AckermannTh. (ed) Wind Power in Power Systems
Подождите немного. Документ загружается.
//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 15_CHA14 .3D – 311 – [299–330/32]
20.12.2004 7:46PM
a maximum wind energy penetration, above which the economic return from adding
wind power begins to decrease. In addition, managers of larger systems usually take a
cautious approach because of generally high fluctuations in wind power.
As indicated by the dotted line in Figure 14.7, the level of wind energy penetration can
be expected to increase significantly in the future. Thus the challenge of na tional (and
transnational) systems will be to increase penetration to levels already existing in smaller
isolated systems, which themselves seem to be well placed to increase their wind energy
penetration to levels typical for just slightly smaller systems.
Obviously, great care has to be taken in the process of introducing wind power into
isolated power systems, as many failures have occurred in the past because of over-
ambitious system designs with a high degree of complexity and limited background
experience in project development. The recommended approach seems to be a gradual
increase, starting at the dashed line of Figure 14.7 and moving upwards towards the
dotted line in a step-by-step approach applying simple, robust, reliable and well-tested
concepts.
14.4 Systems and Experience
The present state-of-the-art of wi nd turbines is given in Chapter 4 of this book as well as
in a number of studies (see for example Hansen et al., 2001). Even though there is rapid
development in wind turbine technology there seems to be a pretty clear picture. The
present state-of-the-art of power systems with wind energy is more difficult to assess
because of the great variety of concepts and applications (for an impression of the
100
80
60
40
20
Wind Penetration (%)
10 100 1 kW 10 kW 100 kW 1 MW 10 MW 100 MW 1 GW
Installed system power
Sal
Mindelo
Dachen
Foula Island
La Desirade
Masabit
Cape Clear
Rathlin Island
Frøya Island
Denham
10 GW
Denmark
(1998)
100 GW
Denmark
(2030)
Today
Future
1 TW
Micro systems
Village power systems
Island power systems
Large interconnected systems
Figure 14.7 Present and expected future development of the wind energy penetration vs. the
installed system capacity, Lundsager et al. (2001), reproduced with permission
Wind Power in Power Systems 311
//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 15_CHA14 .3D – 312 – [299–330/32]
20.12.2004 7:46PM
current state of the technology, see Baring-Gould et al., 2003; Lundsager and Madsen,
1995; Lundsager et al., 2004; Pereira, 2000).
14.4.1 Overview of systems
The source of the system concepts applied can be found in the quite large number of
more or less experimental systems installed at research facilities throughout the world.
Table 14.2 presents a summary of these systems from Pereira (2000), where a complete
overview can be found.
Table 14.3 gives an overview of larger AC-based hybrid power systems installed
worldwide over the last decade. All the systems in the table produce power to their
community, but most of them are installed in the framework of demonstration or
validation projects with some degree of public co-financing (more details of the systems
can be found in Baring-Gould et al. (2003); Pereira, 2000).
14.4.2 Hybrid power system experience
The most recent account of systems and experience was given at the September 2002
wind diesel conference (Baring-Gould et al., 2003). During the conference, 12 presenta-
tions were given providing experience on the operation of more than 17 wind–diesel
power systems from around the world. The associated website (DOE/AWEA/Can-
WEA, 2002) contains links to all the presentations. We have selected five of these
projects that show the varied options for using diesel retrofit technology. These systems
Table 14.2 Selected list of hybrid power systems at research facilities
Laboratory Country Year Rated power
(kW)
Special features
NREL USA 1996 160 Multiple units, AC & DC buses,
battery storage
CRES Greece 1995 45 —
DEWI Germany 1992 30 Two wind turbines
RAL UK 1991 85 Flywheel storage
EFI Norway 1989 50 Wind turbine simulator and wind turbine
RERL–UMASS USA 1989 15 Rotary converter
IREQ Canada 1986 35 Rotary condenser
AWTS Canada 1985 100 Multiple units, rotary condenser
Risø Denmark 1984 30 Rotary condenser
Source: Pereira, 2000.
Note: NREL ¼ National Renewable Energy Laboratory; CRES ¼ Center for Renewable Energy
Resources; DEWI ¼Deutsches Windenergie-Institut; RAL ¼Rutherford Appleton Laboratory; EFI
¼Electric Power Research Institute; RERL–UMASS ¼ Renewable Energy Research Lab at UMass
Amherst; University of Massachusetts; IREQ ¼ Institut de Recherche d’Hydro Que
´
bec; AWTS ¼
Atlantic Wind Test Site
312 Isolated Systems
//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 15_CHA14 .3D – 313 – [299–330/32]
20.12.2004 7:46PM
also depict installations in very remote locations without easy access to highly developed
infrastructure and advanced technical support.
