Title: Energy and the New Reality, Volume 2: C-Free Energy Supply Chapter 3: Wind Energy L. D. Danny Harvey harvey@geog.utoronto.ca
1Energy and the New Reality, Volume 2C-Free
Energy Supply Chapter 3 Wind EnergyL. D.
Danny Harveyharvey_at_geog.utoronto.ca
Publisher Earthscan, UKHomepage
www.earthscan.co.uk/?tabid101808
- This material is intended for use in lectures,
presentations and as handouts to students, and is
provided in Powerpoint format so as to allow
customization for the individual needs of course
instructors. Permission of the author and
publisher is required for any other usage. Please
see www.earthscan.co.uk for contact details.
2Figure 3.1a Annual additions to wind energy
capacity
3Figure 3.1b Growth in total wind energy capacity
4Figure 3.2a Breakdown of installed capacity at
the end of 2009
5Figure 3.2b Capacity (MW) installed in 2009
6Figure 3.3 Wind farm at Pincher creek, Alberta
Source Garry Sowerby
7Table 3.1 Market shares of the worlds leading
wind turbine manufacturers. Source BTM Consult
Press Release, March 2007
8Components of a Wind Turbine
- Foundation
- Tower
- Rotor
- Nacelle
- Gearbox (usually)
- High speed shaft
- Generator
- Control system, cooling unit, anemometer
- Yaw mechanism
9Turbine characteristics
- Rotor diameter up to 120 m
- Hub height up to 120 m
- Peak electrical power output up to 5 MW now, up
to 10 MW foreseen - Cut-in wind speed (typically 3-4 m/s)
- Rated wind speed (typically 15 m/s)
- Cut-out wind speed (typically 25 m/s)
10Figure 3.4 Progression of rotor sizes over time
11Figure 3.5a Rotor diameter vs rated power
12Figure 3.5b Hub height vs rated power
13Figure 3.6 Minimum hub height vs rotor diameter
14Figure 3.7a Power curves for wind turbines with
80-m, 87-m, and 90-m rotors and a 2.0-MW generator
15Figure 3.7b Power curves for wind turbines with
different rotor-generator combinations
16Wind turbine aerodynamics
- Lift, not a pushing force, is what makes the
rotor rotate - Thus, the aerodynamics of a wind turbine have
much in common with the aerodynamics of an
airplane wing
17Figure 3.8 Airflow Past Wing
18Figure 3.9 Forces acting on a turbine rotor blade
Source Danish Wind Turbine Manufacturers
Association
19Efficiency of a wind turbine this is the ratio
of the electrical power produced (W) to the power
of the wind passing through the area swept by the
rotor blades. It is the product of three factors
- Aerodynamic efficiency (ratio of mechanical power
of the rotor to wind power) - Mechanical efficiency (ratio of mechanical power
of the generator axis to the mechanical power of
the rotor axis) - Electrical efficiency (ratio of electrical power
fed into the grid to the mechanical power of the
generator axis)
20- The maximum possible aerodynamic efficiency, as
given by Betz Law, is 59.3, and occurs if the
turbine slows the wind down to 2/3 of its
original speed. The aerodynamic efficiency of a
real turbine varies with wind speed, having a
typical peak value of 44 and a typical value
averaged over all wind speeds of 25 - A typical mechanical efficiency is 96-99
- A typical electrical efficiency is 96-97
- Multiply the efficiencies (expressed as a
fraction) to get the overall efficiency
21Figure 3.10 Variation of power output and
efficiency with wind speed for the Nordex N90-2.3
turbine
22Turbine generators
- Synchronous
-
- Asynchronous (induction)
- Variable speed
-
23Synchronous generators
- Common in fossil fuel powerplants, but rare in
wind turbines - Rotation speed is synchronized with the grid
frequency
24Recap Volume 1 Figure 3.2 Two Pole Synchronous
Generator
25Recap Volume 1 Figure 3.1 Three Phase AC Current
26Asynchronous (induction) generators
- If the rotor were to rotate at the same frequency
as the electric field in the stator, no
electricity would be produced - When the rotor of the generator rotates faster
than the stator, a strong current is induced in
the rotor - The harder one cranks on the rotor, the more
power that is transferred as electromagnetic
force to the stator, converted to electricity,
and fed to the grid - The difference in the rotation speed between no
power and peak power is about 1, but this slip
reduces stress on the rotor and smoothes out
power variations
27Variable Speed Generators
- Becoming more common
- Rotation rate of rotor varies with wind speed
from 8 rpm to 16 rpm - Results in less stress on the structure and more
uniform variation in power output - Requires more complex electronics and gearbox to
always produce electricity at the fixed grid
frequency
28Characteristics of wind
- Variation of mean wind speed with height
- Variation of turbulence intensity with height
- Weibull probability distribution function for
wind speed
29Figure 3.11 Logarithmic velocity profile
- U plots as a straight line on semi-log paper,
with slope u/?. zo is the height at which U
extrapolates to zero
30Figure 3.12 Effect of surface roughness on
velocity profiles
- Wind speed 100-200 m above the surface is fixed
(governed by the large scale pressure patterns) - Rougher surface the air feels the surface to
a greater height, so wind speeds are slower at
all heights within the first 100-200 m.
