Depósi o de in es igación de la Uni e sidad de Se illa
h ps://idus.us.es/
Es a es la e sión acep ada del a ículo publicado en: This is an accep ed manusc ip o a
pape published in:
Ene gy and Buildings (2016): 28/01/2025
DOI: h ps://doi.o g/10.1016/j.enbuild.2016.06.074
Copy igh : © 2016 Else ie L d. All igh s ese ed.
El acceso a la e sión publicada del a ículo puede eque i la susc ipción de la e is a.
Access o he published e sion may equi e subsc ip ion.
“This is an Accep ed Manusc ip o an a icle published by Else ie in Ene gy and Buildings on
2016, a ailable a h ps://doi.o g/10.1016/j.enbuild.2016.06.074”
An app oach o modelling en elope ai igh ness in mul i- amily social housing in
Medi e anean Eu ope based on he si ua ion in Spain
Fe nández-Agüe a, Jesica* ; Domínguez-Ama illo, Samuel; Send a, Juan José; Suá ez, Ra ael
Uni e sidad de Se illa
*j e nandezague [email protected]
Abs ac
Building hyg o he mal pe o mance, indoo ai quali y and ene gy consump ion depend
hea ily on en elope ai igh ness. O e he las h ee decades, single- amily dwellings ha e
been amply s udied in his espec by esea che s in No h Eu ope, he Uni ed S a es and
Canada. Howe e , e y ew s udies ha e been conduc ed on ai igh ness in mul i- amily
housing in wa m clima es such as Medi e anean Eu ope. This pape aims o enhance he
unde s anding o building ai igh ness in ea ly wen y- i s -cen u y mul i- amily buildings in
sou he n Spain. Blowe Doo es s we e conduc ed in 45 uni s in se en such buildings. The
main ai igh ness pa ame e alues ound a e epo ed and compa ed o he da a o o he
buildings in sou he n Eu ope. The pape includes a s a is ical analysis o he indings,
cha ac e ises building ypes and desc ibes he p o ocol used o iden i y and quan i y ai
leakage pa hways. The conclusion d awn is ha al hough a ough p edic i e model can be
de eloped, he esul s a e widely sca e ed due o he impac o he andom componen o
manual cons uc ion, e en in buildings wi h iden ical cons uc ion cha ac e is ics and ypes.
The alues eco ded a e none heless consis en wi h he indings o o he Eu opean su eys.
Keywo ds
Ai igh ness, esiden ial buildings, Blowe Doo es , ai in il a ion, sou he n Eu ope
Abb e ia ions
 V50: ai leakage a e a 50 Pa ac oss he building en elope, including he low a e
h ough join s, issu es and su ace po es (m3/h)
 Qen : ai low (m3/h) ac oss he building en elope
 Δp: p essu e di e en ial (Pa)
 n: ai low exponen
 Cen : ai low (o leakage) coe icien .
 N: numbe o dwellings
 Z: clima e zone
 FA: açade a ea (m2)
 S: g oss loo a ea (m2)
 V: olume (m3)
 WA: window a ea (m2)
 WP: window con ou (pe ime e ) (m)
 WIN: win e se e i y
 SUM: summe se e i y
 n50: ai change a e a 50 Pa (h-1)
 w50: speci ic leakage a ea a 50 Pa (m3/(h·m2))
 50: ai leakage a e pe uni o açade a ea a 50 Pa (m3/(h·m2))
 p50: ai leakage a e pe uni o window pe ime e a 50 Pa (m3/h·m)
 U- alue: he mal ansmission (W/(m2·K))
 T: Mon hly mean empe a u e (OC)
 TM: Mon hly mean o daily maximum empe a u es (OC)
 Tm: Mon hly mean o daily minimum empe a u es (OC)
 H: Mon hly mean ela i e humidi y (%)
 I: sunshine hou s (h)
1. In oduc ion
A his ime, he esiden ial sec o accoun s o 17 % o Spain’s inal ene gy and 25 % o i s
elec ic powe consump ion. App oxima ely 48 % o ha demand is es ima ed o be
a ibu able o hea ing and cooling sys ems, alues ha a e on he ise in Medi e anean a eas
[1]. As Spain is one o he la ges coun ies, a ea-wise, in Medi e anean Eu ope, he
hyg o he mal pe o mance, ene gy consump ion and indoo ai quali y o i s housing s ock a e
su icien ly ep esen a i e o be ex apola ed o simila egions in sou he n Eu ope. In he
wake o he ansposi ion in o Spanish Regula ion (CTE, 2006) [2] o he Eu opean Ene gy
Pe o mance o Buildings Di ec i e (EPBD) and i s p o isions on he mal en elope
equi emen s, ai igh ness may be ega ded as one o he ac o s wi h g ea es impac on
hyg o he mal pe o mance, ene gy consump ion and indoo ai quali y in mode n homes.
Ai igh ness alues and sola adia ion-induced he mal loads (in bo h win e and summe ) a e
ins umen al in de e mining he ac ual ene gy demand in Medi e anean mul i- amily
buildings. This applies o hea ing as well as cooling [3, 4, 5] and consequen ly o building
ene gy s a egies and pe o mance labelling. Iden i ica ion o ai leakage pa hways is a
p e equisi e o p oposing cons uc ion solu ions o imp o e building ai igh ness in compliance
wi h he EU’s 20-20-20 objec i es [6].
Building en elope ai igh ness has become a s anda d a ea o esea ch in he USA [7], whe e
he mos comp ehensi e da abases and wides a ie y o in il a ion models a e a ailable [8, 9,
10, 11]. In pa icula , he LBNL ai leakage measu emen s da abase de eloped by Chan e al.
[12] and he DOE’s Mul i amily En elope Leakage Model [12] a e e y widely used. D awing
om an ex ensi e e iew o measu emen s aken ac oss he coun y, he o me de eloped a
e ised model o single- amily de ached housing. The la e desc ibed a model o mul i-
amily buildings which is, howe e , scan ly applicable o sou he n Eu ope, whe e building
ypes, cons uc ion echniques and clima e a y immensely om he pa ame e s o which he
model was de eloped. Despi e hese cons ain s, bo h can be used a benchma ks o
compa ison.
