Ci a ion: Aizpu ua-E xeza e a, M.;
Ca eno-Madinabei ia, S.; Ulazia, A.;
Sáenz, J.; Saenz-Agui e, A.
Long-Te m F eezing Tempe a u es
F equency Change E ec on Wind
Ene gy Gain (Eu asia and No h
Ame ica, 1950–2019). Sus ainabili y
2022,14, 5630. h ps://doi.o g/
10.3390/su14095630
Academic Edi o : Domenico Cu o
Recei ed: 5 Ap il 2022
Accep ed: 1 May 2022
Published: 7 May 2022
Publishe ’s No e: MDPI s ays neu al
wi h ega d o ju isdic ional claims in
published maps and ins i u ional a il-
ia ions.
Copy igh : © 2022 by he au ho s.
Licensee MDPI, Basel, Swi ze land.
This a icle is an open access a icle
dis ibu ed unde he e ms and
condi ions o he C ea i e Commons
A ibu ion (CC BY) license (h ps://
c ea i ecommons.o g/licenses/by/
4.0/).
sus ainabili y
A icle
Long-Te m F eezing Tempe a u es F equency Change E ec on
Wind Ene gy Gain (Eu asia and No h Ame ica, 1950–2019)
Maddi Aizpu ua-E xeza e a 1, Sheila Ca eno-Madinabei ia 2,* , Alain Ulazia 1, Jon Sáenz 3,4
and Ai o Saenz-Agui e 1
1Ene gy Enginee ing Depa men , Uni e si y o he Basque Coun y (UPV/EHU), E-20600 Eiba , Spain;
[email p o ec ed] (M.A.-E.); [email p o ec ed] (A.U.); ai o [email p o ec ed] (A.S.-A.)
2Depa men o Ma hema ics, Uni e si y o he Basque Coun y (UPV/EHU), E-01006 Vi o ia-Gas eiz, Spain
3Depa men o Physics, Uni e si y o he Basque Coun y (UPV/EHU), E-48080 Leioa, Spain;
[email p o ec ed]
4Plen ziako I sas Es azioa (BEGIK), Uni e si y o Basque Coun y (UPV/EHU), E-48620 Plen zia, Spain
*Co espondence: sheila.ca [email p o ec ed]
Abs ac :
The pe sis en eezing condi ions in cold egions a e he cause o ice acc e ion on me-
chanical and ins umen al elemen s o wind u bines. Consequen ly, ema kable Annual Ene gy
P oduc ion (AEP) losses a e p one o occu in hose wind a ms. Following global expansion o wind
ene gy, hese a eas ha e had inc eased s udy in e es in ecen yea s. The goal o hese s udies is an
imp o ed cha ac e isa ion o he si e o he ins alla ion o u bines, which could p e en unexpec ed
high AEP losses due o ice acc e ion on hem. In his con ex , his pape p o ides an es ima ion o
he eezing empe a u es equency (FTF) a 100 m o e la i udes and e alua es he changes du ing
he las 70 yea s. To ha end, hou ly su ace empe a u e da a (2 m abo e su ace) om he ERA5
eanalysis is used in he [50
◦
N, 75
◦
N] la i udinal bel o he pe iod 1950–2019. The ob ained esul s
show an a e age educ ion o FTF hou s o 72.5 h/decade o all he domain, eaching a maximum
dec ease o 621 h/decade on he sou heas coas o G eenland and a 60% annual educ ion a a speci ic
loca ion in Scandina ia. In e ms o AEP a maximum gain o mo e han 26% would be p ojec ed, as
ca ego ised by he he In e na ional Ene gy Agency.
