network to which it is
connected depends on the network impedance. With increasing impedance, the level of voltage varia-
tions will increase.
Voltage variations
Due to the influence of the variations of the load, the voltage of an electrical network, which is subject
to various switching on and off operations of the connected electrical equipment, varies all the time.
The voltage variation can be slow or fast, depending on whether it is a progressive variation of the
total load supplied by the grid, or it is an abrupt variation of a large load.
Voltage fluctuation
Note: The boundary between slow and rapid variations is very fuzzy and can be assumed to be
situated between a few seconds and a minute. The slow variations are evaluated as an average
over consecutive intervals of ten minutes.
The amplitude of these sudden variations does not in general exceed 6 % to 8 % of the nominal
voltage. The electrical equipment does not, in principle, malfunction for these types of sudden
voltage variations (which are usually corrected by the voltage regulating equipment such as on-
line tap-changers) except in cases when the initial voltage is very low.
Sudden voltage variations are called voltage fluctuations. These types of variations are caused
by the fluctuating loads such as welding machines, arc furnaces, and crushers.
The rapid variations of voltage can be described as cyclic variations of the r.m.s. value or
random series of sudden voltage changes.
The voltage steps of more than 10 % amplitude, whatever being their duration, are considered
as voltage dips (or swells).
Flicker
Impression of insteadiness of visual sensation induced by a light stimulus whose luminance and
spectral distribution fluctuates with time.
Note: Repetitive type voltage variations cause the flicker phenomenon in lighting equipment. The
amplitude of these variations are usually small and within 10 %. The operation of the electrical
equipment connected to this network is not much affected; but a physiological uneasiness in
vision occurs due to electrical lighting, particularly with incandescent lamps.
Flicker is the perturbation affected to lighting equipment by voltage fluctuations. Very small
variations are sufficient to provoke variations of illumination that are unbearable (irritating) to the
human eye. This sensitivity varies according to the frequency of occurrence of this phenome-
non. The strongest human response to light flicker occurs between 800 and 1200 changes per
minute (7 Hz-10 Hz). In this range human perception thresholds when viewing at an incandes-
cent lamp are as low as a 0.3 % r.m.s. voltage change. However, it is well known that the
threshold light flicker perception and irritation are significantly dependent on the observer [14].
Flicker can be caused not only due to repetitive variations of illumination but also due to more
complex phenomenon linked to low frequency noise superimposed to the fundamental wave:
this is the case for flicker with the use of arc furnaces.
- the characteristic of the source of flicker by the form of the deformation of the sine wave,
- the characteristics of the human eyes by the frequency response of the eye,
- the probability of occurrence.
Switching on or off of large capacity electrical equipment cause sudden voltage changes. Some
kinds of residential low voltage equipment also provoke flicker.
It is important to note that voltage fluctuations caused by large industrial loads could affect a large
number of other consumers connected to the same electrical network. Since the impedance of the
public power supply networks differs from place to place, the voltage fluctuations produced by a
particular piece of equipment vary considerably depending on the point of the public supply net-
work at which the equipment is connected, for example, whether it is situated near the source or far
distant from the source.
From the point of view of their effect on incandescent lamps, voltage dips can be considered as
extremely severe voltage fluctuations. Faults cleared by complex cycles of opening and reclosing
of a circuit breaker are at the origin of series of voltage dips, which appear on incandescent lamps.
This phenomenon, which is real light flicker, is therefore not due to voltage fluctuation of the
The major effect of rapid voltage variations is light flicker (fluctuation of illumination of an electric light
source). The physiological effect caused by this phenomenon depends on the amplitude of the fluc-
tuations, sequence of repetitive voltage variations, and the frequency spectrum and the duration of the
disturbance.
Everybody has experienced a similar feeling in an office building that is lit by a defective fluorescent
tube lamp, causing excessive variations in luminance. However, a similar situation can occur in a
building due to excessive voltage variations.
Note: The flashing in a defective fluorescent lamp, although similar to the effect of flicker, is not due to
voltage fluctuations. This condition is not the subject for discussion in this report.
In high doses, flicker can cause problems to epileptics. It has been shown that video games, due to
the frequency of images which is below that one used in cinema can cause epileptic seizures to some
people prone to this disease. In the armed forces, a flicker frequency of 10 Hz is used to test future
combat pilot candidates for nervous disorders. The flicker levels which are normally encountered in an
electrical network are generally very low. Nevertheless the malaise caused by them is not negligible
particularly for two important domestic leisure activities, namely, reading and watching TV.
Due to their small amplitude, voltage fluctuations generally do not affect the image as produced on the
screen of a TV set. However, in case of an electrical lighting of the room where TV is watched, the TV
viewer may feel disturbed by the flicker of the light source, especially if the light source is at the
periphery of the visual field.
The perceptible visual sensitivity of a human being to light intensity variations caused by an incandes-
cent lamp varies with the frequency and presents a maximum spectral response between 8 Hz and
10 Hz at 0,3 % voltage variations. The upper frequency limit is 35 Hz (any frequency higher than this
limit does not cause luminance variations due to the thermal inertia of the tungsten filament). There
are, however, special cases where higher frequency voltage components can cause flicker.
The luminance fluctuation caused by voltage variations depends upon the type and power of the lamp
and its nominal voltage. In addition, there is a perceptible threshold level below which flicker is not
perceptible. In the case where the waveform of the voltage is periodic such as sinusoidal, rectangular,
sawtooth, etc., the threshold level is a function of the frequency of fluctuation.
The theory of flicker is based on the most commonly used lamp: the incandescent 60 W bulb with a
nominal voltage of 230 V. The choice of a particular bulb as reference is mandatory in the standar-
dising process. The bulbs of higher power ratings have thicker filaments; therefore, they have higher
thermal inertia. For this reason, the flicker caused by these lamps is less for the same amount of
voltage variations. Such is the case for fluorescent lamps. In some situations, replacement of existing
lighting system showing flicker by another type of lighting equipment could be an acceptable and eco-
nomical alternative for mitigation.
To complete this theory, an amendment to the existing standards concerning bulb lamps for 110 V
systems is in progress.
That can be achieved by increasing the short-circuit power of the exterior (outside) network,
installation of a series of capacitors or by moving the PCC to a higher voltage level.
o Decreasing the reactive power flow from the network (which causes voltage dips).
