CALCULATION OF SAG QUANTITIES INCLUDING PROTECTION SETTINGS AND BREAKERFAILURE IN

VOLTAGE SAG PREDICTION.

S. P. J. Rombouts L. Driessen-Mutters
PNEM Netwerk BV
PO box 1856 5200 BW `s-Hertogenbosch, Netherlands
Tel: +31-73-6154099 Fax: +31-73-6154123 E-Mail: [email protected]

ABSTRACT the 150 kV system through Yd-transformers of 35 to 77

MW.
Of all power quality aspects voltage sags are the Dutch
main Power Quality concern. At PNEM, a utility company 150 kV grid (overhead lines)
in the southern part of the Netherlands, a computer
program has been developed which can predict voltage 150/10 kV Yd-transformer
sags at any substation in the medium voltage grid. With the
10 kV Earthing (Z-)coil
use of this program we are able to present our customers a
10 kV block
quantification of the kind, the severity and frequency of
sags to be expected at a specific site. This paper presents
the calculation method and how the results can be used to meshed MV grid
verify the need for and the effect of sag mitigation (cables)

techniques.
radial operated MV grid
(cables)
INTRODUCTION
Figure 1: Typical PNEM MV grid
Although the average Dutch medium voltage customer only
suffers five sags per year, the damage those (comparatively) The protection system of the meshed operated part consists
few dips cause, still aggravates these customers. The of single distance or differential relays. The protection
satisfaction of the customers is depending on the ratio of system of the radial operated part consists of definite time
the experienced sags to the expected sags. overcurrent relays.
Because of the deregulation of the regional electricity
companies (REC’s), this item plays more and more an The 150 kV grid is built as overhead lines. There are
important role. The customers want to know what kind of Petersen coils installed for system earthing.
sags they can expect and they want an estimation of the The 10 kV grid (sub-transmission and distribution parts)
frequency of their occurrence. consist completely of cables. Thus we apply no
PNEM has developed a computer tool for system analysis. autoreclosure in the 10 kV system. Earthing coils are
These analysis are based upon systematic calculation of installed to limit the phase to earth fault currents to a
short circuit currents and voltages on many points of the maximum of 2500 A.
system. The behavior of the protection relays and the
probability of breaker/protection failure are taken into Voltage

account. A part of the results of these analysis predict the

Sag value
Sag depth

severity and frequency of voltage sags due to short circuits.

This paper will first introduce the concept of the MV grids
at PNEM, then the different possibilities of characterizing
sags are shown. After that the calculation method itself will
Sag magnitude

be discussed. Then the results of a specific case study Sag duration

follows. At the end a conclusion will be given.