In ense esea ch is also unde way in no he n and cen al Eu ope [13,14,15], whe e building
ai igh ness is ex ensi ely egula ed. In sou he n Eu ope, dwelling ai igh ness and i s
ela ionship o he main ai inle s has been s udied mo e p o usely in F ance han in any o he
coun y. The e, mul i- amily uni s ha e been shown o be mo e ai igh han single- amily
homes and he cons uc ion sys em used has been iden i ied as one o he majo ac o s
a ec ing he esul s [16, 17]. Legisla ion on ai igh ness is lacking in o he sou he n Eu opean
coun ies, al hough insigh ul pape s ha e been published on condi ions in G eece [18], I aly
[19] and Po ugal [20, 21]. The I alian s udy co e ed 20 buildings o di e en ypes and in
di e en seasons, while in Po ugal he objec o esea ch was social housing, namely wo
mul i- amily buildings su eyed be o e and a e e o i ing. The mean V50 alues ound by he
la e o an a ea compa able o he egion s udied he e we e 8.9 h-1 p io o and 6.8 h-1 a e
e u bishmen .
The mos p ominen in il a ion s udies conduc ed in Spain include esea ch by Meiss and
Feijo, who ound a mean V50 o 6.26 h-1 [22] o he 13 homes measu ed and by Tibe io and
B anchi, who an he mos ex ensi e su ey in he egion o da e, es ing 150 ecen ly buil
dwellings. Al hough he da a we e widely sca e ed, wi h a mean V50 o 3.50 h-1, he
dis ibu ion by low, medium and high ai igh ness showed ha as a ule oday’s buildings lie in
he medium ca ego y [23]. While he da a epo ed in hese pape s a e o in e es om he
s andpoin o cha ac e isa ion, hey e e o no h-cen al Spain, a egion wi h a ai ly cold
con inen al clima e eminiscen o cen al Eu opean cons uc ion echniques gea ed o
ai igh ness, insula ion and e ec i e hea ing sys ems. Building condi ions he e consequen ly
a y widely wi h espec o he wa me a eas o he coun y. In con as , a s udy by Mon oya
e al. add essing he less di ec ly ela ed issue o p o ec ion agains ai pollu ion due o oxic
gas elease, adop ed an app oach ele an o he p esen esea ch [24]. Al hough ha s udy
in ol ed no ield es s and ocused on single- amily homes, i p oposed a egional p edic ion
model ha di e ed om he Chan model and d ew om a da abase o F ench es s (CETE
Lyon). The au ho s also p oposed a me hodology o ex apola ing he da a based on housing
ype simila i ies.
The e o e, he s udies conduc ed in sou he n Eu ope ha e ye o deli e a la ge body o da a
enough o iden i y he ac ual impac o ai igh ness on he ene gy e iciency o he esiden ial
buildings o he a ea, and he he link wi h egional cons uc ion sys ems.
This a icle analyses a comp ehensi e sui e o ield da a wi h he aim o enhancing he
unde s anding o housing s ock ene gy pe o mance in an a ea highly ep esen a i e o
Medi e anean Eu ope. I ocuses on mul i- amily uni s buil in sou he n Spain in he i s
decade o he wen y- i s cen u y. The ield campaign indings a e se ou , ai leakage
pa hways a e cha ac e ised, he sca e obse ed in en elope pe o mance is discussed and a
p edic i e model is p oposed.
2. Con ex
Single- amily dwellings accoun o 64 % o he o al Eu opean housing s ock, al hough in he
sou he n pa o he con inen , pa icula ly in I aly and Spain, mul i- amily buildings p e ail
[25]. These coun ies also ha e a la ge s ock o social housing, mos o which adop s mul i-
amily con igu a ions [26] buil o minimum cons uc ion quali y s anda ds [27]. While social
housing can be de ined in a numbe o ways, he e i is mean in he b oades sense: housing
ha ecei es some manne o go e nmen subsidy o assis ance [28].
Subs an ial imp o emen s will be needed in hese buildings in he yea s o come, bo h o
imp o e indoo habi abili y and o comply wi h he ene gy e iciency s anda ds o nea ly ze o
ene gy buildings laid down in he EPBD [29] and in he EU’s 20-20-20 objec i es [30]. This is
going o call o a huge e o in Spain and in o he Medi e anean coun ies o imp o e he
exis ing housing s ock by means o e ec i e, op imised ene gy in e en ion and e u bishmen
echniques. One o he main aims o e u bishmen policy should be o op imise a ailable
public and p i a e esou ces. Scan ly e ec i e e o i ing should be a oided o a ain genuine
ene gy sa ings while eaching he p ima y aim, he enhancemen o indoo com o .
These conside a ions a e applicable o a signi ican po ion o Eu ope: i s en i e Medi e anean
lank, om sou he n Po ugal o Tu key.
The a ea has a empe a e clima e, wi h mild empe a u es ha luc ua e only na owly
h oughou he yea . The win e s a e no e y cold (wi h empe a u es dipping below eezing
only excep ionally), and al hough wa m, he summe s a e no o id. While he clima e a ies
wi h he lay o he land, i is clea ly dis inc om condi ions in cen al and no he n Eu ope.
The a ea is classi ied as Köppen Csa and Csb [31].
Albei wi h egional a ia ions, mul i- amily building a chi ec u e and cons uc ion a e simila
ac oss he a ea [32] and di e widely om he solu ions in place in cen al and no he n
Eu opean coun ies, especially whe e social housing is conce ned [33]. Many o hese solu ions
a e a legacy inhe i ed om he nine een six ies and se en ies esiden ial cons uc ion boom
[34]. The a ea’s inhabi an s sha e mo e han a simila clima e, howe e , o social and cul u al
mo es a ound housing use and occupancy a e likewise analogous.
The ein o ced conc e e ame and con inuous conc e e slab cons uc ion ha cha ac e ises

medium- ise mul i-dwelling housing, especially social housing, in Medi e anean ci ies, esul s
in e y ai igh in e -s o ey compa men a ion. The ou e enclosu es ypically consis o b ick
o block (hea y- o medium-weigh walls), usually wi h an inne ca i y and se e al laye s o
plas e on he inside. Ligh weigh açades a e uncommon.
