Academic Edi o : Paulo San os
Recei ed: 21 Decembe 2024
Re ised: 13 Janua y 2025
Accep ed: 22 Janua y 2025
Published: 25 Janua y 2025
Ci a ion: Apa icio-Ruiz, P.;
Ragel-Bonilla, J.C.; Ba badilla-Ma ín,
E.; Guadix, J. In eg a ion o he
Adap i e App oach in HVAC Sys em
Ope a ion: A Case S udy. Appl. Sci.
2025,15, 1243. h ps://doi.o g/
10.3390/app15031243
Copy igh : © 2025 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/).
A icle
In eg a ion o he Adap i e App oach in HVAC Sys em
Ope a ion: A Case S udy
Pablo Apa icio-Ruiz * , J. C. Ragel-Bonilla , Elena Ba badilla-Ma ín and José Guadix
G upo de Ingenie ía de O ganización, Escuela Técnica Supe io de Ingenie ía, Uni e sidad de Se illa,
Camino de los Descub imien os S/N, 41092 Se ille, Spain; [email p o ec ed] (J.C.R.-B.); [email p o ec ed] (E.B.-M.);
[email p o ec ed] (J.G.)
*Co espondence: [email p o ec ed]
Abs ac : Al hough di e en in es iga ions ha e been ca ied ou on he analysis o adap-
i e he mal com o in na u ally en ila ed buildings, ewe ha e ocused on mixed mode
ope a ion. Mo eo e , he e is limi ed esea ch as o he implemen a ion o adap i e com-
o models in o he con ol sys em o buildings. The e o e, his pape in es iga es how he
applica ion o a se poin based on adap i e com o con ol (ACC) would a ec occupan s’
com o conside ing mixed mode ope a ion and based on he esul s o a longi udinal
ield s udy in an academic o ice building o a e ia y educa ional ins i u ion in sou he n
Spain. The manusc ip analyses he The mal P e e ence Vo e o e 12 mon hs in a mixed
mode oom wi h an HVAC sys em whose se poin is adjus ed wi h a p e iously calcula ed
adap i e algo i hm o he building. Fo ha , a ho ough analysis was conduc ed in which
use s iden i ied si ua ions ega ding he mal com o and he ope a ion o he condi ioning
sys em was collec ed. The esul s indica e ha i is possible o de elop adap i e com o
models ha ensu e he he mal well-being o occupan s. Mo eo e , his s udy highligh s
he need o u he esea ch o assess he implica ions o ACC in e ms o com o and
ene gy consump ion as well as add essing he u u e imp o emen s and he limi a ions o
he wo k ca ied ou .
Keywo ds: ield s udy; adap i e he mal com o ; mixed mode buildings; adap i e com o
con ol
1. In oduc ion
An essen ial unc ion o buildings is o p o ide com o able indoo en i onmen s,
which o en en ails high ene gy consump ion. In pa icula , he in e ac ion o occupan s
wi h building con ols has been ecognized as an impo an ac o ha in luences commonly
obse ed signi ican di e ences be ween ac ual and p edic ed ene gy consump ion in build-
ings [
1
]. As o he ield s udy o he mal com o , esea ch has adi ionally adop ed wo
main app oaches [
2
]: he hea balance model [
3
] and he adap i e model [
4
–
6
]. In pa icula ,
wi h espec o he la e , he main in e na ional s anda ds [
7
] es ablish admissible indoo
empe a u e h esholds in buildings wi hou mechanical en ila ion based on com o s ud-
ies and o e he pas en yea s, nume ous ield s udies ha e been conduc ed along hese
lines [
8
–
10
]. The e o e, adap i e com o has es ablished i sel as a majo esea ch ield,
de ining s anda ds ha ela e accep able indoo empe a u es o ou doo empe a u es.
These p oposals e lec how clima e in luences ou he mal p e e ences and how body eg-
ula ion mechanisms d i e he mo egula o y beha io s. As J.F. Nicol poin ed ou , i could
be summa ized in he adap i e p inciple: “I a change occu s such as o p oduce discom o ,
people eac in ways which end o es o e hei com o ” [
11
]. J.F. Nicol also highligh s ha he
Appl. Sci. 2025,15, 1243 h ps://doi.o g/10.3390/app15031243
Appl. Sci. 2025,15, 1243 2 o 17
adap i e app oach “ o explain he ange o empe a u es people ound com o able in buildings
wi h he a iable indoo empe a u e’s cha ac e is ic o na u ally en ila ed buildings” o e s a
complemen a y solu ion o he challenge o hea balance.