31An alternative mathematical representation of the
variation of wind speed with height is using a
power relationship,Uh/Uref (H/href)nThe
logarithmic relationship is theoretically valid
in a neutral atmosphere only.The power
relationship has no theoretical basis but
provides a good fit to observed atmospheric wind
profiles
32Figure 3.13 Turbulence intensity (wind speed
standard deviation divided by mean wind speed) vs
height
Source Soker et al (2000, Offshore Wind Energy
in the North Sea Technical Possibilities and
Ecological Considerations - A Study for
Greenpeace)
33Power output from a wind turbine
- Kinetic energy of a moving mass ½ mv2
- Power density of wind ½ ?V3
- The efficiency of a wind turbine is defined as
the ratio of power output to the power of the
wind in the area swept by the rotating rotor.
Thus, - Power output of a wind turbine
- efficiency x swept area x power density of
wind, or - P1/2 ?(pR2) ? V3
34Weibull Distribution Function
- Gives the probability of a wind speed occurring
per unit of wind-speed interval - Thus, the units are 1/(m/s)
- The value of the function times the width of the
interval gives the probability of the wind speed
occurring in that interval - The function is
- f(u)k/c(u/c)k-1exp(-(u/c)k)
-
- where c is the scale parameter and k is the
shape parameter
35Figure 3.15 Weibull wind speed distribution with
c5 m/s and k1.6
36Figure 3.14 Distribution of best-fit Weibull
scale factor (c) and shape factor (k) deduced
from observed wind velocity variations at various
sites
37Figure 3.16 Weibull wind speed probability
distributions
38Because wind power varies non-linearly with wind
speed
- The mean (average) wind power for a given mean
wind speed depends on the shape of the
probability distribution on either side of the
mean wind speed - The mean wind power (based on wind power
computed at many different wind speeds and then
weighted by the probabilities) is about twice the
wind power computed once at the mean wind speed
39Figure 3.17 Mean wind power vs mean wind speed.A
smaller k means a more spread out wind speed
distribution, so more winds at both very high and
very low wind speeds, but the high wind speeds
disproportionately contribute to wind power (due
to the cubic dependence), so the mean wind power
is greater at a given mean wind speed with
smaller k
40Table 3.3. Comparison of wind power computed at
the average wind speed with the average wind
power computed over a distribution of wind speeds
giving the same average wind speed.
41Mean Efficiency
- The power output at any given wind speed is given
by the wind power x swept area x efficiency, so
the efficiencies matter more when the wind power
is larger than when it is smaller - Thus, the appropriate mean efficiency involves
the efficiency at each wind speed times the
probability of that wind speed interval times the
wind power at that wind speed, divided by the
mean wind power
42Figure 3.18a Mean efficiency vs wind speed,
computed from the turbine power curve and the
Weibull wind speed probability distribution using
3 different shape parameters
43Figure 3.18b Mean turbine efficiency vs mean
wind speed for three turbines with similar
generator ratings
44Capacity Factor
- This is the mean (average) power output of the
turbine divided by the peak (or rated) power
output - The mean power output is computed as the power
output in the centre of each wind speed interval,
times the probability of that interval, summed
over all intervals and divided by the total
probability (which is 1.0)
45Figure 3.19a Variation of capacity factor with
wind speed for 3 different Weibull shape
parameters
46Figure 3.19b Variation of capacity factor with
wind speed for three different turbines
47Table 3.4 Average wind turbine capacity factors
in 2001. Source BTM Consult (2002).
48Figure 3.20 Mean wind speed over North America at
a height of 100 m.
Wind speed (m/s)
Source for this and other wind maps Prepared
from data file at power.larc.nasa.gov (go to
Sustainable Buildings, Global Datasets)
49Figure 3.21 Mean wind speed over Europe at a
height of 100 m.
