Wind turbines as yet unsuitable as electricity providers
Wind turbines as yet unsuitable as electricity providers
Wind turbines have been widely accepted as electricity producers thanks to claims that wind comes free of charge, and each kWh thus produced replaces a kWh provided by conventional techniques, i.e., it saves the fossil fuel otherwise needed. However, these assumptions have never been validated in existing power distribution systems even after the installation of as much as 86 GW of wind power in Europe alone.
Europe; 21st century
Tavailable if the wind blows strongly enough.
he wind energy flux varies with the third
power of the wind speed.Wind energy is only
This is expressed by saying that wind energy
is supply driven: it is not available on demand.
This feature makes wind energy as such unsuitable as a
power source for an electricity grid where supply and
demand must be balanced at all times.In order to stabilise the
power of the grid we need either storage or conventional
generators as backup.This has important consequences,as
it reduces the fuel savings promised by the wind industry,
while it doubles the costs of the power systems.
An often-heard objection to this argument is that by
spreading wind turbines over a large area the fluctuations
in wind fields average out. This argument can be
quantified by looking at the correlation length in wind patterns.
Rather than discussing these wind patterns we discuss the
production of 7000 wind turbines in Germany.
In 2004 the electricity company E.ON had some 7 GW of
wind power installed over the whole territory of Germany.
In their ‘Wind report 2005’ [
] they published the wind
electricity produced in 2004 on a day-by-day basis (Fig.1).
The conclusion is that on a 24 hour basis the correlation
in wind energy production is strong over all of Germany.
This effect has been analysed for the case of Germany by
the Aachen University [
]. The result is that the possible
substitution of conventional sources by wind is only a
few per cent of the installed wind capacity. This result
is in contradiction to the statement of Eecen et al. .
Other studies have shown a strong correlation between the
wind energy production in Denmark, Germany and Great
Britain together.The North Sea is often hailed as the future
powerhouse of Europe, but its opening hours are quite irregular…
This implies that wind power is not an alternative for
conventional production capacity.The investment in wind turbines
requires extra capital on top of the cost of a conventional
system, which by itself is already capable to satisfy the total
demand. Moreover this extra capacity also requires extra
connection lines.In Germany the plans to install more wind
turbines require 4000 km of extra high voltage transmission
lines, which otherwise would not be necessary.
The capacity factor and curtailment
The frequency distribution of the wind velocity is
described by the Weibull distribution, which reveals that
wind speeds from zero to the average wind force are as
frequent as all other wind speeds together. An average
wind speed is typically 6 m/s onshore. The average wind
force is larger at sea than onshore.
Windmills start producing electricity at about 5m/sec and
reach full power at 13 m/sec. The capacity factor of a wind
turbine is the fraction of the time it effectively runs at full
capacity.The capacity factors offshore are about 0,4 and about
0,25 in good locations onshore.The measured factors are 0,22
for the Netherlands, 0,17 for Germany and 0,24 for Ireland.
b FiG 1: daily
This means that a wind energy system which has enough capacity to satisfy the demand, covers less than 25% of the total demand during a year.
Is it possible to absorb such an amount of wind energy
in a conventional system?
Here we follow the calculations done by B. Ummels [
The electricity demand follows a curve which is largely
the same everywhere in Europe. Figure 2 represents such
a curve for the weekly demand in the Netherlands in 2007.
Such curves are in use by the producers to balance supply and
demand.Taking a full year of such curves,we split the total
demand in 15-minute periods to arrive at 365 × 24 × 4 = 35040
periods of demand.The computer now sorts the 15-minute
intervals with their demand in decreasing order and puts
them on theY-axis of Fig.3.These points form the top curve
in the graph.It shows the so-called demand-duration curve
for the Netherlands. One can see that the maximum
electricity demand was around 20 GW, and the minimum just
below 10 GW. Next one can obtain the data on wind speed
per 15-minute interval from the Meteorological Institute.
These data allow calculating the actual wind electricity yields
for any desirable total installed capacity.This wind-generated
electricity yield (for each relevant 15 minute time slot during
the year) can be subtracted from the demand,so the
production of wind energy is now represented as a negative demand.
] has done this for installed capacities of 2 to 12
GW in 2 GW steps (see Fig. 3).
The‘must run’capacity is defined as the minimum output
of a conventional system under the requirement that it
can deliver full output on short notice, i.e., during a period
of rising demand and falling wind production. The‘must
run’ output is represented as a horizontal line in figure 3.
