1876-6102 © 2016 The Au ho s. Published by Else ie L d. This is an open access a icle unde he CC BY-NC-ND license
(h p://c ea i ecommons.o g/licenses/by-nc-nd/4.0/).
Pee - e iew by he scien i ic con e ence commi ee o EEIC | CIEE 2016 unde esponsibili y o Gues Edi o s.
doi: 10.1016/j.egyp o.2016.12.101
1s Ene gy Economics Ibe ian Con e ence, EEIC | CIEE 2016, Feb ua y 4-5, Lisbon,
Po ugal, APEEN (www.apeen.o g) and AEEE (www.aeee.es)
Renewables e sus e iciency. A compa ison o Spain
Juan-Manuel Roldan-Fe nandeza
a
*, Manuel Bu gos-Payana,
Jesus-Manuel Riquelme-San osa and Angel-Luis T igo-Ga ciaa
aUni e si y o Se illa, Camino de los Descub imien os, s/n. Se illa 41092, Spain
Abs ac
Along he las decades, enewable ene gy (especially wind) in Spain has unde gone a signi ican de elopmen (lead
by a small g oup o enewable p omo e s suppo ed by ins i u ional policies), con ibu ing signi ican ly o elec ic
gene a ion mix (42.8% enewable in 2014). On he con a y, he p omo ion o ene gy e iciency ac ions
(accomplished by a la ge numbe o indus ial and domes ic consume s ha a e e y poo ly suppo ed by ene gy
policies), a e s ill li le explo ed. Acco ding o ODYSSEE-MURE, ene gy e iciency a he EU-28 le el imp o ed by
1.2%/yea on a e age om 2000 o 2013, while o he case o Spain, he a e o imp o emen was only 0.6 %/yea
on a e age h oughou ha pe iod ( he lowes a e o ene gy e iciency imp o emen in he EU-28).
This wo k seeks o compa e he in eg a ion o enewable p oduc ion wi h ene gy e iciency plans, in o de o ad ance
hei po en ial economic impac in he wholesale ma ke and consume s. To each ha goal, he hou ly ma ke da a
e ie ed om he Spanish/Ibe ian Ma ke Ope a o (OMIE) o 2014 will be used as a base. Then, a se o pseudo-
heu is ic scena ios wi h in eg a ion o enewable p oduc ion and ene gy e iciency (load sa ing) will be elabo a ed
and analyzed o quan i y wha a e expec ed o be he main e ec s on he Spanish elec ici y ma ke and consume s.
The esul s will show ha ene gy e iciency exhibi s he bes pe o mance in e ms o economic e iciency (less cos
o he aded ene gy) and en i onmen al sus ainabili y (g ea e eplacemen o ossil uels).
© 2016 Published by Else ie L d. Selec ion and/o pee - e iew unde esponsibili y o [name o ganize ]
Elec ici y ma ke s; Renewable ene gy; Ene gy e iciency; Me i -o de e ec .
* Co esponding au ho . Tel.: +34 954487283; ax: +34 954487284.
E-mail add ess: jm olda[email p o ec ed]s.
© 2016 The Au ho s. Published by Else ie L d. This is an open access a icle unde he CC BY-NC-ND license
(h p://c ea i ecommons.o g/licenses/by-nc-nd/4.0/).
Pee - e iew by he scien i ic con e ence commi ee o EEIC | CIEE 2016 unde esponsibili y o Gues Edi o s.
Juan-Manuel Roldan-Fe nandez e al. / Ene gy P ocedia 106 ( 2016 ) 14 – 23 15
1. In oduc ion
O e ecen yea s he e has been a signi ican e o in he elec ical sys ems o he EU coun ies and
many o he s wo ldwide, o ul ill he manda o y commi men s de i ed om he Kyo o P o ocol and i s
successo s [1,2]. The eby, he gene a ion lee has di e si ied signi ican ly in e ms o echnology
gene a ion, mainly due o la ge-scale in eg a ion o p oduc ion based on enewable ene gy. Consequen ly,
elec ical sys ems a e e ol ing owa ds a gene a ion lee ha is much mo e di e se, dispe sed and
decen alized, wi h a much la ge numbe o gene a o s, in which he enewable gene a ion has a
signi ican and g owing sha e.