As a ule his social buildings has been concibed as Na u al Ven ila ed based on he
unde s anding o he clima e as benign, and gene ally lacks buil -in hea ing o ai condi ioning
sys ems. Use s mee hei hea ing needs wi h po able elec ical adia o s ha in mos cases
a e ene ge ically ine icien and unable o p o ide o an e en dis ibu ion o com o able
empe a u es [35]. No is his housing i ed wi h duc ed, whole-home cooling sys ems. ,
despi e he wa m summe s ha cha ac e ise he a ea. As a esul , homeowne s end o ins all
oom-size spli AC-uni s a pos e io i.
These buildings we e no gene ally i ed wi h mechanically con olled en ila ion sys ems
un il he en y in o e ec o he Technical Building Code in 2007 [2]. P io o ha da e,
ba h oom and ki chen en ila ion was based on he s ack e ec ac oss s a ic en s. Such
ex ac ion sys ems a e o en ine ec i e. Whe e indoo ai quali y in Medi e anean housing is
no pa icula ly poo i is hanks o he uncon olled in low o ai h ough he building
en elope (in il a ion) o olun a y en ila ion (manual ope a ion o windows).
En elope he mal-pe o mance is a e age in such buildings, wi h medium- o-low he mal mass
and wall he mal ansmi ance alues on he o de o 0.5-1.5 W/m2K. A i icial hea ing and
cooling a e needed o ensu e com o able indoo empe a u es in he win e and summe
unde no mal ou doo condi ions. The absence o buil -in HVAC is one o he majo easons o
he indoo empe a u e dipe sion ypically ound in his housing [35,36].
3. Me hods
Field esea ch o assess building ai igh ness was conduc ed be ween la e 2011 and ea ly 2013
in a numbe o loca ions in sou he n Spain (Andalusia) o ensu e co e age o he a ious in a-
egional clima es. P essu isa ion and dep essu isa ion we e measu ed wi h a Blowe Doo Tes
Sys em, la e supplemen ed wi h in a ed he mog aphy and smoke es s o iden i y speci ic
in il a ion pa hways.
3.1. Sampling
The s udy was conduc ed on 45 homes in se en open galle y- ype buildings in sou he n Spain.
This building ype, cha ac e ised by ou doo walkways on all loo s om which each indi idual
uni is accessed (Figu e 1), was chosen o a numbe o easons.
- I is e y commonly ound in Medi e anean esiden ial buildings, especially in social
housing.
- As mos o he en elope a ound each uni sepa a es i di ec ly om he ou doo s
a he han om indoo s ai cases o hallways, leakage ac oss elemen s o he han he
ou e enclosu es can be uled ou .
- Fo he abo e eason, he açade: loo a ea a io is la ge han in o he building ypes.
As a esul , he enclosu es ha e a g ea e impac on ai igh ness, ende ing hese
homes mo e ulne able o ai leakage han o he s.
Figu e 1: G ound plans o he buildings analysed
The buildings chosen we e loca ed in i e clima e zones wi hin he egion, wi h win e s anging
om e y mild (zone A) o cold (zone C) and summe s om wa m (zone 3) o e y wa m (zone
4). The zones we e classi ied by he clima e ca ego ies se ou in he Spanish building code. In
his scheme, a gi en building’s hea ing/cooling demand is compa ed o i s hypo he ical
demand in a e e ence loca ion o which a clima e p o ile has been es ablished based on
clima e se e i y de ined in e ms o deg ee-days and global sola adia ion, as desc ibed by de
la Flo [37] and he mean clima e da a shown in Appendix A o each ci y.
Figu e 1 shows he g ound plans o he buildings analysed and Table 1 lis s hei loca ion,
clima e zone and yea o cons uc ion.The d awings in Figu e 1 a e shown on a la ge scale in
Appendix A.
Building
Ci y
Clima e zone
Yea o
cons uc ion
Window
ype
Kind o
shu e s
Numbe o
s o eys Ven ila ion
Win e
se e i y
(WIN)
Summe
se e i y
(SUM)
CS1 Se ille B 4 2004 Hinged Rolle
blinds 6
Passi e en s in
ba h oom and
ki chen
CS2 Se ille B 4 2006 Sliding Rolle
blinds 6
Passi e en s in
ba h oom and
ki chen
CS3 Co doba
B 4 2007 Hinged Rolle
blinds 6
Passi e en s in
ba h oom and
ki chen
CS4 Malaga A 3 2007 Sliding Rolle
blinds 6
Passi e en s in
ba h oom and
ki chen
CS5 Alme ia A 4 2008 Sliding Rolle
blinds 6
Passi e en s in
ba h oom and
ki chen
CS6 G anada
C 3 2003 Hinged Rolle
blinds 5/7
Passi e en s in
ba h oom and
ki chen
CS7 Jaen C 4 2011 Sliding Rolle
blinds 6
Passi e en s in
ba h oom and
ki chen
Table 1: Loca ion, clima e zone, yea o cons uc ion and cha ac e isa ion o he buildings
s udied.
The uni s chosen in he se en buildings had simila loo a eas (Table 2) and cons uc ion
solu ions o enhance compa abili y and ensu e ha gene ally applicable conclusions on
pe o mance could be d awn.
The sample consis ed in ein o ced conc e e ame, con inuous slab buildings wi h ce amic o
cemen ile loo ing. Only ki chens and ba h ooms had con inuous suspended ceilings. Ve ical
en elope consis ed in ca i y-insula ed walls: ce amic-b ick wall ou side, ai chambe wi h
insula ion panel, and hin ce amic-b ick plas e ed inne shee ing. Pa y walls consis ed in hin
hollowb ick (app oxima ely 15 cm hick) pa i ions, plas e ed on bo h sides. All windows,
whe he side-hinged o ho izon al-sliding, had aluminium ames and double glazing (Table 1)
and we e i ed wi h buil -in olle blinds (‘monoblock’ sys em). The c oss-sec ion o a ypical
building solu ion is shown in Apendix A. The U- alues o walls and windows we e simila o
all he buildings. As hese uni s had no buil -in ai condi ioning o hea ing, he e was no
ho izon al o e ical duc wo k ou side he condi ioned space. This design minimises in e -
s o ey connec ions, which a e con ined o d ainage and en s in we ooms, no mally
embedded in he pa i ions.