Wind speed (m/s)
50Figure 3.22 Mean wind speed over China and
surrounding regions at a height of 100 m.
Wind Speed (m/s)
51Supplemental Figure Mean wind speed over North
Africa and the Middle East at a height of 100 m.
Wind speed (m/s)
52Supplemental figure Mean wind speed over
southern Africa at a height of 100 m.
Wind speed (m/s)
53Supplemental figure Mean wind speed over
Australia, Indonesia and adjoining regions at a
height of 100 m.
Wind speed (m/s)
54Supplemental figure Mean wind speed over South
America at a height of 100 m.
Wind speed (m/s)
55Windfarms
- Clustering of many wind turbines in a regular
array in a region of good winds - Turbines are typically spaced 5-9 rotor diameters
apart in the along-wind direction, and 3-5 rotor
diameters apart in the cross-wind direction - Clustering reduces costs (economies of scale for
installation), and takes better advantage of the
best wind sites
56Impact of windfarms on weather and climate
- Large wind farms (involving hundreds of wind
turbines at the closest permitted spacing (i.e.,
separated by 7 rotor diameters) would have a
noticeable effect on regional winds and hence on
vertical fluxes of heat and moisture in the
atmosphere, thereby changing the surface air
temperature in the region of the wind farm and
downstream from the wind farm - Interference with the winds might reduce the
overall power output from a wind farm by up to
30 compared to the case where the winds are
assumed to be unaffected by the wind farm
57Scaling Relationships
- Intercepted wind power varies with rotor diameter
squared, so with constant efficiency, power
output would also vary with D2 - If turbines are spaced apart at a constant
multiple of the rotor diameter (say, 7D x 7D),
then the land area also increases with D2 - Thus, the wind farm capacity (that is, the peak
power output at high wind speeds) per unit of
land area will be independent of the rotor
diameter - However, larger turbines also have a greater hub
height and so greater mean wind speed - Wind power varies with V3, so 15 greater wind
speed results in 52 greater wind power - Thus, as turbines have gotten bigger (and
higher), the energy production (kWh/yr) per unit
of land area has increased - Capacity factors would also be larger for larger
turbines, all else being equal, due to the
greater wind speeds at a greater turbine height.
58Figure 3.23 Variation of rotor swept area with
rated power for turbines listed in Table 3.2
59Figure 3.24a Peak wind farm power per km2 for
various turbines, assuming the turbines to be
arranged on a grid with a spacing of 7D x 7D,
where Drotor diameter
60Figure 3.24b Annual electricity production per
km2 for various turbines, assuming the turbines
to be arranged on a grid with a spacing of 7D x
7D, where Drotor diameter
61Offshore wind farms
- Wind turbines mounted on the seabed in water up
to 50 m deep - Can double or triple the cost of the wind turbine
connections to the grid, but there can easily
be twice the electricity production - Net result electricity for about the same to
50 higher price but with twice the capacity
factor (i.e., 40-50 instead of 20-25) - Turbines especially designed for offshore
conditions have been built
62Table 3.7 Additional costs of offshore wind
energy as a function of distance from the north
German coast.
Source Soker et al (2000, Offshore Wind Energy
in the North Sea Technical Possibilities and
Ecological Considerations - A Study for
Greenpeace)
63Figure 3.25 Middelgrunden wind farm, next to
Copenhagen
Source Danny Harvey
64Figure 3.26 European offshore wind atlas
Source Risø National Laboratory, Wind Atlases of
the World (www.windatlas.dk)
65Figure 3.27 Types of foundations
Source Soker et al (2000, Offshore Wind Energy
in the North Sea Technical Possibilities and
Ecological Considerations - A Study for
Greenpeace)
66Floating wind turbines
- Under development using technology transferred
from the North Sea offshore oil industry (which
is winding down) - One concept (WindSea) involves a
triangular-shaped floating platform with a 3.2MW
turbine at each corner on an outward-inclined
tower - One rotor would be on an airfoil-shaper tower
that would act like the tail of an airplane,
serving to continuously and automatically orient
all three rotors perpendicular to the wind
67Figure 3.28 Existing and planned (as of late
2009) offshore wind farms (total capacity
44,600 MW)
Source Richard Harrod, 4COffshore
68Fluctuations in Wind Electricity Production
- Because there might be times when the wind speed
might be less than the cut-in wind speed (so that
no electricity is produced), some amount of
non-wind backup capacity is needed (how much will
be discussed later) - As well, because wind is variable, some power
units that can go up and down to offset the
variations in wind electricity production are
needed - The problem is, the units most able to fluctuate
rapidly (such as simple-cycle natural gas
turbines) tend to be less efficient than the
units that would normally be used (such as coal
steam turbines or combined-cycle natural gas