That can be achieved either by introducing a self-inductance in series with the disturbing
appliance (arc furnace) or by connecting a static var compensator (SVC) in parallel to it.
o Motor starting problem: The most simple and economical method of starting is direct starting.
The motor is designed for normal operation with delta-connected stator windings, rated for the
line voltage.
o Star-Delta resistance-delta starter. In order to reduce the in-rush current during starting, one
can introduce an intermediate stage in a star-delta starter by inserting resistors in series with
the delta-connected windings.
o Insertion of series resistance (metallic or electrolytic) or series inductor with stator windings.
This method consists of reducing the applied voltage to the stator winding while starting, by
inserting series resistor or inductor in each phase.
o Thyristor AC controller. Thyristor AC controllers are used for drives that require low torque at
low speeds. They are used if the drives require frequent starting and small speed variations
(5 % to 10 %) of the nominal speed (motor must be of high resistance cage type).
The results show that the Plt values are around a mean value of 0,45 in urban and 0,5 in rural areas
(Fig. 4.1). These mean values are relatively low compared to the compatibility level. Nevertheless, the
Plt
1 C
0,9 o
0,8 m
p
0,7 a
0,6 t
i
0,5
M b
0,4 M i
e e
0,3 l
a a i
0,2
n n t
0,1
0
1
Urban 2Rural 3
4.7 CONCLUSION
The above figure shows that flicker is not a general problem in public supply networks.
Problems associated with flicker usually remain localised. However, implications may be quite
severe in terms of human discomfort. Owing to the subjective aspect of flicker the perception
of what is an acceptable level can vary from person to person. At the extreme there are some
rare cases of supply networks operating in excess of the compatibility levels without any
customers complaints; equally there are rare cases of customers complaining of unacceptable
flicker when in fact the supply is operating well within the compatibility levels.
Mean values are generally below 0,7 and therefore below the susceptibility threshold.
However, the actually quite high maximum values, which are being experienced in some
single locations, demonstrate, that careful consideration should be given to the proper
management of flicker sources.
5. VOLTAGE DIPS
The following paragraphs are mainly taken from [18].
5.1 DEFINITION
A voltage dip is a two dimensional electromagnetic disturbance, the level of which is deter-
mined by both voltage level and time (duration). Typically, a dip is associated with the occur-
rence and termination of a short circuit or other extreme increases in current on the system or
installations connected to it.
The definition of a voltage dip is a convention, derived from practical experience, but also depending
on the context in which the phenomenon is discussed.
Draft IEC 61000-2-8 [18] describes a voltage dip as a sudden reduction of the voltage at a
particular point on an electricity supply system below a dip threshold followed by its recovery
after a brief interval.
The depth of a voltage dip is the difference between the reference voltage and the residual voltage. It
is determined by the location of the observation point in relation to the site of the short circuit respec-
tively large current variation and the source of supply.
The reference voltage is either a fixed value, such as the nominal or declared voltage, or a
sliding value (in case of HV networks), representing the voltage immediately before the start
of the voltage dip.
Generally, the threshold corresponds to the minimum value of the voltage tolerance band.
The start threshold is an r.m.s. value of the voltage on an electricity supply system specified
for the purpose of defining the start of a voltage dip. The start threshold can be either the
lower limit of the bandwidth of normal variation or a value at a specified margin below that
limit. No specified threshold has been defined, the choice of the threshold depends on the
local conditions but typically values between 0.85 and 0.90 of the reference voltage have
been used for this threshold. For example, EN 50160 [1] uses the value 0.90 of the declared
voltage as the threshold.
The end threshold is an r.m.s. value of the voltage on an electricity supply system specified
for the purpose of defining the end of a voltage dip. The end threshold is equal to the start
threshold plus a hysteresis voltage for measurement reason. Typically, a value of 0 or 0.01 of
the reference voltage above the start threshold has been used for this threshold.
Network design affects the network quality. But excessive network improvements are not justified
by too sensitive customer. So the use of underground cable, adequate protection, arrester,
increasing short-circuit power and separating disturbed busbar to the not perturbed one decrease
the number or the severity of the voltage dips.
Taken from [18], Table 5.1 shows results obtained of one of the measurement campaigns, with details
as follows:
o measurements using 45 dip recorders at 109 measurement sites on an MV overhead line net-
work
o carried out for three years from 1996
o recording the results at each selected site for one full year
o taken phase-ground in general, but phase-phase where necessitated by the available voltage
transformer
o using voltage dip thresholds of 90 % (start) and 91 % (end) of the nominal voltage
85 > u ³ 70 238 93 14 5 1 0 0
40 > u ³ 1 55 113 12 4 1 0 0
1>u³0 0 4 1 6 7 2 3
5.6 CONCLUSIONS
Voltage dips and short interruptions can be expected at any place and at any time. The voltage may
drop down to virtually zero; the reduced voltage may last up to and above one second. The frequency
of their occurrence and the probability of their occurrence at any level are highly variable both from
place to place and from one year to another.
Voltage dips cannot be predicted or eliminated. Although their incidence may be generally reduced by
appropriate measures without large costs, this phenomenon will still exist. Depending on the appli-
cation of and the therefrom resulting requirements for equipment performance, equipment will have to
be correspondingly designed and customers will have to take special attention if necessary.
It is clear that quite high annual rates of voltage dips are possible on overhead networks reflecting the
exposure of these networks to fault causes, especially severe climatic conditions, that are additional to
the range of causes that affect all networks.
Voltage dips have been an intrinsic feature of a public electricity supply system since the earliest
times. Yet in recent decades they have become an increasingly troublesome phenomenon giving rise
to inconvenience and even considerable economic loss. The reason is that some modern electricity
utilisation equipment, either in its own design or because of control features incorporated in it, has
become more sensitive to voltage dips.
There is therefore a need for an increased awareness of the phenomenon among electricity users and
the manufacturers of electrical equipment.
In rural areas supplied by overhead MV lines the number of voltage dips can reach hundreds of even
several hundreds per years, depending in particular on the number of lightning strokes and other
meteorological conditions in the area. On cable MV networks, an individual user of electricity
connected at low voltage may be subjected to voltage dips occurring at a rate which extends from
around ten per year to about a hundred per year, depending on local conditions.