The educ ion o he demand ia he p omo ion o load sa ing p og ams o he enac men ene gy
e iciency policies could also be ano he less explo ed ool o accomplish he commi men s om he
Kyo o P o ocol. In ac , he e exis a ce ain syne gy be ween ene gy e iciency and enewables since
when he demand dec eases, a ixed amoun o enewable p oduc ion gi es place o a g ea e sha e o
enewables on he gene a ion mix. In any case, mos coun ies ha e chosen he de elopmen o enewables
as a means o ul ill i s en i onmen al commi men s.
This is he case o Spain whe e along he las decades, enewable ene gy (especially wind) has
unde gone a signi ican de elopmen . This g ow h has been led by a ela i ely small g oup o enewable
p omo e s suppo ed by ins i u ional policies, and ha e made enewables o be a signi ican con ibu o o
elec ic gene a ion mix (42.8% enewable, 20.4% wind in 2014 [3]). On he con a y, he p omo ion o
ene gy e iciency ac ions, accomplished by a la ge numbe o indus ial and domes ic consume s ha a e
e y poo ly suppo ed by ene gy policies, s ill a e li le explo ed. Acco ding o ODYSSEE-MURE [4],
ene gy e iciency a he EU-28 le el imp o ed by 1.2%/yea on a e age om 2000 o 2013 (abou 15%
o e he pe iod). Howe e , he pace o p og ess has slowed down since he economic c isis: he annual
gain be ween 2000 and 2007 has d opped om 1.3%/yea o 1%/yea be ween 2007 and 2013. Fo he
case o Spain, ene gy e iciency only imp o ed on a e age by 0.6 %/yea h oughou ha pe iod, which is
he lowes a e o ene gy e iciency imp o emen in he EU-28. Mo eo e , ene gy conse a ion policies
seem especially well sui ed o Spain due o i s high le el o ene gy dependence. Acco ding o Eu os a
[5] he Spanish a e o g oss ene gy dependency is always much highe han ha o he EU a e age. Fo
2013, he le el o g oss dependence o Spain was 70.5% in 2013, well abo e he 53.2% o he UE
a e age.
This wo k seeks o compa e hese wo app oaches o he Spanish case: he in eg a ion o enewable
p oduc ion and he de elopmen ene gy e iciency p og ams, in o de o ad ance hei po en ial impac on
he elec ici y ma ke . Wi h his pu pose, i s a quali a i e model, based on he linea iza ion o he
wholesale ma ke a ound he clea ing poin , is used o examine some basic hypo heses. An app op ia e se
o empi ical-based scena ios wi h enewables and ene gy e iciency a e hen gene a ed om he e ie ed
his o ical in o ma ion o he Ibe ian/Spanish Ma ke Ope a o (OMIE) o he yea 2014, in o de o
quan i y he main e ec s on he ma ke . The con en o he pape is as ollows. A e he in oduc ion, he
Spanish/Ibe ian elec ici y ma ke is b ie ly desc ibed and a quali a i e model, based on he linea iza ion
o he ma ke , is used o examine some basic hypo hesis ega ding he expec ed e ec s o enewables and
ene gy e iciency. The hou ly me i -o de gene a ion and demand cu es h oughou 2014, e ie ed om
he a chi e o he Ma ke Ope a o (OMIE) a e hen used as sou ce da a o he gene a ion o ealis ic
enewables and ene gy e iciency scena ios. The main po en ial e ec s o enewables and ene gy
e iciency on he ma ke a e hen quan i ied and analyzed. Finally, he pape closes wi h he main indings
o he compa ison.
16 Juan-Manuel Roldan-Fe nandez e al. / Ene gy P ocedia 106 ( 2016 ) 14 – 23
1.1. The Ibe ian ma ke o elec ici y
OMIE is he Ma ke Ope a o o he Ibe ian Elec ici y Ma ke , which is he Eu opean egional ma ke
o Spain and Po ugal. Al hough OMIE has been in eg a ed in o he Eu opean P ice Coupling (EPC)
since 2014, his does no a ec he ac ual ma ke ules [6]. The ma ke is o ganized as a sequence o
ma ke s: he day-ahead ma ke , he in a-day ma ke , wi h six sessions a day, which ope a es close o eal
ime, and he ancilla y se ices ma ke .
