ArticlePDF Available

Estimating the Impact of Wind Generation in the UK

Authors:

Abstract and Figures

This paper studies the impact of wind generation on market prices and system costs in the UK between 2013 and 2014. The wider effects and implications of wind generation is of direct relevance and importance to policy makers, as well as the system operator and market traders. We compare electricity generation from Coal, Gas and wind, on both the wholesale and imbalance market. We calculate the system cost of wind generation (government subsidies and curtailment costs) and the total energy costs. For the first time in the UK, we calculate the Merit Order Effect on spot price due to the wind component and show a 1.32\% price decrease for every percentage point of wind generation (compared to the "zero-wind" price). The net result of total costs and price savings is roughly zero (slight positive gain).
Content may be subject to copyright.
Estimating the Impact of Wind Generation in the UK
Lisa M.H. Hall, Alastair R. Buckley, Jose Mawyin
Department of Physics and Astronomy, Hounsfield Road, Sheeld S3 7RH, United Kingdom
Abstract
This paper studies the impact of wind generation on market prices and system costs in the UK between 2013 and
2014. The wider eects and implications of wind generation is of direct relevance and importance to policy makers,
as well as the grid operator and market traders. We compare electricity generation from Coal, Gas and wind, on both
the wholesale and imbalance market. We calculate the system cost of wind generation (government subsidies and
curtailment costs) and the total energy costs. For the first time in the UK, we calculate the Merit Order Eect on
spot price due to the wind component and show a 1.67% price decrease for every percentage point of wind generation
(compared to the “zero-wind” price). The net result of total costs and price savings is roughly zero (slight positive
gain). We also consider the eect of not having either an onshore or an oshore wind component. We show that the
Merit-Order Eect savings are heavily reduced, leading to an outgoing cost of wind generation in both cases. It is
therefore important to have a significant total percentage of wind generation, from both onshore and oshore farms.
Keywords: , Energy systems, Energy policy
1. Introduction
Investment in energy systems tends to be as much
political as it is technological or scientific. Despite
calls from the international community to retain invest-
ment mechanisms for onshore wind in the UK (because
on-shore wind is by far the cheapest renewable gen-
eration), in May 2015, the newly-elected government
quickly removed support, citing over-expensive subsidy
costs. Additionally, in January 2015, several UK broad-
sheets led with stories claiming that wind farms are
over-charging the market to curtail generation during
low-demand periods. It is notable that both subsidies
and curtailment costs are typically not discussed in com-
parison with fossil fuel electricity costs, nor in terms of
the financial benefits to the market.
Whilst renewable subsidies might seem costly, the ex-
pected outcome of subsidising such technologies in the
market is two-fold: to decarbonise future energy supply
and reduce the price of energy in the long-term. Dis-
cussing the cost of both subsidies and curtailment costs
without defining the context can be misleading and can
lead to the wrong conclusion.
Email address: lisa.clark@sheffield.ac.uk (Lisa M.H.
Hall)
In Germany, it is widely accepted that a supply of
electricity from renewable generators (predominately
wind and solar PV) reduces the spot price and leads to
a considerable cost saving. Several studies have calcu-
lated this Merit-Order Eect (MOE), though analyses
vary. For example, Senfuß et al show that the MOE in
2006 led to a price reduction of about 8 e/MWh and to-
tal costs saving of about e5 billion in 2006[1]. Whilst
they cite enormous support payments around e5.6 bil-
lion, they report a net profit due to the value of renew-
able electricity and the MOE cost saving.
Similarly, Weigt calculated savings in Germany of
e1.3 billion in 2006, e1.5 billion in 2007, and e1.3
billion in the first half of 2008[2]. Ketterer shows a
1.46% spot price drop as wind increases by 1 percentage
point[3], whilst Cludius et al show a spot market price
reduction of 6 e/MWh in 2010 rising to 10 e/MWh
in 2012[4] due to a combined electricity generation by
wind and PV. Similar results can be shown for the Span-
ish market[5] and Irish market[6], though the eect is
heavily reduced in Ireland.
For the UK market, comparable analyses have not
been undertaken. Indeed, the MOE has not been quan-
tified, nor has the comparative cost of wind curtailment.
The reason for this may lie in the fact that wind gener-
ation has only been a significant factor for a few years
Preprint submitted to Energy Policy November 13, 2015
Figure 1: The Merit Order Eect (MOE). An increased, cheap supply
of renewable energy pushes the curve to the right (i.e. the blue curve
becomes the red curve) and results in a decreased marginal cost of
energy per unit.
and until now, such analysis would be meaningless. Ad-
ditionally, analysis of the electricity market in the UK is
impeded by several issues: access to market data, us-
ability of accessible data and the complexity of the mar-
ket structure.
In this paper, for the first time, we calculate the value
of the Merit Order Eect in the UK and compare it
to subsidy costs. Furthermore, we present the cost of
curtailment of both renewable and conventional gener-
ation (coal and gas) in the UK power markets. Where
possible, we dierentiate between onshore and oshore
wind capacity, as well as a geographic split (England
and Scotland, according to operating capacity during the
relevant timescales).
