CIRED 18th International Conference on Electricity Distribution
SYSTEM STUDIES OF VOLTAGE DIPS RESULTING FROM ENERGISATION OF MV WIND
TURBINE TRANSFORMERS
Ting MA
Power Technology, E.ON UK - United Kingdom
ting.ma@eon-uk.com
SUMMARY
A typical wind farm consists of several wind turbines, each
operating at less than 1kV and connected to the MV grid
system via a turbine transformer. When the turbine
transformers are energised, the grid system will see a
transient phenomenon called inrush current and voltage
dips/fluctuations on the local grid system will result. Severe
voltage dips can cause fluctuations of certain types of lights
resulting in customer complaints. System studies of voltage
dips caused by energisation of turbine transformers for three
proposed MV wind farm sites connected to 33kV distribution
networks have been undertaken using PSCAD software and
are described in this paper. Site measurement data was
obtained for one of the three sites for verification purposes.
The results of the system studies, recommended mitigation
measures to overcome any unacceptable voltage dips and
energisation switching sequence for each wind farm site are
contained in the paper.
Key words: Wind Farms, Turbine Transformers, Inrush
Current, MV Grid System, Energisation, Voltage
Dips/Fluctuations, Point of Connection (POC).
INTRODUCTION
As MV wind farms become more prevalent, utilities are
becoming more concerned about their impact on power
systems. One effect of wind farm generation on a weak
distribution system is voltage fluctuation upon energisation of
the turbine transformers and cables.
A typical wind farm consists of several wind turbines, each
operating at less than 1kV and connected to the MV (10kV to
150kV, 33kV connection is mainly considered in this paper)
grid system via a turbine transformer. When the turbine
transformers are energised, the grid system will see inrush
current and voltage dips/fluctuations on the local grid system.
When many turbine transformers are energised
simultaneously these voltage dips may be severe. Severe
voltage dips can cause the end users equipment to operate
inefficiently, e.g. fluctuations of certain types of lights.
Thus, the voltage fluctuation at the point of connection (POC)
and local grid system may be required to be as low as 3%
depending on the interpretation of requirements such as the
British Engineering Recommendation P28 [1], or other
National requirements used by the Distribution Network
Operator (DNO) concerned.
System studies of voltage dips caused by energisation of
turbine transformers for two proposed wind farm sites (a
2x10-turbine site and a 17-turbine site) and an operational
wind farm have been undertaken using PSCAD software [2,
3] and described in this paper. Each wind turbine operates at
690V and is connected to a 33kV collection system via a
33/0.69kV turbine transformer. The 33kV collection systems
CIRED2005
Session No 2
Turin, 6-9 June 2005
Andrew CADMORE
Renewable Energy Systems Group - United Kingdom
andrew.cadmore@res-ltd.com
within the wind farms are connected directly to the DNOs
33kV distribution systems. Two types of turbine transformer
have been studied for the 2x10-turbine site; a standard design
transformer and an innovative and specially designed wind-
turbine transformer. The specially designed transformer has a
significantly lower magnetising inrush magnitude than the
standard design transformer. The turbine transformers with
low magnetising inrush magnitude were proposed for the 17-
turbine site.
Site measurement data were obtained for the operational wind
farm site for verification purposes to compare with the results
obtained with PSCAD models.
2x10-TURBINE SITE
This wind farm consists of 20 wind turbines, each rated at
1.3MW. Each turbine is connected to the 33kV collection
network via a turbine transformer, each rated at 1.5MVA,
8.3% leakage impedance. Two types of turbine transformers
were proposed for the project. One is a standard design
transformer whose maximum peak inrush current is 7.35pu of
rated current. The inrush current is damped to rated current
within 1.6 seconds after energisation. The other is a special
design turbine transformer which has a significantly lower
magnetising inrush magnitude than the standard design
transformer. The maximum peak inrush current is 4.04pu of
rated current, only 55% of that of the standard design
transformer, and the inrush current is damped to rated current
within 1.0 second.
The wind farm substation consists of two POC busbars
running split, each having 10 turbines connected, and no
33kV interconnections between these two 10-turbine circuits.
The wind farm has two separate 33kV grid connections via
these two busbars. Each of the two grid connections is
independently supplied and controlled. Thus, this wind farm
is referred to as a 2x10-turbine site.
The local distribution network, the wind farm collection
system and the wind farm substation were modelled and
studied using the proprietary software PSCAD. The voltage
dips due to the energisation of the wind farm 33kV collection
system, i.e. turbine transformers and cables, are assessed in
compliance with the British Engineering Recommendation
P28 assuming the lower limit of 3%. Generally, the
maximum voltage dip occurs at the instant of energisation,
referred to as voltage dip measured at 0ms in this paper. It is
recommended in P28 that voltage dips due to direct-on-line
starting of induction motors are measured at 30ms after
energisation to assess the voltage changes due to motor
starting. Since energisation of turbine transformers is a
similar transient phenomenon to the direct-on-line starting of