systems) - Thus, it is desirable to minimize the variations
in the electricity production from wind
69Minimizing rapid (seconds to minutes)
fluctuations in wind output
- Use of variable-speed turbines provides some
smoothing of output on a time scale of seconds - Link together several turbines in a wind farm
provides some cancellation of fluctuations at a
time scale of up to a minute or so - Implement short-term storage of excess energy
- - flywheels, supercapacitors, plug-in hybrid
vehicles with a two-way connection to the grid
70Dealing with longer fluctuations (hours to days
and months)
- Link together wind farms over a broad region
- Use electrolyzers and fuels cells (making and
using H2) - Use flow batteries (regenerative fuel cells)
- Use underground compressed air energy storage
(CAES) - Use hydro-electric reservoirs
- Use heat pumps and thermal energy storage in
district energy systems - Use other flexible end-use electric loads
71Table 3.10 Impact on the statistical properties
of wind energy of spreading wind farms over
increasingly larger areas in and around Europe
Source Czisch and Giebel (2000, Wind Power for
the 21st Century, Kassel)
72Figure 3.29 Amalgamation of dispersed wind farms
Source Czisch and Giebel (2000, Wind Power for
the 21st Century, Kassel)
73Table 3.11 Largest variation in wind power over
different time periods and averaged over
differently-sized regions.
Source EWEA (2005, Large-scale Integration of
Wind Energy in the European Power Supply,
www.ewea.org)
74Making Use of Short-term Wind Forecasts
- The variability in electricity output that
remains after making use of the various
strategies outlined in the previous slides can be
better handled if the variation can be predicted
several hours in advance, as this permits
scheduling of slowly responding backup fossil
fuel power units - Thus, improving local wind forecasts with
high-resolution meteorological computer
forecasting models is an intensive area of
research at present
75Electrolyzers and fuel cells
- Electrolyzers generate hydrogen by splitting
water using electricity, and so could use excess
wind electricity - The hydrogen can be stored as a compressed gas
- When there is a shortage of wind electricity,
additional electricity can be generated by
running the electrolyzer backwards as a fuel cell - Rapid variations in output degrade the
performance and shorten the lifespan of
electrolyzers and fuel cells, so a battery would
likely be used to smooth out the electricity
input to the electrolyzer and smooth the demand
for extra electricity from the fuel cell
76Recap Figures 3.8 and 3.9 from Volume 1
Source (right) www.utcfuelcells.com
77Figure 3.30 Regenerative fuel cell (or flow
battery)
Source Modified from Lotspeich and Holde (2002,
Proceedings of the 2002 ACEEE Summer Study on
Energy Efficiency in Buildings 3, American
Council for an Energy Efficient Economy,
Washington)
78Compressed air energy storage (CAES)
- With electricity generation using a gas turbine,
about half of the turbine power is used to
compress the air needed for combustion, and only
about half is used to drive the generator that
produces electricity - If excess wind energy is used to compress air and
store it underground, the compressed air could be
directly used in a gas turbine to generate
electricity when there is a shortage of wind
power - This would more than double the efficiency of
using natural gas to produce electricity from
about 37 (with a simple-cycle turbine) to 84
79CAES was initially developed in the 1970s and
1980s as a means of absorbing excess nuclear
electricity (the output from a nuclear powerplant
is largely fixed, while electricity demand
varies, so if the plant meets a large fraction of
peak demand, there would be excess power at
times)Only two such plants were built (one in
Germany, one in Alabama)Now there is a revival
of CAES, with many plants under construction or
planned (especially in Texas) for storage of
excess wind energy
80Geological formations suitable for CAES
- Salt domes, salt beds and porous sedimentary
rocks are best - These underlie 75 of the land area of the US,
including many of the best wind regions - Salt domes closely coincide with the best wind
resource regions in Europe - Caverns can also be excavated in hard rock, but
these would be considerably more expensive
81Supplemental Figure Location of geology suitable
for CAES and of good wind resources in the US,
and CAES sites.
Source Succar and Williams (2008, Compressed Air
Energy Storage Theory, Resources, and
Applications for Wind Power, Princeton
Environment Institute, Princeton University,
Princeton, New Jersey)
82Figure 3.31 Wind CAES energy flow
Source Denholm (2006, Renewable Energy 31,
13551370, http//www.sciencedirect.com/science/jo
urnal/09601481)
83Existing and planned CAES facilities require some
supplemental fuel (natural gas at present, but it
could be gasified biomass in the future)A new
system under development is called advanced
adiabatic CAES (AA-CAES)In this system, no (or
very little) supplemental fuel is required.