It is interesting to note that other researchers are now
considering the implications of curtailing versus storing
renewable energy[7], but since we do not consider in-
stantaneous curtailment costs, we do not comment on
the wider consequences for the UK market.
2. UK Spot Market and Balancing Mechanism
Since electricity cannot be stored in large amounts,
supply and demand must be balanced at all times. In
the UK, this is undertaken by a system of trading on
the wholesale market by generators, suppliers and cus-
tomers. Electricity can also be imported or exported
through international interconnectors. Trades are com-
pleted for each half-hour period, called Settlement Peri-
ods (SPs). All trading must be completed one hour be-
fore the relevant half-hour period (called gate closure),
at which time the grid operator (National Grid Electric-
ity Transmission) must balance the supply and demand
to ensure instantaneous match. Typically, the Balanc-
ing Mechanism (BM) makes up around 2-5% of total
electricity trades.
The value of electricity on the wholesale market can
be quantified using the electricity spot price, which is
calculated as a volume-weighted average of all trades
executed through the UK APX Power exchange. The
spot-price is calculated for each Settlement Period, but
does not contain any information regarding fuel type.
At gate closure, all Balancing Mechanism Units (BMUs
=generators and suppliers) notify the grid operator
of their dispatch profile and also submit bids and of-
fers of prices to deviate from their contracted genera-
tion/supply. An oer is a proposal to increase genera-
tion or reduce demand. A bid is a proposal to reduce
generation or increase demand.
The grid operator follows the Balancing and Settle-
ment Code (BSC) in order to allow the grid to be bal-
anced at the best price available to the market. Essen-
tially, the bids and oers are tallied against the imbal-
ance volume and appropriate oers/bids are accepted.
The System Buy Price (SBP) and System Sell Price
(SSP) are calculated according to the most expensive
500MWh of energy supplied to the imbalance mar-
ket (volume-weighted average price). Elexon provides
a useful guidance document to the imbalance pricing,
which details the exact calculation[8].
Within the balancing mechanism, oers are always
positive (it will cost the system to increase generation),
but bids can be be either positive or negative (generators
may pay to reduce their generation, but may decide to
charge the system for curtailing their generation). With
the increase of renewable electricity generation, nega-
tive bidding has become more prevalent as the genera-
tors attempt to counteract lost subsidies. When the grid
operator is required to curtail wind generation in order
to balance the market, such negatively priced bids result
in a cost to the system and the SSP may also become
negative.
On the other side, an increase in renewable energy
supply leads to a lower system price, due to the Merit
Order Eect. This eect is shown pictorially in Fig-
ure 1. An increased, cheap supply of renewable energy
pushes the supply curve to the right and results in a de-
creased volume-weighted average cost of energy. This
is true for the wholesale market, but the eect in the
balancing market is reduced due to negative bidding. In
order to quantify the net gain or loss, it is imperative
to consider both costs and savings, which has not been
done previously in the UK.
2
Fuel Type
Total Generation, TWh Imbalance Volume as percentage of Total, %
2013 2014 2013 2014
Oers Positive bids Negative bids Oers Positive bids Negative bids
CCGT 216.88 236.07 3.047 -2.004 -0.002 2.461 -1.616 0.000
COAL 276.40 224.36 0.361 -1.311 -0.051 0.550 -1.709 -0.101
O. Wind (Eng.) 24.76 24.76 0.001 0.000 -3.823 0.001 0.000 -0.046
O. Wind (Scot.) 0.61 3.46 0.000 0.000 0.000 0.001 0.000 -0.005
On. Wind (Eng.) 1.83 2.46 0.000 0.000 0.000 0.000 0.000 0.000
On. Wind (Scot.) 17.93 20.51 2.911 -0.960 -0.001 1.988 -0.059 -0.032
All fuels 717.32 690.66 1.182 -1.373 -0.029 1.264 -1.375 -0.046
All Wind 45.13 51.20 1.157 -0.436 -0.027 0.797 -0.598 -0.035
Table 1: Breakdown of the annual electricity generation by fuel type. Energy generation is separated into total generation (left-hand columns) and
the imbalance market (right-hand columns). Wind generation is split into onshore generation (labelled “On.” for brevity) and oshore generation
(labelled “O.”). Wind is also split into Scottish and English locality (labelled “Scot.” and “Eng.”) respectively. It can be seen that coal and gas
are curtailed by around 1-3% with predominantly positive bids (cost to supplier), whereas wind is curtailed around 0.5% with both positive and
negative bids (cost to both supplier and grid operator). Source: BMReports
3. Data and Methodology
We have collated electricity generation data from var-
ious sources. Whilst we detail specific data sources be-
low, a qualitative overview of the methodology will first
be given. Using electricity generation data for each half-
hourly period between August 2012 and May 2015, it
is possible to separate the two markets: the wholesale
market (with spot price) and the imbalance market (with
system buy and sell prices). Using half-hourly data from
the wholesale market, it is possible to correlate the spot
price to the total percentage of wind generation, whilst
the imbalance market is useful to understand the cur-
tailment costs of generation by fuel type (i.e. the cost
of turning ospecific generators). The data sources we
utilise give generation volumes according to each gen-
erating unit across the UK and we associate these units
to their associated fuel type. We can therefore identify
half-hourly generation in both the wholesale and imbal-
ance markets according to fuel type. This allows us to
separate not only wind from fossil-fuel generation, but
we can also distinguish onshore and oshore wind gen-
eration. Due to the nature of the data, complications
arise from matching and joining data sources (by both
settlement period and generating unit), though statistical
analysis of the resulting dataset is straightforward. For
example, one complication lies within the definition of
date and time; some sources use local time, some give
time in GMT and others only give Settlement Period.