induction motors, the voltage dips due to energisation of
turbine transformers measured at 30ms are used to compare
with the required limit for this site as well as for the other two
sites in the paper.
CIRED 18th International Conference on Electricity Distribution
The PSCAD multiple run facility was used to vary the
switching angle for point-on-wave to search for the worst
voltage dips. The results of studies for the 2x10-turbine site
are given in Figures 1 (standard design transformers) and 2
(special design transformers). Voltage dips at three points of
the system are shown in the figures, the points being POC 1,
POC 2, and an 11kV primary substation which is located
downstream of (i.e. fed from) POC 2 and referred to as Grid
point_11kV. Voltage dips measured at 0ms (i.e. peak values)
are shown in light lines whereas the voltage dips measured at
30ms after the peak are highlighted in heavy lines.
15 POC 1_0ms
POC 2_0ms
Grid point_11kV_0ms
12 POC 1_30ms
POC 2_30ms
Voltage dip (%)
Grid point_11kV_30ms
9
6
3 magnetising inrush magnitude were proposed for the wind
0 current. The inrush current is damped to rated current within
0 1 2 3 4 5 6 7
Number of transformers energised simultaneously
Figure 1. Voltage dips varying with the number of standard design
transformers energised simultaneously (2x10-turbine site)
15
POC 1_0ms
POC 2_0ms
12 Grid point_11kV_0ms
POC 1_30ms
Voltage dip (%)
POC 2_30ms
9 Grid point_11kV_30ms
6
3 P28, i.e. permitted limit of voltage step changes is 3%.
0 As with the 2x10-turbine site, the PSCAD multiple run
0 1 2 3 4 5 6 7
Number of transformers energised simultaneously
Figure 2. Voltage dips varying with the number of special design
transformers energised simultaneously (2x10-turbine site)
The study results show that the higher the inrush magnitude,
the higher the voltage dips. Voltage dip decreases when the
number of turbine transformers energised simultaneously is
reduced. The voltage dip seen at the 11kV substation is
higher than that seen at POC 2. This is because the substation
is downstream of POC 2. The maximum voltage dip
measured at 30ms after energisation is 9.3%, seen at the 11kV
primary substation when energising 10 standard design
transformers simultaneously, and 7.8% when energising 10
special design transformers simultaneously. The figures also
show that a maximum of two standard design turbine
transformers for circuit 1 (corresponding to POC 1) can be
energised simultaneously but only one for circuit 2
(corresponding to POC 2) to meet the requirement of P28.
With special design transformers, a maximum of two
transformers for both circuits can be energised
simultaneously. Thus, at least 10 MV switches are required
within the wind farm collection system to sequentially
energise turbine transformers to meet the limit of 3%.
The effect on voltage dips of a pre-insertion resistance (PIR)
added to the wind farm main circuit breakers has also been
studied. The results show that to energise all 2x10 turbine
transformers simultaneously voltage dips at POCs and grid
CIRED2005
Session No 2
Turin, 6-9 June 2005
system can be reduced within 3% by an 80 PIR inserted for
10ms, added to each of the two main circuit breakers. No
additional switchgear is required within the wind farm
collection system to facilitate sequential energisation.
Although PIR switchgear is currently not generally available
at 33kV, it is believed that PIR technology is feasible and
would be of a design similar to neutral earthing resistors
commonly used on distribution systems. Such resistances are
rated at well over 1kA for as long as 30 seconds. Switchgear
used to switch in and out PIRs would be similar to typical
distribution switchgear and, in fact, may be cheaper as it
would not need to break fault current.
17-TURBINE SITE
This wind farm consists of 17 wind turbines, each rated at
850kW. Each turbine is connected to the 33kV collection
network via a turbine transformer, each rated at 1.0MVA,
5.9% leakage impedance. The turbine transformers with low
farm. The maximum peak inrush current is 4.63pu of rated
8 9 10 1.0 second.
This wind farm substation consists of one busbar and is
proposed to be looped into an existing 33kV circuit which
runs between two existing substations. These two existing
substations are fed from the same 132kV substation. Under
normal operating conditions, the 33kV circuit is fed from both
ends.
The voltage dips due to the energisation of turbine
transformers and cables of this 17-turbine site are assessed in
compliance with the British Engineering Recommendation
8 9 10 facility was used to vary the switching angle for point-on-
wave to search for the worst voltage dips. The results of
voltage dips varying with the number of turbine transformers
energised simultaneously for the wind farm under normal
operating conditions are shown in Figure 3, for voltage dips at
the point of connection busbar (POC) and a 33kV grid point
(Grid point_33kV) at which the voltage dip is the highest in
the grid system. As shown by the points plotted in the figure,
five cases were carried out, simultaneously energising: 3
transformers with 14 initially energised; 6 with 8 initially
energised; 8 with 0 initially energised; 9 with 0 initially
energised; all 17 transformers.
9
POC_0ms
8 Grid point_33kV_0ms
7 POC_30ms
Voltage dip (%)
6 Grid point_33kV_30ms
5
4
3
2
1
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Number of transformers energised simultaneously
Figure 3. Voltage dips varying with the number of turbine transformers
energised simultaneously (17-turbine site)
CIRED 18th International Conference on Electricity Distribution Turin, 6-9 June 2005