6. UNBALANCE
6.1 DEFINITION
Voltage unbalance is a condition in a polyphase system in which the r. m. s. values of the line-to-line
voltages (fundamental component), or the phase angles between consecutive line-to-line voltages, are
not all equal. The degree of the inequality is usually expressed as the ratios of the negative and zero
sequence components to the positive sequence component. The negative sequence (or zero
sequence) voltages in a network mainly result from the negative sequence (or zero sequence)
currents of unbalanced loads flowing in the network.
where a = exp(j2 p /3) and V1, V2 and V3 are the line-to-neutral voltages.
Very often the unbalance is defined by the unbalance t = V-/V+
The propagation of negative sequence voltages from lower to higher voltage networks occurs with
high attenuation. In the direction from higher to lower level any attenuation depends on the presence
of three-phase rotating machines, which have a balancing effect.
Motors and generators, particularly larger and more expensive ones, may be fitted with protection to
detect this condition and disconnect the machine: if the supply unbalance is sufficient, the "single
phasing" protection may respond to the unbalanced currents and trip the machine. Polyphase
converters, in which the individual input phase voltages contribute in turn to the d.c. output, are also
affected by an unbalanced supply, which causes an undesirable ripple component on the d.c. side,
and non-characteristic harmonics on the a.c. side. Since the main effect of unbalance is heating of
machine windings, higher short-term levels of unbalance may be acceptable, for a few seconds or
even a few minutes.
Belgium
'95% Rate'
3
2,5
2
1,5
1
0,5
0
6.6 CONCLUSIONS
The unbalance is found generally at all network levels to be below the 2 % and therefore satisfy the
related compatibility levels [12]. Exceptions may occur in certain places e.g. where arc furnaces with a
2-phase-connection to the supply network are installed.
7. TRANSIENT OVERVOLTAGES
The following description is taken from [10]:
Several phenomena, including the operation of switches and fuses and the occurrence of lightning
strokes in proximity to the supply networks, give rise to transient overvoltages in low-voltage power
supply systems and in the installations connected to them. The overvoltages may be either oscil-
latory or non-oscillatory, are usually highly damped, and have rise times ranging from less than one
microsecond to a few milliseconds. Their levels and durations can sometimes be limited by the use
of surge arrestors throughout the system, and not only at the point of common coupling.
Switching of capacitor banks is a common cause of transient overvoltages. Typically, their value at
the point of incidence is less than twice the nominal voltage. However, wave reflections and voltage
magnification can occur as the transient is propagated along a line, amplifying the overvoltage
incident on connected equipment. This needs to be taken into account if immunity is being
considered for particular equipment or installations.
No measurement results are available for the present report. However, there is no reason to
believe that there has been any increase in the level of this phenomenon, which is an intrinsic
feature of public electricity networks. There is some anecdotal evidence that modern
electronic devices are not always provided with sufficient immunity from this phenomenon.
8. SUMMARY
The measurement results presented by this report show that essential tasks there are given
for ensuring electromagnetic compatibility between the electricity supply networks on the one
side and the electric and electronic equipment connected thereto. A task which the electricity
industry is ready to carefully manage in cooperation with the manufacturing industry, both
areas working in a competitive environment.
Power quality seems to get an increasingly strategic status, for the competing electricity
industry as well as for the manufacturers. Due to the increasing economic pressure on the
electricity industry within a liberalised electricity market, a decrease of the power quality is
feared to take place in future. Such a decrease could cause high costs e.g. considering
production losses, so power quality could have an impact on the choice of locations of
industries in the future.
The results of different surveys reported about within this report demonstrate that
o any change in standardisation would carefully to be considered, due to a presently
quite comprehensive construction of the present EMC standardisation and the mutual
dependability of different areas (e.g.:
- reference network impedance – emission limits;
- compatibility levels – immunity requirements;
- Basic standards – Generic & Product (family) standards;
- definition of a phenomenon – PQ requirements).
o PQ levels in the supply networks show a different behaviour related to the margin
remaining to the compatibility levels for LV networks, phenomenon by phenomenon.
Concerning harmonics the measurement results show that there is still some small
margin remaining whenever in some cases the compatibility levels appear as getting
exceeded. In any case a tendency of steady increase of harmonic levels is to be
recognised. Due to the cumulative effect and the transfer of harmonics from the LV to
MV and HV networks, the harmonic levels in the HV networks do already exceed the
design (planning) levels.
The fact that in most cases the power quality levels continue to be lower than the
compatibility levels may be acknowledged thanks to improvements of the supply net-
Referring to the different phenomena dealt with within this report, it can be summarised:
Harmonics:
Harmonic pollution, its effect growing with the growth of the harmonic voltages, can
cause considerable cost in power networks, which – regarding the steady increase of
harmonic voltage levels during the past twenty years -- may be counted in hundreds of
B$ worldwide.
In regard to known practical or operational problems caused by harmonics, these are not yet
widespread, but the potential for such problems is much greater today than it was 10 years ago,
and the effects of non-linear loads are becoming a steadily increasing concern.
Monitoring of harmonic pollution of public electricity networks has shown that over the past
twenty years, there has been a significant increase in the distortion of the voltage waveform.
Particular harmonics that are characteristic for the power supplies used in televisions, personal
computers and similar equipment have increased by a factor of 3 and have been shown to
follow patterns of usage of such equipment. Most surveys given in this report show a
th
constant increase of the THD voltage, as well as the predominance of the 5 harmonic
in LV, MV and HV networks. Only in a few countries a relatively stable situation can be
observed during the past years.
The principal harmonics are now found at levels of several percent on distribution net-
works, and at about 2 % on many transmission networks. Some decades ago the levels
were only a fraction of these levels now present. A constant increase of 1 % of the THD
th
voltage as well as of the 5 order harmonic can be observed over each 10-year period.
th
On LV networks, a significant percentage of the 5 harmonic values are already around the
compatibility levels established in IEC 61000-2-2 [12] and 61000-2-12 [13]. In some cases, 5 %
to 20 % of the values exceed these compatibility levels.
There is an urgent need for the effective application of EMC, taking due account of the cumula-
tive effect of multiple devices emitting harmonics simultaneously, that considering also the eco-
nomic impact of related emissions from equipment.
Voltage fluctuations/Flicker:
Measurement results in different countries show that flicker levels are relatively low, in
general. In some cases Plt values are found to exceed the compatibility level. The frequency
distribution of Plt shows maxima within the range of from 0,2 and 0,6.
Generally the flicker level is higher in urban areas than in rural areas.