Since hen, his app oach has gained ecogni ion, consolida ing in o s anda ds such
as ANSI/ASHRAE 55:2023 [
2
], he e e ence o he mal com o in No h Ame ica, and
he Eu opean s anda d UNE-EN 16798-1:2020 [
4
], which co e s bo h he mal com o and
o he indoo en i onmen al pa ame e s. These adap i e app oaches a e mainly applied o
he assessmen o he he mal en i onmen in na u ally en ila ed (NV) buildings whe e,
e.g., he occupan s ha e access o ope able windows and ha e some eedom o adap he
insula ion o hei clo hing. Coun ies such as he Ne he lands [
12
] and B azil [
13
,
14
] ha e
also included adap i e models in hei na ional egula ions and s anda ds.
U ilizing na u al en ila ion in buildings can lead o conside able elec ical ene gy
sa ings o cooling and ans [
15
]. While adap a ion is desc ibed as he g adual dec ease
in he eac ion o en i onmen al s imula ion, and can be beha io al (such as clo hing [
16
],
windows, ans), physiological (acclima iza ion), and psychological (expec a ion) [
17
], in a
p ac ical se ing, a ia ions in he pe cep ion o he he mal en i onmen we e iden i ied
be ween occupan s o na u ally en ila ed o ee- unning, ully ai -condi ioned and mixed-
mode o hyb id buildings [17].
Adap i e com o app oach ha e a isen as a sus ainable and esilien solu ion ha ,
acco ding o Bien enido-Hue as e al. [
18
], could educe he incidence o uel po e y, in
con as o he s a ic model ha elies exclusi ely on ac i e cooling. Ha old Wilhi e [
19
]
desc ibed ully condi ioned buildings as “Buildings as com o capsules” whe e he hea ing,
en ila ing and ai -condi ioning (HVAC) sys em is seen as one o he ools con ibu ing o
he “homogeniza ion o people”, esul ing in “ he mal mono ony”, al hough ea ly HVAC
enginee s s i ed o a oid c ea ing mono onous he mal en i onmen s [
20
]. The building
sec o is looking o sus ainable building solu ions, no only wi h passi e hea dissipa ion
capabili ies [
21
] o om an ae o he mal poin o iew [
22
], bu also passi e ac ions a e
expec ed o con ibu e o pe iods wi h na u al en ila ion, such as he adap a ion o he
clo hing insula ion le el [23].
The applica ion o adap i e com o con ol (ACC) as an al e na i e o ixed se poin
sys ems (o he pa adigm o a single empe a u e se poin in each clima ic season) would
be jus i ied by Foun ain e al. [
24
] who explain ha “a e epea ed exposu e o a ia ion in
en i onmen al condi ions,a pe son’s expec a ions o hose condi ions may become mo e elaxed—
e en an icipa o y o empo al changes” (p. 181). This idea p o ides insigh in o why adap i e
com o con ol is a alid al e na i e o ixed se poin sys ems, as i sugges s ha people’s
expec a ions o en i onmen al condi ions can be adjus ed and become mo e lexible o e
ime. This adap i e capaci y comes om epea ed exposu e o such en i onmen al a ia-
ions, which os e s mo e lexible expec a ions and e en posi i e an icipa ion o empo a y
changes in he en i onmen . This ein o ces he idea ha an adap i e con ol sys em,
by allowing con olled a ia ions in indoo condi ions a he han main aining a ixed
se poin , can no only be mo e ene gy e icien , bu also p omo e a na u al he mal com o
expe ience aligned wi h human expec a ions. As highligh ed in [
25
] (p. 16), he applica ion
o adap i e com o sys ems has been shown o imp o e ene gy sa ings, enhance esilience
o clima e change and inc ease he mal sa is ac ion and well-being.
The adap i e com o con ol app oach could be seen as a solu ion o he phenomenon
desc ibed by De Dea [
26
] in which occupan s o buildings wi h adi ional HVAC sys-
ems de elop an expec a ion o he mal cons ancy due o he s ic main enance o ixed
empe a u es. This expec a ion, as he au ho poin s ou , causes e en small luc ua ions
in empe a u e o be pe cei ed nega i ely, gene a ing a hype sensi i i y ha leads o
equen complain s. Au oma ed con ols ha adjus HVAC se poin s acco ding o he
Appl. Sci. 2025,15, 1243 3 o 17
ou doo clima e and seasonal a ia ions can be an e ec i e solu ion. In ac , Deuble and De
Dea [
27
] sugges ha inc easing indoo clima e a iabili y is key o sus ainable building
design s a egies.