Wi h cons uc ion sys ems based on la ge con inuous in e -s o ey slabs and a pauci y o
in e nal se ice sha s, ho izon al ai leakage was minimal in hese buildings, an ad an age
ein o ced by he ou doo loca ion o s ai cases and ele a o s. In open galle y con igu a ions,
in u n, dwellings a e p ima ily in con ac wi h he ou doo en i onmen , wi h scan
communica ion wi h o he building elemen s as e ical sha s o s ai s. The esul ing lack o
e ical ai pa hways [38] p ac ically elimina ed he s ack e ec . As all he uni s s udied we e
one-s o ey dwellings, he açade would cons i u e he p ima y ou e o po en ial in il a ion,
along wi h he en ila ion sha s, which could be eadily sealed and insula ed du ing he es s.
In pu sui o ep esen a i e sampling, a andom, s a i ied and p opo ional design was
adop ed. In all he buildings, a leas one o he uni s es ed was loca ed on he lowes s o ey,
one on an in e media e s o ey and a hi d unde he oo o s udy he possible impac o
ho izon al enclosu es and he s ack e ec esul ing om di e ences in heigh wi hin a gi en
building [39].
Table 2 lis s he cha ac e is ics o he uni s in he se en buildings, whe e Z ep esen s he
win e clima e, N he numbe o s o eys, S and V he mean loo a ea and olume, espec i ely,
FA he mean açade a ea, WA he mean window a ea and WP he mean window pe ime e .
The s anda d de ia ion (σ) o all he a o emen ioned a iables is also shown.
Building
Z
N
S V FA WA WP
Mean
(m2) σ
Mean
(m3) σ
Mean
(m2) σ
Mean
(m2) 
Mean
(m) σ
CS1 B
7 64.51 6.01
176 16.46
54.29 2.92 8.99 1.87
30.84 3.82
CS2 B
5 70.40 0.00
183 0.00 47.20 0.00 8.20 0.00
28.00 0.00
CS3 B
8 70.00 0.00
185 0.00 46.40 0.00 9.10 0.00
30.30 0.00
CS4 A
4 68.63 21.9
188 60.08
60.60 19.77
12.08 3.33
40.38 10.24
CS5 A
8 68.89 1.72
188 5.79 46.91 0.82 9.00 0.67
29.78 1.44
CS6 C
8 69.88 0.23
175 0.51 45.45 0.63 16.03 0.86
42.78 2.44
CS7 C
5 69.56 0.49
174 1.24 34.06 1.05 8.90 0.00
26.40 0.00
Mean 45
68.80 6.50
181 17.90
47.52 8.50 10.43 3.00
32.71 6.70
Table 2: Cha ac e is ics o he buildings selec ed
Al hough he uni s in he buildings selec ed had simila g oss loo a eas, hei mean window
a eas and pe ime e s and loo a eas di e ed subs an ially, a ou ing s a is ical analysis. Unde
hese a angemen s, he possible e ec o each o hese h ee pa ame e s on ai igh ness
could be de e mined.
3.2. P essu ised an ai igh ness measu emen s
The ield s udy consis ed in p essu isa ion/dep essu isa ion es s pe o med wi h a Blowe
Doo de ice supplemen ed wi h in a ed he mog aphy and smoke es s, a echnique gene ally
accep ed as alid by he scien i ic communi y [40]. The es was conduc ed wi h a Minneapolis
Blowe Doo TM sys em unning Tec i e so wa e. Spaces a ound all manne o openings we e
sealed wi h a sui able ma e ial. Building en elopes we e assessed o Spanish and Eu opean
s anda d UNE-EN 13829: 2002, Me hod B [41]. All he en s, mechanical and o he wise, in
ba h ooms and ki chens, including smoke ex ac o s in he la e , we e sealed and all o he
ou doo openings (windows) closed.
In addi ion o he Me hod B ials, a numbe o o he p essu isa ion/dep essu isa ion es s
we e conduc ed, in which he sealing se -up was a ied (Figu e 9). Al e na i e es s P5
(ba h ooms and ki chens sealed o ) and P6 (only ba h ooms sealed o ) p o ed o be o
pa icula in e es o loca e ai pa hways. The indings we e used o compa e he pe o mance
o he en elope enclosing d y ooms as opposed o he we ooms ha a e mo e p one o
uncon olled c acks and openings a ound se ice duc ways.
Each uni was ega ded o cons i u e a single olume, inasmuch as ene gy pe o mance was
p ac ically e en h oughou . The es consis ed in posi ioning he an on he ou side doo o
emo e (dep essu ise) o o ce (p essu ise) ai om o in o he dwelling o a posi i e o
nega i e p essu e o 50 Pa, and subsequen ly measu ing he ai low a 5 Pa in e als. All
es ima ed.
The cumula i e equency cu es and boxplo s in Figu e 6 show he esul s o he leas
signi ican di e ence (LSD) es , om which he h ee sub-g oups o ends iden i ied o n50,
he i e o 50 and he ou o p50 we e de i ed. In a-sample dependence and sampling e o s
we e uled ou .
Figu e 6: Cumula i e equency cu es and box and whiske plo s o he g oups iden i ied in
he se en buildings s udied
4.3. Dependence on clima e zone
The da ase did no appea o co obo a e he in ui i e assump ion ha as a ule, building
cons uc ion seeks g ea e ai igh ness in colde han in wa me a eas [44].
Hence, he ini ial hypo hesis was ha he olume-weigh ed in il a ion alues (n50) we e no
dependen upon he clima e zone whe e buildings we e loca ed.
The explana ion may be ha all he buildings in he sample we e loca ed in sou he n Spain and
despi e clima e di e ences om one a ea o ano he , sha ed cons uc ion solu ions,
a chi ec u e and usage. Ini ially, hen, loca ion was assumed o ha e no impac on he

p edic ion o pa ame e n50 (null hypo hesis).