Voltage dips:
Unbalance:
The levels of unbalance found are generally below 2 % at all network levels and therefore
continue to be reflected by the related compatibility levels [12]. Exceptions may occur in
certain places where, for example, arc furnaces with a 2-phase-connection to the supply net-
work are installed or where significant segments of the local load are on a single phase.
Transient overvoltages:
A.1.1 INDEX
A.1.1.1 Harmonic levels in European networks. General survey over the past 20 years.
Trends.
3 UK 1980 – 1995
9 Denmark:
10 Spain
11 Czech republic
15 UK: IT Office
19 UK: IT office
Results from measurements taken from 1979 to 1999 at a typical distribution substation in a
residential area, the last survey having been carried out in April 1999.
rd th th
The most prevalent harmonic voltages of 3 , 5 and 7 order as well as the THD are represented in
as the following graphic.
rd th th
Fig. A.1 - Development of 3 , 5 and 7 harmonic and THD in a UK distribution
substation in a residential area, 1979 – 1999
th
The THD increases at an average of 1 % U1 every 10 years, and the 5 harmonic at 1,45 % U1
(» 1,5 % U1 / 10 years). The conclusion is that in the year 2008, U5 will reach 6 % C.L. This hypothesis
is even not correct, taking into account that the annual increase of ITE + TV products is expected to be
higher than linear.
5
4,5
4
3,5
3
Un% 2,5
2
3rd
1,5
1 5th
0,5 7th
0
1979 1980 1981 1982 1983 1984 1985 1988 1999
THD
Year
rd th th
Fig. A.2 - Development of 3 , 5 and 7 harmonic and THD in a UK distribution substation in
a residential estate, 1979 – 1999
Summary: This survey is part of an ongoing harmonic monitoring programme, initiated by the
Electricity Council, in 1979, to assess the growth of harmonic voltage distortion seen at a typical
distribution substation. The last survey was carried out in April 1999.
th
Conclusions: There is a clear indication in the rise of the 5 harmonic, and therefore of the THD,
over the years. This is most probably due to the rise in application of colour television equipment
and other electronically controlled domestic appliances over this period. Future measurements
need to be taken at the same time each year to allow more accurate comparison of results.
rd th th
Fig. A.3 - 3 , 5 and 7 harmonic and THD at a substation, MV side, in UK
th
Comment: The THD increases at an average of 1 % U1 every 10 years, and the 5 harmonic at 1,4 %
U1 (» 1,5 % U1 / 10 years). The conclusion is that in the year 2004, U5 will reach the 6 % compatibility
level.
5
4,5
4
3,5
3
Un%2,5
2 3rd
1,5
1 5th
0,5
7th
0
1980 1981 1982 1983 1984 1988 1995 THD
Year
rd th th
Fig. A.4 - 3 , 5 and 7 harmonic and THD in a UK residential area, MV
Summary: This survey is part of an ongoing harmonic monitoring programme, initiated by the
Electricity Council, in 1979, to assess the growth of harmonic voltage distortion seen at a typical
distribution primary substation. The last survey was carried out in September 1995. Since privatisation
of the UK Electricity Industry in 1990, this programme has suffered from a lack of co-ordination, hence
the recording surveys have not been carried out every year, and not always at the same time of year,
hence comparison of year on year results is not straightforward.
th
Conclusions: There is a clear indication in the rise the 5 harmonic, and therefore THD, over the
years. This is most probably due to the rise in application of colour television equipment and other
electronically controlled domestic appliances over this period. Future measurements need to be taken
at the same time each year to allow more accurate comparison of results.
4
3,5
3 79-80
2,5 81-83
2 84-88
1,5 89-91
1 92-94
95-97
0,5
0
3rd 4th 5th 7th 9th 11th 13th
th
Fig. A.5 - Development of the 5 harmonic in Switzerland 1979 - 1997
The following graph shows the evolution of the 95-%-values for different harmonics measured in a
substation from 1997 to 1999.
th th th th
Fig. A.6: Evolution of the 95% 5 , 7 , 9 and 11 harmonic 1997 - 1999
20 public LV systems, including 16 typical systems and 4 special ones, having been involved.
For the 16 typical LV systems, the only criterion of choice was that each LV system only supplies
one type of customers (residential, industrial or tertiary). Next, for each type of customers, the
choice of LV systems was practically made randomly.
For the 4 special LV systems, the criterion of choice was to have LV systems with a high value for
THD.
th
The results obtained on the whole sample for the 5 harmonic were as follows:
50-%-values
95-%-values
maximum values
th
Fig. A.7 - 5 harmonic in French LV systems, 2000
Comment: For nearly a half of the considered sample, the measured 5th harmonic voltage levels
were high compared with the compatibility level.
Comparison with results obtained in 1991. In 1991, EDF already conducted a survey on 20 public
LV systems for one week. These LV systems were situated in the Paris region and mainly supplied
residential customers. The measured harmonic voltages were mean values over time intervals of
th
5 minutes. The results obtained for the 5 harmonic voltage are given in Fig. A.8 below.
50-%-values,
95-%-values,
maximum values.
th
Fig. A.8 - 5 harmonic in French LV systems, 1991
Comparing these results obtained in 1991 with those obtained in 2000, the average of the 95-%-
values for the 5th harmonic voltage may be found between 2,5 % and 3 % in 1991, and between
3,5 % and 4 % in 2000. So, in 10 years, a mean increase of about 1 % for the 5th harmonic voltage on
public LV systems can be recognised.
9 Denmark [17]
The THD and the individual harmonics of order 3, 5, 7, 9, 11, 13 and 15 were measured in urban and
rural areas involving 200 delivery points in LV.
The THD is shown in Fig. A.9 below. It may be noticed that the harmonic level is higher in urban than
in rural areas.
rd
For the 3 harmonic, values of 3-4 % were found at several points.
th
With respect to the 5 harmonic, in urban areas higher values exceeding the compatibility level of 6 %
were found during some periods of the observation time of a week.
Fig. A.9 - Distribution of THD values in urban and rural areas in Denmark
10 Spain [25]
Average of measurements in 79 MV substations during 1987 to 1989.
6,00%
5,00%
4,00%
minimum
Vn %
3,00% 95%
maximum
2,00%
1,00%
0,00%
2 3 4 5 6 7 8 9 10 11 12 13
Harmonic order
Comment: Due to the higher increase of the Spanish GPD and electricity consumption as well during
th
the nineties with respect to the European rate, a higher increase of the 5 harmonic and the THD per
year is to be expected. Consequently, the tendency is to reach the compatibility level in a similar way
as in other European countries.