14.4.2.1 Wales, Alaska: wind–diesel high-penetration hybrid power system
The average electric load for this community is about 70 kW, although there are also
substantial heating loads for buildings and hot water. The Wales, Alaska, wind–diesel
hybrid power system, which underwent final commissioning in March 2002, combines three
diesel gensets with a combined power of 411 kW, two 65 kW wind turbines and a 130 Ah
battery bank made up of SAFT (SAFT Batteries SA) nickel–cadmium batteries, an NREL-
built rotary power converter and various control components. The primary purpose of
the system is to meet the village’s electric demand with high-quality power, while minimis-
ing diesel fuel consumption and diesel engine runtime. The system also directs excess wind
power to several thermal loads in the village, thereby saving heating fuel.
Limited data are available, but it is expected that the two wind turbines will provide
an average penetration of approximately 70 %, saving 45 % of expected fuel consump-
tion and reducing the operational time of the diesel engines by 25 %. The power system
has experienced some service-related technical problems since its installation, which has
reduced its performance.
Table 14.3 Selected list of hybrid power systems installed throughout the world in the past
decade
Site Country or region Reporting
period
Diesel
power
(kW)
Wind
power
(kW)
Other features Wind
penetration
(%)
Wales Alaska 1995–2003 411 130 Heating, storage 70
St Paul Alaska 1999 300 225 Heating —
Alto Baguales Chile 2001 13 000 1 980 Hydro power 16
Denham Western Australia 2000 1 970 690 Low load diesel 50
Sal Cape Verde 1994–2001 2 820 600 Desalination 14
Mindelo Cape Verde 1994–2001 11 200 900 Desalination 14
Dachen Island China 1989–2001 10 440 185 Low-load diesel 15
Fuerteventura Canary Island 1992–2001 150 225 Desalination, ice —
Foula Island Shetland Islands 1990–2001 28 60 Heating, hydro power 70
La Desirade Guadeloupe 1993–2001 880 144 — 40
a
Marsabit Kenya 1988–2001 300 150 — 46
Cape Clear Ireland 1987–1990 72 60 Storage 70
a
Rathlin Island Northern Ireland 1992–2001 260 99 Storage 70
Kythnos Island Greece 1995–2001 2 774 315 Storage, photovoltaic
power
—
Frøya Island Norway 1992–1996 50 55 Storage 94
Denham Australia 1998–present 1 736 230 — 23
Lemnos Island Greece 1995–present 10 400 1 140 — —
a
Peak.
Sources: Baring-Gould et al., 2003; Pereira, 2000.
Wind Power in Power Systems 313
//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 15_CHA14 .3D – 314 – [299–330/32]
20.12.2004 7:46PM
14.4.2.2 St Paul, Alaska: high-penetration wind–diesel power system
In 1999, the first fully privately funded high-penetration, no-storage hybrid power
system that maximises wind penetration was installed by TDX Power and Northern
Power Systems. The primary components of the St Paul, Alaska, plant include a 225 kW
Vestas V27 wind turbine, two 150 kW Volvo diesel engine generat ors, a synchronous
condenser, a 27 000 l insul ated hot water tank, approximately 305 m (1000 feet) of hot
water piping and a microprocessor-based control system capable of providing fully
automatic plant operation.
The primary electrical load for the faci lity averages about 85 kW, but the system also
supplies the primary space heating for the facility, which it does with excess power from
the wind generators and thermal energy from the diesel plant. When the wind generation
exceeds demand by a specific margin, the engines automatically shut off and the wind
turbine meets the load demand with excess power diverted to the hot water tank, which
in turn is used to heat the complex. When wind power is insufficient to meet the load,
the engines are engaged to provide continuous electric supply as well as provide energy
to the hot water system when needed. The total 500 kW wind–diesel co-generation
system cost approximately US$1.2 million. Its operation has saved US$200 000 per year
in utility electric charges and US$50 000 per year in diesel heating fuel. Since its
installation, the load has continued to grow , and additional thermal heating loops have
been added to the facilities.
