21, rue d'Artois, F-75008 Paris
http://www.cigre.org
C4-107
Session 2004
© CIGRÉ
VOLTAGE DIPS AND SHORT INTERRUPTIONS – DIFFERENT STRATEGIES IN
CONTRACT FOR THE ELECTRIC POWER SUPPLY
Grzegorz MATUSZ, Tomasz MAJ, Zbigniew HANZELKA*, Wladyslaw LOZIAK
AGH – University of Science and Technology
(Poland)
1.
INTRODUCTION
The provisions related to the quality of supply are presently being incorporated into contracts for the
electric power supply in Poland. It is important that their formulation will not create a privileged
position for any party of the electric power sales-buying transaction - the supplier or the consumer. For
the purpose of this task, during seven months have been carried out measurements in one of
distribution companies in the south of Poland - the PDC Cracow. The measurements were carried out
in three measuring points at the coupling of the transmission (220 kV) and distribution system
(110 kV). In effect of these measurements has been created the database which contains recorded
events in voltage: dips, swells and interruptions. These data were used for the analysis of supply
conditions in the considered points of power delivery. The following effects have been investigated:
selection of various factors which describe the voltage dip, selection of the dip threshold voltage to
which the measuring thresholds are referred, selection of threshold values which define the
disturbance and methods for aggregation of the measurement results. The considerations are restricted
to single-event and single-site indices only.
2. THE MEASURING SYSTEM
The PDC operated 110 kV distribution system is
connected with neighbouring operators' distribution
systems and with the 220 kV transmission system by
means of three autotransformers, 160 MVA each,
installed at three substations (a) LUBOCZA – industrial
and household customers, (b) WANDA - predominantly
industrial customers, (c) SKAWINA – near to heat and
power generating plant - Fig. 1. At these points have
been installed instruments for the power quality factors
measurement. The measuring instruments were provided
Fig. 1. Block diagram of the PDC Cracow
with modems, which enabled transmission of recorded
distribution system connections to the transmission
data via public telecommunications network and their
system
acquisition in the main computer, located at the AGH UST.
*
hanzel@uci.agh.edu.pl
3. RECORDED VOLTAGE DIPS AND SHORT SUPPLY INTERRUPTIONS
3.1. Aggregation methods
Tables 1-3 give the summary of recorded voltage dips and short interruptions. Most of them occurred
within a short time interval. Each disturbance is treated separately. Three-phase or two-phase events
are counted as three or two independent disturbances, respectively.
Table 1. Voltage dips without aggregation (WANDA) - L1/L2/L3
Dips [%]10 - 100ms 100 - 500ms500 m s- 1s1 - 3s3 - 20s20 - 1min
10- 15
1/3/0
0/0/0
0/0/0 0/0/0 0/0/0 0/0/0
15 - 30
1/0/4
2/3/3
1/1/1 0/0/0 0/0/0 0/0/0
30 - 60
0/1/3
0/0/1
0/0/1 0/0/0 0/0/0 0/0/0
60 - 90
1/1/1
0/0/0
0/0/0 0/0/0 0/0/0 0/0/0
90 - 100 1/2/3
0/0/0
0/0/0 0/0/2 2/1/3 0/1/0
Number of recorded voltage dips: 44
1-minute aggregation – the number of
recorded events: 25
3-minute aggregation - the number of
recorded events: 23
Phase aggregation - the number of recorded
events: 38
Table 2. Voltage dips without aggregation (SKAWINA) - L1/L2/L3
320s
0/0/0 0/0/0
0/0/0 0/0/0
0/0/0 0/0/0
0/0/0 0/0/0
0/0/0 0/0/0
Dips [%] 10 - 100ms 100 - 500ms 500ms - 1s 1 - 3s
10 - 15
15 - 30
30 - 60
60 - 90
90 - 100
2/3/2
0/0/0
0/0/0
1/1/0
2/4/2
0/1/1
0/0/0
0/0/0
0/0/0
0/0/0
0/0/0
0/0/0
0/0/0
0/0/0
0/0/0
Number of recorded events: 19
20s 1min
0/0/0
0/0/0
0/0/0
0/0/0
0/0/0
1- and 3 minute aggregation - the number
of recorded events: 15
Phase aggregation - the number of recorded
events: 13
Table 3. Voltage dips without aggregation (LUBOCZA) - L1/L2/L3
Dips [%]10 - 100ms 100 - 500ms500ms - 1s1 - 3s3 - 20s20s - 1min
10 - 15
0/0/3
2/2/4
0/0/0 0/0/0 0/0/0 0/0/0
15 - 30 1/0/11
3/2/0
0/0/0 0/0/0 0/0/0 0/0/0
30 - 60
0/1/2
0/1/2
0/0/0 0/0/0 0/0/0 0/0/0
60 - 90
1/0/4
0/0/7
0/0/1 0/0/7 0/0/11 0/0/5
90 - 100 1/0/1
0/0/0
0/0/0 0/0/0 0/1/0 0/0/0
Number of recorded events: 73
1-minute aggregation - the number of
recorded events: 46
3-minute aggregation - the number of
recorded events: 44
Phase aggregation - the number of
recorded events: 71
Figure 2 summarizes the results of the influence of different types of aggregation on the number of
voltage dips in given measuring points. It can be seen that the use of phase and time (1- and 3-minute)
aggregation significantly reduces the number of disturbances. This does not concern the LUBOCZA
substation, where a particularly large number of disturbances were recorded in a single phase (L3).