In his con ex , i becomes ele an o analyze mixed mode (MM) buildings, which
combine na u al en ila ion and HVAC sys ems. Al hough he e is no consensus in he
li e a u e on con ol s a egies o mixed mode buildings [
28
], his s udy p oposes he
analysis o a eal case o in eg a ion o he adap i e app oach in he ope a ion o HVAC
sys ems o in es iga e he implemen a ion o a eal mixed mode si ua ion. Gi en ha an
MM building does no ha e o ope a e as a ully condi ioned building he whole ime, and
al hough some au ho s conside o analyze ha hey mus ope a e in he HVAC mode
abo e a ce ain empe a u e [
27
], he eali y is ha an MM building can be a ee-beha ing
building. The hypo hesis analyzed in his s udy is ha ope a ing a signi ican pe cen age o
he ime in passi e mode, wi h he HVAC sys em se ing as a empo a y backup adap i e
measu e, would be a easible solu ion. This hypo hesis is suppo ed by ecen indings [
29
],
al hough only 20% o case s udies on adap i e se poin s in Building Au oma ion Sys ems
(BAS) ha e been implemen ed in eal-wo ld cases (in Eu opean coun ies, his accoun s
o 36% o he 20%), and despi e he ac ha he ope a ional s a egies in mixed mode
buildings a e sui able o such app oaches (48% a e ca ego ized as changeo e s a egies).
The e o e, he aim o his pape is o expand upon hese s udies, add essing he gap in
li e a u e by analyzing eal-wo ld easibili y. Mo eo e , in ecen s udies such as [
30
], i
has been shown ha in si ua ions when ex eme ou doo empe a u es en ail he use o
HVAC sys ems, adjus ing adap i e se poin s can lead o a educ ion o app oxima ely
35% in cooling demand, he eby con ibu ing o bo h ene gy sa ings and g ea e he mal
com o . The use migh ely on his measu e (HVAC usage) wi hou being awa e o he
building’s se poin (due o i s adap i e na u e), hus, exp essing i s need o imp o e he
empe a u e on an ad hoc basis. Consequen ly, he se poin , as i is adap i e, would ollow
he empe a u e p oposed by he adap i e com o algo i hm, wi h he consequen ene gy
sa ings ha his would en ail, compa ed o a con inuous condi ioning ope a ion based on
a ixed se poin .
In ela ion o he numbe o ield s udies ca ied ou in MM buildings, al hough
a g owing end o esea ch ocused on his a ea can be de ec ed in he li e a u e, he
numbe is lowe han in na u ally en ila ed o ully condi ioned buildings [
29
]. In e ms o
clima ology, building ype and HVAC sys ems, a highe olume o wo k can be iden i ied
in MM buildings in oceanic and humid sub- opical clima es, comme cial o business
buildings and change-o e and concu en ope a ions s a egies, espec i ely.
The e is also a wide a ie y o con ol algo i hms de eloped o MM buildings. Al-
hough some s udies p opose ule-based con ol, conside ing a iables such as ou doo
empe a u e, indoo empe a u e, he mal com o o adap i e beha io s, among o he s,
o he esea ch ocuses on complex compu a ional con ols based on uzzy logic o machine
lea ning algo i hms, highligh ing he di e si y o app oaches in his ield.
Addi ionally, he e is a need o his ype o case s udy, since, as indica ed by [
29
], 80%
o he case s udies a e based on simula ions, which poin s o he need o analyze a la ge
numbe o eal implemen a ions.
This a icle is s uc u ed as ollows: Sec ion 2desc ibes he ma e ial and me hods,
desc ibing he case s udy as well as including p e ious knowledge, ele an backg ound
in o ma ion and miles ones ha ha e enabled he de elopmen o he cu en s udy, as well
as he gene al con ex ual condi ions and clima ology o he case s udy. This sec ion also
discusses he da a collec ion p ocess and he p elimina y analyses conduc ed and p esen s
he p oposed me hodology. Sec ion 3shows he esul s ob ained a e one yea o p ac ical
Appl. Sci. 2025,15, 1243 4 o 17
implemen a ion o he in eg a ion o he adap i e algo i hm in o he con ol sys em and
discusses he indings. Finally, Sec ion 4p esen s he conclusions o he wo k.
2. Ma e ials and Me hods
2.1. Case S udy
The s udy was ca ied ou in a building loca ed in Se ille (37
◦
N, 5
◦
W), in he sou he n
egion o Spain. The clima e is classi ied as Csa (ho summe Medi e anean clima e)
acco ding o he Köppen–Geige classi ica ion [
31
], and as B4 acco ding o he Spanish
Building Technical Code [
32
]. The main cha ac e is ics o he clima e in Se ille a e ho and
d y summe s, and mild, humid and empo a y ainy win e s. I is also cha ac e ized by
one o he highes a e age annual empe a u es in Eu ope, equen ly exceeding 40
◦
C.
Du ing he sample pe iod, spanning om July 2023 o June 2024, hese empe a u es, as
shown in (Table 1), exceeded his h eshold.