14.4.2.3 Alto Baguales, Chile: wind–hydro–diesel power system in Coyhaique
The system supplies energy to the regional capital of Coyhaique, Chile, providing a
maximum power of 13.75 MW. In the autumn of 2001, three 660 kW Vestas wind
turbines wer e installed to supplement the diesel and hydro production. The Alto
Baguales wind energy project is exp ected to provide more than 16 % of the local electric
needs and displace about 600 000 liters (158 500 gallons) of diesel fuel per year.
The turbines are operated remotely from the diesel plant, with no additional control
capabilities, and operate at around a 50 % capacity factor because of strong winds at the
turbine site. To date, the highest recorded penetration, based on 15 minute instant-
aneous readings taken at the power station, is 22 % of total demand. In the summer of
2003, additional hydro capacity will be installed, allowing the utility to provide the
whole load with wind and hydropower, completely eliminating diesel production.
The primary challenge of the project implementation was in obtaining the proper
installation equipment, including a crane that had to be brought in over the mountains
from Buenos Aires, Argentina.
14.4.2.4 Cape Verde: the three major national power systems
The Archi pelago of the Republ ic of Cape Verde consists of 10 major islands off the
Western cost of Africa. Three medium-penetratio n wind–diesel systems have success-
fully provided power to the three main communities of Cape Verde – Sal, Mindelo and
Praia – since the mid-1990s. These power systems are very simple in nature, containing
only a dump load and a wind turbine shutdown function to keep minimum diesel load
314 Isolated Systems
//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 15_CHA14 .3D – 315 – [299–330/32]
20.12.2004 7:46PM
conditions. The three 300 kW NKT wind turbines at each site are connected to the
existing diesel distribution grid in a standard grid-connected fashion. The average loads
for the communities vary from 1.15 MW for the smallest, Sal, to 4.5 MW for the largest,
Praia (the nation’s capital).
At present, the power systems operate at monthly wind energy penetrations of up to
25 %, depending on the system and time of year. Yearly penetrations of up to 14 % for
Sal and Mindelo have been experienced. The maximum monthly wind power penetra-
tion of 35 % was reached in Sal, with no adverse system impact. The overall experience
of the three wind plants has been perceived by the utility ELEKTRA as quite positive
and this has resulted in the initiation of a second phase, financed through the World
Bank and managed by Cape Verde, in which wind energy penetration of the three power
systems will be almost doubled.
Risø National Laboratories of Denmark has conducted a detailed assessment of the
phase 2 systems using WINSYS modelling software (Hans en and Tande, 1994). The
expansion of these systems is expected to increase the wind energy penetration in the
first year of operation to levels of up to 30 % (Mindelo). The average annual diesel fuel
consumption is expected to be reduced by an additional 25 %.
14.4.2.5 Australia: wind–diesel power stations in Denham and Mawson
The Denham wind–diesel power station is operated by Western Power and is located on
the Western coast of Australia, north of the regional capital of Perth. The power system
has a maximum demand of 1200 kW, which is supplied by 690 kW of wind from three
230 kW Enercon E-30 wind turbines, four diesel engines with a total rati ng of 1720 kW,
and a recently installed low-load diesel. The low-load diesel is a standard diesel gen-
erator that is modified to allow prolonged operation at very low or negative loads. The
device installed in Denham has a load range of þ250 kW and 100 kW. The power
system was installed by PowerCorp of Australia and is controlled from a highly auto-
mated control centre located in the powerhouse using control logic, also developed by
PowerCorp. The power system has the capability to operate in a fully automated mode
with minimal technical oversight. The system control allows all of the standard diesels to
be shut off, resulting in an average penetration of 50 %.
The power system has been operati ng for more than three years, supplying utility-
quality power and saving an estimated 270 000 l (71 300 gallons) of fuel per year.
In a separate but similar project, PowerCorp recently installed two Enercon E30
300 kW wind turbines with a 200 kW dynamic grid interface into the Mawson diesel
power system in Antarctica. This system, which does not contain stora ge, has achieved
an average penetration of approximately 60 % since its commissioning in January 2003.
14.5 Wind Power Impact on Power Quality
In systems with turbines connected to the AC bus, the critical consideration is the
variation of the power output from wind turbines and its impact on the operation of
the power system and on power quality. Thi s impact increases as the level of penetration
increases.
Wind Power in Power Systems 315
//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 15_CHA14 .3D – 316 – [299–330/32]
20.12.2004 7:46PM
The influence on power quality is mainly noted in the level and fluctuation of system
voltage and frequency, the shape of the voltage signal (power harmonics) and its ability
to manage a reactive power balance. It is also important that the voltage of the
distribution network remains constant, especially when interconnected to the wind
turbines.