CHARACTERISTICS OF THE PNEM 10 kV SYSTEM Time

At PNEM large customers are connected to the MV-grid Figure 2: Characterization of sags
(10 kV). A part of the 10 kV grid is meshed built and
Due to the relative high short circuit power (between 25
operated: this is in fact a sub-transmission grid.
and 300 MVA) and the relative low motor powers (< 1
An other part is meshed built but radial operated: the
MW), most sags are due to short circuits and not to heavy
distribution grid. The sub-transmission grids are fed from
starting loads.
CHARACTERISTICS OF SAGS grids, the transport of reactive power also produces only
small problems in the calculations. Later on we also applied
Sags can be characterized in a number of ways. See figure the program for our 150 kV system. This was possible
2. One way is the duration of the sag. Another possible and because of the relative short distances in this grid.
often used variable is the magnitude of the remaining
voltage (sag magnitude). The complete analysis software is based on the systematic
For voltage sag mitigation devices (UPS/DVR), the calculation of short circuits in every part of the grid.
complement of the sag magnitude, the sag depth is more In every situation we first perform a short-circuit
meaningful. However the relation between these two values calculation to determine the currents in every cable or line
should be seen vectorial. See figure 3. This is particular so and the remaining voltage on every node.
when we talk about equipment that is sensitive to jumping The characteristics of the distance and differential relays
of phase angles [1,2]. are known and the settings are given in the input of the
Another important program.
characteristic can be
Sagdepth
given by the ‘sag We also take into account the reactions of the relays on
value’. This is the phase to earth faults, the over-current, the under-impedance
drop in rms voltage and the under-voltage behavior and the possibility for trip
Voltage under integrated over the commands between one relay and another.
‘normal’ conditions
sag duration. This
Sag must be calculated for With the voltages and currents from the short circuit
magnitude every portion of the calculations we can determine the reaction of every
time that the sag differential, distance or overcurrent relay in the system.
depth changes due to For distance relays the (electrical) distance to the fault is
switching actions. In calculated.
Figure 3: Relation between depth combination with the
and magnitude of a sag ratings of the Calculation of sag duration, depth and value:
equipment to be
protected, this value gives an indication of the energy that When all the reaction times of the relays are known, the
is needed by the sag mitigation devices to restore the relays which will give the first trip command can be
voltage. determined. This is the duration of the first part of the sag.
The above mentioned characteristics: duration, magnitude, The depth of the sag on every node in the system until this
depth and value are all characterizations of the sags time is also known. Multiplication of all these depths with
themselves. However to determine whether a specific node the duration gives the sag values for all the nodes.
in the system is sensitive for the occurrence of sags there Next the cables that have received a trip command are
are two more figures to be determined. taken out of service. Now a new short circuit calculation is
First there are the so called ‘critical distances’ or fault performed. Then the reaction of all the relays in the system
positions [3,4]. For every node in the system it is possible giving us the values for the following part of the sag are
to calculate tables in which we can see what the severity of determined. We keep recalculating until there are no more
a short circuit (depth and/or value of the sag) would be, relays that produce trip commands. The deepest sag depth
should it occur on a particular spot in the system. In this during these calculations will be registered in the tables for
manner a sort of map of the grid where short circuits cause the ‘critical distances’. When there are no more currents
problematic sags can be created. flowing other than the currents due to the loads in the
Second the so called ‘sensitive length’ can be calculated. system and the node or cable in which the short circuit was
The length of all the (parts of the) cables where a short- simulated is isolated, all the quantities for this particular sag
circuit could cause a problem is added. Problems can be are determined. When the short circuited cable or node is
caused by sag depth or sag value. not isolated from the system, a (severe) situation occurs in
which the protection scheme has failed to isolate the fault.
In this case it is assumed that the final result of this short
CALCULATION METHOD circuit will eventually cause the whole MV-system (the
150/10 kV transformer) to go down. This means that in this
Originally the software was developed for the analysis of case there is no sag for all of the nodes in this part of the
the configuration of our MV grids and its protection system!
schemes. Different schemes and configurations can easily
be compared. The used short circuit calculation algorithm is Determination of the chance of occurrence:
very simple. Therefore it is possible to obtain relative short
calculation times. The loads in the system are assumed to Short circuits are simulated on every node and many times
have a constant impedance. This is not correct but the in every cable, e.g. every 3 % of the length of the cables. It
deviations are small enough to be neglected for our is assumed that the probability of a short circuit in a cable
purpose. Because of the relative short distances in our is directly coupled with the length of the cables. The value
of this probability is known from the fault registration From this same registration system we also know the
system of PNEM [5]. probability of a short circuit in a cable. In the calculations a
The breakers in the medium voltage grids have the bad probability of 0.021 occurrences/km/year is used. For the
habit of failing sometimes. Sometimes due to failures in the probability of breaker/protection failures a value of 5 % is
breakers, sometimes due to failures in the protection taken. Unfortunately this also is a realistic number for the
system. This figure also follows from the same registration MV system.
system. With the following procedure the effect of failure
of breaker/protection is taken into account:
After the calculations of the sag properties due to all the 150 kV transmission system
simulations of short circuits in a specific cable (or on a
specific node) the software determines which breakers have
10 kV distribution grid
received a trip command during any of these simulations.
For these breakers, one by one, a failure of the a f g h n
breaker/protection system is simulated by setting the 13 km 0.6 km
reaction time of the relay to infinity. Then all the short j k