The daily ma ke in ol es he scheduling o elec ici y ansac ions o he day ahead, which is
pe o med h ough he submi al o elec ici y sale and pu chase bids by ma ke agen s. The daily ma ke
is composed o 24 hou ly ma ke s ha clea once a day. On he supply side, he Ibe ian elec ici y
p oduce s submi hei bids speci ying he p ice a which hey a e willing o p oduce a gi en amoun o
ou pu om each o hei p oduc ion uni s, one day ahead. Simila ly, demand agen s submi hei bids
speci ying he p ice a which hey a e willing o buy a gi en amoun o ene gy. Once he supply and
demand bids ha e been submi ed by hei agen s, he Ma ke Ope a o (OMIE) c ea es a me i -o de
dispa ch, o e e y hou o he day ahead, by o de ing he supply bids in ascending p ice o de and
demand bids in descending o de . The hou ly equilib ium p ice and he gene a ion dispa ch a e
de e mined h ough ma ke clea ing, i.e. by compu ing he in e sec ion be ween he supply and demand
cu es. Condi ional on being dispa ched, he p ice o be ecei ed o paid by he ma ke pa icipan s is se
acco ding o a uni o m-p ice auc ion. I espec i e o hei bids, he p ice p oduce s ecei e, o he p ice
paid by demand uni s is se equal o he highes accep ed supply bid: he so-called sys em ma ginal p ice.
Two kinds o bids a e conside ed in he Ibe ian Elec ici y Ma ke : simple and complex bids. Simple
bids a e jus simple p ice and amoun o ene gy bids. Complex bids a e bids ( o gene a o s only) which
include any o he ollowing condi ions: indi isibili y o blocks o ene gy, minimum income,
p og ammed s ops and load amps. Fi s , he simple ma ching p ocedu e o he Ma ke Ope a o (OMIE)
inds a solu ion o simple bids (Fig. 1). An op imiza ion algo i hm, EUPHEMIA, hen inds a new
solu ion which akes complex o e bids in o accoun [6]. Subsequen ly, he inal clea ing p ice and aded
ene gy a e se o he conside ed hou . Finally, he Sys em Ope a o (REE – Red Eléc ica de España)
alida es he schedule o he uni s while conside ing he elec ical sys em’s echnical cons ain s.
Fig. 1. Me i o de gene a ion and demand cu es o a peak hou (20:00 h) in a win e wo king day (Tuesday, Feb ua y 10, 2015)
co esponding o he Ibe ian ma ke , OMIE. Example o simple ( ade ene gy, Wis = 36.81 GWh; clea ing p ice, pis = 49.57 €/MWh)
and complex ma ching ules ( aded ene gy, Wi = 34.18 GWh; clea ing p ice, pi = 71.00 €/MWh) [7]
0
40
80
120
160
200
010 20 30 40 50 60 70
Gene a ion
simple bids
Demand
simple bids
Gene a ion
clea ed
Demand clea ed
pG(W)
p
D
(W)
Ene gy (GWh)
P ice (€/MWh)
Wi
piComplex ma ching
Simple ma ching
Juan-Manuel Roldan-Fe nandez e al. / Ene gy P ocedia 106 ( 2016 ) 14 – 23 17
As o en happens in op imiza ion p oblems, an inc ease in he numbe o es ic ions leads o a
deg ada ion o he op imali y o he solu ion. In he ma ke case, complex o e bids (signi ican ly)
inc ease he inal clea ing p ice and (sligh ly) educe he aded ene gy as shown in Fig. 1.
2. Quali a i e Analysis o he Wholesale Ma ke
In Fig. 2 can be seen bo h he me i -o de gene a ion cu e, pG = pG(W), and he demand cu e,
pD = pD(W), as well as he aded ene gy (Wi = 34183.4 MWh) and he ma ching clea ing p ice (pi = 71.00
€/MWh) o a peak hou (20:00 h) on a win e wo kday (Tuesday, Feb ua y 10, 2015) co esponding o
he wholesale Ibe ian ma ke [7]. By i s own na u e, he supply cu e, pG = pG(W), has a posi i e slope,
and he demand cu e, pD = pD(W), has a e y nega i e slope.