Table 7 lis s he s a is ics o n50 by clima e zone and Figu e 7 ep oduces he cumula i e
equency cu es, likewise by clima e zone.
Z
N
umbe
o uni s
Mean
(h-1)
Median
(h-1)
S anda d
de ia ion
Coe icien o
a ia ion
Lowes
Highes
A
12
4.08
4.1
0.51
12.65%
3.20
5.30
B
20
6.59
6.7
1.64
24.93%
3.90
8.70
C
13
5.86
5.3
1.59
27.14%
3.70
8.30
Mean
45
5.71
5.30
1.73
30.36%
3.20
8.70
Table 7: S a is ics o n50 (h-1) by clima e zone
Figu e 7: Cumula i e equency cu es o n50 (h-1) by clima e zone
Las ly, a chi-squa e es was un (wi h a es alue o 65.97) o de e mine whe he he
assump ion o independence be ween n50 and clima e zone could be ejec ed. Since he p-
alue was g ea e han 0.01 (0.22), independence be ween he wo a iables could no be
uled ou a 99 % con idence. In ligh o hose esul s and he dis ibu ion o he alues, hen,
n50 was no necessa ily dependen upon clima e zone.
4.4. Mul iple linea co ela ion app oach o de eloping a p edic i e model
While acknowledging he s ochas ic na u e o hese alues due o hei associa ion wi h
cons uc ion p ocesses, a p edic i e model could be p oposed, subjec o a band o
unce ain y. This model may p o e use ul when o ecas ing expec ed beha iou and
es ablishing a amewo k o compa ison.
Pea son’s bi a ia e co ela ion coe icien s we e ound o he geome ic pa ame e s
associa ed wi h he n50 alues o asce ain he exis ence o o he wise o in e - a iable
dependence. The co ela ion coe icien s anged om -0.98 o 0.75. Fo V, hey deno ed a
s a is ically signi ican p- alue o 0.003, which was somewha highe o WA and e en highe
o WP.
In ligh o his possible dependence, which e en i p esen would no p o e causali y, a se ies
o ANOVAs and linea eg ession analyses we e conduc ed o de e mine whe he in il a ion in
he g oup o dwellings s udied could be p edic ed wi h a model based on hei geome ic
cha ac e is ics, as he ini ial hypo hesis would appea o indica e.
A ac o ial analysis was un on he independen a iables in he linea eg ession model. This
p ocedu e, which excludes he leas signi ican a iables in he model, educed he numbe o
he ou wi h he highes s a is ical signi icance (Figu e 8). All six a iables, S, V, FA, WA, WP
and Z, we e included in he ac o ial analysis o ob ain a be e i . As expec ed, FA and Z
(clima ic zone) exhibi ed low s a is ical signi icance and we e dis ega ded.
Figu e 8: Componen + esidual plo o n50 and di e en pa ame e s
The p edic i e model p oposed can be exp essed as shown below (Equa ion 3), wi h he
coe icien s lis ed in Table 8.
𝑛 = 𝛼 × 𝑆 − 𝛽× 𝑉 − 𝛾 × 𝑊𝐴 + 𝛿 ×𝑊𝑃 Eq.(3)
Pa ame e Coe icien Es ima e S anda d
e o
S
𝛼
1.019 0.101
V
𝛽
-0.353 0.044
WA
𝛾
-1.763 0.360
WP
𝛿
0.549 0.161
Table 8: P edic i e linea eg ession model: coe icien s
The model p oposed a o ded a good app oxima ion, explaining 98.12 % o he a ia ion in he
dependen a iable (n50). Al hough he es ima ed e o was signi ican (0.849 h-1), i p o ided a
mo e accu a e i han he s anda d de ia ion o he measu emen s (app oxima ely hal ing he
wid h o he luc ua ion band). Ba ing any p io knowledge o geome ic ac o s, he bes
p edic ion was ha he ai low in a dwelling would be a ound 5.721.74 h-1 (Table 5).
This model is p oposed o p edic en elope ai - igh ness in he social housing s ock buil in
sou he n Spain, based on a sample o 45 galle y- ype buildings. The model was de eloped
using mul iple linea eg ession.
4.5. Typical ai leakage pa hways
The p ima y ai leakage pa hways [47] we e iden i ied wi h in a ed he mog aphy o smoke
du ing he Blowe Doo es . These es s we e un du ing he win e wi h he hea ing on and
he ooms dep essu ised (ISO 18434-1: 2008) [48].
In he building ype s udied, he en elopes in all he uni s we e bounded in pa by he
ou doo s and in pa by o he uni s. Buildings we e es ed in de ail, sealing o he pa s o he
uni whe e ai could low in om he pa y walls, o de e mine he pe cen age o in il a ion
a ibu able o he ou doo enclosu e (Figu e 9).
Figu e 9: Sealing p o ocols o loca e ai leakage pa hwaysThe indings o he a ious ypes o
es (p o ocols 1 o 9) we e analysed o de e mine he ela i e weigh o each dwelling
componen in o e all ai in il a ion. The end iden i ied in his sui e o es s ep oduced he
beha iou obse ed in he de ailed analysis conduc ed on building CS2, albei wi h a signi ican
a e o a ia ion (Figu e 10).
In il a ion be ween adjacen uni s was ound o be scan ly signi ican : jus 3.6 % o he o al.
Figu e 10: Dwelling ai igh ness, con ibu ion by componen
These es s also showed ha ai leakage ook place p ima ily ac oss he e ical en elope,
pa icula ly in d y ooms whe e mos o he windows a e loca ed. None heless, plumbing,
elec ical and en ila ion duc ways also con ibu ed o in il a ion. In con as , no signi ican
in il a ion was ound ac oss ho izon al elemen s o pa y walls, no we e di e ences obse ed
be ween unde - oo uni s and o he dwellings.