circuit simulations for that particular cable (or node) are b c d i l substation B
repeated. Now the back-up system will react (or fail) as in 0.6 km e m
reality. We only assume one breaker to fail at the same
time. The ‘critical distance’ for these situations are not
substation A
(re)determined. First because in most cases this value will
not be affected, secondly because in this value it is not Figure 4: Simplified MV grid
possible to account for the probability of protection/breaker
failure.
Basically the program gives three important results: sag
In the MV system the probability of a short circuit on a profiles, critical distances and sensitive length.
substation is neglected with respect to the probability of a
short circuit in a cable. This is not the case in the HV Sag profiles
system.
When all the simulations are completed and the reaction to The software produces sag frequency profiles or voltage
short circuits anywhere in the system is known, the sag coordination charts [10]. For the substations A in our
‘sensitive length’ for each node can be determined by case study the table 1 was calculated. We can also present
simply adding all the lengths of the cables where the sag these figures in the form of a 3D-plot (figure 5):
depth or sag value due to a short circuit in that part of the Table 1: Sag frequency in #/year for substation A
cable is higher than a certain value. These values were
Sag Sag duration in seconds
registered in the tables for the ‘critical distances’. This is depth >3s >2½-3 s >2-2½ s >1½-2 s >1-1½ s >½-1 s >0.1-½ s <=0.1 s
done for every node in the system. 10 - 20 % 0.0662 0 0.0015 0.0064 0.2852 0.3321 0.2769 0.2479
20 - 30 % 0.0099 0 0.001 0.0037 0.06 0.1743 0.1717 0.0624
30 - 40 % 0 0 0.0016 0.0045 0.0719 0.2665 0.2708 0.042
40 - 50 % 0 0 0.0014 0.0037 0.0049 0.1057 0.3839 0.0646
RESULTS 50 - 60 % 0.0001 0 0.0018 0.0056 0.0364 0.1108 0.2043 0.2453
60 - 70 % 0.0068 0 0.0017 0.0008 0.0543 0.0849 0.1494 0.0946
Let us now take a closer look at the results of calculations 70 - 80 % 0.0322 0 0.0369 0.0083 0.0538 0.097 0.1691 0.367
80 - 90 % 0 0 0.0014 0.0007 0.0745 0.0709 0.0751 0.1795
that have been made on the grid of figure 4. It is to be noted >= 90 % 0.0001 0 0.0011 0.0148 0.0719 0.0341 0.1211 0.0851
that in figure 4 not all the cables that are present in the
system are drawn.
Sag profile for node A
In the input for the calculations (as in reality) there were 26
cables with an average length of 3.7 km in the meshed part,
protected by differential and distance relays and 153 cables 0.6

with an average length of 1.5 km in the radial operated part, 0.5

protected by 98 overcurrent relays. 0.4

# of sags

0.3

First the contribution of the sags in the 150 kV grid are <=0.1 s
>0.1-½ s
0.2

calculated for the node to which the transformer is 0.1

>½-1 s
>1-1½ s
>1½-2 s
Sag duration
connected. In substations 10 % of the short circuits are 0
>2-2½ s
>2½-3 s

assumed to be multi phase faults. For the overhead lines 10 - 20 -

20% 30%
30 - 40 -
40% 50%
50 - 60 -
>3s
70 - 80 -
60% 70%
this is 15 %. Single phase faults in the HV system have no Sag depth
80% 90%
>=
90%

effect because of the Peterson coil. These figures are added

to the figures of all the nodes in the MV grid. In the MV Figure 5: Sag profile substation A
grid a combination of 35% single phase faults and 65%
multi phase faults is used, numbers that originate from the For a even better insight in the figures it is better to present
earlier mentioned fault registration system of PNEM. a graph of the cumulated frequencies [6]. See figure 6 for
substation A.
 Cumulated sag frequency for node A Sagvalue frequency node B

1.8

6
1.6

5 1.4

# of sags
1.2
4
# of sags

3 0.8

>0s 0.6
2
> 0.1 s
> 0.5 s 0.4
1 > 1.0 s
Sag duration 0.2
> 1.5 s
0 > 2.0 s 0

>240 - 250 %s

>230 - 240 %s

>220 - 230 %s

>210 - 220 %s

>200 - 210 %s

>190 - 200 %s

>180 - 190 %s

>170 - 180 %s

>160 - 170 %s

>150 - 160 %s

>140 - 150 %s

>130 - 140 %s

>120 - 130 %s

>110 - 120 %s

>100 - 110 %s

> 90 - 100 %s

> 80 - 90 %s

> 70 - 80 %s

> 60 - 70 %s

> 50 - 60 %s

> 40 - 50 %s

> 30 - 40 %s

> 20 - 30 %s

> 10 - 20 %s
>=

<= 10 %s
>= > 2.5 s

> 250 %s
>= >=
10 20 >=
30 >= >=
% 40 50 >=
% % 60 >=
% 70 80
% % 90
% % %
Sag depth
Sagvalue