I he supply and demand cu es we e con inuous (and no s epped) and mG = dpG(W)/dW > 0 and
mD = dpD(W)/dW << 0 we e, espec i ely, he slopes o he supply (mG = 1.4 €/GWh) and demand cu es
(mD = –13.5 €/GWh) a he ini ial clea ing poin (A in Fig. 2), hen bo h he supply and demand cu es
could be linea ly app oxima ed, in he su oundings o he ini ial clea ing poin (Wi,pi), as:
() ( )
GiGiiGiGGiG
pW p mWW p mW mW p mW  
() ( )
DiDiiDiDDiD
pW p mWW p mW mW p mW  
Whe e he espec i e o dina es a he o igin a e
Gi i G i
ppmW 
and
Di i D i
ppmW 
The o al ini ial income o he p oduce s o he o al ini ial cos o consume s, C(Wi), de i ed om
he ene gy ading in he wholesale ma ke , conside ing only he ma ke ules, is C(Wi)= Wi · pi:
Fig. 2. Me i o de gene a ion, pG = pG(W), and demand, pD = pD(W), cu es as well as he aded ene gy, Wi = 34.18 GWh, and he
ma ching clea ing p ice, pi = 71.00 €/MWh, o a peak hou (20:00 h) in a win e wo king day (Tuesday, Feb ua y 10, 2015)
co esponding o he wholesale Ibe ian ma ke [7]. Linea iza ion o he ma ke a ound he ini ial ma ching poin
Wi h his linea ized ma ke model a ound he clea ing poin , a compa ison be ween he in luence o
he in eg a ion o enewable gene a ion and DSM (ene gy e iciency) can be e alua ed in o de o
ad ance o he po en ial e ec s o hese measu emen s on he pe o mance o he ma ke . I mus be
ema ked ha since he wholesale ma ke is a ma ginal ma ke , he clea ing poin is he mos impo an
0
20
40
60
80
100
010 20 30 40 50
p
G
(W) ≈p
Gi
+ m
G
W
Gene a ion
me i -o de
simple bids
Demand
me i -o de
simple bids
Gene a ion clea ed
Ene gy (GWh)
P ice (€/MWh)
Demand clea ed
Wi
piA(Wi,pi)
pD(W) ≈pDi + mDW
p
G
(W)
pD(W)
18 Juan-Manuel Roldan-Fe nandez e al. / Ene gy P ocedia 106 ( 2016 ) 14 – 23
ac o o know he impac o hese measu emen s. Al hough he linea app oxima ions o he me i -o de
cu es (pG ≈ pGi + mGW and pD ≈ pDi + mDW) di e om he ac ual p oduc ion and demand cu es
(pG = pG(W) and pD = pD(W)), bo h se s o cu es can lead o he same (o e y close) clea ing poin . This
inally de e mines he p ice ha e e y buye has o pay o p oduce s and he amoun o ene gy ha
gene a o s ha e o deli e o he cus ome s.
2.1. In eg a ion o enewable ene gy
The Ibe ian ma ke egula ion cu en ly equi es he Ma ke Ope a o o include all bids ecei ed om
enewable gene a o s as long as hey cause no echnical di icul y o he ope a ion o he sys em.
The e o e, he in eg a ion o new enewable gene a ion bids (∆ER > 0) a e y low (o e en null) ma ginal
p ice yields a mainly igh -hand-side shi o he ini ial me i -o de gene a ion cu e. The linea
app oxima ion o his new supply cu e, pGR = pGR(W), is shown in Fig. 3 as a s aigh line pa allel o he
p imi i e gene a ion cu e:
()() ()
GR G R Gi G R Gi G R G GRi G
pW pW E p mW E p mE mWp mW ' ' ' 
Whe e he o dina e a he o igin is:
()
GRi Gi G R i G i R
ppmEpmWE '  '
Wi h his linea ized ma ke model, he new clea ing p ice and aded ene gy (B shown in Fig. 3) can be
ob ained by equa ing he new gene a ion cu e wi h he demand cu e:
() ()
GR GRi G D Di D
pW p mWpW p mW 
Fig. 3. Me i o de gene a ion, pG = pG(W), and demand, pD = pD(W), cu es and ma ching clea ing p ice (pi = 71.00 €/MWh) and
aded ene gy (Wi = 34.18 GWh) o a peak hou (20:00 h) in a win e wo king day (Tuesday, Feb ua y 10, 2015) co esponding o