Figure 4 (a) shows the 3-ph wind farm energisation currents
when the first 8 transformers are energised simultaneously.
Figure 4 (b) shows the 3-ph energisation currents when 6
transformers are energised simultaneously with the first 8
initially energised transformers. These figures show the
effects of initially energised transformers on the decay of the
energisation inrush current. The energisation inrush current is
more rapidly damped when there are no initially energised
transformers (Figure 4 (a)) than with initially energised
transformers (Figure 4 (b)). The slow decay of the inrush
current with initially energised transformers prolongs the
voltage dip, as indicated in Figure 3. The voltage dips
measured at 0ms increase regularly with the increase in
number of transformers energised, but the voltage dips
measured at 30ms, when energising 6 transformers with 8
initially energised transformers, are equal to these when
energising 8 transformers without initially energised
transformers.
Main : Graphs
I_Windfarm_A I_Windfarm_B
0.80
0.60
0.40
0.20
0.00
kA
simultaneously, voltage dips at POC and grid system can be
reduced within 3% by a 40 PIR inserted for 10ms. The
wind farm can be energised under both normal operating
conditions and minimum fault level conditions with the 33kV
grid circuit single-end fed.
VERIFICATION OF PSCAD MODEL
Verification of the PSCAD model is based on measured data
from an existing wind farm having a comparable number of
turbines to the two proposed projects. Each turbine is
connected to the 33kV collection network via a turbine
transformer, each rated at 1.5MVA, 5.9% leakage impedance.
Standard design transformers were used for the wind farm.
The maximum peak inrush current is 7.35pu of rated current
and the inrush current is damped to rated current within 1.6
seconds. The wind farm substation consists of one busbar
and is directly connected to a 33kV substation via a dedicated
overhead line
I_Windfarm_C This wind farm has been commissioned and is in operation;
site measurement data has been obtained. Figures 5 (a) to 5
(c) are the voltage, current and rms voltage waveforms plotted
from the site measurement data. Figures 6 (a) to (c) are the
corresponding waveforms obtained by the PSCAD modelling
study. Table 1 shows a comparison of the voltage dips
obtained by site measurement with the PSCAD modelling