Table 1. Mon hly a ia ion in ou doo empe a u e (T ou ) and a e age ela i e humidi y (RH).
July
2023
Augus
2023
Sep embe
2023
Oc obe
2023
No embe
2023
Decembe
2023
Janua y
2024
Feb ua y
2024
Ma ch
2024
Ap il
2024
May
2024
June
2024
Tou min (◦C ) 18.9 18.9 13.4 11.1 3.9 1.6 2.2 4.4 5.0 7.2 7.5 14.4
Tou max (◦C ) 41.1 42.8 36.1 36.7 26.6 20.0 22.8 22.8 27.2 31.1 37.8 37.8
Tou a g (◦C ) 31.7 32.6 25.7 23.3 16.8 12.0 13.3 14.7 16.0 20.2 23.9 26.1
RHa g (%) 35 30 53 59 70 79 81 73 68 53 39 47
No e:
Tou min
(
◦
C ),
Tou max
(
◦
C ) and
Tou a g
(
◦
C ) ep esen he minimum, maximum, and a e age ou doo
empe a u es, espec i ely. RHa g deno es he a e age ela i e humidi y.
This ield s udy was ca ied ou in an o ice oom o a e ia y, non- esiden ial building.
Following he p oposal o Mo gan and De Dea [
33
], among he condi ions o he adap i e
app oach, “I is essen ial ha building occupan s a e ee o adap hemsel es,p ima ily h ough
clo hing adjus men , o he a iable indoo clima e egime p e ailing inside such buildings” (
p. 267
),
so he p esen s udy allows use s o make adap i e adjus men s, al hough i ocuses on he
analysis o he inclusion o an adap i e algo i hm in o he building managemen sys em.
The adap i e com o algo i hm applied in his a ea o he building was p esen ed
by Ba badilla e al. [
34
], in which a single adap i e model was p oposed o bo h he NV
and HVAC ope a ion modes. The algo i hm was based on an analysis o mo e han i e
housand su eys, leading o he ollowing equa ion (Equa ion (1)).
Tc=0.2427 ×T m +19.284 (1)
whe e Tcis he indoo com o empe a u e and T m is he ou doo unning mean empe a u e.
Figu e 1shows he adap i e ange co esponding o he beha io o his algo i hm
o e a ull yea .
Rega ding Figu e 1, in [
34
] a compa a i e analysis o he adap i e con ol algo i hm
conside ed o he de elopmen o he p esen case s udy was ca ied ou wi h espec
o o he models, showing ha he neu al empe a u es ob ained we e lowe han hose
p oposed using he ASHRAE-55:2023 and UNE-EN 16798-1:2020 models o NV buildings.
Mo eo e , acco ding o [
9
] (pp. 6–7), whe e a summa y o di e en adap i e eg ession
models and hei coe icien s was ca ied ou , mos o he linea adap i e linea eg ession
models a e wi hin he limi s ecommended in ASHRAE 55.
As o he implemen a ion o he adap i e con ol algo i hm, in [
35
] a me hodology
is p esen ed o implemen ing an adap i e con ol algo i hm in a Building Managemen
Sys em (BMS). In he p esen s udy, he BMS was adap ed o allow use s o swi ch be ween
na u al en ila ion and hea ing, en ila ion, and ai condi ioning modes. Du ing he s udy
Appl. Sci. 2025,15, 1243 5 o 17
pe iod, use s had he op ion o no using HVAC, wi h a key ea u e o his adap a ion being
ha , in NV mode, occupan s could eely open o close windows a hei disc e ion.
Appl. Sci. 2025, 15, 1243 6 o 18
Figu e 1. Daily a e age ou doo empe a u e, daily ope a ing empe a u e and adap i e com o
ange based on [34].
This p e ious app oach is based on he “change-o e ” mixed mode, commonly ap-
plied in office buildings, whe e na u al en ila ion and HVAC sys ems a e used in he
same space bu a diffe en momen s [10,36]. In he p esen s udy, HVAC mode ope a ion
equi ed all windows o emain closed, and windows could only be opened i he HVAC
sys em had been u ned off o inac i e o a leas 5 min [37].
To make use o he HVAC sys em, a limi ed one-off measu e in building e ofi , i
was p oposed o include a 2-h inac i i y ime in he BMS; i no in e ac ion occu ed du -
ing his pe iod, he sys em was au oma ically swi ched off. Howe e , he sys em could be
s opped a any ime. This design no only imp o es ene gy use bu also gi es he use
pe sonal con ol o e he clima e con ol sys em. This pe cei ed con ol has an impo an
psychological effec , as i allows people o eel ha hey can di ec ly influence hei en i-
onmen , hus, imp o ing hei com o and sa is ac ion [38]. The decision o allow he use
o he HVAC sys em, wi h au oma ic shu -off unc ionali y a e 2 h o inac i i y o when
eaching he empe a u e o he adap i e com o se poin , aligns wi h he findings o [38].
