Supplementary memorandum by E.ON UK
REQUIREMENT FOR THERMAL GENERATION TO BACK-UP
1. Following the recent oral evidence session
this supplementary note provides further explanation for the statement
that a high % of wind capacity needs to be "backed up"
by thermal plant to meet winter peak demand.
Variability in Wind Generation
2. Wind generation and thus its contribution
to meeting electricity demand varies with wind speed. A typical
single 3MW wind turbine generates no electricity output when wind
speed is less than about 3 metres per second (m/s), reaches its
maximum output at about 15 m/s, and shuts down when wind speed
reaches around 25 m/s to preserve its physical integrity.
3. A single wind farm with a number of turbines
will smooth or average this effect as wind speed will vary somewhat
across the area of the wind farm. For example at an average wind
speed of about 3 m/s there will be some output because the wind
will be above that speed at some locations and below it at others.
On the same basis, it is less likely that the wind speed will
be high enough across the entire area to deliver maximum output
from all wind turbines. For a small part of the year (less than
1%) an individual wind farm is close to full output, but for much
of the year it generates far less and for 15-20% of the year it
generates no output.
4. The effect of operating a portfolio of
wind farms is to smooth this effect further, with less variation
in output across the year. The portfolio never reaches full output
and peak output is around 80% of full capacity, but there is only
a very small portion of the time when there is virtually no output.
Correlation between wind output, season and time
of day, and with electricity demand
5. The extent to which wind speed, and thus
output from wind generation, correlates with periods of high electricity
demand is important in assessing the extent to which we can rely
on wind generation to meet winter peak electricity demand. Winter
is generally windier than the summer, with the median output for
a winter day higher than in the summer. However, on the coldest
days (with temperatures below zero), there tends to be little
to no wind, corresponding to winter anti-cyclones. There is an
increased risk of very low wind speeds, with wind generation output
less than 10% of theoretical maximum, on high demand days.
6. It is also important to recognise that
the output of windfarms are correlated with each other, so that
if a particular windfarm is suffering a lack of wind it is very
likely that those nearby are too, and even the most distant windfarms
are less likely to be generating. This is a result of weather
systems such as windless anticyclones being large enough to affect
all of the UK.
7. The precise correlation between UK wind
generation and wind speed is complex and needs further analysis
but, overall, we conclude that the relationship between the level
of UK wind power output and UK electricity demand is very weak
and, at best, the availability of wind generation is no better
during high demand periods than in periods of lower demand.
Assessment of the extent to which we can rely
on wind to generate to meet winter peak demand
8. From the point of view of the system
operator (National Grid) who must schedule sufficient capacity
in order to meet winter peak demand with a very high degree of
probability, an assessment must be made of how likely it is that
the capacity available will in fact operate at the time required.
For this purpose, planned outages and weekend maintenance can
be ignored as they are highly predictable and scheduled for low
9. Excluding these factors, the weekday
availability (actual availability/maximum total availability)
of thermal (ie burning coal, gas, oil or biomass) plant over the
winter period is about 95% with breakdowns accounting for 5% of
maximum total availability. Crucially, breakdowns are generally
not correlated with each other. There are exceptions and these
have to be taken into account (such as loss of gas supply at a
number of stations, or type faults), but they are generally a
small effect. So, if a unit at Station X is unavailable, there
is no reason to suppose that another unit at a different station
is going to be unavailable. This means that conventional units
are very effective at backing each other up, especially when the
portfolio contains a mixture of plant types and fuels.
10. To assess the extent to which investment
in wind capacity will be able to replace thermal plant on the
system while ensuring that peak demand can be met at the same
level of reliability, we need to assess how much wind capacity
on the system can be relied on to meet peak demand at a dependability
of 95%. Our assessment of winter wind generation data in 2007
indicates that the system operator could rely on 8% of total UK
wind capacity to meet winter peak demand at the same level of
dependability as thermal plant. On this basis, if the UK required,
say, 40,000MW of wind capacity to meet its renewable target by
2020, only 8% of this renewable capacity (3,600MW) could be relied
on to meet winter peak demand. This would avoid the need to build
3,600MW of new thermal plant but the remaining 36,400MW of renewable
capacity would need to be "backed-up" by thermal plant
to meet winter peak electricity demand in 2020. This effect could
be to some extent mitigated by more extensive electricity interconnections
with continental Europe (which would enable "back-up"
power to be imported), the longer term development of new electricity
storage technologies at a significant scale (which would be able
to store power from the grid and produce it when required), or
more demand side management capability which would enable demand
to be varied in relation to the level of wind generation.
11. This assessment is consistent with other
studies carried out in the UK and Germany,
where there is extensive experience of operating grids with large
volumes of onshore wind. However, for the UK, the calculation
should be refined in the light of a more detailed assessment of
the actual wind portfolio likely in 2020, further analysis of
the correlation between wind speed and demand and an analysis
using data over a longer period, but we believe the broad implications
for future generation requirements will remain valid. We recommend
that the issue is explored in more detail as part of the Government's
forthcoming consultation on delivery of the UK's share of its
5 Based on E.ON and E.ON contracted windfarms but assumed
to be representative of all of UK. Back
DENA Grid Study: "Planning of the Grid Integration of Wind
Energy in Germany Onshore and Offshore up to the Year 2020"
Cologne, February 2005 Back