On the Use of Average versus Marginal Emission Factors
Wouter Schram
a
, Ioannis Lampropoulos
b
, Tarek AlSkaif
c
and Wilfried van Sark
d
Copernicus Institute of Sustainable Development, Utrecht University, Princetonlaan 8A, Utrecht, The Netherlands
Keywords: Marginal Emission Factors, Merit Order, Co
2
Price.
Abstract: In this paper, we propose using marginal emission factors instead of average emission factors for determining
the impact of adding variable renewable electricity to the generation mix. Average emission factors assume
constant emissions over time, which does not reflect reality. Therefore, they cannot be used for e.g. accurately
determining the mitigated CO
2
emissions by renewables, or for scheduling shiftable loads in order to have the
lowest CO
2
emissions. To solve this, we provide a method to construct the marginal emission profiles via the
merit order and demonstrate the method by composing these for the case of the Netherlands. Using this
method, we re-evaluate the CO
2
impact in 2014 of photovoltaic-generated electricity to be 0.42 Mt – compared
to 0.36 Mt using the average emission factor - and for wind-generated electricity to be 3.6 Mt instead of 2.9
Mt CO
2
(an increase of 14.3% and 24.2%, respectively). Furthermore, we show the impact of CO
2
price on
the merit order and show that even high CO
2
prices of 50 to 75 €/tCO
2
are not sufficient to phase-out new
coal-fired power plants.
1 INTRODUCTION
Two options to mitigate greenhouse gas (GHG) emis-
sions are to add variable renewable electricity to the
electricity mix, or to decrease electricity demand
through energy conservation measures. There are two
main methods to determine the impact of such
measures on the CO
2
emissions of a country. Mostly,
an average emission factor (AEF) is used to estimate
the emissions of the replaced electricity generation.
This makes the implicit assumption that a decrease in
conventional electricity generation, such as from
coal- and gas-fired power plants, is evenly distributed
over all generation facilities. However, this is not in
line with the functioning of electricity markets, since
in practice a decrease in requested supply results in
decreased electricity generation of facilities operating
at the margins.
The AEF is defined as the total direct CO
2
emis-
sions of the electricity generation sector, divided by
the total electricity generation over a certain period –
usually one year (Mancarella and Chicco, 2009). The
concept of marginal emission factors (MEF) focuses
a
https://orcid.org/0000-0003-3407-7893
b
https://orcid.org/0000-0001-8566-4970
c
https://orcid.org/0000-0002-1780-4553
d
https://orcid.org/0000-0002-4738-1088
on the notion that renewably-generated electricity re-
places the electricity generated by the price-setting
power plants of a specific settlement period used in the
market, e.g. 1-hour or 15-minutes trading interval
(Siler-Evans et al., 2012). This is generally seen as a
superior method over the use of AEFs, because the lat-
ter disconnects the actual contribution to CO
2
emis-
sions and the abatement scenario by implicitly assum-
ing constant CO
2
intensity (Harmsen and Graus, 2013).
Several studies have shown that the use of MEFs
leads to increased accuracy of estimations of CO
2
sav-
ings. In England and Wales the use of the AEF led to
an underestimation of CO
2
savings when determining
the impact of energy efficiency measures (Bettle et al.,
2006). Similar results were obtained for the case of
California (Marnay et al., 2002). In addition, the envi-
ronmental impact of increased wind power generation
in Great Britain could be estimated more accurately by
using MEFs (Thomson et al., 2017). In general, AEFs
are lower than MEFs and therefore result in underesti-
mation of CO
2
savings (Hawkes, 2010).
The contribution of this paper is threefold. First,
we provide a straightforward method of designing
marginal emission profiles. Second, we survey and
Schram, W., Lampropoulos, I., AlSkaif, T. and van Sark, W.
On the Use of Average versus Marginal Emission Factors.
DOI: 10.5220/0007765701870193
In Proceedings of the 8th International Conference on Smart Cities and Green ICT Systems (SMARTGREENS 2019), pages 187-193
ISBN: 978-989-758-373-5
Copyright
c
2019 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved
187
report the data of Dutch generation facilities. We also
show the actual CO
2
mitigation based on the marginal
emission factors, and compare it with the average
emission factor, which has not been done before for
the Netherlands. Third, we show the impact of CO
2
price on the merit order, hereby offering guidance for
policy makers in determining options to meet CO
2
emission targets.