4,5
4
3,5
3
Underground
2,5
Administrative complex
2
Commercial center
1,5
Electric transport
1
0,5
0
THD 100% THD 95% 5th 100% 5th 95%
th
Fig. A.11 - THD and 5 harmonic in different locations with specific type of loads
th
Fig. A.12 - 5 harmonic on French HV busbars 1990, 1994
For 5 substations, the level of harmonic increased between 1990 and 1994. For 4 substations, the
levels remained unchanged. For three of them, the levels decreased due to the fact that the Short
Circuit Power has been increased.
5
5th harmonic
0
21/06/00 0:00 22/06/00 0:00 23/06/00 0:00 24/06/00 0:00 25/06/00 0:00 26/06/00 0:00 27/06/00 0:00 28/06/00 0:00
th
Fig. A.13 - 5 harmonic at 400V-, 30-kV and 110-kV busbar over a week, in Austria, 2000
th
It may be assumed that this evening peak of the 5 harmonic results from the decrease of the load in
th
conjunction with an increase of the share of equipment with a significant 5 harmonic current (e.g. TV
th
sets). The rapid decrease of the 5 harmonic at 10 pm may be assumed to be caused by the switch-
th
off of a large number of equipment with a significant 5 harmonic current on the one side and by the
switch-on of ohmic heating equipment by ripple control as e.g. heating radiators and water boilers.
Residential Estate
5
4,5
4
3,5
3
Un%2,5
2 3rd
1,5 5th
1
7th
0,5
0 THD
00
00
00
00
00
00
00
00
00
00
00
00
:
:
01
03
05
07
09
11
13
15
17
19
21
23
Hour
rd th th
Fig. A.14 - 3 , 5 and 7 harmonic and THD in UK residential estate over 24 hours
Summary: This monitoring survey is part of an ongoing survey at this particular substation. The
survey programme was initiated by the Electricity Council to provide details on the background level of
harmonic distortion present on a typical distribution substation serving a residential environment. This
particular survey was carried out over one week in April 1999. The graph shown represents a typical
set of results over one day of the survey.
th
Conclusions: As expected the dominant harmonic is the 5 harmonic, this follows the typical daily
rd
load pattern. It has its peak of 4.4 % at 19:00 (THD = 4.52 %). There is no coincident rise in the 3
th th
and the 7 harmonics, indicating that the source of the rise in the 5 harmonic might be on an
th
adjacent section of the network. The most likely cause for the rise in the 5 harmonic would be TV
load.
th
The 5 harmonic reaches 4,4 % at “TV peak” hours. According to an averaged increase of 1,5 % /
10 years, U5 would reach the compatibility level of 6 % in 2009.
9: 0
11 5
12 00
13 15
14 30
16 45
17 00
18 15
19 0
21 45
22 00
23 15
0
0
1
3
4
0
3
4
:3
:3
1:
8:
:
:
:
:
:
:
:
:
:
Time of day
rd th th
Fig. A.15 - 3 , 5 and 7 harmonic and THD in UK residential area over 24 hours
This office is an international IT support centre, that operates 24 hours per day seven days per week
and is located on an industrial estate. The load is mainly comprised of IT equipment and fluorescent
lighting. The survey was prompted by a complaint from the consumer to the local Electricity Supply
Company of “VDU screen ‘wobble’ and excessively hot neutral conductor”. The neutral current was
approximately 150 % of the highest phase current; this was due to third harmonic current and severe
out of balance loading. The VDU interference was caused by stray neutral current returning to earth
via the structure of the building; producing local EM fields around the building supports.
th
Conclusions: The level of the 5 harmonic and hence THD, is high throughout the day, and often
exceeds the 6 % compatibility limit for that harmonic order. In this instance the consumer paid for
neutral uprating, load balancing and anti EMF hoods for affected VDUs.
2,5
2 3rd
Un% 5th
1,5
7th
1 THD
0,5
0
00
15
30
45
00
15
30
11 5
12 0
13 5
14 0
16 5
17 0
18 5
19 0
21 5
22 0
23 5
0
:0
:1
:3
:4
:0
:1
:3
:4
:0
:1
:3
4
1:
2:
3:
4:
6:
7:
8:
9:
Time of day
rd th th
Fig. A.16 - 3 , 5 and 7 harmonic and THD in a UK plastic extrusion factory
over 24 hours
Summary: This factory is located within a large housing estate. The site load is comprised of several
extrusion machines that are driven by dc motors, supplied via six pulse drives. The prompt for this
survey came from the consumer applying to the local Electricity Supply Company for permission to
connect a new dc drive. The monitoring instrumentation was connected to the metering VT and CTs.
Conclusions: The factory starts production at 07:00, this is characterised by the rise in the
th th
5 harmonic at about this time. The 5 harmonic continues to rise throughout the day, reaching a peak
of 3.19 %, at about 21:00, this is most likely due to TV load within the surrounding residential houses.
The planning limit used at 11 kV in the UK is 3 % THD. Therefore it was agreed to allow the connec-
tion of the new drive on the basis that a system monitoring survey is to be completed after the instal-
lation, and if deemed necessary the consumer will install filters to reduce the harmonic emissions.
4 ,5
3 ,5
U n %
2 ,5
1 ,5
3 rd
1
5 th
0 ,5 7 th
THD
0
01:00
03:00
05:00
07:00
09:00
11:00
13:00
15:00
17:00
19:00
21:00
23:00
hour
rd th th
Fig. A.17 - 3 , 5 and 7 harmonic and THD in UK residential housing estate over
24 hours
Summary: This monitoring survey is part of an ongoing survey at this particular substation. The sur-
vey programme was initiated by the then Electricity Council to provide details on the background level
of harmonic distortion present on a typical distribution substation serving a residential environment.
This particular survey was carried out over one week in June 1999. The graph shown represents a
typical set of results over one day of the survey.
th
Conclusions: As expected, the dominant harmonic is the 5 harmonic, this follows the typical daily
rd
load pattern. It has its peak of 4.4 % at 19:00 (THD = 4.76 %). There is no coincident rise in the 3
th th
and the 7 harmonics, indicating that the source of the rise in the 5 harmonic might be on an
th
adjacent section of the network. The most likely cause for the rise in the 5 harmonic would be TV
load.