-0.20 study.
-0.40 Figures 5 and 6 and Table 1 show that the PSCAD modelling
results are very similar to those obtained by site
measurements, which verifies the PSCAD model. The
0.000 0.050 0.100 0.150 0.200 0.250 0.300 voltages obtained with PSCAD modelling contain more
harmonic components within the first 100ms after
Figure 4 (a) Inrush current when energising eight transformers
energisation. Site measured currents decay rather more
(with no initially energised transformers) quickly than those obtained with PSCAD modelling. This
implies higher damping within the actual system than the
Main : Graphs PSCAD model.
I_Windfarm_A I_Windfarm_B I_Windfarm_C The site measured voltage waveforms also indicate that the
0.60
0.50 three phases of the turbine transformers were energised
0.40 non-simultaneously. The PSCAD modelling results shown in
0.30 Figures 6 were obtained when C-phase was closed at voltage
0.20 zero (1800), A-phase closed 1.1ms after C-phase and B-phase
0.10 closed 1.6ms after B-phase. This mimics the site
kA

0.00
-0.10
-0.20
-0.30
-0.40
0.000 0.050 0.100 0.150 0.200
Figure 4 (b) Inrush current when energising six transformers
(with eight initially energised transformers)
The maximum voltage dip measured at POC 30ms after
energisation is 3.9% when energising all 17 transformers
simultaneously, exceeding the required limit of 3%. Voltage
drops at the wind farm substation and the local grid system
are within 3% at normal system operating conditions if
transformers are energised in the sequence of 8 turbine
transformers energised first, then 6 and finally 3.
The effect on voltage dips of a PIR added to the wind farm
main circuit breaker has also been studied for this site. The
results show that to energise all 17 turbine transformers
measurements.
Figure 7 and Table 2 are the results of studies on three-phase
non-simultaneous energisation. Each point on the black
B-phase voltage dip curve in Figure 7 is the maximum
0.250 0.300 voltage dip when the B-phase is energised at a certain angle
(X-axis, varying between -900 to 300, 1800 is the 00 reference
for the angle) and the A and C phases are energised between
0 to 10ms after the B-phase is energised. The corresponding
energisation time for the A and C phases at which the
maximum voltage dip is obtained is referred to as delta (t) ms
and is plotted as a pink curve in Figure 7. Figure 7 shows that
the maximum voltage dip occurs when one phase is energised
before voltage zero (-230) and the other two phases are
energised about 5ms later. Table 2 shows that the voltage dip
at the POC is 19.5% when all three phases of all transformers
are energised simultaneously. Then the conclusion can be
drawn that the voltage dip increases if the voltage of one
phase is around zero (around -600 to 150) when the system is
energised and the three phases are energised
non-simultaneously.

CIRED2005

Session No 2
CIRED 18th International Conference on Electricity Distribution Turin, 6-9 June 2005

Main : Graphs Main : Graphs

Test4_V_A Test4_V_B Test4_V_C V_wfbus_A V_wfbus_B V_wfbus_C
40 40
30 30
20 20
10 10
0 0
kV

kV
-10 -10
-20 -20
-30 -30
-40 -40
0.000 0.025 0.050 0.075 0.100 0.125 0.150 0.175 0.200 0.000 0.025 0.050 0.075 0.100 0.125 0.150 0.175 0.200

(a) Voltage waveforms (a) Voltage waveforms

Main : Graphs Main : Graphs

Test4_I_A Test4_I_B Test4_I_C I_Windfarm_A I_Windfarm_B I_Windfarm_C
2.00 2.00
1.50 1.50
1.00 1.00
0.50 0.50
0.00 0.00
kA

kA
-0.50 -0.50
-1.00 -1.00
-1.50 -1.50
-2.00 -2.00

0.000 0.050 0.100 0.150 0.200 0.250 0.300 0.350 0.400 0.000 0.050 0.100 0.150 0.200 0.250 0.300 0.350 0.400

(b) Current waveforms (b) Current waveforms

Main : Graphs Main : Graphs

Vrms_T4_A Vrms_T4_B Vrms_T4_C Vwfb_rmsA Vwfb_rmsB Vwfb_rmsC
1.050 1.050
1.025 1.025
1.000 1.000
0.975 0.975
0.950 0.950
p.u.

p.u.