2 METHODS
To compose marginal emission profiles, first the
merit order needs to be constructed. This is the elec-
tricity generation mix sorted from lowest to highest
marginal operating costs (i.e. the costs of increasing
generation by one unit of energy, in this case MWh
e
).
Marginal costs MC (in €/MWh) of facility j is the
sum of the fuel costs, the emission costs and the var-
iable operating costs (Biggar and Hesamzadeh,
2014), and thus can be determined as follows:
𝑀𝐶
𝑗
= 𝐹𝑃
𝑗
η
𝑗
⁄
+ 𝐸𝐹
𝑗
η
𝑗
⁄
∗ CP + 𝑉𝑂𝐶
𝑗
(1)
where 𝐹𝑃 denotes the price of fuel (in €/MWh
t
), η is
the conversion efficiency, 𝐸𝐹 is emission factor of
the fuel (in tCO
2
/MWh
t
), CP denotes the EU Emis-
sion Trading System (ETS) CO
2
price (in €/tCO
2
),
and 𝑉𝑂𝐶 is the Variable Operational Costs (in
€/MWh).
We follow IPCC (2006) and focus on CO
2
for
GHG emissions in power generation. The marginal
emissions 𝑀𝐸 (in tCO
2
/MWh) of facility j are deter-
mined as follows:
𝑀𝐸
𝑗
= EF
𝑗
η
𝑗
⁄
(2)
Subsequently, there are two options for construct-
ing a marginal emission profile. First, one can com-
pose a generation mix based on the electricity demand
in a specified time period, and take the marginal emis-
sions of the price-setting facility. Second, one can
take the day-ahead market (DAM) clearing prices, de-
termine from this which facility was operating at the
margin. Then, the emissions of this facility can be
taken for the marginal emission profile. The latter is
more accurate, as it reflects what historically hap-
pened. The former is more suitable when looking at
future scenarios.
For constructing the merit order and marginal
emission profiles, the following assumptions had to
be made:
e
Unless otherwise specified, MWh
electric is meant
To determine the marginal operating facility, we
assumed that the facility with marginal costs clos-
est to the spot price was the marginal operating
facility.
All facilities were assumed to operate at their
maximum efficiency; efficiency losses of operat-
ing at partial load are not considered.
There are several methods to allocate CO
2
emis-
sions in the case of combined heat and power pro-
duction (Graus and Worrell, 2011). Here we chose
to allocate CO
2
emissions to power generation.
Co-firing of biomass in coal-fired power plants
was not included.
The marginal operating facility was assumed to be
located in the investigated country.
No assumptions about future scenarios are made.
Bid strategies of retailers were not considered.
In the following section, the proposed method is elu-
cidated by applying it on a case study for the Nether-
lands.
3 RESULTS
3.1 Merit Order and Marginal
Emissions Netherlands
To construct the marginal emission profile, various
data were required as input. First, the generation port-
folio of the Netherlands was established, using the da-
tabases of ENTSO-E (ENTSO-E, 2019b). For every
facility, the installed capacity and the efficiency were
determined. Table 1 shows all these values and the
accompanying sources. Because of data availability,
we chose 2014 as base year for obtaining data. Fuel
price for coal was based on data from Statistics Neth-
erlands; on average the price of coal was 9.1 €/MWh
t
for 2014 (CBS, 2017) and of natural gas prices 24.3
€/MWh
t
(Schoots et al., 2017). VOC was assumed to
be 1.2 €/MWh for gas-fired power plants, and 3.0
€/MWh for coal-fired power plants (Brouwer et al.,
2015). For 2014, CP was assumed constant at the av-
erage of 6.9 €/ tCO
2
(Investing, 2019).
The EF of bi-
tuminous coal and natural gas were determined to be
0.341 tCO
2
/MWh
t
and 0.204 tCO
2
/MWh
t
, respec-
tively (IPCC, 2006). Velsen-24 was a special case; a
peak-load facility that uses a mixture of blast furnace
gas from nearby steal production and natural gas. Fol-
lowing the position of the Dutch government, we at-
tribute an EF of 1.25 times the EF of natural gas for
SMARTGREENS 2019 - 8th International Conference on Smart Cities and Green ICT Systems
188
Figure 1: Merit order based on marginal costs (left y-axis, green) and marginal emissions (right y-axis, red) of the facilities
opearting in 2014 (see Table 1). Each marker represents a power plant; the distance on the x-axis between two markers reflects
the size of the facility.
this (Afman and Wielders, 2014). Figure 1 shows the
resulting merit order and accompanying marginal
emissions.