Keywo ds:
wind ene gy po en ial; global wa ming; ice acc e ion; annual ene gy p oduc ion; ERA5;
empe a u e; applied ma hema ics
1. In oduc ion
Wind ene gy is one o he as es -g owing global enewable elec ici y gene a ion
echnologies. An inc easing numbe o wind a ms a e e ec ed, b eaking eco ds in new
wind powe ins alla ions since 2015. Global ins alled wind powe capaci y eached 744 GW
in 2020, which is capable o gene a ing 7% o he wo ld’s elec ici y demand, acco ding o
da a p o ided by he Wo ld Wind Ene gy Associa ion [
1
]. Many o he a ou able a eas o
wind a m cons uc ion a e in cold egions o high la i udes, such as Eas e n and No he n
Eu ope, No h Ame ica, and Asia [
2
]. Acco ding o s a is ics om he In e na ional Ene gy
Agency (IEA), by he end o 2015, app oxima ely 30% o all wind u bine ins alla ions we e
loca ed in cold egions [3].
These a eas p esen g ea po en ial o wind ene gy exploi a ion o se e al easons.
On he one hand, om he poin o iew o global a mosphe ic ci cula ion, pola a eas a e
cha ac e ised by high p essu e due o su ace cooling; as a esul o he subsidence in hem,
pola egions end o be qui e s able. Howe e , in egions close o he Equa o ha a e
he eby a ec ed by he Fe el cell (a ound 60
◦
N in ou case), he e exis many low-p essu e
a eas due o ba oclinic ins abili y. These ex a opical cyclones p oduce wind egimes o
ema kable alue o he exploi a ion o wind ene gy (e.g., he Bal ic Sea and he UK). I is
in hese la i udes whe e he synop ic su ace wes e ly egime mani es s i sel wi h maximal
Sus ainabili y 2022,14, 5630. h ps://doi.o g/10.3390/su14095630 h ps://www.mdpi.com/jou nal/sus ainabili y
Sus ainabili y 2022,14, 5630 2 o 15
in ensi y in acco dance wi h gene al a mosphe ic ci cula ion pa e ns [
4
]. On he o he
hand, in cold clima e a eas, wind ene gy po en ial is 10% highe han ha in o he egions
due o inc eased ai densi y a low empe a u es wi hin seasonal changes, mainly om
summe o win e [
5
–
9
]. Nume ous loca ions in cold clima e a eas, such as he Swiss Alps,
he no he n Scandina ian coas line, many pa s o China, no he n USA, and Canada o e
high wind-ene gy po en ial, especially du ing he win e mon hs [10].
Howe e , mo e han 50% o wind u bines loca ed in cold egions a e a ec ed by
a ious icing e en s in win e [
11
]. Tempe a u es below he eezing poin o wa e and
we en i onmen s po en ially lead o ice o ma ion and icing pe sis ence on s uc u es
exposed o wind, esul ing in conside able powe losses. P ocesses o ice o ma ion, ype,
du a ion, and se e i y depend on a subse o condi ions [
2
]. Tempe a u e, wind speed,
liquid wa e con en , and d ople size dis ibu ion a e he mos impo an me eo ological
condi ions o conside when s udying ice o ma ion. Ne e heless, he e is a subs an ial
lack o in o ma ion, especially on he la e wo a iables, which can be used o assess and
p edic he se e i y and equency o he phenomenon [
12
]. In addi ion, iden ical wea he
condi ions do no always cause he same icing e ec s on he same u bine o e en di e en
u bines. A di e en ope a ing speeds, depending on he o o , angle o a ack, o a ional
speed, and o he ac o s, icing e ec s a e di e en [2].
This is why i is so complex o ca y ou a de ailed s udy on ice o ma ion and
p oduc ion losses. E en so, se e al s udies in es iga ed he ice acc e ion p ocess
[12–17]
on
wind u bine blades, and i s de imen al e ec on he pe o mance and powe p oduc ion o
wind u bines [
2
,
10
,
18
,
19
]. Mapping all po en ial isks and si e-assessmen s udies a e keys
o success in a wind a m in es men [
20
]. The main d awback associa ed wi h ice acc e ion
on wind u bines om which wind u bine manu ac u e s su e is he conside able Annual
Ene gy P oduc ion (AEP) loss expe ienced by wind u bines loca ed in cold clima e a eas.