We now detail the exact data sources and specific
methodology. Spot price data was obtained by aca-
demic license from APX Group and is also available
per settlement period. Wholesale electricity generation
disaggregated by fuel type is available from two sep-
arate data providers, originating from the sources: (a)
the Elexon Portal publishes data for each settlement
period, bundled in quarter-yearly periods and (b) the
BMReports website publishes the Final Physical Noti-
fication (FPN) data, which can be linked to fuel type
by individual generation plant information. Informa-
tion regarding the Balancing Mechanism was sourced
from the BMReports website, using the system prices
(SYSPRICE data) and detailed system prices (DET-
SYSPRRICE data). It should be noted that the FPN
and Balancing Mechanism data is published online (to
BMReports) shortly after each settlement period and is
never updated. However, the information at time of pub-
lication is not always complete (some transactions occur
post-publication). The complete (and therefore more
correct) data is maintained by Elexon, though they do
not widely publish as much disaggregated data. The
dierence between data is minimal and should not al-
ter any overarching results. However, it is important
to acknowledge that there is a compromise between the
use of (slightly) incomplete data sets and breakdown of
information. For example, the Elexon data does not dis-
tinguish between onshore or oshore wind generation,
so we must rely on the incomplete BMReports data.
The fuel types categories are not standardised be-
tween the datasets, so we limit our analysis to just
CCGT (Combined Cycle Gas Turbine), Coal and Wind,
which are consistent across the sets. We have chosen
CCGT and Coal as they are the predominant fossil fuel
generators. The date availability of data also diered be-
tween the datasets. For most of our analysis, we there-
fore consider timescales for which we have a complete
annual data (two full years: 2013 and 2014), but extend
the timescale when we analyse the Merit Order Eect
(from August 2012 until March 2015).
We have correlated the three datasets and analysed
3
them statistically using the SAS JMP program. When
considering the mean of a dataset, we always calculate
the volume-weighted mean, such that insubstantial vol-
umes do not skew the analysis. Additionally, when dis-
cussing the balancing mechanism data, it is imperative
to split the oers and bids into three (not two) separate
categories: oers, positive bids and negative bids. This
is undertaken to ensure that cancelling eects do not oc-
cur.
It will be necessary later to use the Transmission Loss
Multiplier (TLM). However, the BMReports data asso-
ciated with each oer/bid is not complete with respect
to the these values (if the oer/bid is not part of the
SSP/SBP price calculation, then the TLM is not pub-
lished). Equally, the BMReports publish a separate his-
toric TLM dataset, which does not fully correspond to
the oer/bid TLM data. Furthermore, the TLM value
depends on whether the BMU is a producer or con-
sumer and this data is not readily available (we have ac-
cess to only an incomplete list for all BMUs). Notably,
a producer has a “delivering” TLM (less than unity)
and a consumer has an “o-taking” TLM (greater than
unity). It is therefore necessary to utilise data from three
datasets (BMReports historic TLM, BMReports DET-
SYSPRICE data and Elexon historic TLM) by follow-
ing a given procedure:
If the BMU is a known producer/consumer, then:
If the Elexon TLM data exists for that SP, use
the Elexon TLM data.
If the Elexon TLM data does not exist for that
SP, use the BMReports historic TLM dataset.
If it is not known whether the BMU is a producer
or consumer, utilise the published DETSYSPRICE
TLM data value:
If the DETSYSPRICE TLM value is <1,
then assume the BMU is a producer (and use
the above rules)
If the DETSYSPRICE TLM value is >1,
then assume the BMU is a consumer (and use
the above rules)
This procedure defines all the oers/bids within our
dataset.
For calculation of the wind subsidies we utilise data
published by the UK energy regulator, Ofgem, on the
Renewable Obligation Certificates (ROCs) and Feed-In
Tari(FIT) scheme.