Acco ding o his s udy, he pe cei ed abili y o use s o con ol hei he mal en i on-
men imp o es hei pe cep ion o he mal com o due o he psychological influence o
pe cei ed con ol. Fu he mo e, when aced wi h ex eme he mal condi ions, occupan s
ha e a g ea e desi e o adjus hei en i onmen , and his pe sonal con ol is e en mo e
effec i e.
The sys em design no only p o ides pe cei ed con ol ha educes he mal discom-
o e en wi h sligh imp o emen s in he mal condi ions, bu also ensu es ha he he -
mal demands o he occupan s a e adequa ely add essed. This can ul ima ely minimize
complain s abou he he mal en i onmen , op imizing he mal com o as well as ene gy
efficiency.
The s udy g oup consis ed o 10 people aged 24 o 50 yea s, wi h 7 men and 3 women,
in a oom wi h 5 wo ks a ions (Figu e 2), whe e he desks allowed use s o exchange
h oughou he wo k day (“ho desk”). The office hou s conside ed we e om 8:00 o 20:00
h, and he pa icipa ion o he subjec s was olun a y, as is he case in many he mal com-
o field s udies. The s udy a ea depic ed in Figu e 2 shows a oom wi h senso s and
Figu e 1. Daily a e age ou doo empe a u e, daily ope a ing empe a u e and adap i e com o
ange based on [34].
This p e ious app oach is based on he “change-o e ” mixed mode, commonly applied
in o ice buildings, whe e na u al en ila ion and HVAC sys ems a e used in he same space
bu a di e en momen s [
10
,
36
]. In he p esen s udy, HVAC mode ope a ion equi ed all
windows o emain closed, and windows could only be opened i he HVAC sys em had
been u ned o o inac i e o a leas 5 min [37].
To make use o he HVAC sys em, a limi ed one-o measu e in building e o i ,
i was p oposed o include a 2-h inac i i y ime in he BMS; i no in e ac ion occu ed
du ing his pe iod, he sys em was au oma ically swi ched o . Howe e , he sys em
could be s opped a any ime. This design no only imp o es ene gy use bu also gi es
he use pe sonal con ol o e he clima e con ol sys em. This pe cei ed con ol has an
impo an psychological e ec , as i allows people o eel ha hey can di ec ly in luence
hei en i onmen , hus, imp o ing hei com o and sa is ac ion [
38
]. The decision o
allow he use o he HVAC sys em, wi h au oma ic shu -o unc ionali y a e 2 h o
inac i i y o when eaching he empe a u e o he adap i e com o se poin , aligns wi h
he indings o [
38
]. Acco ding o his s udy, he pe cei ed abili y o use s o con ol hei
he mal en i onmen imp o es hei pe cep ion o he mal com o due o he psychological
in luence o pe cei ed con ol. Fu he mo e, when aced wi h ex eme he mal condi ions,
occupan s ha e a g ea e desi e o adjus hei en i onmen , and his pe sonal con ol is
e en mo e e ec i e.
The sys em design no only p o ides pe cei ed con ol ha educes he mal dis-
com o e en wi h sligh imp o emen s in he mal condi ions, bu also ensu es ha he
he mal demands o he occupan s a e adequa ely add essed. This can ul ima ely mini-
mize complain s abou he he mal en i onmen , op imizing he mal com o as well as
ene gy e iciency.
The s udy g oup consis ed o 10 people aged 24 o 50 yea s, wi h 7 men and 3 women,
in a oom wi h 5 wo ks a ions (Figu e 2), whe e he desks allowed use s o exchange
Appl. Sci. 2025,15, 1243 6 o 17
h oughou he wo k day (“ho desk”). The o ice hou s conside ed we e om 8:00 o 20:00
h, and he pa icipa ion o he subjec s was olun a y, as is he case in many he mal com o
ield s udies. The s udy a ea depic ed in Figu e 2shows a oom wi h senso s and ables,
wi h an a ea o 14 m2wi h windows wi h double glass and manual and indoo blinds.
Appl. Sci. 2025, 15, 1243 7 o 18
ables, wi h an a ea o 14 m2 wi h windows wi h double glass and manual and indoo
blinds.
Figu e 2. Floo plan o he field s udy oom and senso s’ loca ions.