Instead, the heat that is produced when air is
compressed is stored in the form of hot ( 650ºC)
molten salts in an insulated tank, and used
instead of fuel along with the compressed air to
generate electricity when needed
84Use of hydro-electric reservoirs
- Most hydro-electric reservoirs are not running at
full capacity all the time, because there is not
enough water - Thus, when there is excess wind, the water flow
and hence electricity production can be reduced,
and when there is a shortage, greater water flow
than would otherwise be the case (using the saved
water) can be allowed - This entails no energy loss, and in fact can
slightly increase the annual hydro-electric
energy production from the same annual water
flow, because the average reservoir level will be
greater (hydro-electric power production depends
on flow rate x elevation drop)
85- Many areas of the world that are lousy for CAES
(such as the hard pre-Cambrian rocks of the
Canadian Shield) have excellent hydro-electric
resources or excellent potential hydro-electric
resources
86Pumped hydro
- When there is no reservoir to hold back the river
flow, build a dam in some mountain valley and
pump water up behind the dam using excess wind
electricity, creating a reservoir - Let water drain the reservoir through a
conventional hydro-electric turbine at times of
wind electricity deficit - This is used in Europe (in the Alps, for example)
87Use of heat pumps
- One way to provide heat with electricity is
through electric resistance heating, but this is
not efficient (only 1 unit of heat per unit of
electricity used) - A better method is to use electric heat pumps,
which provide 3-4 units of heat per unit of
electricity used - If we have well-insulated buildings (such that
they can drift for a few hours without the
heating system on), then heat pumps could be used
when there is excess wind and the heating turned
off altogether when wind energy drops - Similarly, heat pumps (or chillers) can be used
for cooling purposes (in the summer) at times of
excess wind and turned off at times of deficit - In effect, excess wind energy is being stored as
thermal energy (heat or coldness)
88- There is no loss of energy in this way, so the
storage efficiency is 100 - If, however, heat or coldness is stored in a
large insulated tank outside the building (as in
some district energy systems), there would be
some loss of stored heat or coldness to the
environment - The storage efficiency in this case is lt 100,
perhaps 95
89Other dispatchable loads using dynamic demand
- The power industry talks about dispatchable
power sources those that can be quickly brought
on line and varied in output to meet fluctuating
electricity demand - The other side is to have dispatchable demand
demand that can be reduced by the power utility
to compensate for surges in demand elsewhere or
for the loss of power units - This is already done with things like electric
resistance water heaters a signal can be sent
from the utility to temporarily turn them off
90- When electricity demand exceeds electricity
supply, the voltage and frequency drop (and vice
versa when supply exceeds demand) - Equipment with compressors (such as refrigerators
and air conditioners) can sense changes in
frequency and can now be designed to
automatically shut down when the frequency drops
below some threshold - This could compensate for sudden drops in wind
power until the wind power resumes or backup
systems come on line - This is called dynamic demand
91Present round-trip efficiency of various energy
storage option (energy taken out vs energy put in)
- Electrolyzer/hydrogen/fuel cell system 32-42
- Flywheels (store for 1 day) 45
- Pumped hydro 65-80
- Flow batteries 75
- CAES 70-75
- AA-CAES 70
- Flywheels (store for seconds) 85
- Batteries 85-90
- Capacitors 95
- Thermal energy storage 95-100
- Hydro-electric reservoirs 100
92Improving system stability through the addition
of wind turbines
- Until recently, wind turbines had been thought of
only as a liability as far as system stability is
concerned (due to the fluctuating and partly
unpredictable nature of wind) - However, modern wind turbines can be used to
improve overall system stability by - - compensating for shifts in reactive power
(related to the phase shift between voltage and
current oscillations) caused by other supply
sources or loads in the system - - maintaining connection to the grid and
continuing to produce power when faults elsewhere
cause large transient variations in voltage (this
is called fault ride through (FRT) capability,
and is still under development) - - varying their output in response to changes in
grid frequency (this comes with a cost output
has to be restrained slightly under normal
conditions)
93Transmission
94 Transmission basics
- Transmitted power Voltage (V) x Current (I)
- Resistance loss varies with I2, so,