Figure 2: The imbalance market: the cash flow summation of oers
and bids (positive and negative) for 2013 and 2014. Cash flow is tak-
ing from the perspective of the grid operator (positive flow is paid out,
negative flow is cash income). Source: BMReports
4. Results
The total energy generation statistics are shown in Ta-
ble 1, as well as the breakdown into CCGT, Coal and
Wind (onshore and oshore, split into geographic re-
gions), for the years 2013 and 2014. Alongside these
numbers, we have calculated the corresponding imbal-
ance volumes as a percentage of the total generation. It
can be seen that the imbalance volume typically makes
up 2-3% of the total generation volume. The imbalance
volume has been split into oers and bids, which shows
that CCGT and Coal make up the majority of the oers
and positive bids, whilst wind generation accounts for
much of the negative bid volume (curtailment).
Within public dissemination articles (media and pub-
lic engagement), the high curtailment costs of wind gen-
eration has been commonly used as an argument against
wind subsidies and renewable technology. The main ar-
gument is that wind generators often submit very signif-
icant negative bids when compared to the other genera-
tors and are therefore expensive to curtail. We will now
explore this statement.
The breakdown of the cash flow associated with
the oers and bids detailed in Table 1 is shown in
Figure 2. The cash flow for each component (i=
oers,positive bids,negative bids) is calculated as:
Cash flowi=X
j
VPTLM (1)
4
CCGT
COAL
OFFSHORE WIND (England)
OFFSHORE WIND (Scotland)
ONSHORE WIND (England)
ONSHORE WIND (Scotland)
Bid Price (negative), £/MWh
-400
-350
-300
-250
-200
-150
-100
-50
0
Figure 3: The breakdown of negative bid prices for CCGT, COAL
and Wind generation in 2013 and 2014. The horizontal line within
each box represents the median sample value, the ends of the box
represent the 25th and 75th quartiles and the whiskers (extending from
each box) are determined by 1st(3rd) quartile -(+) interquartile range.
Source: BMReports
where jcorresponds to the set of settlement periods, V
is the bid volume, Pis the oer/bid price and T LM is
the associated Transmission Loss Multiplier.
We now consider just the set of negative bids. From
Figure 2 it can be seen that Scottish onshore wind gen-
erators comprise the huge majority of cash flow relating
to negative bids (nearly £90M) over the two year period
(2013-2014). The negative bids cash flow can be further
broken down into the variance of negative bids (shown
in Figure 3) and the dierence between the bid price
(per bid) and the volume-weighted mean bid price (per
settlement period) (shown in Figure 4). The latter statis-
tic is a measure of how close (or far away) each single
negative bid is from the corresponding bids during the
same settlement period. As can be seen (Figure 2), the
mean wind negative bid prices well exceed the lost sub-
sidy of 55 £/MWh (i.e. negative bid price significantly
lower than -55 £/MWh). Additionally, whilst CCGT
and Coal negative bids roughly correlate with the mean
Settlement Period bid price (consistent with a zero dif-
ference), the wind bids are statistically more negative by
a significant level (Figure 4).
At this point, it is also worth considering the level
of subsidies paid to wind generators. There are several
subsidy policies available in the UK, which include the
Renewable Obligation Certificates (ROCs) and Feed-In
Taris (FITs). In this simple analysis, we assume that
the significant majority of the subsidies arise from the
ROCs. The number of ROCs and the final cost (broken
down into obligation periods between 2012 and 2015)
is shown in Table 2. Roughly speaking, the annual wind
subsidies lie around £1.8 billion. We compare this to
the estimated cost of the FIT scheme in 2013-2014 of
CCGT
COAL
OFFSHORE WIND (England)
OFFSHORE WIND (Scotland)
ONSHORE WIND (England)
ONSHORE WIND (Scotland)
Difference, £/MWh
Figure 4: The variance of dierence between the bid price (per bid)
and volume-weighted mean bid price (per settlement period) within
2013 and 2014. The horizontal line within each box represents the
median sample value, the ends of the box represent the 25th and 75th
quartiles and the whiskers (extending from each box) are determined
by 1st(3rd) quartile -(+) interquartile range. Bids from wind gener-
ators tend to be heavily below the volume-weighted mean bid price
within a settlement period. Source: BMReports
Figure 5: The eect of increasing wind generation (as percentage of
energy mix) on System Sell Price (SSP), System Buy Price (SBP) and
Spot Price. Linear fits are quoted in the table above and below the
“knee”, where a fit is assumed to be y=y0+mx and x0is the x-axis
intercept. Data is taken from August 2012 until March 2015. Source:
APX group and BMReports
around £97M (£56M in 2012-2013 and only £6M in
2011-2012).
The next point to consider is the Merit Order Eect,
shown in Figure 5. The eect of increasing the percent-
age of wind generation in the total energy mix is clearly
seen on the SSP, SBP and Spot Price. All prices de-
crease with every percentage of wind energy produced.
5
Number of ROCs (millions) Cost of ROCs (£bn)
2012-13 2013-14 2014-15 2012-13 2013-14 2014-15
Oshore wind 15.69 23.94 25.37 0.639 1.006 1.099
Onshore wind 12.21 18.71 17.73 0.497 0.786 0.768
TOTAL 27.9 42.65 43.1 1.136 1.792 1.866
Table 2: The number of Renewable Obligation Certificates (ROCs) issued per obligation period (April to March) between 2012 and 2015 and the
associated level of costs. Source: Ofgem
x0y0m
SSP 59.34 46.88 -0.79
SBP 64.62 60.10 -0.93
Spot Price 60.06 52.85 -0.88
Table 3: Best linear fit for Merit Order Eect (up to 25%) in Figure 5.