he wholesale Ibe ian ma ke [7]. Changes in he ma ke -clea ing p ice and ene gy aded due o he in eg a ion o new enewable
gene a ion and due o he educ ion o he demand de i ed om he imp o emen o ene gy e iciency
The aded ene gy, W = W(ΔER), can now be exp essed as:
()
() ()
Di GRi i D i i G i R G
RiRiR
GD GD GD
pp pmWpmWE m
WE W E W WE
mm mm mm
'
' ' ''
 
0
20
40
60
80
100
010 20 30 40 50
p
DL
(W) = p
D
(W - ΔE
L
)
p
G
(W)
p
GR
(W) = p
G
(W + ΔE
R
)
p
D
(W)
W(ΔE
L
)W(ΔE
R
)
Gene a ion
me i -o de
simple bids
Demand
me i -o de
simple bids
Gene a ion clea ed
Ene gy (GWh)
P ice (€/MWh)
Demand clea ed
ΔE
L
= ΔE
R
ΔW(ΔE
L
)
ΔW(ΔE
R
)
W
i
A(W
i
,p
i
)
Δp(ΔE
L
)= Δp(ΔE
R
)
p
DL
(ΔE
L
)= p
GR
(ΔE
R
)
p
i
C(p
L
,W
L
)B(p
R
,W
R
)
ΔE
L
ΔE
R
0
20
40
60
80
100
010 20 30 40 50
p
DL
(W) = p
D
(W - ΔE
L
)
p
G
(W)
p
GR
(W) = p
G
(W + ΔE
R
)
p
D
(W)
W(ΔE
L
)W(ΔE
R
)
Gene a ion
me i -o de
simple bids
Demand
me i -o de
simple bids
Gene a ion clea ed
Ene gy (GWh)
P ice (€/MWh)
Demand clea ed
ΔE
L
= ΔE
R
ΔW(ΔE
L
)
ΔW(ΔE
R
)
W
i
A(W
i
,p
i
)
Δp(ΔE
L
)= Δp(ΔE
R
)
p
DL
(ΔE
L
)= p
GR
(ΔE
R
)
p
i
Δp(ΔE
L
)= Δp(ΔE
R
)
p
DL
(ΔE
L
)= p
GR
(ΔE
R
)
p
i
C(p
L
,W
L
)B(p
R
,W
R
)
ΔE
L
ΔE
R

Juan-Manuel Roldan-Fe nandez e al. / Ene gy P ocedia 106 ( 2016 ) 14 – 23 19
As a esul , he inc emen o aded ene gy can be app oxima ed as:
() 0
G
RR
GD
m
WE E
mm
'' '!

The clea ing p ice, pGR (W) = pD(W), becomes:
( ( )) ( ( )) ( )
GD
GR i R GRi G i R i R i R R
GD
mm
pWW WE p mW WE p E p p E
mm
''  ''  ' ''

whe e he p ice a ia ion ( educ ion) is:
() 0
GD
RR R
GD
mm
pE E
mm
'' '

I should be obse ed ha , since he slope o he supply is smalle han ha co esponding o he
demand cu e, in absolu e alues (0 < mG << |mD|), he inc emen in aded ene gy is signi ican ly less
han he inc emen in enewable ene gy bids in eg a ed in o he ma ke :
()
01
G
R
RGD
m
WE
Emm
''
'
This means ha he clea ing o a ce ain amoun o enewable ene gy bids, ΔER, by he Ma ke
Ope a o , lea es ou almos he same amoun o ene gy bids om o he mo e expensi e and p obably
pollu ing p oduc ion echnologies, ΔW(ΔER) << ΔER, and leads o a educ ion o he hou ly clea ing p ice
p opo ional o he amoun o clea ed enewable ene gy. This is he base o he so-called me i -o de
e ec o enewable ene gy [8-12].
2.2. Ene gy e iciency
The consume s ha apply load sa ing o ene gy e iciency p og ams can p o i om a educ ion in he
elec ici y ene gy bill, mainly due o a educ ion o he demanded ene gy. The cu ailmen o ce ain
amoun o demand bids (∆EL > 0) a high ma ginal p ice, whe e he ene gy e iciency ac ions would be
mo e cos -e icien , mainly esul s in a le -shi ing o he ini ial me i o de demand cu e. The linea
app oxima ion o his new demand cu e, pDL = pDL(W), is shown in Fig. 3 as a s aigh line pa allel o he
p imi i e demand cu e:
()() ()
DL D L Di D L Di D L D DLi D
pW pW E p mW E p mE mW p mW ' ' ' 
Whe e
()
DLi Di D L i D i L
ppmEpmWE '  '
The new clea ing p ice and aded ene gy (C in Fig. 3) can now be ob ained by equa ing he new
( educed) demand cu e wi h he p imi i e gene a ion cu e. I is wo h no ing ha , since he slope o he
supply is less inclined han ha co esponding o he demand cu e, in absolu e alue (0 < mG << |mD|),
he educ ion in aded ene gy is simila o (bu smalle han) ha o he sa ed load bids:
()
01
LD
LGD
WE m
Emm
''
!!