Table 4. Number of dips depending on
the selected threshold value
(% 110 kV)
Threshold
90 80 70 60
SKAWINA 19 6 6 6
WANDA
44 35 24 21
LUBOCZA 74 59 45 42
Dependence of the number of dips on the selected threshold
value, which defines the disturbance is presented in table 4 and in
Figure 3. For each location can be seen the reduction of the
number of disturbances with an increase in the threshold voltage.
Figure 4 shows the diagram of the network supplying WANDA and LUBOCZA substations. If a
disturbance occurs in the point, 1 it will be also transferred to points 2 and 3. In this case the use of
place aggregation can be taken into consideration, that is, voltage dips occurring simultaneously on
different lines supplying the same user can be aggregated.
2
comparison of aggregation methods
80
70
80
60
60
number of dips
number of dips
70
50
40
30
20
10
50
40
30
20
0
no aggregation
phase aggregation
1-minute timeaggregation
3-minute timeaggregation
10
3-minute timeaggregation and
phase aggregation
0
Skawina
aggregation method used
Wanda
Lubocza
Substations
Skawina
Lubocza
90%
Wanda
Fig. 2. Dependence of the number of dips on the
selected method of aggregation
1
220kV
GPZ Wanda
110kV
110kV
2
80%
70%
60%
Fig. 3. Dependence of the number of dips on the
selected threshold value, which defines the disturbance
The sum of aggregated dips for the substations
WANDA and LUBOCZA – after phase and time
(3-minute) aggregation – is: 16 + 41 = 57.
Assuming the place aggregation for these two
GPZ Lubocza substations, the total number of contractual
disturbances will be reduced to 44, for the 13
simultaneous dips have been recorded on these
substations.
3
Fig. 4. Diagram of the network supplying the
measuring points WANDA and LUBOCZA
Number of dips recorded in phase-to-phase
voltages: WANDA – 32; SKAWINA - 14;
LUBOCZA – 52.
3.2. Reference voltages for voltage dips recording
The residual voltage is normally expressed in percent or per unit values, with respect to the nominal or
declared voltage at the considered point of a system. On HV systems, it is more advantageous to
measure voltage dips with respect to the voltage UX, preceding the disturbance, determined in a
continuous manner in the assumed measuring window. The measurements were performed at the
assumed, constant reference voltage 110 kV. Only those disturbances where the voltage decreased
below 90% of 110 kV were recorded. It was also checked whether these disturbances might have been
recorded if the declared voltage (120 kV), or the voltage determined in the last measuring window,
would have been used for the reference voltage. That’s why the number of dips for the reference
voltage 120 kV and that for the voltage preceding the disturbance, are different; they are denoted with
asterisk (*).
Table 5. The influence of various reference voltages on the
number of dips (WANDA/LUBOCZA/SKAWINA)
% of the nominal voltage 110 kV
Reference
voltage [kV]
90
80
70
60
110k
44/74/19 35/59/6 24/45/0 21/42/0
120
44*/74*/19* 41/67/17 30/54/4 21/42/0
UX
44*/74*/19* 40/64/12 29/50/4 21/42/0
With the lower thresholds defining a dip (80
and 70 % UN), the difference in numbers of
recorded dips can be noticed. No reference
voltage less than 110kV, measured
immediately before a dip, has been recorded
over the entire period of measurements
4. VOLTAGE DIPS INDICES (SELECTED)
4.1. Voltage-sag energy [12, 13]
In the event the residual voltage (U) and duration (T) of a dip are known, and assuming that the r.m.s.