Table 1 lists the main characteristics for all Dutch
centralized thermal power plants
f
, their installed capac-
ity, efficiency and the resulting MC and ME. Figure 1
shows the merit order, with for every individual facility
the accompanying marginal emissions. At around 4.5
GW of installed capacity, we see the substantial gap of
around 13 €/MWh between the most expensive coal-
fired power plant, and the cheapest gas-fired power
plant. These coal-fired power plants have much higher
marginal emissions: around 850 tCO
2
/MWh, com-
pared to around 350 tCO
2
/MWh for gas-fired power
plants.
Figure 2 illustrates the marginal emission profile of a
randomly chosen day, i.e. 11 January 2018. DAM-
prices are taken from (ENTSO-E, 2019a) and CO
2
price was 9.28 €/t CO
2
(Investing, 2019). During most
f
Must-run facilities are excluded
of the day, gas-fired power plants operate at the mar-
gin, while during the night coal-fired power plants
Figure 2: Marginal costs and emissions on 11-01-2018.
On the Use of Average versus Marginal Emission Factors
189
Table 1: Overview and characteristics of all thermal power plants in the Netherlands, using a CO
2
price of 7 €/t CO
2
.
Facility Name
Main
Fuel
Installed
capacity
(MW)
In Opera-
tion
Marginal
costs
(€/MWh)
Marginal
emissions
(kg CO2
/MWh)
Source
Centrale Maasvlakte
Coal
1.070
2016
27.0
728
(D66, 2015)
Eemshavencentrale
Coal
1.560
2015
27.5
743
(D66, 2015)
Engie Centrale
Rotterdam 11
Coal
730
2015
27.5
743
(D66, 2015)
Amer Bio WKC
Coal
600
1994
31.2
852
(D66, 2015)
Centrale Hemweg
Coal
630
1994
31.2
852
(D66, 2015)
Maasvlakte-2
Coal
520
1974
31.9
874
(D66, 2015)
Maasvlakte-1
Coal
520
1973
32.7
897
(D66, 2015)
Gelderland-13
Coal
602
1982
32.7
897
(D66, 2015)
Amer-8
Coal
645
1981
33.5
921
(D66, 2015)
Diemen-34
Gas
435
2012
43.9
345
(Nuon, 2018)
Centrale Hemweg (gas)
Gas
435
2012
43.9
345
(Nuon, 2018)
Sloecentrale-10
Gas
432
2010
44.1
347
(Sloecentrale, 2018)
Sloecentrale-20
Gas
432
2010
44.1
347
(Sloecentrale, 2018)
Maximacentrale FL5
Gas
439
2010
44.3
348
(Seebregts et al., 2009)
Maximacentrale FL4
Gas
438
2010
44.3
348
(Seebregts et al., 2009)
Magnum Eemshaven 10
Gas
440
2013
44.6
351
(Nuon, 2018)
Magnum Eemshaven 20
Gas
440
2013
44.6
351
(Nuon, 2018)
Magnum Eemshaven 30
Gas
440
2013
44.6
351
(Nuon, 2018)
Moerdijk-2
Gas
426
2014
44.6
351
(RWE, 2018b)
Enecogen
Gas
870
2011
44.6
351
(Seebregts et al., 2009)
Maasstroom Energie
Gas
427
2010
44.6
351
(Seebregts and
Daniëls, 2008)
Maasbracht-C (Claus)
Gas
1.275
2012
46.2
364
(RWE, 2018a)
Rijnmond Energie
Gas
820
2004
46.2
364
(Seebregts and
Volkers, 2005)
Pergen-1
Gas
260
2007
46.2
364
(Seebregts, 2007)
Eemscentrale EC4
Gas
341
1996
47.0
370
(Siebelink, 2006)
Eemscentrale EC5
Gas
341
1996
47.0
370
(Siebelink, 2006)
Eemscentrale EC6
Gas
341
1997
47.0
370
(Siebelink, 2006)
Eemscentrale EC7
Gas
341
1997
47.0
370
(Siebelink, 2006)
Eemscentrale EC3
Gas
341
1996
48.7
384
(Siebelink, 2006)
Energiecentrale Den Haag
Gas
95
1906
49.6
392
(Enipedia, 2018)
Diemen-33
Gas
266
1995
50.6
400
(Arcadis, 2009)
Lage Weide
Gas
248
1996
57.2
453
(Croezen, 2016)
Eemscentrale EC20
Gas
695
1978
57.2
453
(Siebelink, 2006)
Merwede-12
Gas
225
1990
57.2
453
(Croezen, 2016)
Centrale Bergum CB10
h
Gas
332
1975
59.8
474
(Seebregts and
Volkers, 2005)
Centrale Bergum CB20
h
Gas
332
1976
59.8
474
(Seebregts and
Volkers, 2005)
Centrale RoCa
h
Gas
220
1997
61.2
485
(Seebregts and
Volkers, 2005)
Centrale Swentibold
h
Gas