As would be expec ed in açades o he ype ound on hese buildings, he main ai pa hways
we e a he in e ace be ween windows and opaque enclosu es. The wide a iabili y eco ded
in he es alues may ha e been he esul o he manual cons uc ion p ocedu es deployed in
his ype o buildings, inducing di e ences in dwellings in one and he same building. Hens
iden i ied he cause o such a ia ions o be poo wo kmanship [49]. In his s udy, he indings
o di e en window ypes we e compa ed o co obo a e ha conclusion. No ela ionship
was ound be ween window closu e ype (hinged o sliding) and in il a ion a es, whe eas
beha iou was obse ed o a y widely wi hin each ype o window.
The whiske box in Figu e 11 shows he p opo ion o leakage a ibu able o he we (ki chen
and ba h oom) and d y ooms. No e ha o e hal o he o al leakage was associa ed wi h he
en elope a ound d y ooms (in he Me hod A es , i.e., wi h he dwelling in use). None heless,
hese alues we e subjec o 17 % a iabili y. The p opo ion a ibu able o we ooms was
lowe , and highe in ba h ooms ( om unde 10 % o 20 %) han in ki chens (unde 7 %).
Me hod B, i.e., conside ing measu emen s o he en elope alone (dis ega ding en s and
pe manen openings), yielded he ollowing alues: 72.69 % o he d y oom en elope,
18.85 % o he ba h oom en elope and 8.71 % o he ki chen en elope.

Figu e 11: P opo ion o in il a ion accoun ed o by oom ype
4.6. Compa ison o o he egions
The p esen indings we e compa ed o he esul s epo ed o simila s udies on mul i- amily
dwellings conduc ed in sou he n Eu ope. Resea ch ocusing on single- amily homes o wi h a
small numbe o cases was dis ega ded. When he in o ma ion was p o ided on he da e o
cons uc ion, he da a we e b oken down o sepa a e analysis o he dwellings buil wi hin
mo e o less he same ime ame as he ones s udied he e, in pa icula uni s buil o
legisla ion de i ing om he EPBD. The n50 alues a e ep oduced in Figu e 12, which also
shows he in e als ha de ine he low, medium and high ai igh ness ca ego ies se ou in
Eu opean and in e na ional s anda d EN-ISO 137900 [50] and he numbe o es s un o
es ablish he ep esen a i eness o he alues.
Figu e 12: Values o Eu opean mul i- amily dwellings
The igu e shows in o ma ion on he ollowing da ase s: I alya, o e all o al [19]; I alyb,
dwellings buil in 2001-2010 [19]; Po ugala [20]; Po ugalb, e u bished dwellings [21];
Po ugalc, non- e u bished dwellings [21]; N-Spaina, dwellings buil in 1979-2007 [22]; N-
Spainb, EPBD dwellings [22]; N-Spainc, mainly EPBD dwellings [23]; S-Spain, p esen s udy.
The abo e analysis e ealed ha he S-Spain n50 alues, which lay p edominan ly in he
medium band, we e in line wi h he indings o simila uni s in he Medi e anean a ea. These
alues we e lowe han in o he newe buildings in no he n and cen al Spain, buil a e he
ansposi ion o he EPBD, and lowe as well han I alian buildings e ec ed mo e ecen ly. The
lowes n50 alues (high ai igh ness) we e ound o he oldes buildings (I alya and N-Spaina),
while he highes (low ai igh ness) we e eco ded in he Po uguese dwellings (Po ugala,b,c)
e en a e e u bishmen . The sca e obse ed in he n50 alues, while signi ican , was simila
o he indings epo ed o o he Medi e anean egions.
The n50 alues eco ded he e we e highe han equi ed in no he n and cen al Eu opean
coun ies’ design s anda ds, whe e housing ai igh ness has long been he objec o egula ion.
The e e ence n50 alue in Ge many and No way is 3 h-1 and in Finland, 4 h-1. While some o he
dwellings in his s udy exhibi ed alues nea ly as low as hose benchma ks, o he s, in keeping
wi h esul s epo ed by he ea lie s udies analysed he e, doubled he sco e.
5. Conclusions
The ai pe meabili y eadings a 50 Pa o 45 mul i- amily uni s wi h open galle ies buil in
sou he n Spain a e 2000 a e aged 5.72 h-1. This alue was highe han ene gy e iciency
s anda ds and ecommenda ions o Eu opean buildings, al hough simila o he esul s
epo ed o no he n Spain and o he sou he n Eu opean coun ies such as I aly and Po ugal.
The s udy e ealed a wide sca e in he ai igh ness alues eco ded and highligh ed he
di icul ies in ol ed in p oposing an accu a e p edic i e model. Tha he sca e o pa ame e
n50 was g ea e han he mean alues is an indica ion o laws in building en elope con inui y.
The alues o he en i e sample anged om 3.2 o 8.7 h-1 and in some cases he high alue in
a gi en building doubled he low alue in he same building. The alue o exponen n anged
om 0.56 o 0.61 in all he measu emen s pe o med. Values on he o de o 0.6 a e
associa ed wi h leakage a in e aces be ween openings and hei opaque su ounds.
Despi e he unce ain y gene a ed by he a isanal açade cons uc ion me hods ypical o his
housing s ock in sou he n Eu ope, he model de eloped was able o p edic he deg ee o
ai igh ness in hese dwellings wi h accep able accu acy. Tha model may be pa icula ly use ul
o compa e ai igh ness alues o buildings e ec ed in he same ime ame. I should be
ega ded as a egional ool, ailo ed o sou he n Eu opean cons uc ion and a chi ec u e ha
aims o be an al e na i e o he mo e uni e sal models de eloped o da e. In addi ion, he
p obabilis ic model o beha iou p oposed he e is applicable o he assessmen and
cons uc ion o models o analysing he housing s ock.
In ligh o he di icul ies desc ibed o build a eliable p edic i e model, no design solu ions o
buildings and hei en elopes ha would enhance ai igh ness can be pu o wa d.
Consequen ly, e o s should ocus on cons uc ion pe se, seeking mo e assembly-sensi i e
solu ions and ein o cing he con ol o wo kmanship in openings and in e aces wi h he
opaque plane. Tes s ha measu e in il a ion a es mus also be conduc ed. No in il a ion was
obse ed a he in e ace be ween he deck slab and e ical membe s in he dwellings es ed,
wi h no pe cep ible di e ences be ween unde - oo and in e media e uni s, no was any
de ec ed a pa y walls. The e o e, imp o emen s should ocus on he join s a ound açade
openings.