The wo easons behind hese losses a e he educed ae odynamic pe o mance o iced
wind- u bine blades and he unexpec ed s oppage pe iods o wind u bines due o high
ae odynamic deg ada ion.
Howe e , acco ding o he In e go e nmen al Panel on Clima e Change (IPCC) epo s,
annual wa ming [21] in he no he n hemisphe e is expec ed wi h global wa ming. E e y
decade since 1880, he empe a u e o Ea h has wa med by 0.14
◦
C [
22
]. Mo eo e ,
as shown in Collins e al.
[23]
, he numbe o os days (numbe o days below 0
◦
C) is
signi ican ly educed globally. Acco ding o Yan e al. [24], os days, cold days, and cold
nigh s a e showing a dec easing end in China. This sugges s ha empe a u es below
he eezing poin o wa e o a mosphe ic ice o ma ion will occu less equen ly in he
coming yea s.
In his con ex , he es ima ion o AEP losses be o e he ins alla ion o wind u bines
is necessa y o wind u bine manu ac u e s, especially in cold clima e a eas. The mos
sui able loca ion o he ins alla ion o wind u bines can be selec ed on he basis o AEP
es ima ions, and pos e io unexpec ed losses ha migh comp omise he con ac s can be
minimised. Ou analysis examines he empo al e olu ion o eezing empe a u es as a
necessa y condi ion o he o ma ion o me eo ological ice, and subsequen AEP losses
om 1950 o 2019 o iden i y clea pa e ns in i s e olu ion. As a esul , a ounda ion o
he si e assessmen , ope a ion, and main enance o wind u bines in no he n la i udes was
de eloped on he basis o hei expec ed AEP losses due o icing.
The pape is o ganised as ollows: Sec ion 2p esen s da ase s and he me hod used
in his s udy. Sec ion 3desc ibes he esul s, Sec ion 4discusses he esul s, and Sec ion 5
concludes his s udy.
2. Da a and Me hodology
2.1. Da a
In his s udy, empe a u es es ima ed by ERA5 [
25
] eanalysis a 2 m o e he su -
ace in high la i udes o he no he n hemisphe e we e used. ERA5 is he mos ecen
eanalysis [
26
] de eloped a he Eu opean Cen e o Medium-Range Wea he Fo ecas s
Sus ainabili y 2022,14, 5630 3 o 15
(ECMWF) and i is eely a ailable h ough he Cope nicus Clima e Da a S o e. I is he i h-
gene a ion eanalysis by he ECMWF. The a ailable ERA5 da a pe iod was om 1950 o he
p esen ; his eanalysis a chi es da a a an hou ly ou pu esolu ion and a ecommended
0.25
◦×
0.25
◦
spa ial esolu ion. Di e en esea che s success ully alida ed ERA5 in wind
po en ial assessmen s o bo h wind speed and densi y, and consequen ly empe a u e,
e en sugges ing ha i can lead o wind ene gy models [7,27,28].
Acco ding o ERA5 da a obse a ion documen a ion [
29
], empe a u e da a om
buoys a e no assimila ed. Fo his eason, in his s udy, wo buoys we e alida ed. One
co esponded o S a ion 44078, OOI I minge Sea Su ace Moo ing managed by he Na ional
Oceanic and A mosphe ic Adminis a ion (NOAA) in G eenland (68.47
◦
N, 9.26
◦
W), which
was used o pe o m alida ion du ing he pe iod o 10 Sep embe 2014–31 Decembe 2019.
The second buoy is loca ed on he Iceland coas (59.95
◦
N, 39.57
◦
W) and is a ailable
om he Pangaea P ojec (PANG) [
30
]. These da a we e used o alida ion du ing he
pe iod o 23 No embe 2007–21 Augus 2009. Valida ion was ca ied ou by means o a
Taylo diag am [
31
] because i shows he co ela ion, cen ed oo mean s anda d e o , and
s anda d de ia ion ( ep esen ed by he ho izon al axis, adial dis ance, and ho izon al axis,
espec i ely) in he same plo . The Taylo diag am is a success ul ool o he e alua ion o
model simula ions agains obse ed da a. I is possible o isually compa e h ee di e en
quali y indica o s simul aneously. ERA5 eanalysis da a a e ep esen ed by a g ey poin in
he x axis in he diag am (see Figu e 1), and isual inspec ion allows o easily de e mining
how close he model da a (g een do s) we e o he buoy da a (i.e., p esen lowe e o ).