voltage value during the dip is constant, the disturbance can be described using the relation (1), where
3
Uref is the reference voltage. For multi-channel events the voltage-sag energy is defined as the sum of
the voltage-sag energy in the individual channels: EUS = ( EUS − L1 + EUS − L 2 + EUS − L3 ) . Eexample values
of indices for the substation WANDA are shown in Figure 5.
wartość wskaźnika Evs 3 faz [s
35
30
Fig. 5. Example sag energies for the
three-phase system – WANDA
25
20
15
⎡ ⎛
U
EUS = ⎢1 − ⎜
⎢ ⎜ U ref
⎣ ⎝
10
5
0
0
5
10
15
20
25
30
35
40
numer zaburzenia
⎞
⎟
⎟
⎠
2⎤
⎥T
⎥
⎦
(1)
The values of energy indices are varying within a broad range: from several thousandth to over thirty.
This kind of index may be dominated by long-duration dips: a single long disturbance can be
equivalent to many short-duration dips.
For description of the measuring point in the assumed time-interval have been proposed the so-called
Sag Energy Index (SEI), that is the sum voltage-sag energies of all qualified disturbances at the given
measuring point and the given time-interval – n: SEI =
n
∑E
US −i
. I These indices are usually
i =1
determined over a one-month or one-year interval.
The Average Sag Energy Index (ASEI) is the
average of voltage sag energies for all
qualified events measured at a given site
during a given period - Table 5. The value of
n
the average sag energy index indicates how
1
ASEI =
EUS −1 1.25421
2.49405
0.001948
deep and long voltage dips most frequently
n i =1
occur at a given point of the network.
The ASEI is dependent on the triggering of the monitor. A sensitive setting will result in a large
number of shallow events (with a low sag energy) and this in a lower value for ASEI. The SEI on the
other hand will increase for sensitive setting of the monitor. To compare results from site to site and
from one period to another, a standardized trigger setting needs to be defined.
Most of all long-duration and deep voltage dips recorded at the substation LUBOCZA, hence the ASEI
value is so high. The least number of relatively shallow dips have been at the substation SKAWINA,
this fact manifests itself in the value of the index.
Table 5. Summary of sag energy indices
Indices
WANDA LUBOCZA SKAWINA
SEI
90.30355 179.572
0.140264
∑
4.2. The EPRI indices
The EPRI indices are summarized in Table 6; their values for the considered substations and for the
entire period if measurement are given in Table 7. The value of voltage dips weighting coefficient has
been assumed unity for all locations.
Table 6. The comparison of the indices EPRI
< 90 %
< 80 %
< 70 %
< 50 %
< 10 %
0.5 cycle
– 60 s
SARFI90
SARFI80
SARFI70
SARFI50
SARFI10
0.5 cycle – 0.5 s – 3 s
0.5 s
SIARFI90 SMARFI90
SIARFI80 SMARFI80
SIARFI70 SMARFI70
SIARFI50 SMARFI50
SMARFI10
3 s – 60 s
STARFI90
STARFI80
STARFI70
STARFI50
STARFI10
Table 7. Aggregated indices of voltage dips acc. to
Table 6 (LUBOCZA/ SKAWINA/ WANDA)
0.5 cycle – 0.5 cycle – 0.5 s – 3 s – 60
Dips
60 s
0.5 s
3s
s
< 90 % 74/19/44
50/17/33 7/2/4 17/0/7
< 80 % 59/6/35
35/4/24
7/2/4 17/0/7
< 70 % 45/6/24
21/4/14
7/2/3 17/0/7
< 50 % 41/6/18
17/4/9
7/2/2 17/0/7
< 10 %
3/0/14
2/0/7
1/0/7
4
Assuming different voltage dips weighting coefficient (e.g. depending on the number of users, their
powers, the importance of loads at the given point, etc.) these coefficient can be assigned an additional
meaning, which is significant in terms of a contract.
4.3. The index Se referred to the reference characteristic [13]
The index (Se), defined by the relation (2), is determined from the known: the voltage p.u. value and
duration of disturbance where: U – residual voltage during the voltage dip of duration d; Uref (d) - the
voltage dip magnitude on the reference characteristic (e.g. CBMEA, ITIC, SEMI) for the disturbance
duration d.
For the disturbances on the curve, the index assumes the value 1 – Fig. 6. For the disturbances above
the characteristic, the index value is less than 1, below - greater than 1. For the disturbances of the
voltage value (per unit) equal 1, the index assumes the value zero. The longer is the duration of a
disturbance and smaller the residual voltage, the greater is the index value. The values of coefficients
in Table 8 represent the numerical description of the situation at given substations. Figure 7 shows
example values of the coefficient Se , determined for the substation WANDA. The values of index are
represented on vertical axes, on horizontal axes are ordinal number of subsequent disturbances.