- For a given energy flow, the resistance loss
varies with 1/V2 - Thus, the key to minimizing resistance losses is
to transmit electricity at high voltage - There were be some offsetting losses in the
transformers from low to high and back to low
voltage - Typical voltages for long distance transmission
- 500-800 kV, compared to 30 kV for local
distribution
95HVDC (high voltage DC)
- Less expensive transmission lines with smaller
resistance losses, but more expensive
transformers with greater losses - Thus, HVDC costs less and entails less overall
loss only for transmission beyond some minimum
distance, namely, - HVDC costs less beyond about 750 km distance, and
entails less loss beyond about 250 km distance
(the exact break-even distance for cost depends
on the terrain and local market conditions)
96Other pros and cons of HVDC compared to HVAC
- HVDC
- - has a much narrower right of way
- - generates negligible magnetic fields
(concern over which has been one source of public
opposition to new transmission lines) - - has better reactive power control and full
control of where the power flows (unlike AC mesh
grids) - - an offshore grid for offshore wind farms would
permit the wind turbines to operate at a greater
range of speeds, which in turn would permit more
efficient operation (no need for synchronization
of the power output with the land AC grid) - However,
- - branching of DC lines is difficult, as is the
construction of multiple terminals, although
these problem should be solvable in a few years
97Figure 3.32 Typical DC and AC Transmission Pylons
Source GAC (2006, Trans-Mediterranean
Interconnection for Concentrating Solar Power,
Final Report, GAC, www.dlr.de/tt/trans-csp)
98Figure 3.33 Transmission corridors transmitting
10 GW of electric power
Source GAC (2006, Trans-Mediterranean
Interconnection for Concentrating Solar Power,
Final Report, GAC, www.dlr.de/tt/trans-csp)
99Economics
100Direct Cost of Wind EnergyThe cost per kWh is
the annual revenue requirement per kW of capacity
divided by the number of kWhs sold per year per
kW of capacity. That is,C (CRFOM)CCwt/(?sfa
8760CF) where CRFcost recovery factor
i/(1-(1i)(-N)) i
interest rate (expressed as a fraction per
year) N number of years over which the wind
project is financed OM annual
operation and maintenance cost as a fraction of
the initial capital cost CCwt, CCwt
initial capital cost given as /kW ( per kW of
turbine capacity)
101 8760 number of hours in a year ?s is an
efficiency that takes into account various losses
that are not accounted for in the turbine power
curve (such as dirt on the blades, imperfect
tracking of the wind direction by the yaw
mechanism, or wake effects in wind farms) fa
is the fraction of time that the turbine is
available CF capacity factor (the average
power output as a fraction of the peak output or
capacity)a 1kW turbine running full out all the
time would produce 1kW x 8760 hr/yr 8760 kWh/yr
of electricity
102Units in the previous equation(yr)-1 (for OM
and CRF) x /kW, divided by kWh/kW/yr
gives/kWh
103Figure 3.34 Total cost (including installation
and grid connection) of wind turbines in various
countries in 2006
Source Krohn et al (2009, The Economics of Wind
Energy, A Report by the European Wind Energy
Association, European Wind Energy Association,
Brussels, www.ewea.org )
104Figure 3.35 Trend in turbine and non-turbine
costs and in the cost of wind-generated
electricity in Denmark
Source Krohn et al (2009, The Economics of Wind
Energy, A Report by the European Wind Energy
Association, European Wind Energy Association,
Brussels, www.ewea.org )
105Figure 3.36 Illustrative costs of wind
electricity for various rates of return (ROI) in
the investment and for various capital costs,
assuming a CF of 0.35, 20-year financing and?s
fa 1.0
106Figure 3.37a Costs of offshore wind farms as a
function of the size of the turbines used in the
wind farm
Source Redrafted from Snyder and Kaiser (2009,
Renewable Energy 34, 1567-1578,
http//www.sciencedirect.com/science/journal/09601
481)
107Figure 3.37b Costs of offshore wind farms as a
function of the size of the wind farm
Source Redrafted from Snyder and Kaiser (2009,
Renewable Energy 34, 1567-1578,
http//www.sciencedirect.com/science/journal/09601
481)
108In spite of the lack of any relationship between
the cost of offshore wind farms and the size of
the wind farm, the purchase price of turbines has
been reduced by up to 45 from the list price for
orders of 500-1600 turbines
109Progress Ratio
- This is the fraction by which the cost of
something is multiplied for every doubling in
cumulative production. - The cost of a wide range of manufactured products
follows this relationship - For wind turbines, the observed progress ratio up
to 2005 was about 0.8, meaning a 20 reduction in
cost for every doubling in cumulative global
production