'
As a esul , he educ ion o an amoun o demand bids esul ing om a ce ain ene gy e iciency
imp o emen o he load, ΔEL, causes he Ma ke Ope a o o lea e ou almos he same quan i y o
gene a ion bids om p oduc ion echnologies o a mo e expensi e na u e, WL(ΔEL) # ΔEL, and his leads
20 Juan-Manuel Roldan-Fe nandez e al. / Ene gy P ocedia 106 ( 2016 ) 14 – 23
o a educ ion o he clea ing p ice and cos o he aded ene gy p opo ional o he amoun o sa ed
ene gy bids. This o ms he base o wha could be called he me i -o de e ec o he ene gy e iciency,
which is e y simila o ha co esponding o enewables. Finally, he ollowing hypo heses a e s a ed:
x The educ ions in he clea ing p ice o he enewable and ene gy e iciency scena ios a e he
same.
x The ene gy e iciency scena io leads o a educ ion in he aded ene gy which is almos equal o
ha o sa ed ene gy, while he enewable scena io leads o a sligh inc emen in he aded
ene gy. This means ha he ene gy e iciency is (sligh ly) mo e e ec i e han enewables in
elimina ing ( eplacing) he mo e expensi e and p obably pollu ing gene a ion echnology.
x The educ ion o he cos o he aded ene gy in he ene gy e iciency scena io is g ea e han
ha co esponding o he enewable scena io, since he educ ion in he e iciency scena io
p o i s om bo h he educ ion in he clea ing p ice and om he educ ion in he amoun o
aded ene gy.
3. Me hodology and Resul s
The hou ly me i -o de gene a ion and demand cu es o he yea 2014 ha e been e ie ed om he
his o ic da a a chi e o he Ibe ian Ma ke Ope a o [7]. Tha means 8760 hou ly ma ke s wi h hund eds
o p oduc ion and demand bids aking pa in each hou ly ma ke . This as amoun o hou ly ma ke bids
handled o each scena io made desi able a simpli ica ion in he complex clea ing ules o he Ma ke
Ope a o in o de o simula e new scena ios. As a consequence, a simpli ied p ocedu e simila o ha
used by he Ma ke Ope a o (OMIE) is applied wi h he in en ion o pe o ming a quan i a i e analysis o
he scena ios.
Fo each new scena io analyzed, new hou ly me i -o de gene a ion and/o demand cu es a e
p oduced by i s loca ing he bids o he ype o agen unde in e es (i.e. enewables o consume s bids)
in he hou ly me i -o de (gene a ion and/o demand) cu es. The bids unde in e es a e hen modi ied
(amoun o ene gy and p ice) by aking in o accoun he app op ia e cha ac e is ics o he scena io unde
conside a ion. The modi ied bids co esponding o he new scena io a e subsequen ly inse ed p ope ly
in o he e ie ed hou ly me i -o de cu es: upwa d p ice o de o bids in he gene a ion cu es, and
downwa d o de o he bids in he demand cu es. When new ma ke scena ios a e de eloped h ough
he modi ica ion o he supply and/o demand cu es, i is necessa y o conside ha he bids o he o he
agen s would emain he same when inc easing o educing new gene a ion o demand bids. Fo una ely,
his hypo hesis o a pe ec ma ke can be conside ed ul illed since each ma ke agen should elabo a e
hei bids wi hou any knowledge o he bids o he o he agen s. Finally, he ma ching poin
co esponding o he new scena io is de e mined as he in e sec ion o he modi ied gene a ion and
demand cu es.
Fo example, in o de o conside he case o he in eg a ion o an amoun o enewable ene gy, ΔER,
i s he o iginal hou ly me i -o de gene a ion (pG = pG(W)) and demand (pD = pD(W)) cu es a e
e ie ed, as shown in Fig. 4. Since enewable gene a o s o e ene gy a a e y low (e en null) p ice, he
o iginal bids co esponding o he enewable gene a ion uni s can be ound in he ini ial la egion ( i s
s eps) o he me i -o de gene a ion cu e. The o iginal enewable bids a e now p opo ionally inc eased
un il he a ge ed amoun (ΔER) is eached.