I also andomly c ea ed 1000 new se ies using he boo s ap echnique o asce ain he
sensi i i y o he poin posi ion in he Taylo diag am o he a ailable sample.
2.2. Annual F eezing F equency Tempe a u e a 100 m
In o de o ob ain p elimina y in o ma ion on he p esence o a ou able condi ions
o icing in wind ene gy zones, a new a iable was de ined, namely, annual eezing
empe a u e equency (FTF), o which he annual hou s o empe a u es below 0
◦
C a
each g id poin a he hub heigh o he u bines we e coun ed.
Fu he mo e, be o e ob aining he annual FTF, i is necessa y o calcula e empe a u e
a hub heigh , which was 100 m he e. Acco ding o he Ame ican Me eo ological Socie y’s
Glossa y o Me eo ology [
32
], empe a u e in e sion co esponds o a laye o he a mosphe e
a which empe a u e inc eases wi h heigh . This is no consis en wi h he mos equen
beha iou , which is ep esen ed by a educ ion in empe a u e wi h heigh . Al hough em-
pe a u e in e sions a e ypical a high la i udes due o su ace cooling p ocesses, his s udy
app oxima ed a s anda d a mosphe e o he whole domain whe e empe a u e dec eased
h ough a cons an e ical empe a u e g adien (Γ) in K/m (Equa ions (1) and (2)).
Γ=−0.0065 K m−1(1)
100m = 2m +Γ(100 m −2 m)(2)
2.3. Annual FTF Mean Da a and Decadal T ends
Wi h he annual FTF da a desc ibed in he p e ious sec ion, he mean annual alue
a each g id poin was calcula ed o he pe iod o 1950–2019, and esul s a e shown in
Sec ion 3.2 as a pe cen age o annual hou s.
In o de o ob ain a gene al decadal end o he whole domain, annual FTF esul s
we e g ouped o each decade, and linea eg ession was calcula ed wi h he median o
each decade. Las ly, in o de o analyse he empo al e olu ion o his a iable in mo e
de ail, linea eg ession wi h annual da a was calcula ed a each g id poin . The objec i e
was o iden i y a eas wi h he s eepes slope whe e he mos impo an changes occu ed.
Fo his pu pose, he Theil–Sen es ima o was used [
33
,
34
], which obus ly i s simple linea
eg ession by choosing he median o slopes o all lines c ossing each pai o poin s [
35
].
Coe icien s o linea eg essions we e calcula ed, ob aining he 95% con idence in e al.
Sus ainabili y 2022,14, 5630 4 o 15
In his wo k, he mblm() unc ion o he R-c an [
36
] lib a y o he same name was used.
This unc ion om he mblm package [
37
] is used o i linea models on he basis o he
Theil–Sen me hod.
2.4. AEP Loss App oxima ion
The calcula ion o he AEP losses o a wind u bine is based on he me hodology
p oposed by he IEA (see Table 1). This able ela es he du a ion o me eo ological icing
e en s wi h an ice class and associa ed po en ial AEP losses on he basis o expec ed
ae odynamic pe o mance educ ion and s oppage pe iods due o he ice acc e ion ha he
me eo ological icing would cause. Ice classes a e based on measu emen s, and an es ima e
o possible bes and wo s cases unde gi en icing condi ions [38].
Table 1. Ice classi ica ion acco ding o IEA.
IEA Ice Class Me eo ological Icing AEP Loss Due o Icing
[−]% o Yea %
1 0–0.5 0–0.5
2 0.5–3 0.5–5
3 3–5 3–12
4 5–10 10–25
5>10 >20
This IEA ice classi ica ion p o ides a i s idea o he anscendence o eezing and i s
consequences on he AEP o a gi en loca ion. This ice classi ica ion e e s o long- e m
condi ions; o indi idual yea s o win e s, esul s may lie ou side hese classes. I he ini ial
ca ego isa ion o ice is he second o highe , he egula ion ecommends an ice measu emen
campaign and de ailed s udy o ela ed losses [3].