5
Un%
4
3
2
1
0
00
00
00
00
00
0
:0
:0
:0
:0
:0
:0
:0
1:
3:
5:
7:
9:
11
Houry
13
15
17
19
21
23
rd th th
Fig. A.18 - 3 , 5 and 7 harmonic and THD in a UK IT insurance office over 24 hours
Summary: This survey was carried out by the local Electricity Company for the reason of identifying
the level of harmonic distortion that is present on a typical distribution substation. The main load on
this substation is an insurance office, within which there is a large amount of IT equipment and fluo-
rescent lighting. The duration of the survey was one week, and it was carried out during February
2000.
Conclusions: The office starts work at 08:00, this is characterised by the rise in harmonic distortion
rd th
seen at this time. The coincident rise in the 3 and 5 harmonics indicates that the source of the dis-
tortion is connected to the LV network of this substation, the daily peak is reached at approximately
rd
14:00. The office continues working until about 7 p.m., again this aligns with the recordings of the 3
th th
and 5 harmonics. However the 5 harmonic continues to remain high as a result of domestic load
th
connected to this and adjacent substations. The level of the 5 harmonic is high throughout the
working day, and often exceeds the 6 % compatibility level. However the nature of harmonic distortion
is such that it could be some time before a high level of distortion manifests itself in equipment failure
or other problems occur.
19 UK IT office [20]
IT Office 3rd
6
5th
5
7th
4 THD
Vn% 3
2
1
0
00
15
30
45
00
15
30
11 5
12 0
13 5
14 0
16 5
17 0
18 5
19 0
21 5
22 0
23 5
0
4
:0
:1
:3
:4
:0
:1
:3
:4
:0
:1
:3
Hour
1:
2:
3:
4:
6:
7:
8:
9:
rd th th
Fig. A.19 - 3 , 5 and 7 harmonic and THD in a UK IT office over 24 hours
Summary: This office is supplied via a discrete pole mounted transformer rated at 200 kVA, hence
the low fault level. The office is situated in residential housing estate in a semi-urban area. The survey
was prompted by the consumer complaining to the local Electricity Company of interference on their
telephone lines. The office is the head office of a national and international equipment supplier. The
load is mainly composed of IT equipment and fluorescent lighting. This survey was carried out during
one week in early March 1995.
th
Conclusions: The office starts work at about 08:00 and this is characterised by the rise in the 5 and
th th
7 harmonics. Throughout the day the 5 harmonic continues to increase until the office closes for the
day at about 18:00. The sudden rise at around 18:00 is thought to be due to additional lighting being
th
switched on and the work of cleaning staff. The back- ground level of the 5 harmonic increases in the
evening as a result of the surrounding residential TV load. Replacing the PABX system by one with a
higher immunity rating, cured the problem of telephone interference.
Flicker
Pst 99% Plt Plt 95%
A 0.39 0.32 0.22
B 0.53 0.46 0.37
C 0.62 0.53 0.48
D 2.55 2.03 1.67
0.37 0.32 0.27
F 0.84 0.63 0.55
G 0.22 0.32 0.18
H 0.99 0.78 0.51
I 0.94 0.8 0.62
J 2.28 1.66 0.93
Austria
12,0
8,0
6,0
4,0
2,0
0,0
0,2 % -
0,4 % -
0,6 % -
1,2 % -
< 0,2 %
> 1,4 %
0,8 % - 1
1 % - 1,2
Pst
0,4 %
0,6 %
0,8 %
1,4 %
< 0,2 0,2-0,4 0,4-0,6 0,6-0,8 0,8-1,0 1,0-1,2 1,2-1,4 > 1,4
%
Pst ranges
Comment: The unbalance in MV and in 110 kV networks is low. The magnitude of unbalance k95
always satisfies the criteria of EN 50160 despite of the fact that in individual cases (see diagrams) the
unbalance k95 amounts up to 1.6 % (22 kV) or 1.4 % (110 kV). These exceptional cases concern the
points of the network situated in the vicinity of supplying the a.c. single-phase railway traction or,
maybe, also when combined with a small short-circuit power or a non-standard scheme of the net-
work. However, the unbalance is even higher - up to 3.6 % or 3.3 % respectively for the rest 5 % of
the week interval but these values are being tolerated by EN 50160 [1].
Therefore, in 1981 IEC published a report on a reference network impedance [28]. This was
necessary for the definition of the emission limits, in particular for mass product equipment
whose final conditions of application are not foreseeable (neither for manufacturers nor for
network operators), especially for flicker and voltage variations. For industrial or large
commercial installations, the reference network impedance is not appropriate; if problems
occur, these are usually treated on a case by case basis taking into account the real network
impedance - measured or assessed - at the PCC.
B.2 DEFINITION
The value recommended by IEC 60725 [28] was as follows for three-phase four-wire 230/400 V
supplies:
Tab. B.1 - Percentage of consumers with less than the stated impedance
in Ohms at 50 Hz
A more recent study was made in Spain at 1440 LV PCCs, of which 1301 were mainly situated in a
citizen area (A) and 139 were mainly situated in rural areas (B). The choice of PCCs was made
randomly.
Spain 98 % 95 % 90 % 85 %
A 1301 PCCs 1,58 + j 0,45 0,86 + j 0,31 0,58 + j 0,23 0,44 + j 0,20
B 139 PCCs 2,12 + j 0,61 1,59 + j 0,40 1,22 + j 0,23 1,03 + j 0,19
A+B 1440 PCCs 1,75 + j 0,46 1,05 + j 0,32 0,67 + j 0,23 0,51 + j 0,20
Tab. B.2 - Percentage of consumers with less than the stated impedance in Ohms
A study made in France in 1987 shows a different situation for overhead networks, cables networks
and twisted networks. The following main results were obtained:
Tab. B.3 - System impedance in Ohms in overhead, cable and twisted networks
Over all, the future Section 30 of IEC 61000-4 [9] will define the methods for measurement and inter-
pretation of results for power quality parameters in 50/60 Hz a. c. power supply systems.
C.1 HARMONICS
IEC 61000-4-7 [29] defines the measurement instrumentation intended for testing individual items of
equipment related to emission limits given in product standards as well as for the measurement of
harmonic currents and voltages in supply systems.
Two classes of accuracy (I and II) are considered, to permit the use of simple and low-cost
instruments.