110Figure 3.43 Reduction in cost with a progress
ratio of 0.81and growth in capacity decreasing
linearly from20/yr in 2008 to 0/yr in 2050
111However, prices of onshore wind turbines in
Europe increased by 75 between 2005-2008, while
that of offshore turbines increased by almost
50. There are two reasons for this
- Demand greater than supply (perhaps growth has
been too rapid global installed capacity had
been growing by 26/yr from 2000-2007) - Spikes in the costs of steel and copper
(affecting fossil fuel and nuclear power plant
costs too) - Prices have subsequently fallen a little
112Indirect costs of wind turbines
- Reduced electricity output by non-wind generators
(which increases the unit cost of their
electricity), partly offset by a reduction in the
need for non-wind generators - Wasted wind electricity generation potential
113It is often thought that the addition of wind
energy does not allow any reduction in the amount
(capacity) of other power sources, because there
could be zero wind production near times of peak
demand. That is, the capacity credit of wind is
often assumed to be zero. However, this is not
correct.Instead, the amount of non-wind
capacity that is needed is calculated so as to
have the same loss-of-load probability as when
there is no wind capacity with, instead, the full
non-wind capacity
114The result is that, for small wind penetration
(that is, small wind capacity compared to the
total capacity), the capacity credit is roughly
equal to the capacity factor. So, if 100 MW of
wind power capacity is added to a very large
system and the CF (average output as a fraction
of peak output) of the wind turbines is 20, then
the non-wind capacity can be reduced by 20 MW
while still having the same overall reliability.
115As the wind penetration increases,the capacity
credit as a fraction of the capacity factor
becomes progressively smallerSo, at 1000 MW and
the same 20 capacity factor, the capacity credit
might be only 10 instead of 20, so the non-wind
capacity can be reduced by only 100 MW The
capacity credit for wind is non-zero only because
the backup fossil fuel powerplants are themselves
not 100 reliable, as seen in the next table
116Table 3.16 Outage rates for various electricity
generators in the US.
117Figure 3.38 Capacity credit for wind as a
function of the wind penetration and capacity
factor
118Figure 3.39 Capacity credit for wind as a
function of wind penetration and the degree of
geographical dispersion of the wind turbines
119To recap,
- The addition of wind means that the existing
fossil fuel powerplant is used less, which
increases the unit cost of that portion of the
electricity from the fossil fuel plant - However, less fossil fuel powerplant is needed
(which is not helpful if the fossil capacity has
already been built) due to the non-zero capacity
credit from wind - Other indirect costs of wind include
- - wasted wind electricity potential due to the
need to maintain a minimum fossil fuel output - - reduction in the efficiency of the fossil fuel
powerplant when wind is added (either because it
is operating at lower average load and thus less
efficiency, or because of larger swings in output)
120Figure 3.40 Wasted wind energy potential as a
function of wind energy penetration for Danish
conditions
Source Redlinger et al (2002, Wind Energy in the
21st Century Economics, Policy, Technology and
the Changing Electricity Industry, Palgrave,
Basingstoke)
121Cost of transmission
- Direct related to the investment cost for
transmission equipment and annual operation and
maintenance costs - Indirect related to the loss of electricity
during transmission. This has to be made up by
generating more electricity than if there were no
transmission losses. The indirect cost is equal
to the required extra electricity generation x
the cost of the electricity, which includes the
powerplant and the direct transmission costs
122- If Tloss is the fractional electricity loss when
the transmission line is transmitting at its full
capacity, and CFtr is the transmission line
capacity factor (average power transmitted over
peak power transmission capacity), then the
average loss is CFtrTloss - The amount of electricity that needs to be
generated in order to deliver X kWhs at the end
of the transmission line is thus - X / (1- CFtrTloss)
- Similarly, if C is the cost before transmission
losses (but including the cost of the
transmission equipment), then the cost after
accounting for transmission losses becomes - C / (1- CFtrTloss) C / (1 e), where e
CFtrTloss
123- Cost C/(1 e)
- C(1 e e2 e3 .)
- (this is a Taylor Series expansion)
- C eC (1 e e2 .)
- C eC/(1 e)
- (using (1 e e2 e3 ) 1/(1-e))
- Thus, the extra cost due to transmission losses
is - eC/(1 e)
- where, remember, C is the cost of electricity
before accounting for transmission losses, and
includes the transmission line capital cost.