Linear fit assumed to be y=y0+mx and x0is the x-axis intercept. m
has units of £/MWh/%.
0
5
10
15
20
25
Offshore Wind Percentage, %
0 5 10 15 20 25
Onshore Wind Percentage, %
SpotPrice(£/MWh)
<= 10
<= 20
<= 30
<= 40
<= 50
<= 60
> 60
Figure 6: The correlation between onshore and oshore wind gener-
ation (by percentage of total generation) to Spot Price. The colour
relates to the right-hand colorbar and represents the mean spot price
/MWh). The black line around the contour delineates the location
of data points (i.e. all data lies within this outer edge). The diago-
nal black line represents the equality of onshore and oshore wind
percentages. The majority of data points lie slightly above this line,
which shows that oshore wind generation is typically higher than
that for onshore wind. Data is taken from August 2012 until March
2015. Source: APX group and BMReports
There is an observed “kink” in all the trendlines around
a wind percentage of 30%, which is not seen in studies
of the foreign markets. Naively, it would appear that the
price increases again above this percentage. However,
we note that this “kink” is due to a natural skew in the
data: datapoints for which the wind percentage is high
(i.e. above 30%) occur during the early morning (low
SPs) when total demand for electricity (and therefore
price) is also low. If total demand of electricity were
to be plotted against the wind percentage, a kink would
also be seen at the same position. The linear fits are de-
tailed in Table 3, which present both intercepts (x- and
y-axis) and the gradients.
From the graph, it is reasonable to assume that if wind
generation were to be supported in the future, such that
higher percentages were reached, the trend would fol-
low the initial lines above the “kink”. From Table 3,
it can be seen that the price decrease in this range is
around £0.8 - £1.0 per MWh per extra percent of wind
(i.e. min the table). Compared to the zero-wind price,
a decrease of £0.88 per MWh per % (i.e. spot price
gradient) corresponds to a 1.67% price decrease for ev-
ery percentage point of wind generation. This figure
broadly agrees with Ketterer who quotes an electricity
price drop of 1.46% for a percentage point wind in-
crease in Germany[3].
In Figure 6, the correlation between onshore and o-
shore wind generation is compared with the spot price
of energy. The o-centre nature of the contours is due
to the imbalance between oshore and onshore com-
ponents; typically, the UK always produces more elec-
tricity from oshore wind and hence the location of the
contours follow the population of data points. There is
not enough data long the y=0 or x=0 lines to esti-
mate the Merit-Order Eect for only onshore or oshore
wind components. However, it can be seen that the spot
price decreases in roughly concentric circles (centred on
zero), which implies that the spot price decreases with
total wind generation and does not strongly favour ei-
ther an oshore or onshore component. We therefore
conclude that it is the total wind percentage that is the
important factor when reducing energy prices.
It is possible to consider the case in which total elec-
tricity demand remains the same, but wind generation
is minimal. In this way, comparison between the two
cases would present a measure of the cost savings made
on energy prices due to wind generation. It is also pos-
sible to estimate the savings cost due to the presence of
either an onshore or oshore wind component. For this,
we first calculate the actual cost of electricity in 2013
and 2014. Since we do not have the cost breakdown for
each transaction on the APX market, we instead assume
that the price of every unit generated during every set-
tlement period is the spot price. Since the spot price is a
volume-weighted average, this is a reasonable assump-
6
Year Total Actual “No Onshore Wind” “No Oshore Wind” “Low Wind” Cost increase Cost increase
Cost, £bn Cost, £bn Cost, £bn Cost, £bn £bn %
2013 36.69 38.24 38.45 39.04 2.35 6.4
2014 29.40 30.67 31.11 31.36 1.95 6.6
Table 4: Annual cost of energy generation: actual cost and predicted volume-weighted total “no onshore”, “no oshore” and “low wind” costs per
annum. Due to the reduction of energy prices with significant total wind generation (the Merit-Order Eect from both onshore and oshore wind),
the total annual energy bill is heavily reduced by around £2bn per annum. Source: Elexon and APX Group
Figure 7: The monthly cost of energy for actual generation (blue line)
and predicted cost with no onshore wind (red line), no oshore wind
(green line) and low-wind generation (purple line). Source: APX
group and BMReports
tion and the total cost should be the correct value.
In order to calculate the savings for each option
(“low-wind”, “no oshore” and “no onshore”), we need
to simulate the price for each settlement period for each
option according to the level of wind generation. Ide-
ally, we would use the existing dataset to simulate exact
fits, but the availability of data is too sparse for this.
