Study of Wind Generators in Rural Distribution Feeders: K. Mitchell M. Nagrial J. Rizk

STUDY OF WIND GENERATORS IN RURAL DISTRIBUTION FEEDERS
K. Mitchell
M. Nagrial
J. Rizk
School of Engineering and Industrial Design

University of Western Sydney
Locked Bag 1797, Penrith South DC
NSW 1797
AUSTRALIA
E-mail: k.mitchell@uws.edu.au
Abstract
This paper is an update on a research project commenced in 2002 to develop technical
planning guidelines for the grid connection of wind generation sources in weak rural
distribution feeders, based on data from an existing wind farm in the central tablelands
district of NSW. Results obtained after about 4 years monitoring are discussed together
with some preliminary planning rules.
1.
INTRODUCTION
Wind energy has now become well established in

many areas of the world, particularly in the Atlantic
seaboard countries. However, there are still a number
of issues concerning the safe levels of penetration of
wind power [1], and a number of planning guidelines
e.g. [2] and others, have been developed as a result.
However, most overseas developments are associated
with large wind farms connected to transmission
systems, where the grid is relatively short and solid.
In Australia, with its wide-flung rural areas, there are
many potential uses for grid connected wind power, as
grid extension costs are relatively high. A generator
embedded deep in the distribution network does not
have to compete with base-load generation plant, but
rather with the delivered cost of power at the
distribution level. Wind generation is now becoming
competitive at this level. This is especially so along
the NSW tablelands regions, where the wind resource
is good. However, in these areas, the grid is weak or
often non-existent and this has hampered
development. The electricity supply industry in
particular is concerned about the impact of wind farms
on rural distribution lines and numerous seminars and
conferences have been devoted to the topic.
The Australian Greenhouse Office set up a project in
2002 using the existing wind farm at Hampton, in the
central tablelands of NSW, as a test case. This is a
small wind farm connected to two rural feeders, which
are further complicated by the presence of in-line
voltage regulators. The Hampton Wind Farm was
chosen as it is an existing small wind farm and thus
would represent a typical embedded rural wind
generation application.
An active program of turbine real and reactive power

outputs has been under way for 4 years. A report was
made in the 2003 Aupec conference [3] and
preliminary findings discussed, based on 10 months
data. It was noted in that paper that about 25% of
rated turbine capacity was available on the ten highest
load days over the year 2000 winter season.
2.
SUPPLY ISSUES IN RURAL GRIDS
One of the major issues surrounding wind and solar

sources is their stochastic nature. Whilst being
increasingly recognized as useful energy sources, they
have been regarded as of little value as power sources
at time of peak load. Some European data suggests
that wind farms can only be counted upon to provide
about 10% of their rated output as a base load
equivalent to t he grid [4]. Furthermore, in South
Australia, which has the largest wind penetration of
any Australian state, has reported significant
variations in total wind output over periods as fast as
one hour [5]. Not surprisingly, a major issue is what
the relevant figure would be for other Australian wind
farms, noting that wind regimes can vary considerably
from season to season, site to site.
As part on the on-going research for the Hampton
wind farm, data has been collected continuously at 15minute intervals from 2002 to the current day. In this
paper, the patterns for the 2003, 2004 and 2005 winter
season are compared to the original 2002 results.
Main interest has been focussed on each winter season
as the local area is in the central tablelands of NSW,
Australia, which experiences cold winters with a cold
day winter system peak. It is worth noting that the
results published would only apply to tablelands type
climates; coastal ones in particular being quite

different and being subject to hot weather, summer
peaks.
The previous paper [3] discussed the range of issues
with embedded wind generators, including:
Voltage rise, particularly when connected

towards the end of the feeder
Voltage fluctuations, again particularly when
connected towards the end of the feeder
Islanding during feeder faults, (if synchronous
type)
Harmonic distortion due to starting and stopping
transients
Impact on in-line voltage regulators, causing
excessive tap-changes
Need for feeder capacity increases to
accommodate generation inflow
Difficult to put value on peak energy or benefit
to grid support during peak load times
In this paper, the 4 years data has been used to better

put a value on planning capacity and peak output
benefit to the network (last item).
3.
COLLECTED DATA
The original results for winter and summer 2002 from

[1] are shown in figure 1(a) and 1 (b) respectively, and
those for winter and summer 2003, 2004 and winter
2005 in figures 2(a) and 2(b) to 4 respectively, in the
Attachment. The graphs show the daily load pattern
for the top 10% of net load on one of the 11,000-volt
HV distribution lines that traverses the area, the wind
turbine being near the extremity of the feeder. Shown
on each diagram are the raw feeder load, wind turbine
output and net feeder load, in amps. The winter 2003
graphs show raw wind speed as well.
4.
DISCUSSION OF RESULTS
The effect of the wind input, in terms of proportion of

turbine rated output, is summarised for each year in
table 1.
Table 1
Effective Levels of Wind Generation at Top 10% of
Load Days
Year/Season Winter
Summer
2002
25
n.a.
2003
23
43
2004
40
15
2005
41.5
20
Examination of table 1 reveals two main effects.
Firstly, there are significant variations between
individual years, with wind inputs varying from 25%
to over 40% of rated output a variation of 60%.