124Figure 3.41a Transmission energy loss using GAC
(2006) data
125Figure 3.41b Absolute transmission investment cost
126Figure 3.42 Cost of transmission assuming a line
capital cost of 560/kW and 8.6 loss at full
capacity
127Strategies for Baseload Wind Energy
- Oversized wind farms compared to the transmission
link can give capacity factors at the receiving
end of the link of 0.6-0.7 - Compressed air energy storage
- Use of dispatchable loads (such as reverse
osmosis for desalination or heat pumps in
district heating and cooling systems)
128Figure 3.44 Oversizing Concept
129Figure 3.45a Wind farm capacity factor as a
function of mean wind speed for various degrees
of over-sizing
1
2
3
4
130Figure 3.45b Wasted wind energy potential as a
function of mean wind speed for various degrees
of wind farm over-sizing
4
3
2
131Figure 3.45c Transmission line capacity factor as
a function of mean wind speed for various degrees
of wind farm over-sizing
4
3
2
1
132Figure 3.46 Cost of electricity from 2000-km
distant wind farms oversized by various factors
4
3
2
1
133Figure 3.47 Contribution to the cost of
electricity
Capacity Factor 0.57 0.72
0.77 0.81
134Figure 3.48 Comparison of electricity costs from
local and distant oversized wind farms vs wind
speed
135Capital Cost Estimates
- Natural gas combined cycle 660/kW
- Wind natural gas hybrid 1640/kW
- Wind natural gas CAES 2270/kW
- (the last is cheaper than many recent nuclear
power plants)
136Figure 3.49 Cost with electricity from natural
gascombined cycle and from wind with natural gas
or CAES
Source Greenblatt et al (2007, Energy Policy 35,
14741492, http//www.sciencedirect.com/science/jo
urnal/03014215)
137Desalination of seawater
- The most efficient method is reverse osmosis,
using electricity (the energy requirement is 2-3
kWh/m3) - Desalination is a load that could come on when
there is excess wind, thereby allowing a bigger
wind plant in order to meet a larger fraction of
the non-desalination loads, without wasting
energy - However, the desalination equipment capacity
factor would be small, increasing the capital
cost contribution to the cost of desalinated
water - But since the wind that is used would otherwise
be wasted, it would have to be sold at a deep
discount
138Figure 3.50a Cost of desalinated water using
wind-generated electricity and reverse osmosis
139Because most of the cost of desalinated water is
from the capital cost rather than the energy
cost, even making the electricity free would not
offset the impact of a low utilization of the
desalination equipment (low desalination capacity
factor)Instead, it is better to have dedicated,
even over-sized, windfarms to power the
desalination equipment
140Figure 3.50b Cost of desalinated water
usingelectricity from oversized windfarms
141Energy Payback Time
- This is the time required for the amount of
primary energy saved by the wind turbine to
offset the total primary required to produce the
turbine - Saved primary energy per year electrical energy
produced per year divided by the efficiency of
the powerplant that would otherwise be used to
produce electricity
142- Generally speaking, calculated payback times for
wind turbines are 2-8 months - The payback time would be significantly longer if
the components need to be transported 1000 km or
more by truck
143Noise and impacts on birds and bats
- Bird mortality is miniscule compared to many
other human causes of bird deaths (see table to
follow) - Noise level at a distance of 350 m is less than
the typical background level in a home - Impacts on bats need further study
144Table 3.24 Main human-related causes of bird
deaths in the US.
Source GWEC (2006, Global Wind Energy Outlook
2006, www.gwec.net)
145Wind energy potential and global scenario
146Figure 3.51 Wind energy potential in various US
states. The total potential from the 10 states
shown here is more than twice the total US
electricity demand of 4200 billion kWh in 2004.
147Table 3.26 Land areas, percent of area with Class
3 or better winds (6.49 m/s) at a height of 80
m, and annual electricity production using 1-MW
and 5-MW wind turbines compared with current
demand (in 1000s TWh/yr).
Based on Archer and Jacobson (2005, Journal of
Geophysical Research 108, D9, 4289)
148Figure 3.52a Scenario whereby the global wind
powerplant grows to a capacity of 12500 GW
following a logistic growth curve with a growth
parameter of 0.2 (20/yr initial growth)
149Figure 3.52b Scenario whereby the global wind
powerplant grows to a capacity of 12500 GW
following a logistic growth curve with a growth
parameter of 0.2 (20/yr initial growth)
15012500 GW 12.5 TWAssume an average capacity
factor of 0.312.5 TW x 8760 hr/yr x 0.3 32350
TWh/yrCurrent world electricity demand is
18500 TWh/yr