Figure 6: Cumulated sag frequency substation A Figure 8: Sag value frequency for substation B
In this plot we can see what the minimal specifications of a
These figures are very useful to determine the effect of sag
sag mitigation technique must be in order to mitigate a
mitigation techniques [5] in order to make a cost benefit
certain number of sags [5].
analysis.
As previously mentioned there is also the possibility to
present the sag values instead of sag depth. Now we have
one table (table 2) or 2 2D-plots (figures 7 and 8) for the Critical distances
figures of two substations A and B.
Let us now look at the tables for the critical distances. Only
Table 2: Sag value frequencies the columns of the cables that are numbered in figure 4 are
Sag value frequencies shown (table 3).
Sag value in # of sag/year The way to read this table is as follows: When a three phase
procent*sec Node A Node B
> 250 %s 0 0 fault occurs on a distance of 10.5% from the beginning of
>240 - 250 %s 0 0 cable ‘a’, there will be a sag depth of 85% on substation A.
>230 - 240 %s 0.0003 0
>220 - 230 %s 0.0006 0
>210 - 220 %s 0.0006 0
>200 - 210 %s 0.0004 0 Table 3: Critical distances (sag depth) substation A
>190 - 200 %s 0.0003 0
Critical distances (sag depth in % for multiphase fault) for node A:
>180 - 190 %s 0.0108 0.0001
Distance¦ Cable
>170 - 180 %s 0.0085 0.0047 [%] ¦ a b c d e f g h i j k l m n ¦
--------¦----------------------------------------------------------------------¦
>160 - 170 %s 0.0026 0.0025 1.5 ¦ 97 100 100 99 81 99 99 99 69 74 74 73 49 100 ¦
4.5 ¦ 92 100 99 97 80 97 96 98 68 73 73 73 49 99 ¦
>150 - 160 %s 0.0008 0.0007 7.5 ¦ 88 100 98 95 80 95 94 97 67 73 73 72 49 99 ¦
10.5 ¦ 85 100 97 94 79 93 92 96 65 73 73 71 49 99 ¦
>140 - 150 %s 0.0285 0.0256 13.5 ¦ 82 100 97 92 79 91 90 95 64 73 73 70 49 99 ¦
16.5 ¦ 80 99 96 91 79 90 88 94 63 73 73 69 49 98 ¦
>130 - 140 %s 0.0542 0.0264 19.5 ¦ 78 99 95 89 78 88 86 93 62 72 73 69 49 98 ¦
22.5 ¦ 76 99 94 88 78 86 84 92 61 72 72 68 49 98 ¦
>120 - 130 %s 0.0649 0.0312 25.5 ¦ 75
28.5 ¦ 74
99
99
94
93
87
86
77
77
85
84
82
81
91
90
60
59
72
72
72
72
67
66
50
50
98 ¦
97 ¦
>110 - 120 %s 0.0543 0.0245 31.5 ¦ 73
34.5 ¦ 73
99
99
92
92
84
83
77
76
82
81
80
78
89
88
58
58
72
71
72
72
66
65
50
50
97 ¦
97 ¦
>100 - 110 %s 0.0435 0.0161 37.5 ¦ 72
40.5 ¦ 72
99
99
91
91
82
82
76
76
80
79
77
76
87
86
57
56
71
71
72
72
64
64
50
50
97 ¦
96 ¦
43.5 ¦ 72 99 90 81 76 78 75 85 55 71 72 63 50 96 ¦
> 90 - 100 %s 0.0678 0.0525 46.5 ¦ 72 99 89 80 75 77 74 84 55 71 72 63 50 96 ¦
49.5 ¦ 72 99 89 79 75 76 74 84 54 71 72 62 50 96 ¦
> 80 - 90 %s 0.0991 0.0647 52.5 ¦ 73 99 88 78 75 75 73 83 54 71 72 62 51 96 ¦
55.5 ¦ 73 99 88 78 75 75 72 82 53 71 72 61 51 96 ¦
> 70 - 80 %s 0.0768 0.0593 58.5 ¦ 74 99 87 77 74 74 72 81 53 70 72 60 51 95 ¦
61.5 ¦ 75 99 87 77 74 73 72 81 52 70 72 60 51 95 ¦
> 60 - 70 %s 0.1406 0.1623 64.5 ¦ 76 99 86 76 74 73 71 80 52 70 72 59 51 95 ¦
67.5 ¦ 77 99 86 76 74 72 71 80 51 70 72 59 52 95 ¦
> 50 - 60 %s 0.1223 0.1725 70.5 ¦ 78 99 85 75 74 72 71 79 51 70 72 59 52 95 ¦
73.5 ¦ 80 99 85 75 74 71 71 78 51 70 73 58 52 95 ¦
> 40 - 50 %s 0.24 0.1867 76.5 ¦ 81 99 84 74 74 71 71 78 51 70 73 58 52 95 ¦
79.5 ¦ 83 99 84 74 73 71 71 77 50 70 73 57 53 95 ¦
> 30 - 40 %s 0.5848 0.2405 82.5 ¦ 85
85.5 ¦ 87 100
99 83
83
74
74
73
73
71
70
71
71
77
77
50
50
70
70
73
73
57
57
53
53
95 ¦
95 ¦
> 20 - 30 %s 0.8719 0.4032 88.5 ¦ 89 100
91.5 ¦ 92 100
83
82
74
73
73
73
70
70
72
72
76
76
50
50
70
70
74
74
56
56
53
54
95 ¦
95 ¦
> 10 - 20 %s 1.1384 1.5118 94.5 ¦ 94 100
97.5 ¦ 97 100
82
81
73
73
73
73
70
70
73
73
75
75
50
49
70
70
74
74
55
55
54
55
95 ¦
95 ¦
<= 10 %s 1.8006 1.6574 --------¦----------------------------------------------------------------------¦