Interestingly, 2002, the year of the original report had
the least wind input. It shows that at least 10 years of
good wind data would be required to obtain a useful
history. In Australia that would equate to a history
encompassing at least one El Nino - non El Nino
cycle. Secondly, despite variations in strength, overall
seasonal patterns are consistent from year to year. In
this case, spring is always the windiest season,
followed by winter and summer. This seasonal
pattern, would however, be very site and region
specific. The seasonal wind patterns would be quite
different on nearby coastal regions. Thus care needs
to be taken in extrapolating wind patterns from one
site to another.
The results did however strongly indicate that a certain
proportion of generator output can be depended upon
at time of system peak. This is the crux of the
network planning issue. Although wind output can and does regularly fall to zero, the issue is the
correlation between wind output and raw feeder load.
The impact of the wind generator is (as with any
embedded generation) is the amount of wind input that
is present statistically at times of feeder peak. The
figures suggest strongly that days of peak feeder load
in the NSW tablelands are consistently windy days.
This result as stated earlier, is site specific and other
figurers would apply to other areas.
5.
CONCLUSIONS
The results of the earlier report have been confirmed

and a certain percentage of turbine output can be
depended upon in a planning sense over the network
peak in the tablelands area.
In this particular case it has been shown to be 25% of
turbine output although it was as high as 50% in one
season. It will vary according to site and timing of
local feeder peak load. However, to ignore the wind
contribution completely (i.e. use zero) would be being
overly pessimistic.
The level of input has been shown to vary
considerably from season to season; so at least 10
years data is required to determine the minimum level
for planning purposes with some confidence. It is
planned to continue work and collect this amount of
data at the Hampton site.
6.
REFERENCES
[1] Outhred H (2003), Wind Energy and the National

Electricity Market, with particular reference to
South Australia, Aust Greenhouse Office Report,
Jan 2003
[2] Arnott I (2002), The participation of wind energy

in the National Electricity Market, Australian
Wind Energy Conference, Melbourne, 29-30
April.
[3] Mitchell K, Nagrial M, Rizk J, (2003), The
Development of Technical Planning Rules for
Embedded Wind Generators in Rural Distribution
Feeders,
AUPEC
Annual
Conference,
Christchurch, September 2003.
[4] Baitch A (2006), International Developments:
CIGREs Involvement in Wind Energy Systems,
Electric Energy Society of Australia Seminar on
Successfully Integrating Wind Energy into
Electricity Networks, Sydney, May 2006.
[5] Macaulay C (2006), Integration of Wind
Generation Into the NEM, Electric Energy
Society of Australia Seminar on Successfully
Integrating Wind Energy into Electricity
Networks, Sydney, May 2006.
ATTACHMENT
Selected Max Days W 2002
Selected Max Days S2002/3
80
60
Load
40
Wind
20
2200
1915
1630
1345
80
60
Load
40
Wind
20
Fig 1(a) Results for Winter 2002

Net Load
20
100
15
75
10
50
2200
2000
1800
1600
Time
2200
1915
120
100
80
Load
60
Wind
40
Net Load
20
140
120
100
80
60
40
20
0
2200
1915
1630
1345
1100
815
530
Time
S elected Max D ays W 2004
Fig 2(b) Results for Summer 2003/4

Load
W ind
Amps at 11kV
Time
100
80
Load
60
Wind
40
Net Load
20
Time
Load
Wind
Time
Fig 4 Results for Winter 2005
2100
1800
1500
1200
900
600
300
Net Load
2200
1915
1630
1345
1100
815
530
245
Selected Max Days W 2005

160
140
120
100
80
60
40
20
0
Selected Max Days S2004/5

120
2100
1800
1500
1200
900
600
300
Net Load
Amps at 11kV
1630
Selected Max Days S 2003/4
245
1400
1200
1000
200
800
600
400
25
Amps at 11kV
25
125
Amps at 11kV
1345
W Speed
Speed, km/h
Wind
Time
150
Load, 11kV Amps
Load
1100
815
530
Time
245
1100
815
530
245
100
Amps at 11kV
Amps at 11kV
100

Study of Wind Generators in Rural Distribution Feeders: K. Mitchell M. Nagrial J. Rizk

Uploaded by

Copyright:

Available Formats

Study of Wind Generators in Rural Distribution Feeders: K. Mitchell M. Nagrial J. Rizk

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Study of Wind Generators in Rural Distribution Feeders: K. Mitchell M. Nagrial J. Rizk

Uploaded by

Copyright:

Available Formats

STUDY OF WIND GENERATORS IN RURAL DISTRIBUTION FEEDERS

School of Engineering and Industrial Design

Wind energy has now become well established in

An active program of turbine real and reactive power

SUPPLY ISSUES IN RURAL GRIDS

One of the major issues surrounding wind and solar

climates; coastal ones in particular being quite

Voltage rise, particularly when connected

In this paper, the 4 years data has been used to better

The original results for winter and summer 2002 from

The effect of the wind input, in terms of proportion of

to over 40% of rated output a variation of 60%.

The results of the earlier report have been confirmed

[1] Outhred H (2003), Wind Energy and the National

[2] Arnott I (2002), The participation of wind energy

Selected Max Days S2002/3

Fig 1(a) Results for Winter 2002

Fig 2(a) Results for Winter 2003

S elected Max D ays W 2004

Fig 2(b) Results for Summer 2003/4

Fig 3(a) Results for Winter 2004

Fig 4 Results for Winter 2005

Fig 3(b) Results for Summer 2004/5

Selected Max Days W 2005

Selected Max Days S2004/5

Selected Max Days S 2003/4

Fig 1(b) Results for Summer 2002/3

Load, 11kV Amps

You might also like