The height of this line is a subject of debate, but the choice
of 10 GW is the result of the following arguments:
A. The difficulty of running large installations below
50% of full load.
B. Industrial electricity production in combined heat and
power installations is beyond control of the electricity
distribution network. The same holds for Combined Heat
and Power (CHP) installations used for district heating.
FeatUres Wind tUrbines
C. Cycling of coal-fuelled generators can lead to unex- The saving of fossil fuel
pected consequences, as detailed in the Bentek report Many studies worldwide indicate that the fuel saved by
about the Colorado electricity supply [
] wind is far less than proponents assume. [
]. Most of
those studies are based on models by lack of operational
The surface under the horizontal line represents the data with two exceptions: The Bentek report on wind in
amount of wind energy which cannot be accommodated in Colorado and Texas [
] and Udo’s analysis of the
operathe grid.With 12 GW wind power installed one could theo- tional data of EirGrid [
retically provide 20% of the total demand. However, the But let us first review the factors affecting the actual
perdemand-duration curves demonstrate that about half the formance of the electricity-generating system when wind
time there is surplus wind power.The total surplus amounts turbines are added.
to 40% of the supply by wind.Not only is a solid 40% of the The widely published argument in favour of wind energy
wind production lost,but also the conventional system has is: “The conventional units produce less energy, so they
to run at minimum capacity during exactly half the time. consume less fuel”. Several factors influence this simple
Export is still touted as the solution for wind energy over- result in a negative sense.
supply. This argument is not valid because, when we have 1. Conventional units are forced to operate at less than
high winds in one country in Western Europe, it also blows optimal power due to supply of wind electricity. As
in the surrounding countries. Such is the nature of our a consequence, they have a lower thermal efficiency.
weather systems. They consume more fuel per kWh compared to the
Let us turn again to the great example for the wind pro- consumption without wind.
ponents, i.e., the Danish situation. The Danes produce 2. Cycling conventional units – ramping up or down
20% of their electricity by wind, but half of that cannot – in response to fluctuating wind contribution uses
be accommodated locally. This oversupply is exported to more fuel than running at constant power.
Norway and Sweden, where it is used to replace hydro 3. Idling of conventional generators synchronous with
electricity. Since 2010 the Scandinavian Power market the grid but not delivering electric power costs 6 – 8%
rewards electricity oversupply with a negative price of of the fuel required for running at design capacity.
up to € 200 per MWh, or € 0,2 per kWh. This is necessary to step in, at sudden lulls of the wind
CEPOS, an independent Danish scientific institute, over wind turbine arrays.
has issued a report entitled: “Wind energy, the case of 4. Extra cold starts of conventional units easily require
]. The authors are quite explicitly negative more fuel than wind developments produce. For
exabout the practice of supplying free (subsidized) wind ample: one cold start of a highly efficient 360 MW
electricity to their Scandinavian neighbours at times of CCGT (Combined Cycle Gas Turbine) consumes the
oversupply, and the purchase of electricity at high cost fuel equivalent to the electric energy which twenty 3
during periods when the wind cannot deliver. MW wind turbines produce during 40 hours running
We quote from page 29: “The very fact that the wind on average output.
power system, that has been imposed so expensive- 5. Construction and installation of wind turbines
rely upon the consumers, cannot and does not achieve quires a quantity of energy equivalent to about 10%
the simple objectives for which it was built, should of their life-time production [
be warning the energy establishment, at all levels, of 6. Capital costs and energy involved in connecting
the considerable gap between aspiration and reality.” wind turbine parks to the grid, including adaption
Negative electricity prices are the only way to stop the tur- and transforming, have to be added to the wind park
. FiG 2: typical bines, but this phenomenon endangers the exploitation costs and subtracted from yields [
totbaelhealevcitoruicriotyf of the conventional power generation system, unless the 7. Self-consumption of electric energy by wind turbines
demand over a government (i.e., the citizens) pays additional subsidies and their electronics while idling and at rest, i.e.,
heatperiod of a week to keep the backup systems running. ing during cold periods and electricity consumed by
the power electronics for grid adaption .
8. Increased use of inefficient, fast ramping OCGTs
(Open Cycle GAS Turbines) instead of twice as
efficient CCGTs to meet sudden wind variations.
9. Energy costs related to extra wear and tear of
conventional units due to frequent forced ramping in
response to the wind variations.