The general structure of the instrument is based on the use of the Discrete Fourier Transform DFT
technique. It comprises an input circuitry with anti-aliasing filter, A/D-converter including sample and
hold-unit, synchronisation and window-shaping-unit if necessary, DFT-processor providing the Fourier
coefficients and it is complemented by the special parts devoted to current assessment and/or voltage
assessment.
Assessment of compliance with emission limits is performed by statistical handling of the data accor-
ding to the conditions given in the relevant standards, such as IEC 61000-3-2 [30].
The measurement of voltage dips involves recording the value of the voltage during the dip, of
the duration of the event, and of the number of events within the measurement period. In
order to record these values and to enable results to be compared, it is necessary to adopt
certain rather arbitrary conventions concerning reference voltage and threshold.
In case of multi-phase measurements, voltage dips whose durations overlap in time have
conventionally been counted as a single event . In some cases the practice has been to
measure the duration from the instant at which the first phase or line voltage falls below the
start threshold to the instant at which the last phase or line voltage rises to or above the end
threshold.
In order to deal with the complexity of non-rectangular dips it would be possible to designate
several gradations on the voltage scale and to record the duration for which the voltage is at
or below each such mark. This would result in the recording of several residual data pairs of
voltage level and duration for describing each voltage dip.
Table C.1 shows the tabulation developed by UNIPEDE. For a given measurement site, each
cell is intended to contain the number of voltage dips of the corresponding depth and duration
occurring within a specified period, usually one year.
Duration (seconds)
Residual Voltage
u% of U ref 0.01 < Dt 0.02 < Dt 0.1 < Dt £ 0.5 < Dt £ 1 < Dt 3 < Dt 20 < Dt £ 60 < Dt £
£ 0.02 £ 0.1 0.5 1 £3 £ 20 60 180
90 > u ³ 85
85 > u ³ 70
70 > u ³ 40
40 > u ³ 10
10 > u ³ 0
e Measurement results in the first column and first row are likely to be inflated by transients and load
fluctuations, respectively.
A similar table is used to compile the results from all the sites in the measurement campaign.
In this case each cell can contain
- a percentile (typically 95) of the number of dips recorded in that cell for all sites
- the maximum number recorded in the cell
- the average number for the cell for all sites
- or other statistic.
If several types of networks are involved, it is appropriate to produce a separate table for
each type. For example, measurement results from overhead lines should be distinguished
from underground networks.
C.4 UNBALANCE
To apply the symmetrical component method.
D.1 HARMONICS
Compatibility levels are given in
- IEC respectively ENV 61000-2-2 for LV networks [12] and
- IEC 61000-2-12 for MV networks [13]
- IEC 61000-2-4 [32], which specifies compatibility levels for harmonic voltages at the IPC (in-
plant point of coupling) for industrial installations, establishing three classes according to
required immunity of equipment.
These are set by experience in order to give guidance for setting appropriate emission limits and
immunity requirements, also aiming at reduction of complaints about mal-operation (both of loads and
of system components) to a minimum. By definition, compatibility levels do not represent rigid limits,
but are expected to be exceeded with a low probability and only to a small percentage of places.
rd th
Compatibility levels for the 3 respectively 5 order of harmonic voltage are 5 % respectively 6 % and
8 % for the Total Harmonic Distortion Factor (THD).
Note: Additionally to the compatibility levels, so-called planning levels have been defined; they repre-
sent levels, set by the network operator for their single supply network, which is aimed at not
getting exceeded, being lower than the standardised compatibility level.
IEC 61000-3-6 [33] establishes planning levels for HV and EHV networks as well as the con-
ditions for connecting disturbing loads producing harmonics.
- IEC 61000-3-2 [30], which specifies limits for harmonic current emissions applicable to electri-
cal and electronic equipment having an input current up to and including 16 A per phase, and
intended to be connected to public low-voltage distribution systems. The tests according to
this standard are type tests.
- IEC 61000-3-12 [34]: the recommendations of this -- at present -- draft standard, are
applicable to electrical and electronic equipment with a rated input current exceeding 16 A, up
to 75 A per phase and intended to be connected to public low-voltage ac distribution systems.
- IEEE 519 [35]: establishes limits on harmonic currents and voltages at the point of common
coupling (PCC) or at the point of metering.
- IEC 61000-4-13 [36], which proposes different immunity levels in accordance with different
performance criteria.
A flicker value is also given in EN 50160 [1], with Plt lower or equal to 1 during 95 % of the time.
For emission, the limits for equipment connected to the LV supply network are given in:
The immunity levels as well as the performance criteria and the immunity measurement method are
specified in:
“Voltage dips are generally caused by faults occurring in the customers’ installations or in the public
distribution system. They are unpredictable, largely random events. The annual frequency varies
greatly depending on the type of supply system and on the point of observation. Moreover, the
distribution over the year can be very irregular.
Indicative values: Under normal operating conditions the expected number of voltage dips in a year
may be from up to a few tens to up to one thousand. The majority of voltage dips have a duration
less than 1 s and a depth less than 60 %. However, voltage dips with greater depth % of Un can
occur very frequently as a result of the switching of loads in customers’ installations.”
Also IEC 61000-2-2 “Compatibility levels for low-frequency conducted disturbances and signalling in
public low-voltage power supply systems” [12] deals with the voltage dip issue.
As far as voltage dips are concerned, this standard will be completed by the future standard IEC
61000-2-8 “Voltage dips and short interruptions on public electric supply with statistical results” [18].
The present report reflects mainly this future standard.
EN 61000-6-1 “Generic immunity standard: Residential, commercial and light industry” [40] and
EN 61000-6-2 “Generic standard – Immunity for industrial environments” [41] specify minimum
immunity requirements:
- For a 10 ms duration and 30 % depth of a dip, the apparatus shall continue to operate as
intended after the test. No degradation of performance or loss of function is allowed below a
performance level specified by the manufacturer. During the test, degradation of performance is
however allowed. No change of actual operating state or stored data is allowed. This is called
performance criterion B.
- For a 100 ms duration and 60 % depth, temporary loss function is allowed, provided the function
is self-recoverable or can be restored by the operation of the controls. This is called per-
formance criterion C.
D.4 UNBALANCE
EN 50160 [1] gives values for unbalance in LV and MV supply networks.
Due to these three standards, for LV and MV systems, in normal operating conditions, 95% of
the 10 minutes values of t (over the week) shall be less than 2%. In some areas where single or
bi-phase lines are existing, the limits could be 3 %.