12
wartość współczynnika Se
10
8
6
4
2
0
0
5
10
15
20
25
30
35
40
45
50
numer porządkowy zaburzenia
Fig. 6. The index of the voltage dip „severity”, referred to
the reference characteristic (solid line) for disturbances of
various duration and residual voltage [2]
GPZ
Σ Se
Table 8. Total value of index Se
LUBOCZA SKAWINA WANDA TOTAL
191.023
11.71262 107.542 310.2776
Fig. 7 The values of the coefficient Se at the
substation WANDA
Se =
1−U
1 − U ref (d )
(2)
4.4. Sag Score [14]
The sag score is determined on the basis of so-called qualified dips. The threshold of dip recording is
75% of the nominal voltage and is aggregated over 15-minute period. The sag score calculated from
the formula: sag score = 1 – (UA + UB + UC)/3 where: UA, UB, UC – p.u. value of respective phase
voltages during the disturbance.
The following values of sag score target (total of the all sag scores for the given measuring point
compared with the value set forth in a contract) have been obtained: WANDA - sag score target:
3.222527 (8 qualified dips); LUBOCZA - sag score target: 4.880257 (22 qualified dips); SKAWINA:
- sag score target: 0.084377 (1 qualified dip).
On the basis of the sag score value, which varies over the range 0.0833 to 1, it could be only possible
to estimate very roughly how “dangerous” a given voltage dip was. The range of aggregation time is
too long for actual assessment of a given voltage dip: for instance, the same value can be obtained for
a dip with duration of several tens milliseconds and a dip with duration of several seconds, or at least,
several dips within a specified interval of time.
5
4.5. Weighting coefficients [7, 10, 14]
Tables 9 and 10 present the conception of equivalent, weighted voltage dips (calculated on annual
basis), which are the basis for determination of mutual financial commitments of the supplier and
consumer of electric power. The actual, aggregated voltage dips index, determined on this basis, (total
of weighted dips acc. to the tables) can be subtracted from the value set forth in the contract and the
difference is multiplied by the agreed compensation rate.
Table 9. The concept of the voltage dips
weighting coefficients (No 1)
Dips
Weight SKA. LUB. WAN
[%]
coeff.
.
0
1
0/0*
4/4
6/6
<50%
1
0/0 38/38
3/3
50-70
0.7
0/0
4/2.8 5/3.5
70-80
0.4
10/4 17/6.8 16/6.
4
80-90
0.1
9/0.9 11/1.1 4/0.4
TOTAL
5.1
52.7
19.7
*
number of dips /weighted dips
Table 10. The concept of weighting coefficients
(No 2)
Voltage dip
magnitude
[%Un]
Aver.
20100ms
0.06
100500ms
0.3
Duration
500ms1-3s
1s
0.75
0.875
3-20s
20-60s
0.875
0.875
15 - 10
0.125
0.0075
0.0375
0.0938
0.1094
0.1094
0.1094
30 - 15
60 - 30
99 - 60
0 - 99
0.225
0.45
0.795
0.995
0.0135
0.0270
0.0477
0.0597
0.0675
0.1350
0.2385
0.2985
0.1688
0.3375
0.5963
0.7463
0.1969
0.3938
0.6956
0.8706
0.1969
0.3938
0.6956
0.8706
0.1969
0.3938
0.6956
0.8706
Total of
weighted dips
SKAWINA:
0.9571
LUBOCZA:
21.0633
WANDA:
10.0611
4.6. Tabele ESKOM [8, 9, 14]
Table 11. Classification of voltage dips according to
the ESKOM table
1
2
3
4
5
6
Number of dips per year
Voltage
Voltage dip window class
magnitude
Z
T
S
X
Y
110kV 8/0/8* 15/0/9 3/2/8 17/4/7 15/13/9
* LUBOCZA/SKAWINA/WANDA
The threshold value of a voltage dip is 0.9 of the
declared voltage. The depth of a voltage dip equals
the maximum voltage change during the disturbance,
and its duration is the maximum time in the most
disturbed phase. Voltage dips are graphically
represented in the co-ordinate system: the residual
voltage value vs. dip duration – Table 11.
The limitation of a voltage dip duration to 3 seconds (according to the original ESKOM proposal) has
significantly influenced the number of dips. Additional “voltage dip windows”, which are essential for
the parties to the contract, can be included into contractual provisions.