The app oxima ion o FTF o he me eo ological icing du a ion is app op ia e gi en
ha ai empe a u e below 0
◦
C is a necessa y condi ion o icing [
18
,
39
]. The e o e,
p o ided maps and associa ed ends wi h his a iable cons i u e a ough es ima e o
po en ial icing condi ions, and should no be used as he only indica o when i comes o
s udying eal icing condi ions o he design o wind a ms in cold clima es. Fo ha case,
di ec measu emen s o icing pa ame e s (liquid wa e con en , medium olume d ople ,
and wind speed) a e i al and can p o ide mo e accu a e in o ma ion han ice maps can.
In his pape , on he basis o annual FTF app oxima ion da a, he yea ly du a ion
o me eo ological icing condi ions was es ima ed, and ice ca ego isa ion and AEP loss
associa ed wi h ha me eo ological icing can he eby be compu ed, ollowing Table 3-1 o
wind ene gy p ojec s in cold clima es [3].
2.5. AEP Gain
The gene al educ ion in eezing days due o global wa ming p oduces changes
in AEP du ing he en i e pe iod (
∆AEP
) ha can be quan i ied using alues in Table 1.
The alues o he able we e linea ly in e pola ed o es ablish a con inuous ela ionship
be ween me eo ological icing du a ion and AEP loss in e als, and a e ages o e decades
we e hen calcula ed. ∆AEP is AEP loss in pe cen age in he en i e pe iod (Equa ion (3)).
∆AEP =AEP1950−1959 −AEP2010−2019, (3)
gi en ha i is expec ed ha AEP2010−2019 <AEP1950−1959.
3. Resul s
3.1. ERA5 2mValida ion on Two Buoys
The nea es g id poin o
2m
om ERA5 eanalysis and buoy da a was compa ed
h ough he Taylo diag am in Figu e 1. The alida ion o he Iceland and G eenland
buoys had a co ela ion o abo e 0.8 and oo mean squa e e o o less han 1
◦
C and
Sus ainabili y 2022,14, 5630 5 o 15
2
◦
C. The s anda d de ia ion o he model was e y simila o he obse a ion in Iceland,
bu in G eenland, ERA5 showed g ea e a iabili y han ha o he buoy measu emen s,
almos doubling i s s anda d de ia ion om 2 o 3.6
◦
C. Al hough wo se alida ion da a
we e expec ed nea e he A c ic (i.e., G eenland) because o he sys ema ic e o ound
by
Wang e al. [40]
a empe a u es below
−
25
◦
C and he lack o da a in he A ic a ea due
o he ha sh en i onmen al condi ions (see Sec ion 4), alida ion was e y good in bo h
cases in e ms o co ela ion and ela i e e o s.
This was an expec ed esul , gi en ha ERA5 pe o ms be e in e ms o accu acy
han p e ious models and ega ding eanalysis, such as ERA-In e im, e en in he emo e
wea he s a ions o he An a c ic, wi h ewe da a assimila ion sou ces han hose in he
no he n hemisphe e and he A c ic. ERA5 is highly accu a e, and i s highe spa ial and
empo al esolu ion signi ican ly educes cold coas al biases iden i ied in ERA-In e im,
inc easing accu acy ep esen ing he wind [41].
G eenland
S anda d de ia ion
S anda d de ia ion
0 1 2 3 4 5
0 1 2 3 4 5
1
2
3
4
0.1 0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
0.95
0.99
Co ela ion
Obse ed
Iceland (59.95 ºN, 39.57 ºW)
S anda d de ia ion
0 1 2 3 4 5
0 1 2 3 4 5
1
2
3
4
5
0.1 0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
0.95
0.99
Co ela ion
Obse ed
Iceland G eenland
Figu e 1. Taylo diag ams o 2m be ween ERA5 and Iceland and G eenland buoys.