IEC 61000-4-27 [31] defines the immunity levels of equipment from unbalance as well as the
measurement method.
Note: This standard was submitted for enquiry concerning a possibly given need for revision. As at
present, available experience does not seem to justify any modification, it has been decided to
leave that standard unmodified.
[3] UNIPEDE DISNORM 12, “Definitions of the Physical Characteristics of Electrical Energy
Supplied by Low and Medium Voltage Public Systems”, NORMCOMP, September 1989
[4] UNIPEDE Application Guide to the European Standard EN 50160 on “Voltage Charac-
teristics of Electricity Supplied by Public Distribution Systems”, 230.02 NORMCOMP,
Ref.: 23002Ren9530, July 1995
[5] UNIPEDE Report “Measurement Guide for Voltage Characteristics”, NORMCOMP, 1995
[6] UNIPEDE Report “Voltage Dips and Short Interruptions in Electricity Supply Systems”,
(50.DISDIP), Ref.: 91 en 50.02
[7] UNIPEDE Report on “Expected Harmonic Distortion in LV Networks”, NORMCOMP,
October 1995
[8] UNIPEDE: “Harmonic Distortion in Public Electricity Distribution Systems: Contribution
of Class D Equipment,” February 1997
[9] Draft IEC 61000-4-30 Ed. 1.0): “Electromagnetic compatibility (EMC) – Part 4-30: Testing and
measurement techniques -- Power quality measurement methods” (77A/356/CDV)
[10] IEC 77A/367/FDIS (revision of 61000-2-2)
[11] Study of the extra cost due to harmonics in HQ networks and equipment. R. Bergeron. 1998.
[12] IEC/ENV 61000-2-2:1990-05: “Compatibility levels for low-frequency conducted disturbances
and signalling in public low-voltage power supply systems” (second edition under way (RDIS))
[13] Draft IEC 61000-2-12 Ed. 1.0: “Compatibility levels for low-frequency conducted disturbances
and signalling in public medium-voltage power supply systems” (77A/321/NP, Approved new
work)
[14] UIE - Guide to Quality of Electrical Supply for Industrial Installations. Part 5 : Flicker and Voltage
Fluctuations
[15] IEC 61000-4-15 (1997-11): “Electromagnetic compatibility (EMC) – Part 4: Testing and
measurement techniques – Section 15: Flickermeter – Functional and design specifications”,
superseding IEC 60868 and IEC 60868-0 [see 17]
Draft Amd. 1 Ed. 1.0 under way „Amendment to add specifications for flickermeters for 120 V
systems and more closely define the specification of a filter in the instrument according to
standard IEC 61000-4-15” (77A/341/CVD)
[16] IEC/TR2 60868 (1986-09): “Flickermeter – Functional and design specifications” (see also [16])
+
IEC/TR2 60868-am1 (1990-06), IEC/TR2 60868-0 (1991-05) “Flickermeter – Part 0: Evaluation
of flicker severity”
[17] The Power Quality in Denmark. P. Jorgensen et al. CIRED 1999
[18] Draft IEC 61000-2-8 TR3 Ed. 1.0: “Voltage dips and short interruptions on public electric power
supply system with statistical measurement results” IEC 77A/329/CD (future
[19] UIE-Guide to the quality of electrical supply for industrial installations. Part 4: Voltage
unbalance.
[34] Draft IEC 61000-3-12 Ed. 1.0: “Electromagnetic compatibility (EMC) – Part 3-12: Limits for
harmonic currents produced by equipment connected to public low-voltage systems with input
current < 75 A per phase and subject to restricted connection” (77A/325/CD)
[35] IEEE 519 “Recommended Practices and Requirements for Harmonic Control in Electric Power
Systems”
[36] Draft IEC 61000-4-13 Ed. 1.0 “Electromagnetic compatibility (EMC) – Part 4-13: Testing and
measurement techniques – Harmonics and interharmonics including mains signalling at a.c.
power port, low frequency immunity tests “ (77A/323/CDV)
[37] IEC 61000-3-3 (1994-12): “Electromagnetic compatibility (EMC) – Part 3: Limits –Section 3:
Limitation of voltage fluctuations and flicker in low-voltage supply systems for equipment with
rated current £ 16A”
+
IEC 61000-3-3-am1 (2001-01)
[38] IEC 61000-3-11 (2000-08): “Electromagnetic compatibility (EMC) – Part 3-11: Limits – Limitation
of voltage changes, voltage fluctuations and flicker in public low-voltage supply systems –
Equipment with rated current £ 75 A and subject to conditional connection”
[39] IEC 61000-4-14 (1999-02) + IEC 61000-4-14-am1 (2001-07): „Electromagnetic compatibility
(EMC) – Part 4-14: Testing and measurement techniques – Voltage fluctuation immunity test“
[40] IEC 61000-6-1 (1997-07): “Electromagnetic compatibility (EMC) – Part 6: Generic standards –
Section 1: Immunity for residential, commercial and light-industrial environments”
- Phase 1 Report on European Residential Harmonics, 2001, The Low Frequency Emission
Industry (LFEIC)
- “Why it is necessary to control harmonic and flicker emission”. E.A. G. Finlay. July 1999. Ref.
TF5a-9925
- Indices for Assessing Harmonic Distortion from Power Quality Measurements: Definitions and
Benchmark Data. D.D. Sabin et al. CC02-9821 (1998).
- Results of EEI Harmonic Survey for IEC
- Draft IEC 61000-1-4 Ed. 1.0: „Electromagnetic compatibility (EMC) – Part 1-4: Rationale for
limiting power-frequency conducted harmonic and inter-harmonic current emissions from
equipment, in the frequency range up to 9 kHz” (77A/339/CD)
- Draft IEC 61000-3-1 TR Ed. 1.0): „Electromagnetic compatibility (EMC) – Part 3-1: Overview of
emission standards and guides” (Approved new work)
- „Empfindlichkeit von elektrischen Geräten gegenüber Spannungseinbrüchen und Kurzzeit-
Unterbrechungen“, Studie, im Auftrage des Verbandes der Elektrizitätsunternehmen Öster-
reichs (VEÖ) durchgeführt an der TU Wien, Institut für Elektrische Anlagen und Energiewirt-
schaft, November 2001
- ERA Technology “How to improve Voltage Dip Immunity in Industrial and Commercial Power
Distribution System”. ERA Report 99-0632R, August 1999