4.7. Characteristics of voltage dip severity
The useful tool for comparing measuring points in terms of voltage dips are IEEE severity
characteristics. They are created in the co-ordinate system, where the severity indices (SI), determined
as the product of the depth and duration of a voltage dip, are indicated on the vertical axis. The
horizontal axis gives information on how many of the recorded disturbances has the magnitude, which
exceeds the severity indices being assumed as threshold values, given as an example in Table 12.
These characteristics allow for fast interpretation of results of voltage dips recording.
5. CONCLUSIONS
The paper presents several selected strategies for formulating the issue of voltage dips and short
supply interruptions in a contract for the electric power supply. A database, obtained as a result of
seven month monitoring of power quality factors at the metering points at the coupling of transmission
and distribution system, has been used for illustration of the procedures proposed in bibliography. In
the authors’ opinion, a large part of the proposed procedures of contractual assessment of voltage dips
is weakly founded, excessively complicated, and their practical usefulness has not been recognized
yet. Sometimes it is proposed to categorize the voltage dips in transmission systems, taking into
consideration only their magnitude (residual voltage), assuming that in most cases the disturbance
duration is constant, and results from the settings of short-circuit protection times. The obtained results
of measurements do not confirm this thesis.
6
Table 12. The number of events exceeding the
assumed threshold of severity index (SI)
LUBOCZA
WANDA
0.01
19
0.01
74
0.01
44
0.1
19
0.1
74
0.1
31
1
19
1
74
1
9
5
18
5
72
5
5
10
17
10
62
10
3
20
5
20
55
20
2
50
2
50
40
50
0
100
1
100
34
100
0
200
0
200
32
200
0
500
0
500
24
500
0
1000
0
1000
22
1000
0
10000
0
10000
7
10000
0
100000
0
100000
0
100000
0
100000
wartości indeksu uciążliwości (SI)
SKAWINA
Severity Number Severity Number Severity Number
Index of events Index of events Index of events
10000
1000
100
10
1
0,1
0,01
1
11
21
31
41
51
61
71
Liczba zapadów powyżej SI
Lubocza
Wanda
Skawna
Fig. 8. Comparison of severity indices (SI) at individual
substations
6. ACKNOWLEDGMENTS
The authors would like to thank the PDC – Cracow for giving their consent to the measurements and
for help in carrying out the measurements, the LEM Company and its representative in Poland –
SEMICON for technical advice and making the measuring instruments available free of charge, and
EPRI for encouraging the experiment and making the software available free of charge.
7. BIBLIOGRAPHY
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
Bollen M.H.J.: Understanding power quality problems – voltage sags and interruptions. IEEE Press Series
on Power Engineering 2000.
Bollen M.H.J.: On going standard work on statistical presentation of voltage dips. (material obtained from
CIGRE).
G Beaulieu, MHJ Bolen,R Koch, S Malgarotti, X Mamao, J Sinclair. Power Quality Indices and objectives.
CIGRE WG36.O7/CIRED Progress Paper for the CIRED Conference in May 2003-01-13
Brooks D.L., Gunther E.W., Sundaram A.: Recommendations for tabulating rms variation disturbances with
specific reference to utility power contracts. CIGRE 36.05/CIRED 2 CC02 9923.
Dettlof A., Sabin D.: Power quality performance component of the special manufacturing contracts between
power provider and customer. 0-7803-6499-6/00IEEE.
Gosbel V.J., Robinson D., Petra S.: The analysis of utility voltage sag data.
McGranaghan M., Gunther E.: The economics of custom power. Power Quality and EMC in Power Quality.
Colloquium of Czech National Committee and 36 Study Committee of CIGRE, Dep. of Power Eng., CTU
Prague, Sep. 28, 2001, Prague.
NRS 048-2: 1996 Electricity supply – quality of supply. Part 2: Minimum standards. ISBN 0-626-11239-7.
Published in Republic of South Africa by the South African Bureau of Standards.
NRS 048-4:1999 Electricity supply – quality of supply. Part 4: Application guidelines for utilities. Published
in Republic of South Africa by the South African Bureau of Standards.
Robert A., Hoeffelman J., De Jaeger E.: CIRED 2001 – Session 2: Power Quality & EMC, Presentations
and Contributions for Thursday 21 June 2001 (material obtained from the authors).
Transmission Power Quality Benchmarking Methodology. EPRI Final Report, December 2000.
Voltage sag indices – draft 4, working document for IEEE P1564, August 2003.
Voltage sag indices – draft 2, working document for IEEE P1564, November 2001.
Understanding premium power grades. EPRI Final Report, Nov. 2000.
7