2.2. Acquisi ion, P ocessing and P elimina y Analysis o Da a
Th oughou he s udy, objec i e measu emen s (indoo and ou doo clima e) and
subjec i e measu emen s ( he mal p e e ence) we e collec ed simul aneously. The s udy
was ca ied ou o e 12 mon hs (July 2023 o June 2024) in o de o cha ac e ize all he
seasons. An a ay o wi eless echnology senso s eco ded he beha io o he oom. The
newly acqui ed senso s we e e ified p io o he s udy. A o al o 685,152 indoo en i-
onmen al eco ds (loca ed in Figu e 2) and 35,136 ou doo eco ds (Table 2) we e col-
lec ed.
Measu emen s inside he oom ( empe a u e, humidi y, e c.) we e ob ained a 10 min
in e als. All senso s we e new, calib a ed, compa ed wi h simila senso s unde he same
condi ions, and showed no significan diffe ences in hei measu emen s. Senso s we e
placed app oxima ely 1 m om he occupan s’ wo ks a ions in o de o desc ibe hei im-
media e he mal en i onmen unde ypical wo king condi ions. Ou doo wea he obse -
a ions we e collec ed om a wea he s a ion close o he building [39] and loca ed abou
100 m om i , o which ou doo empe a u e was o he g ea es in e es .
Table 2. Da a acquisi ion sys ems and cha ac e is ics.
Senso
Numbe
o Senso s
Ope a i e
Range
Accu acy
Uni s
Concen a ion o pa icle ma e
1
0 o 500
±10%
µg/m
3
CO
2
Concen a ion
1
0 o 10.000
±45 + 3%
ppm
Rela i e humidi y
5
0 o 100
±2%
%
Ai empe a u e
5
−25 o 70
±0.3
°C
Globe empe a u e
6
−25 o 70
±0.3
°C
Tempe a u e in he wea he s a ion
1
−50 o 60
±2%
°C
Rela i e humidi y in he wea he s a ion
1
0 o 100
±2%
%
Figu e 2. Floo plan o he ield s udy oom and senso s’ loca ions.
2.2. Acquisi ion, P ocessing and P elimina y Analysis o Da a
Th oughou he s udy, objec i e measu emen s (indoo and ou doo clima e) and sub-
jec i e measu emen s ( he mal p e e ence) we e collec ed simul aneously. The s udy was
ca ied ou o e 12 mon hs (July 2023 o June 2024) in o de o cha ac e ize all he seasons.
An a ay o wi eless echnology senso s eco ded he beha io o he oom. The newly
acqui ed senso s we e e i ied p io o he s udy. A o al o 685,152 indoo en i onmen al
eco ds (loca ed in Figu e 2) and 35,136 ou doo eco ds (Table 2) we e collec ed.
Table 2. Da a acquisi ion sys ems and cha ac e is ics.
Senso Numbe o Senso s Ope a i e Range Accu acy Uni s
Concen a ion o pa icle ma e 1 0 o 500 ±10% µg/m3
CO2Concen a ion 1 0 o 10.000 ±45 + 3% ppm
Rela i e humidi y 5 0 o 100 ±2% %
Ai empe a u e 5 −25 o 70 ±0.3 ◦C
Globe empe a u e 6 −25 o 70 ±0.3 ◦C
Tempe a u e in he wea he s a ion 1 −50 o 60 ±2% ◦C
Rela i e humidi y in he wea he s a ion 1 0 o 100 ±2% %
Measu emen s inside he oom ( empe a u e, humidi y, e c.) we e ob ained a
10 min
in e als. All senso s we e new, calib a ed, compa ed wi h simila senso s unde he
same condi ions, and showed no signi ican di e ences in hei measu emen s. Senso s
we e placed app oxima ely 1 m om he occupan s’ wo ks a ions in o de o desc ibe
hei immedia e he mal en i onmen unde ypical wo king condi ions. Ou doo wea he
obse a ions we e collec ed om a wea he s a ion close o he building [
39
] and loca ed
abou 100 m om i , o which ou doo empe a u e was o he g ea es in e es .
The 2859 wo king hou s du ing which he use s we e wo king in he o ice we e
selec ed, a e elimina ing weekends, holidays, and in ensi e wo king days in July (when
Appl. Sci. 2025,15, 1243 7 o 17
he oom was no used in he a e noon). A o al o 1320 alid su eys we e collec ed and
we e g ouped acco ding o he ime o day, esul ing in he e alua ion o 483 decla ed
he mal p e e ences.
A he end o he s udy, he BMS egis e was ob ained wi h he HVAC sys em s a es.
The base se poin alue o he HVAC sys em applied he adap i e com o algo i hm o
he building, in which he unning mean empe a u e was calcula ed wi h an alpha o
0.8. The ange associa ed wi h he adap i e com o law is ep esen ed in Figu e 1by
wo g een dashed lines, which indica e he o al wid h o he com o zone o he sample
pe iod. The same igu e shows he beha io o he indoo and ou doo empe a u es o
he building, oge he wi h he com o zone based on he implemen ed adap i e law. The
ange is indica ed in i s mos es ic i e o m, wi h a scale o +2/
−
3
◦
C, building ca ego y I
acco ding o UNE-EN 16798-1 [
4
]. This mo e es ic i e ange was adop ed o p io i ize
pe cei ed he mal com o in his analysis, conside ing ha e en small a ia ions can
gene a e signi ican esponses.