Therefore, we bin the data by total wind generation in
5% categories (0-5%,5-10%,10-15% etc. with no over-
lap) and then calculate the volume-weighted mean spot
price for each settlement period for each bin. This leads
to a dataset of volume-weighted mean spot prices for
each settlement period for each binned category (e.g.
volume-weighted mean 0-5% wind price for each half-
hour period). There are naturally some settlement peri-
ods for which there is not enough data to allow for an
accurate mean value. For these few datapoints (1%
of total data), we assume the actual spot price for that
settlement period, which should always provide a con-
servative estimate.
From this new dataset, it is possible to estimate the
spot price for each settlement period for the three op-
tions: “low-wind”, “no oshore” and “no onshore”. In
order to calculate the “low-wind” cost, we note that
there isn’t enough data to consider a true “zero-wind”
data set. Therefore, we consider the subset for which
wind is less that 5% and utilise this price to calculate
the total cost of energy, as if the wind component didn’t
exist (but assume that the total demand were the same).
For the “no onshore” wind option, for each settlement
period we assume that the total wind generation percent-
age is only produced by the oshore wind component
(once again assuming the total generation is the same)
and use the binned prices calculated above to simulate
the cost of energy in this period. For the “no oshore”
wind option, we similarly assume that the total wind
generation comes from the onshore wind component.
From these costs, the total savings can be simulated for
each option.
The simulated monthly costs for each option are
shown in Figure 7. As fully expected, the monthly cost
of energy without wind, or with only one component, is
higher than the actual cost, due to the Merit Order Ef-
fect. Furthermore, the “low-wind”, “no-onshore wind”
and “no-oshore wind” costs increase during the winter,
when demand is high and actual total wind generation
is high. The total annual savings due to the wind com-
ponents can be calculated as the dierence between the
curves. The statistics are detailed in Table 4.
It can be seen that the cost of providing 717.32 TWh
of energy in 2013 was £36.69 billion, but with “low-
wind” that increased to £39.04 billion. Therefore, the
cost saving due to a wind generation component is es-
timated to be £2.35 billion in 2013 (6.4%). The cost
saving in 2014 is £1.95 billion (6.6%).
Looking at the complete picture, for the 2013-2014
period, we can see that wind subsidies cost around £3.7
billion, wind curtailment cost £86.2M but the total sav-
ing is around £4.3 billion due to the Merit Order Eect.
The net saving is therefore around £514M and (within
the errors of our statistical analysis) can be considered
at best as a positive eect or conservatively as net zero.
Finally, we can consider the implication of remov-
ing a single component of wind (either onshore or o-
7
shore). If there had been no oshore wind, we would
have reduce the ROC costs to £1.554 billion over the
past two years, but the savings would be reduced to
only £840M, resulting in a total outgoing loss of around
£797M. If there had been no onshore wind, the cost of
ROCs would have been £2.105 billion and whilst the
savings are £1.49 billion, the result still is a loss of
around £618M. This implies that wind generation from
both onshore and oshore farms is important to the total
economic gain: if either one component is removed, the
system would result in a net loss of finance.
5. Conclusions
We have analysed electricity generation datasets
within the UK market to study the eect of wind genera-
tion and costs (subsidies and curtailment) to the market.
This is of direct relevance in the UK currently, due to
the decreasing subsidies for onshore wind.
We have analysed the imbalance markets for rela-
tive curtailment costs of wind generation and show that
these generators regularly submit negative bids (costs
to the system) to curtail generation. Whilst all gener-
ators submit negative bids on occasion, we show that
onshore wind generation accounts for the large major-
ity of negative bid cash flow. Additionally, these bids
fall outside of the average bid range for corresponding
bids (when compared within each settlement period, as
a like-for-like comparison). The mean negative bid for
wind generators is significantly higher than the lost sub-
sidy (around £55/MWh).
We also calculate the savings on the market due to
wind generation and the Merit Order Eect. Compared
to the “zero-wind” price, we observe a 1.67% price de-
crease for every percentage point of wind generation,
which agrees with analyses of the German market. (Us-
ing the Elexon data to calculate the same eect, we ob-
serve a 1.67% price decrease for every percentage point
of wind generation). Due to this eect, we estimate the
total cost savings to be around £2 billion per annum,
whilst wind subsidies (due to ROCs) cost only around
£1.8 billion per annum. Due to the estimates in this pa-
per (specifically ignoring other subsidies such as FITs
which account for around £100M per annum in 2014),
we therefore argue that there is an approximate net zero
cost to the system due to wind generation and curtail-
ment.
We consider two additional cases, in which there is
either no oshore or no onshore wind component. In
both these cases, we find that the Merit-Order Eect is
significantly reduced, such that cost savings are also re-
duced. The result in both options is a significant final
cost to the economy. We conclude that it is best to have
a significant total percentage of wind generation in order
to reduce energy prices and both oshore and onshore
wind generation are equally important in this eect.
6. Acknowledgements
We thank APX Group for access to their datasets un-
der an academic license. Thanks also go to Thomas
Routier from Elexon, for his considerable help in ex-
plaining the imbalance market data to us. We are hugely
grateful to Rt Hon Ed Davey and Good Energy Ltd
for their useful comments and suggestions to enhance
this work. This work has been financially supported by
EPSRC Grant EP/I032541/1 (“Photovoltaics for Future
Societies”).