14.5.1 Distribution network voltage levels
Isolated power systems considered in this context are usually characterised by having
only one power station. The transmission and distribution network is thus usually quite
simple and weak, meaning it is suscept ible to low-voltage issues. When the wind turbines
are connected to the grid, they will often be at locations where consumers are also
connected, not close to the power station.
The voltage at the point of common connection will depend on the output from the
wind turbines and on the consumer load. Situations where the voltage becomes high as a
result of an overabundance of wind power production and a low consumer load may
occur. Measurements completed at the system in Hurghada in Egypt (in the context of
the study described in Lundsager et al., 2001) illustrate the dependence of the voltage on
wind power production and consumer load. Several other chapters in this book provide
examples on the influence of voltage levels of both wind turbine (farm) output and
consumer loads, see Chapters 12 and 19 for instance.
14.5.2 System stability and power quality
In all power systems that produce AC power, four critical parameters must be main-
tained:
.
System frequency: the balance of energy being produced and energy being consumed.
If more energy is being produced, the frequency will increase.
.
Active (real) and reactive power: depending on the types of loads and devices being
used, a balance of active and reactive power must be maintained. In addition, some
loads will require large amounts of reactive power, such as any inductive motor, and
the equipment operating the system will need to provide this reactive power.
.
System voltage: stable voltage from the source is required to ensure the proper
operation of many common loads. Control is generally maintained by modulating
the field of a rotating generator or through the use of power electronics.
.
Harmonic distortion of voltage waveform: different generation devices will provide
voltage of different harmonic quality. The loads being powered by the device will also
impact the harmonic nature of the power, particularly the use of power electronics such
as in the ballast of most fluorescent lighting or power converters. Rotating equipment
provides the voltage waveform with the least amount of harmonic distortion.
Different devices can be used to control different aspects of system power quality. The
need for different devices will depend on the instantaneous penetration of renewable
(wind) energy that is expected. The higher the penetration, the more one must worry
316 Isolated Systems
//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 15_CHA14 .3D – 317 – [299–330/32]
20.12.2004 7:46PM
about how this will impact power quality and the more devices that must be added to the
system to ensure high-quality electrical power.
14.5.3 Power and voltage fluctuations
The key driver of many of these power quality issues is the power fluctuations from the
wind turbines. The power fluctuations create fluctuations in system voltage and impos e
fluctuations in the diesel output. Voltage fluctuations can create disturbances such as
light flicker and low intensity. The power fluctuations depend not only on the amount of
installed wind power capacity but also on the number of wind turbines. The decrease in
power fluctuations with increasing number of wind turbines is a result of the stochastic
nature of the turbulent wind: the power fluctuations from the individual wind turbines
will, to some extent, be independent of each other and they will therefore even out some
of the higher frequency fluctuations. Keeping power fluctuations small is a desirable
feature in an isolated power system with a high penetration level.
When diesel generators are operating in parallel to wind turbines then, clearly, as the
standard deviation of the wind power increases so will the standard deviation of diesel
power since the diesel is required to fill any difference between the required load and the
wind generation (see Figure 14.8). This is an especially important point when considering
the retrofit of existing diesel plants with wind energy as care will have to be taken to ensure
that the current diesel controls can cope with the expected wind power fluctuations.
14.5.4 Power system operation
The operation and performance of an isolated system with wind power depends on
many factors, too numerous to mention. Generally speaking, however, the complexity
0.00
2.00
4.00
6.00
8.00
10.00
12.00
14.00
500.0 550.0 600.0 650.0 700.0 750.0 800.0 850.0 900.0 950.0
Average output power (kW)
STD (%)
Figure 14.8 Standard deviation (STD) of wind farm power in Mindelo, Cape Verde, as a
percentage of rated power
Source: Hansen et al., 1999, reproduced with permission.
Wind Power in Power Systems 317
//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 15_CHA14 .3D – 318 – [299–330/32]
20.12.2004 7:46PM
of operation, as well as the expected reduction of diesel fuel consumption is dependent
on the level of wind penetration. The higher the penetration, the higher the fuel savings
but also the higher the complexity. Low-penetration systems usually do not require any
automated supervisory control whereas high-penetration system must have very sophis-
ticated control systems to ensure proper ope ration.