When needed, he o iginal hou ly-clea ed cu es a e comple ed wi h he single-bid cu e om he
ini ial clea ing poin onwa ds, as shown in Fig. 4. In his way, a new me i -o de gene a ion cu e,
pGR = pGR(W), is c ea ed h ough he p ope in eg a ion (upwa d p ice) o hese modi ied bids in he
gene a ion cu e. This new me i -o de gene a ion cu e essen ially esul s in a igh -hand-shi ing o he
ini ial clea ed supply cu e by an amoun equal o ΔER. Finally, he new clea ing poin (B in Fig. 4) can
Juan-Manuel Roldan-Fe nandez e al. / Ene gy P ocedia 106 ( 2016 ) 14 – 23 21
be de e mined as he c ossing poin o he new igh -hand-shi ed gene a ion cu e, pGR = pGR(W), wi h
he o iginal demand cu e, pD = pD(W).
Fig. 4 also shows he case o he educ ion o he demand in an amoun o load, ΔEL, due o, o
example, ene gy sa ing o small consume s. Since small consume s demand ene gy a he highes p ice,
hei p imi i e bids a e always ound in he ini ial la egion o maximum p ice o he me i -o de
demand cu e. The o iginal bids o he small-consume ade s a e hen p opo ionally educed un il he
a ge ed amoun (ΔEL) is eached. In his way, a new me i -o de o educed demand cu e, pDL = pDL(W),
is c ea ed h ough he p ope in eg a ion (downwa d p ice) o hese modi ied bids in he demand cu e.
This new me i -o de demand cu e essen ially esul s in a le -hand-shi ing o he ini ial clea ed demand
cu e by an amoun equal o ΔEL, as shown in Fig. 4.
Finally, he new clea ing poin (C in Fig. 4) can be de e mined as he in e sec ion o he new le -hand-
shi ed educed-demand cu e, pDL = pDL(W), wi h he o iginal gene a ion cu e, pG = pG(W).
Th ee scena ios a e gene a ed o 0.5%, 1% and 2% o load demand educ ion as well as he
symme ical scena ios wi h 0.5%, 1% and 2% g owing o enewable ene gy. Tha gi es place o a o al o
6·8760 = 52560 hou ly ma ke s scena ios which mus be simula ed using he me hodology desc ibed
abo e. The ob ained esul s a e summa ized in Table 1 and 2.
Fig. 4. Hou ly me i -o de gene a ion, pG = pG(W), and demand, pD = pD(W), cu es and c ea ion o he new ( igh -shi ed)
enewable gene a ion cu e, pGR = pGR(W), and o he new (le -shi ed) educed demand cu e, pDL = pDL(W).
Mo e p ecisely, Table 1 summa izes he mean alues o he a ia ions o he annual aded ene gy,
hou ly clea ing p ice and annual cos o he aded ene gy, while Table 2 shows he mean alues o he
a e o a ia ion wi h he ene gy-sa ing and enewable bids o he annual aded ene gy, hou ly clea ing
p ice and annual cos o he aded ene gy. As an icipa ed by he linea model app oxima ion, he esul s
lead o he ollowing main conclusions:
x Ene gy e iciency: The mean alues o he yea ly aded ene gy, he hou ly clea ing p ice and
he annual cos o he aded ene gy a e always smalle han he co esponding o he base case,
and hei educ ions g ow almos linea ly wi h he amoun o load-sa ing bids.
x Renewables: The aded ene gy is always sligh ly g ea e ha he co esponding o de base case
and i s inc emen g ows wi h he amoun o enewable bids. On he con a y, he clea ing p ice
and he cos o he aded ene gy a e smalle han o he base case, and hei educ ions g ow
wi h he quan i y o enewable bids.