Sagvalue frequency node A

Let us assume that we have equipment installed on
2
substation A which could withstand a sag depth of 75% or
1.8 less. In this table we see that the equipment is vulnerable
1.6
for multi phase short circuits anywhere in cables ‘b’, ‘c’
1.4
and ‘n’ and for almost the whole length of cable ‘h’.
# of sags

1.2

1 Remarkable is the lower sag depth in the middle part of

0.8
cable 'a', due to the meshed operation. With the help of this
0.6

0.4
table we can mark the vulnerable locations in figure 4 to
0.2 obtain figure 9.
0
> 90 - 100 %s

> 80 - 90 %s

> 70 - 80 %s

> 60 - 70 %s

> 50 - 60 %s

> 40 - 50 %s

> 30 - 40 %s

> 20 - 30 %s

> 10 - 20 %s
>240 - 250 %s

>230 - 240 %s

>220 - 230 %s

>210 - 220 %s

>200 - 210 %s

>190 - 200 %s

>180 - 190 %s

>170 - 180 %s

>160 - 170 %s

>150 - 160 %s

>140 - 150 %s

>130 - 140 %s

>120 - 130 %s

>110 - 120 %s

>100 - 110 %s

<= 10 %s
> 250 %s

Sagvalue

Figure 7: Sag value frequency for substation A

 150 kV transmission system Table 4: Critical distances (sag value) substation A
Critical distances (sag value in %s for multiphase fault) for substation A:

Distance¦ Cable
[%] ¦ a b c d e f g h i j k l m n ¦
10 kV distribution grid --------¦----------------------------------------------------------------------¦
1.5 ¦ 14 0 103 110 48 20 20 20 7 7 114 7 5 18 ¦
4.5 ¦ 14 0 103 109 48 19 16 20 7 7 115 7 5 10 ¦
a f g h n 7.5 ¦ 9 0 102 107 48 16 16 20 7 7 115 7 5 10 ¦
10.5 ¦ 8 0 101 94 48 16 16 16 7 7 116 7 5 10 ¦
13 km 0.6 km 13.5 ¦ 8 0 100 92 49 16 16 16 6 7 117 7 5 10 ¦
16.5 ¦ 8 188 100 90 49 16 16 16 6 7 118 7 5 10 ¦
j k 19.5 ¦ 8 187 95 89 49 16 16 16 6 7 119 7 5 10 ¦
22.5 ¦ 8 187 94 87 49 16 16 16 6 7 120 7 5 10 ¦
25.5 ¦ 7 186 94 86 50 16 16 16 6 7 121 7 5 10 ¦
b c d i l substation B 28.5 ¦ 7 186 93 84 50 16 16 16 6 7 122 7 5 10 ¦
31.5 ¦ 7 186 92 83 50 16 16 16 6 7 123 7 5 10 ¦
0.6 km m 34.5 ¦ 7 185 92 81 50 15 16 16 6 7 124 7 5 10 ¦
e 37.5 ¦ 7 185 91 80 51 15 16 16 6 7 22 6 5 10 ¦
40.5 ¦ 7 185 91 79 51 15 16 16 6 7 22 6 5 10 ¦
43.5 ¦ 7 184 89 78 17 15 16 16 6 7 22 6 5 10 ¦
46.5 ¦ 7 184 88 76 17 15 16 16 5 7 22 6 5 10 ¦
49.5 ¦ 7 183 88 75 17 16 16 16 5 7 22 6 5 10 ¦
substation A 52.5 ¦ 7 183 87 74 17 16 16 16 5 7 22 6 5 10 ¦
55.5 ¦ 7 183 86 73 17 16 16 16 5 7 22 6 5 10 ¦
58.5 ¦ 7 182 86 72 18 16 16 16 5 7 22 6 5 10 ¦
61.5 ¦ 7 182 85 71 18 16 17 16 5 7 22 6 5 10 ¦
Figure 9: Marked cable parts for sag depth on 64.5 ¦ 8 182 85 70 66 16 17 16 5 7 22 6 5 10 ¦
67.5 ¦ 8 181 84 69 65 16 17 16 5 7 86 6 5 10 ¦
substation A > 75 % 70.5 ¦
73.5 ¦
8 181
8 181
83
83
69
68
65
64
16
16
17
17
16
16
5
5
7
7
85
84
6
6
5
5
10 ¦
9 ¦
76.5 ¦ 8 180 82 67 64 16 18 16 5 7 84 6 5 9 ¦
79.5 ¦ 8 180 81 66 64 16 18 179 5 7 83 6 5 9 ¦
82.5 ¦ 8 180 81 65 63 16 130 177 5 7 83 6 5 9 ¦
Multi phase faults in the dotted parts of the cables result in 85.5 ¦ 9 179 80 65 63 141 127 175 5 7 82 6 5 9 ¦
88.5 ¦ 9 179 80 64 62 139 125 173 5 7 82 6 5 9 ¦
a sag depth on substation A of less than 75 %, while faults 91.5 ¦
94.5 ¦
9 179
9 178
79
79
63
63
62 137 123 171
62 135 121 169
5
5
7
7
81
81
6
6
5
5
9 ¦
9 ¦
in the solid drawn cables result in a sag depth of 75 % or 97.5 ¦ 10 178 78 62 61 133 120 167 5 7 80 6 5 9 ¦
--------¦----------------------------------------------------------------------¦
more.
We can go through the same procedure for substation B. The zero values in cable ‘b’ indicate that, when a three
The result is shown in figure 10. phase fault occurs in the first 13.5 % of cable ‘b’,
substation A will be isolated from the grid, thus there will
150 kV transmission system
be no sag. This is caused by the settings of the protection
relays. When we mark the cables for sag values > 100 %s
10 kV distribution grid
we obtain the drawings in figure 11 and 12.