230
2000
61.2
485
(Seebregts and
Volkers, 2005)
Merwede-11
h
Gas
103
1985
62.6
497
(Seebregts and
Volkers, 2005)
Velsen-24
BF-Gas
i
459
1975
74.2
728
(Seebregts and
Volkers, 2005)
g
Planned to reopen in 2020
h
Estimation based on average similar plants
i
Blast furnace gas mixed with natural gas. Based on emis-
sion factor natural gas times 1.25 (Afman and
Wielders, 2014)
SMARTGREENS 2019 - 8th International Conference on Smart Cities and Green ICT Systems
190
operate at the margin. This shows that scheduling de-
mand to optimize on costs mainly leads to increasede-
lectricity generation by coal-fired power plants, and
thus to increased emissions.
When applied to the Netherlands in 2014, we end
up with an average MEF of 629 kg/MWh (standard
deviation of 278 kg/MWh), compared with an AEF of
503 kg/MWh. Our results are in line with the official
Dutch statistics, where only the total yearly MEF is
provided; an MEF of 636 kg/MWh (CBS, 2018b). We
applied the marginal emission profile to determine the
mitigated CO
2
emissions in the Netherlands in 2014
by wind- and PV-generated electricity, assuming
these are linearly dependent on wind speed and solar
irradiation, respectively (CBS, 2018a; KNMI, 2019).
This results in an updated CO
2
impact of wind
from 2.9 Mt CO
2
when using the AEF to 3.6 Mt CO
2
when using MEFs (+24.2%) and from 0.36 Mt CO
2
to 0.42 Mt CO
2
(+14.3%). From this, one can also
conclude that compared to PV, wind is producing
more during hours when coal-fired power plants are
operating at the margin, e.g. at night.
3.2 Impact of CO
2
Price
Figure 3 shows the merit order of all Dutch thermal
power plants for various CO
2
prices. A CO
2
price of
25 €/tCO
2
does not lead to changes in the merit order,
apart for the higher prices (figure 3b). The break-even
price between the most efficient coal-fired power
plants (46-47%) and the most efficient gas-fired
power plants lies around 50 €/tCO
2
, as can be seen in
Figure 3c.
With this price, the old coal-fired power
plants become more expensive than many gas-fired
power plants, whereas the new coal-fired power
Figure 3: Impact CO
2
price on merit order of all Dutch thermal power plants. See Table 1 for overview power plants.
On the Use of Average versus Marginal Emission Factors
191
plants, which went in operation in 2015 and 2016, re-
main base load. This only changes when prices are
increased to 75 €/tCO
2
(figure 3d).Hence, stating one
CO
2
price that is needed for a shift from coal to gas is
too simplistic; it is depending on the entire generating
mix, and the relative age of the different facilities. This
varies from country, providing a strong argument for
national policies for CO
2
taxes on top of the European
level policies. Furthermore, this shows that even high
CO
2
prices are not able to push all coal-fired power
plants out of the merit order, despite their higher
emissions. Either very high CO
2
prices (from 100
euro per tonne), or additional measures are needed if
policy makers decide these emissions should be de-
creased.
4 CONCLUDING REMARKS
In this paper, we presented a method for designing
marginal emission profiles for a specific country
based on its generation merit order and applied this
for the case study of the Netherlands. The value of
this approach can be understood from two perspec-
tives. From a bottom-up perspective, consumers may
reconsider the scheduling of their electricity demand.