This s udy e ealed he need o indi idualise ai igh ness su eys o be able o iden i y he
main ai in il a ion pa hways ac oss he en elope and hence apply p o ocols bo h o
cons uc ion quali y and asce ainmen o he condi ion o buildings wi h a iew o ene gy
e u bishmen . P essu isa ion echniques ha e p o en o be a cos -e ec i e app oach o
assessing and imp o ing he ene gy pe o mance o buildings in gene al and esiden ial
buildings in sou he n Eu ope in pa icula .
Gi en he impac o in il a ion on ene gy demand and consump ion in esiden ial buildings,
ai igh ness should be ega ded as an ins umen al ac o o he ene gy assessmen o he
exis ing housing s ock and i s e u bishmen o mee nea ly ze o ene gy building (NZEB)
aspi a ions in he sou he n Medi e anean.
Acknowledgemen s
This s udy was unded by he Spanish Minis y o Economy and Compe i i eness unde p ojec
BIA2012-39020-C02-01 and BES-2013-063097 suppo .
Appendix A
Figu e A.1: Plan iew o CS1.
Figu e A.2: Plan iew o CS2.
CS5 8 24.46
7.64 2.76 11.30% 19.80 28.50 8.70 0.04
CS6 8 19.48
11.33 3.37 17.28% 14.60 26.60 12.00
2.02
CS7 5 50.62
11.08 3.33 6.57% 47.80 54.50 6.70 -1.47
Mean 45 33.26
187.38 13.69 41.15% 14.60 55.90 41.30
-1.91
Table B.3: S a is ics o p50 (m3/hm)
P edic i e
model
R
2
:
0.981
Adjus ed R
2
: 0.981
Es ima ed s anda d e o
0.84
Table 11: P edic i e linea eg ession model: summa y
Re e ences
[1] Minis e io de Indus ia, Tu ismo y Come cio. PROYECTO SECH-SPAHOUSEC: Análisis del
consumo ene gé ico del sec o esidencial en España IDAE, 2011.
[2] Real Dec e o 314/2006: España, Real Dec e o 314/2006, po el que se ap ueba el Código
Técnico de la Edi icación, de 17 de ma zo. Bole ín O icial del Es ado, 28 de ma zo de 2006, n.
74, p. 11816.
[3] J.J. Send a, S. Domínguez, A.L. León, J. Na a o, e al, P oyec o E icacia: Op imización
ene gé ica en la i ienda colec i a, Uni e sidad de Se illa, 2011
[4] S. Domínguez, J.J. Send a, A.L. León, P.M. Esqui ias, Towa ds Ene gy Demand Reduc ion in
Social Housing Buildings: En elope Sys em Op imiza ion S a egies, Ene gies 5 (2012) 2263-
2287.
[5] A.L. León, S. Domínguez, M.A. Campano, C. Ramí ez-Balas, Reducing he Ene gy Demand o
Mul i-Dwelling Uni s in a Medi e anean Clima e Using Sola P o ec ion Elemen s Ene gies 5
(2012) 3398-3424.
[6]Di ec i a 2012/27/UE del Pa lamen o Eu opeo y del Consejo, de 25 de oc ub e de 2012,
ela i a a la e iciencia ene gé ica, po la que se modi ican las Di ec i as 2009/125/CE y
2010/30/UE, y po la que se de ogan las Di ec i as 2004/8/CE y 2006/32/CE
[7] M.H. She man, W.R. Chan, Building Ai igh ness: Resea ch and P ac ice, E nes O lando
Law ence Be keley Na ional Labo a o y, 2004.
[8] W. R. Chan, W. W. Naza o , P. N. P ice, M. D. Sohn, A. J. Gadgil, Analyzing a da abase o
esiden ial ai leakage in he uni ed s a es, A mosphe ic En i onmen 39 (2005) 3445-3455.
[9] J. McWilliams, M. Jung, De elopmen o a Ma hema ical Ai -Leakage Model om Measu ed
Da a, E nes O lando Law ence Be keley Na ional Labo a o y, 2006.
[10] M.H. She man, D. J. Dicke ho , Ai Tigh ness o U.S. Dwellings, E nes O lando Law ence
Be keley Na ional Labo a o y, 2002.
[11] M. H. She man, N. Ma son, Ai Tigh ness o New U.S. Houses: A P elimina y Repo ,
Depa men o Ene gy, E nes O lando Law ence Be keley Na ional Labo a o y, 2002.
[12] W.R. Chan, J. Joh, M.H. She man, Analysis o ai leakage measu emen s o US houses,

Ene gy and Buildings 66 (2013) 616–625.
[13] J. Jokisalo, J. Ku ni ski, M. Ko pi, T. Kalamees, J. Vinha, Building leakage, in il a ion, and
ene gy pe o mance analyses o Finnish de ached houses. Building and En i onmen 44
(2009) 377-387.
[14] D. Sinno , M. Dye , Ai igh ness ield da a o dwellings in I eland. Building and
En i onmen 51 (2012) 269-275.
[15] W. Pan, Rela ionships be ween ai igh ness and i s in luencing ac o s o pos -2006 new-
build dwellings in he UK, Building and En i onmen 45 (2010) 2387-2399.
[16] R. Ca ié, R. Jobe M. Fou nie , S. Be haul , Pe meabili e´ a` l’ai de l’en eloppe des
ba imen s. Gene ali es e sensibilisa ion, CETE de Lyon, 2006.
[17] F. R. Ca ié, B. Rosen hal, Augus An o e iew o na ional ends in en elope and duc wo k
ai igh ness, AIVC In o ma ion Pape 29 (2008).
[18] A. S akianaki, K. Pa lou, M. San amou is, I. Li ada, M. Assimakopoulos, P. Man as, e al.,
Ai igh ness measu emen s o esiden ial houses in A hens, G eece, Building and En i onmen
43 (2008) 398-405.
[19] F.R. d´Amb osio Al ano, M. DellÍsola, G. Ficco, F. Tassini, Expe imen al analysis o
ai igh ness in Medi e anean buildings using he an p essu iza ion me hod, Building and
En i onmen 53 (2012) 16-25.