0
20
40
60
80
100
120
140
160
180
200
010 20 30 40 50 60 70
O e
Simple bids
Demand
Simple bids
O e
Ma ched
Demand
Ma ched
p
G
(W)
p
D
(W)
Ene gy (GWh)
P ice (€/MWh)
p
GR
(W)
p
DL
(W)
A
B
C
W
i
p
i
ΔE
R
ΔE
L
22 Juan-Manuel Roldan-Fe nandez e al. / Ene gy P ocedia 106 ( 2016 ) 14 – 23
Fo he same amoun o load-sa ing and enewable bids (ΔER = ΔEL = ΔE):
x The clea ing p ice (and i s a ia ion) o ene gy e iciency is almos he same han o he
co esponding enewable scena io, Δp(ΔER = ΔE) | Δp(ΔEL = ΔE).
x The in ensi y o he a ia ion o he aded ene gy (absolu e alue), ΔW(ΔE)/ΔE, and cos o he
aded ene gy, ΔC(ΔE)/ΔE, is always s onge o ene gy e iciency han o he co esponding
enewable scena io. Mo e p ecisely, he in ensi ies a e (ΔW(ΔER)/ΔER)/(°ΔW(ΔEL)°/ΔEL) ≈ 2.4
and (ΔC(ΔER)/ΔER)/(ΔC(ΔEL)/ΔEL) ≈ 1.6, espec i ely.
Table 1. Mean alues o he a ia ions o he yea ly aded ene gy, hou ly clea ing p ice and yea ly cos o he aded ene gy
Spain 2014
Yea ly
mean
W = 221 TWh/y
p = 42.13 €/MWh
C = 9346 M€/y
ΔE = 0.5% Load
(0.90 TWh/y)
ΔE = 1% Load
(1.81 TWh/y)
ΔE = 2% Load
(3.62 TWh/y)
Small
consume s
Uni s
E iciency
ΔEL*
Renewable
ΔER*
E iciency
ΔEL*
Renewable
ΔER*
E iciency
ΔEL*
Renewable
ΔER*
ΔW(ΔE)
GWh
-610.96
246.81
-1237.44
478.10
-2476.90
954.19
Δp(ΔE)
€/MWh
-0.36
-0.36
-0.68
-0.68
-1.33
-1.33
ΔC(ΔE)
M€
-104.89
-66.56
-204.17
-128.09
-401.06
-251.22
*ΔER = ΔEL = ΔE
Table 2. Mean alues o he a io o a ia ion wi h he load sa ing o enewable bids o he annual aded ene gy, hou ly clea ing
p ice and yea ly cos o he aded ene gy
Spain 2014
Yea ly
mean
W = 221 TWh/y
p = 42.13 €/MWh
C = 9346 M€/y
ΔE = 0.5% Load
(0.90 TWh/y)
ΔE = 1% Load
(1.81 TWh/y)
ΔE = 2% Load
(3.62 TWh/y)
Small
consume s
Uni s
E iciency
ΔEL*
Renewable
ΔER*
E iciency
ΔEL*
Renewable
ΔER*
E iciency
ΔEL*
Renewable
ΔER*
ΔW(ΔE)/ΔE
-
-0.68
0.28
-0.67
0.28
-0.67
0.28
Δp(ΔE)/ΔE
€/MWh2
-3.81·10-3
-3.81·10-3
-3.34·10-3
-3.34·10-3
-3.05·10-3
-3.05·10-3
ΔC(ΔE)/ΔE
€/MWh
-116.54
-73.95
-112.80
-70.77
-110.79
-69.40
4. Conclusions
Renewables and ene gy e iciency pu a downwa d p essu e on he clea ing p ice and he o al cos o
he aded ene gy in he wholesale ma ke . These educ ions a e mainly ela ed o he educ ion o he cos
o he ossil uel necessa y o he ma ke ope a ion, de i ed om he ossil uel subs i u ed ( enewables)
o a oided (ene gy e iciency).
This wo k has shown he esul s o compa ing he po en ial economic impac in he Ibe ian elec ici y
ma ke o he in eg a ion o enewable p oduc ion and ene gy e iciency. To each ha goal and o
explo e some basic hypo heses, i s a quali a i e model based on he linea iza ion o he wholesale
ma ke , has been used. Then, he yea 2014 has been analysed by using he hou ly Ibe ian Ma ke
in o ma ion. An app op ia e se o heu is ic-based scena ios ha e been gene a ed including in eg a ion o
enewable gene a ion as well as ene gy e iciency, in o de o o eseen he main quan i a i e e ec s on
he ma ke pe o mance.
The quan i a i e esul s o he heu is ic-based scena ios con i m ha , o he same amoun o
enewable and load sa ing bids, he e iciency scena io exhibi s he bes economic ( educ ion o aded