a f g h n 150 kV transmission system

13 km 0.6 km
j k

i l substation B 10 kV distribution grid

b c d

0.6 km e m
a f g h n
13 km 0.6 km
j k
substation A
b c d i l substation B
Figure 10: Marked cable parts for sag depth on 0.6 km e m
substation B > 75%
substation A
In these pictures we get a very good overview of the critical
locations in the grid. Figure 11: Marked cable parts for sag value on
In stead of the sag depth we can also use the sag value. substation A > 100 %s
Then we get a complete different picture. We present here
table 4 for substation A. The way to read this table is as These results are very different from the previously
follows: When a three phase fault occurs on a distance of obtained drawings. This is caused by the high effect of the
10.5% from the beginning of cable ‘a’, there will be a sag sag length. Because of the settings of the distance relays to
value of 8 %s on substation A. 85 % of the length of the cables, short circuits in the last (or
first) part of the cables have a longer duration then in the
middle part where the relays at the begin and end of the
cables both react rapidly.
 150 kV transmission system The effect of different protection schemes as well as the
effect of breaker failures can be accounted for in the sag
profiles.
10 kV distribution grid
We can draw 'maps' of the meshed and radial power
a f g h n networks to visualize the spots where short circuits have a
13 km 0.6 km significant influence on the sag quantities. The vulnerability
j k
of one node is comparable to that of another one.
b c d i l substation B We are able to quantify the effect of different sag
0.6 km e m mitigation techniques.

substation A

Figure 12: Marked cable parts for sag value on REFERENCES

substation B > 100 %s 1. Bollen, MHJ; Mansoor, A; Collins, ER:
'Characteristics of voltage sags experienced by single
phase and three phase equipment', International
Sensitive length Conference on Power Quality: End-Use Applications
and Perspectives, PQA 1997, Sweden.
For easy comparison of the vulnerability of different nodes
to sag depth or sag value the program produces a simple 2. Bollen, MHJ; Wang, P; Jenkins, N: 'Analysis and
table where the length of the cableparts in which we have a consequences of the phase jump associated with
problem is added. This is the sensitive length. In the table voltage sags’, Power System Computation
below we have the sensitivity values for three phase and Conference, Dresden, Germany, August 1996,316-
one phase faults for sag depths of 70, 50 and 30 % and for 322.
sag values of 10, 100 and 200 %s. 3. Qader, MR; Bollen, MJH; Allan, RN: 'Stochastic
Prediction of voltage sags in the reliability test
system', PQA 1997, Stockholm, Sweden.
Table 5: Sensitive length for substations A and B
4. Bollen, MJH: 'Method of critical distances for
¦ Node A Node B stochastic assessment of voltage sags', IEE Proc.-
70% - 3f ¦ 105.9 41.2 Gener. Transm. Distrib., Vol 145, No 1. January
70% - a ¦ 0.8 0.0 1998.
50% - 3f ¦ 172.6 82.1 5. Driessen-Mutters, L; Rombouts, SPJ; Meeuwsen, JJ:
50% - a ¦ 0.8 0.0
'A Tool for well-founded Cost Benefit Analyses of
30% - 3f ¦ 260.7 186.7 Sag Mitigation Techniques', PQA 1998, Cape Town,
30% - a ¦ 0.8 0.0
South Africa.
10% - 3f ¦ 239.3
10% - a ¦ 52.2
209.4
12.7
6. Bollen, MHJ: 'Prediction and mitigation of voltage
sags', IBC conf on Improving Power Quality in
100% - 3f ¦ 19.4 19.2 Transmission and Distribution, Amsterdam, January
100% - a ¦ 0.0 0.0
1998.
200% - 3f ¦ 0.0 0.0
200% - a ¦ 0.0 0.0 7. Meeuwsen, JJ; Kling, WL; Rombouts, SPJ:
'Computerized determination of protection relay
Total grid length: 323 km.
settings in meshed operated MV networks', Protecting
Electrical Networks and Quality of Supply,
E.g.: The voltage on node A will, for 3 phase short circuits 1997,Heathrow, UK.
on 105.9 km of the grid of 323 km show a sag depth of 8. Meeuwsen, JJ; Kling, WL; Rombouts, SPJ: 'The
more than 70%. influence of protective relay schemes on the reliability
indices of load points in meshed operated MV
networks', ... , 1997, ...
CONCLUSIONS
9. Bloemhof, GA; Hendriks Boers, MHAJ; Knijp, J:
With the help of the discussed computer program we are 'Stochastic quantification of the impact of voltage
able to give a quick and clear overview on the vulnerability dips', PQA 1998, Cape Town, South Africa.
of certain nodes in respect to voltage sags due to short- 10. Conrad, LE; Bollen, MHJ: ‘Voltage sag coordination
circuits in a MV system. We can see that it makes a great for reliable plant operation’, IEEE Transactions on
difference whether we look at the sag depth or at the sag Industry Applications, Vol 33 (1997), p. 1459-1464.
value.