[20] N. M.M. Ramos, R. M.S.F. Almeida, b, A. Cu ado, P. F. Pe ei a, S. Manuel, J. Maia,
Ai igh ness and en ila ion in a mild clima e coun y ehabili a ed social housing buildings –
Wha use s wan and wha hey ge , Building and En i onmen 92 (2015) 97–110.
[21] M. Pin o, J. Viegas, V.P. de F ei as, Ai pe meabili y measu emen s o dwellings and
building componen s in Po ugal, Building and En i onmen , 46 (2011) 2480-2489.
[22] A. Meiss, J. Feijó-Muñoz, The ene gy impac o in il a ion: a s udy on buildings loca ed in
no h cen al Spain, Ene gy E iciency 8 (2015).
[23] A. Jimenez Tibe io, P. B anchi, A s udy o ai leakage in esiden ial buildings,
New Concep s in Sma Ci ies: Fos e ing Public and P i a e Alliances, (2013) 1-4.
[24] M. I. Mon oya, E. Pas o , F. Remi Ca ie, G. Guyo , E. Planas, Ai leakage in Ca alan
dwellings: De eloping an ai igh ness modeland leakage ai low p edic ions, Building and
En i onmen 45 (2010) 1458–1469.
[25] K. Dol, M. Ha ne , Housing s a is ics in he Eu opean Union, Del Uni e si y o
Technology, 2010.
[26] D. Pennes i, The ene gy and en i onmen al equali ica ion o pos -wa housing:
p oblema ics and inno a i e solu ions o he building en elope, P oceedings o SB13 P ague
on Sus ainable Building and Re u bishmen o Nex Gene a ion, 2013.
[27] BPIE, Buildings Pe o mance Ins i u e Eu ope, Eu ope’s buildings unde he mic oscope,
2011.
[28] A. T. Ca swell, The Encyclopedia o Housing, Second Edi ion Uni e si y o Geo gia, 2012.
[29] Di ec i e 2010/31/EU o he Eu opean Pa liamen and o he Council o 19 May 2010 on
he ene gy pe o mance o buildings ( ecas ).
[30] Eu ope’s 20-20-20 objec i es
[31] M. Ko ek, J. G iese , C. Beck, B. Rudol , and F. Rubel, Wo ld Map o he Köppen-Geige
clima e classi ica ion upda ed, Me eo ol. Z., 15 (2006) 259-263.
[32] S. Fe a i, V. Zano o, Building Ene gy Pe o mance Assessmen in Sou he n Eu ope,
Edi o Sp inge , 2015.
[33] J. Allen, J. Ba low, J. Leal, T. Malou as, L. Pado ani, Housing and Wel a e in Sou he n
Eu ope, Wiley-Blackwell, 2004.
[34] D. Ma eis e al, A o dable Housing Housing P axis o U ban Sus ainabili y, Ipe es o
Edizioni, Ve ona Good, 2008.
[35] J.J. Send a, S. Domínguez-Ama illo, P. Bus aman e, A.L. León, Ene gy in e en ion in he
esiden ial sec o in he sou h o Spain: Cu en challenges, In o mes de la Cons ucción,
65(2013) 457-464.
[36] A.L. León, S. Muñoz, J. León, P. Bus aman e, Moni o ización de Va iables
Medioambien ales y Ene gé icas en la Cons ucción de Vi iendas P o egidas: Edi icio C os-
Pi o ecnia en Se illa, In o mes de la Cons ucción 62 (2010) 67-82.
[37] F. J. S. De la Flo , S. A. Domínguez, J. L. M. Félix, R. G. Falcón, Clima ic zoning and i s applica ion o
Spanish building ene gy pe o mance egula ions, Ene gy and Buildings 40 (2008) 1984–1990.
[38] W.R. Chan, J. Joh, and M. H. She man, Analysis o Ai Leakage Measu emen s om
Residen ial Diagnos ics Da abase, E nes O lando Law ence Be keley Na ional Labo a o y,
2012.
[39] W. Wal he , B. Rosen hal, Ai igh ness Tes ing o La ge and Mul i- amily Buildings in an
Ene gy Pe o mance Regula ion Con ex, In o ma ion Pape no P165, EPBD buildingspla o m,
2009.
[40] H. E. Feus el, Measu emen s o ai pe mabili y in mul izone buildings, Ene gy and
Buildings, 14 (1990) 103-116.
[41] EN 13829:2002 The mal pe o mance o buildings - De e mina ion o ai pe meabili y o
buildings - Fan p essu iza ion me hod. (ISO 9972:1996, modi ied)
[42] F. R. Ca ié, V. Lep ince, Unce ain ies in building p essu isa ion es s due o s eady wind, Ene gy
and Buildings, 116 (2016) 656–665.
[43] D. E he idge, M. Sandbe g, Building en ila ion. Theo y and measu emen , Chiches e ,
England: John Wiley & Sons, 1996.
[44] UNE-EN 15242:2007 Ven ila ion o buildings - Calcula ion me hods o he de e mina ion
o ai low a es in buildings including in il a ion.
[45] M. O me, M. Liddamen , A. Wilson, An Analysis and Da a Summa y o he AIVC's
Nume ical Da abase, Technical No e 44, Ai In il a ion and Ven ila ion Cen e, Co en y,
1994.
[46] AIVC Guide 5, Ven ila ion Modelling Da a Guide, 2002.
[47]EN 13187:1998 The mal Pe o mance O Buildings - Quali a i e De ec ion O The mal
I egula i ies In Building En elopes - In a ed Me hod
[48] ISO 18434-1:2008. Condi ion moni o ing and diagnos ics o machines. The mog aphy. Pa
1: Gene al p ocedu es.
[49] H. Hens, Dwelling ai - igh ness in a 55 yea s old es a e. In: 9 h No dic Symposium on
Building Physics (NSB 2011), Tampe e, Finland, 2011.
[50] EN-ISO 13790:2008 Ene gy pe o mance o buildings. Calcula ion o ene gy use o space
hea ing and cooling.