[1] Sensfuß, F., Ragwitz, M., Genoese, M.. The merit-order eect:
A detailed analysis of the price eect of renewable electricity
generation on spot market prices in germany. Energy Policy
2008;36(8):3086 – 3094. URL: http://www.sciencedirect.
com/science/article/pii/S0301421508001717.
doi:http://dx.doi.org/10.1016/j.enpol.2008.03.035.
[2] Weigt, H.. Germany’s wind energy: The poten-
tial for fossil capacity replacement and cost saving.
Applied Energy 2009;86(10):1857 – 1863. URL:
http://www.sciencedirect.com/science/article/
pii/S0306261908003139. doi:http://dx.doi.org/10.
1016/j.apenergy.2008.11.031.
[3] Ketterer, J.C.. The impact of wind power generation
on the electricity price in germany. Energy Economics
2014;44:270 – 280. URL: http://www.sciencedirect.com/
science/article/pii/S0140988314000875. doi:http://
dx.doi.org/10.1016/j.eneco.2014.04.003.
[4] Cludius, J., Hermann, H., Matthes, F.C., Graichen,
V.. The merit order eect of wind and photovoltaic
electricity generation in germany 2008–2016: Estima-
tion and distributional implications. Energy Economics
2014;44:302 – 313. URL: http://www.sciencedirect.
com/science/article/pii/S0140988314001042.
doi:http://dx.doi.org/10.1016/j.eneco.2014.04.020.
[5] de Miera, G.S., del R´
ıo Gonz´
alez, P., Vizca´
ıno, I.. Analysing
the impact of renewable electricity support schemes on power
prices: The case of wind electricity in spain. Energy Policy
2008;36(9):3345 – 3359. URL: http://www.sciencedirect.
com/science/article/pii/S0301421508001882.
doi:http://dx.doi.org/10.1016/j.enpol.2008.04.022.
[6] Swinand, G.P., O’Mahoney, A.. Estimating the im-
pact of wind generation and wind forecast errors on en-
ergy prices and costs in ireland. Renewable Energy
2015;75:468 – 473. URL: http://www.sciencedirect.com/
science/article/pii/S0960148114006314. doi:http://
dx.doi.org/10.1016/j.renene.2014.09.060.
[7] Barnhart, C.J., Dale, M., Brandt, A.R., Benson, S.M.. The en-
ergetic implications of curtailing versus storing wind- and solar-
generated electricity. AGU Fall Meeting Abstracts 2013;:A905.
[8] Routier, T.. Imbalance pricing guidance: A guide to electricity
imbalance pricing in great britain. Tech. Rep.; ELEXON; 2014.
8
ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
We present a theoretical framework to calculate how storage affects the energy return on energy investment (EROI) ratios of wind and solar resources. Our methods identify conditions under which it is more energetically favorable to store energy than it is to simply curtail electricity production. Electrochemically based storage technologies result in much smaller EROI ratios than large-scale geologically based storage technologies like compressed air energy storage (CAES) and pumped hydroelectric storage (PHS). All storage technologies paired with solar photovoltaic (PV) generation yield EROI ratios that are greater than curtailment. Due to their low energy stored on electrical energy invested (ESOIe) ratios, conventional battery technologies reduce the EROI ratios of wind generation below curtailment EROI ratios. To yield a greater net energy return than curtailment, battery storage technologies paired with wind generation need an ESOIe > 80. We identify improvements in cycle life as the most feasible way to increase battery ESOIe. Depending upon the battery's embodied energy requirement, an increase of cycle life to 10000-18000 (2-20 times present values) is required for pairing with wind (assuming liberal round-trip efficiency [90%] and liberal depth-of-discharge [80%] values). Reducing embodied energy costs, increasing efficiency and increasing depth of discharge will also further improve the energetic performance of batteries. While this paper focuses on only one benefit of energy storage, the value of not curtailing electricity generation during periods of excess production, similar analyses could be used to draw conclusions about other benefits as well.
Article
Full-text available
This paper studies the impact of wind generation on system costs and prices in Ireland. The importance of wind power and potential impacts on system costs is of interest to power system planners and policy makers globally. However, the impact of wind generation on system costs has been only studied with limited actual data from power systems with increased wind penetration. The paper uses a unique dataset of half-hourly system demand, generation, forecast and actual wind generation, along with Irish system marginal price (SMP) data from 2008 to autumn 2012. An econometric time-series model of SMP as a function of forecast and realized demand and wind generation yields results which suggest that each 1% increase in wind generation reduces SMP in Ireland by about 0.06%, while each 1% wind forecast error increases SMP about 0.02%. In absolute terms, though, at the mean the impact of wind forecast errors is small, or about 0.4€cent/MWh-wind generated.