The operation of the power system must be carefully considered as part of the power
system design. Additionally, the higher the degree of system complexity the more
important it is that the system has the ability to operate under simpler struc tures. This
concept of ‘failure safety’ promotes the use of different levels of operation that can fall
back to simpler operation upon the failure of specific components. In the extreme case,
the diesels should always be able to supply power to the grid even if everything,
including the supervisory control, fails.
One concept that is critical to the operation of remote power systems is that in such
systems the levels of penetration that can be reached are much higher than in standard
grid-connected wind turbines. For example, systems have been installed where the wind
turbine can provide three to four times the existing load, something completely unheard
of in the grid-connected market. For this reason, an understanding of the diurnal and
seasonal fluctuations in consumer load must be considered in the design of all higher
penetration wind–diesel power systems.
0 5 10 15 20
0
50
100
150
200
250
Time of day (hour)
Active load (kW)
Diurnal load pattern, three corner hotel, Spring
0 5 10 15 20
0
20
40
60
80
100
Time of day (hour)
Reactive load (kVA)
Diurnal reactive load pattern, three corner hotel, Spring
(a)
Figure 14.9 Diurnal patterns for (a) a hotel and (b) residential load, Hurghada, Egypt
Source: Hansen et al., 1999, reproduced with permission.
318 Isolated Systems
//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 15_CHA14 .3D – 319 – [299–330/32]
20.12.2004 7:46PM
Owing to their impac t on the operation strategy of the diesels, several other aspects of
the load are also important. Briefly summarising, the minimum load, together with the
allowed technical minimum load of the diesel, determines the amount of wind power
that can be absorbed by the system without dissipation of surplus power to a dump load
or the battery bank. The maximum load determines the necessary installed diesel
capacity. The variations in loads and wind power – active and reactive – together with
the rate of change of the load and the rate of change of the wind power, determine the
necessary spinning capacity in order to ensure adequate power quality and reliability in
terms of loss of load probability.
The operation and performance of an isolated system with wind power generally
depends on the characteristics of the load. The load is the sum of many individual loads
which in some cases may be categorised. Figures 14.9(a) and 14.9(b) show the diurnal
load patterns of a hotel and residential load, respectively, measured as part of a study in
Egypt (described by Hansen et al., 1999; Lundsager et al., 2001).
Clearly, there is a distinct diurnal pattern for the residential load whereas the hotel load is
quite constant over the day. It is also seen that the pattern for the reactive power is very
similar to that of the active power. Day-to-day variations of loads shown by the minimum to
0 5 10 15 20
0
50
100
150
200
Time of day (hour)
Active load (kW)
Diurnal load pattern, residential area, Spring 1998
0 5 10 15 20
(b)
20
40
60
80
100
120
Time of day (hour)
Reactive load (kVA)
Diurnal reactive load pattern, residential area, Spring 1998
Average Max Min
Figure 14.9 (continued)
Wind Power in Power Systems 319
//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 15_CHA14 .3D – 320 – [299–330/32]
20.12.2004 7:46PM
maximumbandaresmallintheresidentialloadcase,butinthehotelloadcasethevariation
is larger. This is also seen in the standard deviation of the loads shown in Figure 14.10.
Last, in all higher power systems, significant consideration must be given to uses of
surplus energy such as space or water heating and water purification. In many cases, the
savings of displacing the secondary fuel used in these other processes can play a key role
in making a project financially successful.
Owing to their impact on the operation strategy of the diesels, several aspects of the
load are important. The minimum consumer load and the allowed technical minimum
load of the diesels determine the amount of wind power that can be absorbed by the
system without dissipation of surplus power. The maximum consumer load determines
the necessary installed diesel capacity. The variations in consumer loads and wind power
determine the necessary spinning capacity in order to ensure adequate power quality
and reliability in terms of loss of load probability.
14.6 System Modelling Requirements
Numerical modelling is an important part of the design, assessment, implementation
and evaluation of isolated power systems with wind power. Usually, the performance of
such systems is predicted in terms of the technical performance (power and energy
production) and the economical performance, typically the cost of energy (COE) and
the internal rate of return (IRR) of the project.
Technically, the performance measures may also include a more detailed design-based
analysis of:
.
power quality measur es;
.
load flow criteria;
.
grid stability issue s;
.
scheduling and dispat ch of generating units, in particular diesel generators.
0 5 10 15 20
0
5
10
15
20
25
30
35
Time of day (hour)
STD (kW)
Hotel
Residential
Figure 14.10 Standard deviation (STD) of the load categories in Figures 14.9(a) and 14.9(b)
Source: Hansen et al., 1999, reproduced with permission.
320 Isolated Systems