Al hough in he UNE-EN 16798-1 [
4
] indoo com o ange o he design and e alua-
ion o he ene gy pe o mance o buildings would be +3/
−
4
◦
C, in Ca lucci e al. [
40
] a
compa a i e analysis o di e en adap i e he mal com o models can be ound, in which
he mos es ic i e accep abili y pe cen age co esponds o 90%. These pe cen ages a e
jus i ied on he g ounds ha cu en models canno accu a ely p edic how an indi idual
will eel on a speci ic day, so i is no easonable o expec all use s o expe ience he mal
com o in he same en i onmen , e en when i complies wi h cu en s anda ds.
Based on he BMS egis e , he HVAC sys em was equi ed as a punc ual measu e
in 379 h (13.25% o he o al ac i e hou s o o ice use). O hese, 206 h was su icien o
achie e he mal com o wi hin he planned ope a ion ime, while in 192 h, he planned
ope a ion ime was insu icien , equi ing an ex ension o main ain com o . A o al o 166 h
o punc ual ac i i y occu ed in summe , om June o Sep embe , and 113 h in win e , om
Decembe o Ma ch. In he s udy, he HVAC sys em emained o wi h passi e adap a ion
measu es o 2480 h (92.7% o he o al hou s o use).
2.3. Me hodology
One o he main challenges o he mal com o s udies is o keep people mo i a ed o
pa icipa e o e ime. This is especially ele an in app oaches ha equi e ac i e in e ac ion
o e long pe iods o ime.
The p esen p oposal was, he e o e, aimed a analyzing HVAC as an adap i e measu e
in MM and, gi en ha he ield s udy was ca ied ou o e one yea , a cen al ques ion
on he mal p e e ence was used: “How would you p e e o be now?” [
41
]. The scale issue
abou he he mal p e e ence a ose “ o o e come he ambigui y o he mal condi ion accep abili y
conside ing he 3-poin McIn y e p e e ence scale” [
42
], bu based on ISO 10551:2019 [
41
], “A
7-deg ee scale should be applied in he case o en i onmen s judged o be close o neu ali y; a
9-deg ee scale should be applied in he case o en i onmen s judged o be mo e in ense”, al hough
s udies can be ound wi h an 11-poin scale [
42
]. In [
43
] i is indica ed ha he scale o
he he mal p e e ence o e (TPV) ep esen s how an occupan would p e e o adjus
hei he mal en i onmen and i is, he e o e, an accu a e measu e o de ining an op imal
indoo he mal en i onmen . Acco ding o he au ho s, unlike o he measu es, TPV di ec ly
sugges s a change in cu en condi ions, making i an e ec i e ool o imp o e he p edic ion
o he HVAC sys em and op imize ene gy e iciency.
In o de o conduc his analysis, i is necessa y o ha e a meaning ul da a se . Figu e 3,
ela ing o a e age TPV and ope a i e empe a u e, depic s how he a e age he mal
pe cep ion o pa icipan s a ies wi h ope a ional empe a u e. I shows ha 78.8% o
Appl. Sci. 2025,15, 1243 8 o 17
he a ia ion in a e age TPV can be explained by ope a ional empe a u e, so, gi en he
s a is ical signi icance ob ained, he s udy se can be analyzed.
Appl. Sci. 2025, 15, 1243 9 o 18
a ia ion in a e age TPV can be explained by ope a ional empe a u e, so, gi en he s a-
is ical significance ob ained, he s udy se can be analyzed.
Figu e 3. Rela ionship be ween A e age TPV and Ope a i e Tempe a u e.
A e applying he adap i e com o equa ion (Equa ion (1)) and he associa ed em-
pe a u e anges o he one-yea pe iod, he indoo he mal com o is analyzed hou by
hou (Figu es 4 o 6), conside ing he he mal p e e ence o he use s. Fo his analysis,
non-wo king hou s, holidays and weekends a e excluded. Subsequen ly, e alua ion me -
ics a e calcula ed o examine he accu acy o he sys em by defining he ollowing indi-
ca o s om [42], bu adap ed o he case o he analysis wi h a com o law applied and
on he basis o use s’ p e e ences:
Figu e 4. Rep esen a ion o indica o s.
T ue Com o (TC): This alue is iden ified when he com o law applied co ec ly
p edic s he mal com o , ha is, when he he mal p e e ence o he use s is neu al,
Figu e 3. Rela ionship be ween A e age TPV and Ope a i e Tempe a u e.