Article
Generation from renewable energy sources in Germany has experienced a considerable uptake in recent years. Mainly responsible for this development is the German Renewable Energy Sources Act (Erneuerbare Energien Gesetz, EEG). This paper considers redistributive implications of the EEG for different electricity consumers. Using time-series regression analysis, we show that electricity generation by wind and PV has reduced spot market prices considerably by 6 €/MWh in 2010 rising to 10 €/MWh in 2012. We use these results to build a near-term forecasting tool for merit order effects, projected to reach 14-16 €/MWh in 2016. On the other hand, the costs of the EEG are passed forward to consumers in the form of a surcharge. Our findings highlight significant redistributive transfers under the current design of the EEG. In particular, some energy-intensive industries are benefiting from lower wholesale electricity prices whilst being largely exempted from contributing to the costs of the scheme. We also highlight implications of our results for other areas for reform of the EEG, such as a gradual shift to a more market-based method for remuneration of renewables under the scheme. More generally, these findings suggest that policy makers need to integrate distributional assessments into policy design and implementation.
Article
This paper investigates the relationship between intermittent wind power generation and electricity price behaviour in Germany. Using a GARCH model, I evaluate the effect of wind electricity generation on the level and the volatility of the electricity price in an integrated approach. The results show that variable wind power reduces the price level but increases its volatility. This papers’s results also indicate that regulatory change has stabilised the wholesale price. The electricity price volatility has decreased in Germany after a modification of the marketing mechanism of renewable electricity. This gives confidence that further adjustments to regulation and policy may foster a better integration of renewables into the power system.
Article
Wind energy has become the major renewable energy source in Germany with an installed capacity of more than 20Â GW and an annual output of about 40Â TWÂ h in 2007. In this paper we analyze the extent to which wind energy can replace fossil capacities based on wind injection and demand data for 2006 through June 2008. The results indicate that the wind potential in Germany will not allow a significant reduction of fossil capacities. We also assess the potential savings due to wind energy. The German market is modeled with and without wind input to estimate the net savings of fossil fuels in the observation period. We find that the cost-saving potential for electricity production is quite significant and exceeds the subsidies.
Article
The German feed-in support of electricity generation from renewable energy sources has led to high growth rates of the supported technologies. Critics state that the costs for consumers are too high. An important aspect to be considered in the discussion is the price effect created by renewable electricity generation. This paper seeks to analyse the impact of privileged renewable electricity generation on the electricity market in Germany. The central aspect to be analysed is the impact of renewable electricity generation on spot market prices. The results generated by an agent-based simulation platform indicate that the financial volume of the price reduction is considerable. In the short run, this gives rise to a distributional effect which creates savings for the demand side by reducing generator profits. In the case of the year 2006, the volume of the merit-order effect exceeds the volume of the net support payments for renewable electricity generation which have to be paid by consumers.
Article
It is sometimes argued that renewables are “expensive”. However, although it is generally true that the private costs of renewable electricity generation are certainly above those of conventional electricity, that statement fails to consider the social benefits provided by electricity from renewable energy sources (RES-E), including environmental and socioeconomic ones. This paper empirically analyses an additional albeit usually neglected benefit: the reduction in the wholesale price of electricity as a result of more RES-E generation being fed into the grid. The case of wind generation in Spain shows that this reduction is greater than the increase in the costs for the consumers arising from the RES-E support scheme (the feed-in tariffs), which are charged to the final consumer. Therefore, a net reduction in the retail electricity price results, which is positive from a consumer point of view. This provides an additional argument for RES-E support and contradicts one of the usual arguments against RES-E deployment: the excessive burden on the consumer.
URL: http://www.sciencedirect.com/science
Applied Energy 2009;86(10):1857 – 1863. URL: http://www.sciencedirect.com/science/article/ pii/S0306261908003139. doi:http://dx.doi.org/10. 1016/j.apenergy.2008.11.031.
Imbalance pricing guidance: A guide to electricity imbalance pricing in great britain
  • T Routier
Routier, T.. Imbalance pricing guidance: A guide to electricity imbalance pricing in great britain. Tech. Rep.; ELEXON; 2014.
The merit-order effect: A detailed analysis of the price effect of renewable electricity generation on spot market prices in germany URL: http://www.sciencedirect. com/science Germany's wind energy: The potential for fossil capacity replacement and cost saving
  • F Sensfuß
  • M Ragwitz
  • M H Genoese
Sensfuß, F., Ragwitz, M., Genoese, M.. The merit-order effect: A detailed analysis of the price effect of renewable electricity generation on spot market prices in germany. Energy Policy 2008;36(8):3086 – 3094. URL: http://www.sciencedirect. com/science/article/pii/S0301421508001717. doi:http://dx.doi.org/10.1016/j.enpol.2008.03.035. [2] Weigt, H.. Germany's wind energy: The potential for fossil capacity replacement and cost saving. Applied Energy 2009;86(10):1857 – 1863. URL: http://www.sciencedirect.com/science/article/ pii/S0306261908003139. doi:http://dx.doi.org/10. 1016/j.apenergy.2008.11.031.