The operation of shiftable loads, such as electric ve-
hicles, wet appliances and stationary storage devices
can be scheduled considering the minimization of
CO
2
emissions, in addition to cost, if the demand can
be shifted to periods with cleaner periods. This can be
from coal to gas in the nearby future, but also in a
more distant future from periods with fossil-fuel
based power plants operating at the margin to periods
where renewables are operating at the margin. From
a top-down perspective, the approach might help to
better determine the impact of implementing renewa-
bles in the generation mix, and for determining ade-
quate CO
2
prices to enforce a shift from coal to gas.
ACKNOWLEDGEMENTS
This project is part of the PVProsumers4Grid Project,
which received funding from the European Union’s
Horizon 2020 research and innovation programme
under grant agreement No 764786. Furthermore, this
work has received funding in the framework of the
joint programming initiative ERA-Net Smart Grids
Plus as part of the CESEPS project, as well as from
TKI Urban Energy (Project: B-DER, contract number
1621404).
REFERENCES
Afman, M. R., and Wielders, L. M. L. (2014).
Achtergrondgegevens stroometikettering 2013. Delft.
Arcadis. (2009). Passende beoordeling Diemen 34 Natura
2000-gebied Markermeer & IJmeer. Retrieved from
http://www.commissiemer.nl/docs/mer/p20/p2097/209
7-026vergunningaanvraag003.pdf
Bettle, R., Pout, C. H., and Hitchin, E. R. (2006).
Interactions between electricity-saving measures and
carbon emissions from power generation in England
and Wales. Energy Policy, 34(18), 3434–3446.
https://doi.org/10.1016/j.enpol.2005.07.014
Biggar, D. R., and Hesamzadeh, M. R. (2014). The
Economics of Electricity Markets. The Economics of
Electricity Markets (Vol. 9781118775). https://doi.org/
10.1002/9781118775745
Brouwer, A. S., van den Broek, M., Seebregts, A., and
Faaij, A. (2015). Operational flexibility and economics
of power plants in future low-carbon power systems.
Applied Energy, 156, 107–128. https://doi.org/
10.1016/j.apenergy.2015.06.065
CBS. (2017). Ketelkolen ; invoerprijs uit niet EU-landen.
Retrieved April 19, 2018, from http://statline.cbs.nl/
StatWeb/publication/?VW=T&DM=SLNL&PA=3721
5&D1=2&D2=a&HD=120417-1100&HDR=T&STB=
G1
CBS. (2018a). Hernieuwbare elektriciteit; productie en
vermogen. Retrieved from http://statline.cbs.nl/
Statweb/publication/?DM=SLNL&PA=82610ned&D1
=0%2C7&D2=a&D3=a&HDR=T&STB=G1%2CG2&
VW=T
CBS. (2018b). Rendementen en CO2-emissie elektriciteits-
productie 2016. https://doi.org/10.1023/A:102023833
0817
Croezen. (2016). Analyse conclusies en uitgangspunten
berekening Stadsverarming. Delft.
D66. (2015). Kolendeal Energieakkoord : Emissiereductie
en verduurzaming kan en moet. Retrieved August 8,
2018, from https://d66.nl/content/uploads/sites/2/2015/
03/DELTA-Kolendeal-Energieakkoord.pdf
Enipedia. (2018). Den Haag Powerplant. Retrieved August
9, 2018, from http://enipedia.tudelft.nl/wiki/
Den_Haag_Powerplant
ENTSO-E. (2019a). Day-ahead Prices. Retrieved March 8,
2019, from https://transparency.entsoe.eu/transmis
sion-domain/r2/dayAheadPrices/show
ENTSO-E. (2019b). Installed Capacity Per Production
Unit. Retrieved March 30, 2018, from
https://transparency.entsoe.eu/generation/r2/installedC
apacityPerProductionUnit/show
Graus, W., and Worrell, E. (2011). Methods for calculating
CO2 intensity of power generation and consumption: A
global perspective. Energy Policy, 39(2), 613–627.
https://doi.org/10.1016/j.enpol.2010.10.034
Harmsen, R., and Graus, W. (2013). How much
CO2emissions do we reduce by saving electricity? A
focus on methods. Energy Policy, 60, 803–812.
https://doi.org/10.1016/j.enpol.2013.05.059
SMARTGREENS 2019 - 8th International Conference on Smart Cities and Green ICT Systems
192
Hawkes, A. D. (2010). Estimating marginal CO2emissions
rates for national electricity systems. Energy Policy,
38(10), 5977–5987. https://doi.org/10.1016/j.enpol.