The Bentek study used the actual exhaust data of CO2
and other gases before and after the introduction of
wind developments. The authors corrected for import
and export, change in demand and other relevant factors.
The conclusion is: wind does not save fuel and does not
reduce emissions in Colorado and Texas.
Udo’s review of the actual Irish data revealed that a 12%
wind contribution in the Irish grid system without
largescale storage resulted in a fuel saving of 5%. In the Bentek
research the factors 5, 6 & 9 are not taken into account.
The inclusion of 7 is dubious. In the EirGrid data, the
basis of Udo’s analysis, the factors 4, 5, 6 & 9 were not
taken into account. The inclusion of 3 and 7 is uncertain.
The cost of wind energy
In the article of Eecen et al.  two claims are made about
the costs of wind energy:
A. “The cost of onshore wind energy reaches the cost of
Our comments are the following:
1. The product of a windmill is supply driven, so it does
not have the same value as conventional electricity.
Formerly this was partly expressed by the imposition
of imbalance costs on the producers of wind energy.
These costs are now paid by the consumers. Experts
estimate the true value of wind energy between 1 and
3 cents per kWh. [
2. The interest rates have been exceptionally low in the
last years, while the dominant part of the cost of wind
energy is capital. This partly explains the recent
decrease in cost of onshore wind energy.
3. The environmental costs are omitted in the price of
onshore wind. Land occupation: The distance between
turbines of a wind farm must be large due to the wake
effects described in ref..The result is that the
capacity density cannot be more than 9 MW/km2, so 9 GW
of wind capacity will occupy around 1000 km2 of land
area. This territory will be unfit for human
occupation because of continuous noise and flicker effects.
Lowering of estate values. Surveys have shown that
houses within sight of wind parks lose an appreciable
percentage of their value. In highly populated areas
this depreciation is larger than the amount of capital
invested in wind turbines.
B.“…it is expected that the relatively young offshore wind
technology will experience a strong learning curve, leading
to significant cost reductions”
We note: From 2005 to 2010 the cost of installing offshore
wind farms has increased from 2 million euro/MW to 4
million euro/MW . This is a strong learning curve indeed.
Quantifying the decrease in efficiency of the electric power
system and the extra fuel consumption induced by wind
developments is by no means a simple matter. To our
knowledge there are presently not sufficient data in the
public domain to substantiate a definite answer to the question
how much fuel and CO2 emission is saved. It depends on the m FiG 3:
actors, the equipment, the kind of fuel, the amount of wind cduermveasnidn-dthueration
penetration,the behaviour of the regional wind,the amount absence of wind
of storage, the interconnection of regional grids etc. power (upper curve)
and with the wind
Decisions to install large-scale wind-powered electricity power contributions
generation are based more on the expectation to save indicated
significant amounts of fossil fuel and CO2 emission than
on any evidence that this is indeed the case.
Wind technology is not suited for large-scale application
without a good buffer and storage system.We propose to
stop spending public money on large-scale use of wind.
This money should be spent on R&D of future power
systems.We expect that wind will not play an important
role in these future systems. n
 The British Consumer Society,WHICH, discovered that small wind
turbines for individual houses use more electricity than they
produce, due to self-consumption of their power electronics.
 KEMA-ECF “Power perspectives 2030”, full report page 32.
 E. ON: Wind Report 2005 , page 7.
 Ibid. page 9.
 B. Ummels , Thesis, Delft University of Technology, 2009 .
 “How less became more”:Wind, Power and unintended consequences in the Colorado Energy Market, 2010 (www .bentekenergy.com).
 CEPOS 2009 : www .cepos.dk/english/analyses-and-notes/
 Kent Hawkins , MasterResource Nov . 2009 “ Wind integration” . www.masterresource.org/ 2009 /11/wind -integration-incremental-emissions-from-back-up-generation-cycling-part-i-aframework-and-calculator/
 W. Post , The Energy collective: www .theenergycollective.com/willempost/64492/wind-energy -reduces-co2-emissions-few-percent?
 W. Boone MasterResource Sept 2010 . www.masterresource. org/ 2010 /09/windpower-overblown-part-1/
 F. Udo : www .clepair.net/Udo-okt-e.html .
 Wind proponents usually claim it is 3 to 6 months. C. le Pair & K. de Groot: www.clepair.net/windefficiency.html , note 13 .
 C. le Pair : www .clepair.net/windSchiphol.html , 4b & 4c .
 WalterConsultancy Ir.R.N.Walter - Oosterhout , September 2011 .