0.925 0.925
0.900 0.900
0.875 0.875
0.850 0.850
0.825 0.825
0.000 0.050 0.100 0.150 0.200 0.250 0.300 0.350 0.400 0.000 0.050 0.100 0.150 0.200 0.250 0.300 0.350 0.400

(c) rms voltage waveforms (c) rms voltage waveforms

Figure 5. Site measurement results Figure 6. PSCAD modelling results

TABLE 1 Voltage dips obtained by site measurement and PSCAD 30

Voltage dip (%) and delta (t) ms

B-phase voltage dip (%)

modelling study delta (t) ms
25
Voltage dip at POC (%)
20
Method measured at 0ms measured at 30ms
15
A-ph B-ph C-ph A-ph B-ph C-ph
Site measurement 6.71 17.1 18.1 3.33 9.13 9.15 10

PSCAD 7.67 17.4 18.4 3.55 9.05 9.20 5

0
TABLE 2 PSCAD modelling results maximum voltage dip -90 -75 -60 -45 -30 -15 0 15 30
Percentage percentage of Voltage dip at B-phase angle (degrees)
of initially POC (%)
transformer energised 3 phases Figure 7. Maximum voltage dips due to non-simultaneous energisation
energised transformers energised 0 ms 30 ms of 3- phase turbine transformers

100% 0 S 19.54 9.73

100% 0 NS 24.91 11.90 CONCLUSIONS AND RECOMMENDATIONS
15% 0 S 8.08 5.15
Three wind farm sites have been digitally modelled and
15% 0 NS 10.16 6.25
studied to assess the maximum voltage dips occurring at the
10% 0 NS 7.49 4.95 point of connection and on the local grid system. The
10% 10% NS 7.76 5.63
PSCAD model has been verified with site measurement data.
5% 95% NS 4.23 3.63 The voltage drops at the points of connection and on the local
Notes S: Simultaneously grid system, due to energisation of all turbine transformers
NS: Non-simultaneously.
simultaneously at each of the two proposed wind farm sites

CIRED2005

Session No 2
CIRED 18th International Conference on Electricity Distribution
both exceed the permitted limits. It is therefore recommended
to either install a pre-insertion resistance (PIR) at the wind
farm main circuit breaker to achieve the required voltage dip
limits, or install adequate additional 33kV switches within the
wind farm collection system to energise the turbine
transformers with the switching sequences mentioned above.
Since switchgear incorporating PIR is currently not generally
available at 33kV its application needs to be addressed by the
equipment manufacturers.
Voltage dips are affected by the transformer inrush
magnitude, number of turbine transformers energised
simultaneously, point-on-wave switching angle, initially
energised transformers and three-phase non-simultaneous
energisation. These factors should be considered when a
study is undertaken.
For newly proposed wind farms to be connected to
distribution networks, a system study is necessary to assess
voltage dips due to energisation of turbine transformers to
ensure that voltage dips are within the required limits.
Mitigation measures, either reducing the number of the
turbine transformers energised simultaneously or adding a
PIR, should be studied and recommended.
CIRED2005
Session No 2
Turin, 6-9 June 2005
ACKNOWLEDGEMENTS
The authors would like to acknowledge with thanks E.ON UK
and Renewable Energy Systems Group, UK for their
permission to publish this paper.
REFERENCES
[1] System Utilisation Consultancy Group, September
1989, Engineering Recommendation P28 Planning
limits for voltage fluctuations caused by industrial,
commercial and domestic equipment in the United
Kingdom, United Kingdom.
[2] Manitoba HVDC Research Centre, PSCAD, EMTDC
V4 Users Manuals, Copyright Manitoba HVDC,
Canada.
[3] N Watson; J Arrillaga, Power system electromagnetic
transients simulation, The Institution of Electrical
Engineers, London, United Kingdom.
[4] Edited by H M Ryan, High voltage engineering and
testing, 2nd edition, The Institution of Electrical
Engineers, London, United Kingdom.