2010.05.053
Investing. (2019). Carbon Emissions Futures Historical
Data. Retrieved March 8, 2019, from https://www.
investing.com/commodities/carbon-emissions-historic
al-data
IPCC. (2006). Volume 2: Energy - Chapter 2: Stationary
Combustion. 2006 IPCC Guidelines for National
Greenhouse Gas Inventories, 2.1-2.47. https://doi.org/
10.1016/S0166-526X(06)47021-5
KNMI. (2019). Klimatologie: Uurgegevens van het weer in
Nederland. Retrieved from http://www.knmi.nl/klimato
logie/uurgegevens/selectie.cgi
Mancarella, P., & Chicco, G. (2009). Global and local
emission impact assessment of distributed cogeneration
systems with partial-load models. Applied Energy,
86(10), 2096–2106. https://doi.org/10.1016/j.apenergy.
2008.12.026
Marnay, C., Fisher, D., Murtishaw, S., Phadke, A., Price,
L., & Sathaye, J. (2002). Estimating Carbon Dioxide
Emissions Factors for the California Electric Power
Sector. Retrieved from https://cloudfront.escholarship.
org/dist/prd/content/qt3kc8r2n1/qt3kc8r2n1.pdf
Nuon. (2018). Gasgestookte centrales. Retrieved August 9,
2018, from http://www.nuon.com/activiteiten/produ
ctie/gas/gasgestookte-centrales/
RWE. (2018a). Clauscentrale. Retrieved August 9, 2018,
from http://www.rwe.com/web/cms/nl/1772100/rwe-
generation-se/brandstoffen/overzicht-locaties/nederlan
d/clauscentrale/
RWE. (2018b). Moerdijkcentrale. Retrieved August 9,
2018, from http://www.rwe.com/web/cms/nl/1772076/
rwe-generation-se/brandstoffen/overzicht-locaties/ned
erland/moerdijkcentrale/
Schoots, K., Hekkenberg, M., and Hammingh, P. (2017).
Nationale Energieverkenning 2017. Ecn-O--14-036.
https://doi.org/ECN-O--16-035
Seebregts, A. J. (2007). Beoordeling nieuwbouwplannen
elektriciteitscentrales in relatie tot de WLO SE- en GE-
scenario’s: een quickscan. Petten.
Seebregts, A. J., and Daniëls, B. W. (2008). Nederland
exportland elektriciteit? Nieuwe ontwikkelingen
elektriciteitscentrales en effect Schoon & Zuinig, 1–39.
Seebregts, A. J., Snoep, H. J. M., Van Deurzen, J., Lensink,
S. M., Van Der Welle, A. J., and Wetzels, W. (2009).
Brandstofmix elektriciteit 2020 Inventarisatie,
mogelijke problemen en oplossingsrichtingen,
(December). Retrieved from https://www.vemw.nl/~/
media/VEMW/Downloads/Public/Elektriciteit/Rappor
t ECN - merit order en brandstofmix.ashx
Seebregts, A. J., and Volkers, C. H. (2005). Monitoring
Nederlandse elektriciteitscentrales 2000-2004. ECN
Beleidsstudies, (November). https://doi.org/ECN-C--
05-090
Siebelink, J. (2006). Stroom uit vliegtuigmotor. Retrieved
August 9, 2018, from https://www.digibron.nl/
search/detail/012dc30cc887fbe8c2a21e72/stroom-uit-
vliegtuigmotor
Siler-Evans, K., Azevedo, I. L., and Morgan, M. G. (2012).
Marginal emissions factors for the U.S. electricity
system. Environmental Science and Technology, 46(9),
4742–4748. https://doi.org/10.1021/es300145v
Sloecentrale. (2018). Onze centrale. Retrieved August 9,
2018, from https://www.sloecentrale.nl/nl/#Centrale
Thomson, R. C., Harrison, G. P., and Chick, J. P. (2017).
Marginal greenhouse gas emissions displacement of
wind power in Great Britain. Energy Policy,
101(December 2016), 201–210. https://doi.org/
10.1016/j.enpol.2016.11.012
On the Use of Average versus Marginal Emission Factors
193