3.2. Annual FTF Mean Da a and Decadal T ends
Annual FTF da a a each g id poin in high la i udes (50
◦
N, 75
◦
N) o he no he n
hemisphe e du ing 1950–2019 we e quan i ied, and he mean o he esul s is shown
in Figu e 2. The igu e shows ha eezing is highe o e land han ha o e he sea,
and inc eases a highe la i udes. In he A c ic Ocean, high alues o annual FTF we e
obse ed. A below abou 60
◦
N la i ude, annual FTF was ze o in o sho e a eas, especially
in a eas a ec ed by he anspo a ion o ene gy due o he A lan ic Ocean Gul S eam.
Maximal annual FTF was ound in inland a eas in G eenland, whe e all hou s o he yea
a e below 0 ◦C.
In ela ion o decadal annual FTF changes, o a domain co e ing all high la i udes
(50
◦
N, 75
◦
N), a
−
72.5 h/decade slope is shown in Figu e 3. This slope had a la ge
95% con idence in e al [
−
84.33,
−
50.21] h/decade. Because he s udy a ea is la ge, i is
di icul o assign a gene al beha iou o i . Howe e , lowe and uppe limi s o he slope’s
con idence in e al we e nega i e, and his con i med an o e all descen in he his o ical
e olu ion o annual FTF. This descen was highe in he las h ee decades (1990–2019).
In o de o iden i y zones wi h di e en beha iou s, he slope o annual FTF a each g id
poin was calcula ed o ob ain mo e accu a e in o ma ion (Figu e 4).
Sus ainabili y 2022,14, 5630 6 o 15
20
40
40
60
60
60
60
60
60
60
60
80
180°E
0°E
60°N
Figu e 2.
Mean alue o annual FTF a 100 m o ERA5 in 1950–2019. Resul s in pe cen age o hou s
pe yea .
0
2500
5000
7500
1950−1959 1960−1969 1970−1979 1980−1989 1990−1999 2000−2009 2010−2019
hou s/yea
AFF
Figu e 3.
Decadal box plo o annual FTF in 1950–2019 om ERA5 da a. Red dashed line, Theil–Sen
es ima o o median o annual alues.
Sus ainabili y 2022,14, 5630 7 o 15
180°E
0°E
60°N
−600 −500 −400 −300 −200 −100 0 100 200
slope (h/decade)
Figu e 4.
Map o decadal ends in annual FTF o 1950–2019. Slope ob ained using Theil–Sein
es ima o o decadal ERA5 da a. Whi e alues, nonsigni ican .
Mos o he esul s we e signi ican a he 95% con idence le el, bu 7% we e no
signi ican and a e plo ed wi h whi e. The highes educ ion up o
−
621 h/decade is
loca ed on he sou heas coas o G eenland. S eep declines a e also obse ed in Iceland,
No way, he Bal ic Sea and wes o Alaska. In o sho e a eas whe e he mean annual
FTF alues (Figu e 2) a e ze o, he end is a ound ze o o sligh ly posi i e. Howe e ,
in he o sho e a eas o highe la i udes, whe e he e a e high annual FTF mean alues,
he nega i e slope is e y p onounced, being maximum o e he sea.
Finally, inc easing he de ail o e he Scandina ian a ea, loca ion o wind a ms and
annual FTF in o ma ion a e combined. Figu e 5shows he loca ion o wind a ms egis e ed
up o 2016, wi h a o al capaci y o 515 MW. These da a we e ex ac ed om he Wind
Powe Da abase [
42
]. The annual FTF slopes a e also shown in Figu e 5. In he a ea shown
(4.5
◦
E–40
◦
E, 52.5
◦
N–75
◦
N) an a e age slope o
−
119 h/decade is ob ained. The maximum
slope,
−
310 h/decade is loca ed in sou hwes coas o No way a 5.75
◦
E, 62
◦
N poin .
The dec ease in he annual FTF a his poin is analysed by aking he alues o he annual
hou s o bo h ex emes 1950 (3623 h) and 2019 (1447 h). A educ ion o 60% is ob ained in a
place whe e wind a ms a e al eady placed.
Sus ainabili y 2022,14, 5630 8 o 15
−50
−50
−150
−150
−150
−150
−150
−150
−150
−350 −300 −250 −200 −150 −100 −50 0
slope (h/decade)
10° 20° 30° 40°
60° 60°
70° 70°
Figu e 5.
Map o he decadal ends in annual FTF o pe iod 1950–2019 ob ained using Theil-Sein
es ima o and ERA5 da a in Scandina ian a ea. The g een do s indica e he loca ion o wind a ms.
3.3. AEP Losses App oxima ion and Gain
The empo al e olu ion o he IEA ice ca ego isa ion app oxima ion, see Table 1, o he
analysed egions and du ing he analysed decades is p esen ed in Figu e 6. Each IEA ice
ca ego y is highligh ed in he maps in Figu e 6wi h a di e en colo . The changes o he
colo s o e he decades is indica i e o he ans o ma ion o he eezing se e i y decade
by decade in he egion.
180°E
0°E
60°N
180°E
0°E
60°N
180°E
0°E
60°N
180°E
0°E
60°N
1 2 3 4 5
IEA Ice class
1950−1959 1960−1969
2000−2009 2010−2019
Figu e 6.
Decadal dis ibu ion o IEA ice class acco ding o Table 1. Ca ego isa ion based on annual
FTF app oxima ion da a a 100 m.
Sus ainabili y 2022,14, 5630 9 o 15
Analysis o he empo al e olu ion o he empe a u e o e he yea s mainly showed
ha i led o mild a ou able condi ions, especially in he wes e n coas o Canada, he Bal ic
Sea, he Uni ed Kingdom, and he A lan ic Ocean nea Iceland and No way. This educ ion
in u n caused a educ ion in en i onmen al ha shness o he ope a ion and main enance o
wind u bines in cold clima e condi ions, and imp o ed hei ene gy p oduc ion. In he las
decade (2010–2019), hese condi ions we e weakened. In he Bal ic Sea and Sco land, eez-
ing se e i y dec eased om Ca ego y 5 o Ca ego y 4. The No wegian Sea is likewise one o
he mos a ec ed places, oge he wi h wes e n Canada (Figu e 7). This weakening would
be accompanied by a dec ease in AEP losses om wind u bines, ha is, an AEP gain.
180°E
0°E
60°N
1->2 2->1 2->3 3->2 4->2 4->3 4->3 4->5 5->2 5->3 5->4
IEA Ice class ansi ion
Figu e 7.
T ansi ions in IEA ice class acco ding o Table 1 om i s (1950–1959) o las decade
(2010–2019). Ca ego isa ion based on annual FTF app oxima ion da a a 100 m.
The empo al e olu ion o ice ca ego isa ion hus shows a clea pa e n o a educ ion
in eezing o e he decades, especially in a eas wi h a cold clima e due o hei p olonged
exposu e o icing condi ions. On he basis o he long- e m end o he me eo ologi-
cal icing, AEP losses o a wind u bine loca ed in a cold clima e a ea a e hus expec ed
o dec ease.
Gi en he esul s in he p e ious maps and he well-known wind ene gy po en ial o
he loca ions, he Gul o Alaska and No he n Eu ope we e selec ed o deepe s udy.
3.3.1. AEP Gain in Gul o Alaska
A milde clima e causes ewe icing e en s, and i s e ec on wind u bine pe o mance
is lowe . Fo ins ance, in he Gul o Alaska, in domain 180
◦
W, 120
◦
W, 50
◦
N, 65
◦
N, a mean
educ ion in AEP losses o 5.19% was obse ed by compa ing decades om 1950 o 2019
(Figu e 8). This is a clea indica o (unde cu en ly expec ed clima ological ends) o he
a ou able long- e m p ospec i e o he ins alla ion o wind u bines in cold clima e a eas.