A e applying he adap i e com o equa ion (Equa ion (1)) and he associa ed em-
pe a u e anges o he one-yea pe iod, he indoo he mal com o is analyzed hou by
hou (Figu es 4–6), conside ing he he mal p e e ence o he use s. Fo his analysis, non-
wo king hou s, holidays and weekends a e excluded. Subsequen ly, e alua ion me ics
a e calcula ed o examine he accu acy o he sys em by de ining he ollowing indica o s
om [
42
], bu adap ed o he case o he analysis wi h a com o law applied and on he
basis o use s’ p e e ences:
Appl. Sci. 2025, 15, 1243 9 o 18
a ia ion in a e age TPV can be explained by ope a ional empe a u e, so, gi en he s a-
is ical significance ob ained, he s udy se can be analyzed.
Figu e 3. Rela ionship be ween A e age TPV and Ope a i e Tempe a u e.
A e applying he adap i e com o equa ion (Equa ion (1)) and he associa ed em-
pe a u e anges o he one-yea pe iod, he indoo he mal com o is analyzed hou by
hou (Figu es 4 o 6), conside ing he he mal p e e ence o he use s. Fo his analysis,
non-wo king hou s, holidays and weekends a e excluded. Subsequen ly, e alua ion me -
ics a e calcula ed o examine he accu acy o he sys em by defining he ollowing indi-
ca o s om [42], bu adap ed o he case o he analysis wi h a com o law applied and
on he basis o use s’ p e e ences:
Figu e 4. Rep esen a ion o indica o s.
T ue Com o (TC): This alue is iden ified when he com o law applied co ec ly
p edic s he mal com o , ha is, when he he mal p e e ence o he use s is neu al,
Figu e 4. Rep esen a ion o indica o s.
Appl. Sci. 2025,15, 1243 9 o 17
Appl. Sci. 2025, 15, 1243 12 o 18
Figu e 5. Rep esen a ion o hou ly he mal p e e ence du ing a yea o s udy wi hin wo king hou s.
When analyzing he esul s in Figu e 6, whe e he e ms TC, FC, TD and FD a e as-
socia ed wi h he me ics shown in Equa ions 2 o 6, i can be obse ed in Table 4 ha a
PD o 6.5% du ing he wo k yea and an accu acy o 92.96% ha e been ob ained. This
indica es ha he model co ec ly iden ifies 92.96% o cases in which he mal com o is
p edic ed wi hin he com o ange. This high alue sugges s ha he p oposed empe a-
u e ange o he mal com o is qui e ep esen a i e o he eal condi ions pe cei ed as
com o able by he occupan s.
Figu e 6. Rep esen a ion o com o p e e ences wi hin he defined adap i e com o anges.
I is impo an o no e ha he p opo ion o TC, FC, TD and FD a ies h oughou
he yea due o seasonal ac o s. In pa icula , om July o Sep embe , a highe le el o
he mal discom o (TD and FD) is obse ed. This is mainly due o empe a u e cha ac-
e is ics du ing hese mon hs, which can exceed he limi s o he p oposed he mal com-
o ange.
Figu e 5. Rep esen a ion o hou ly he mal p e e ence du ing a yea o s udy wi hin wo king hou s.
Appl. Sci. 2025, 15, 1243 12 o 18
Figu e 5. Rep esen a ion o hou ly he mal p e e ence du ing a yea o s udy wi hin wo king hou s.
When analyzing he esul s in Figu e 6, whe e he e ms TC, FC, TD and FD a e as-
socia ed wi h he me ics shown in Equa ions 2 o 6, i can be obse ed in Table 4 ha a
PD o 6.5% du ing he wo k yea and an accu acy o 92.96% ha e been ob ained. This
indica es ha he model co ec ly iden ifies 92.96% o cases in which he mal com o is
p edic ed wi hin he com o ange. This high alue sugges s ha he p oposed empe a-
u e ange o he mal com o is qui e ep esen a i e o he eal condi ions pe cei ed as
com o able by he occupan s.
Figu e 6. Rep esen a ion o com o p e e ences wi hin he defined adap i e com o anges.
I is impo an o no e ha he p opo ion o TC, FC, TD and FD a ies h oughou
he yea due o seasonal ac o s. In pa icula , om July o Sep embe , a highe le el o
he mal discom o (TD and FD) is obse ed. This is mainly due o empe a u e cha ac-
e is ics du ing hese mon hs, which can exceed he limi s o he p oposed he mal com-
o ange.
Figu e 6. Rep esen a ion o com o p e e ences wi hin he de ined adap i e com o anges.
T ue Com o (TC): This alue is iden i ied when he com o law applied co ec ly
p edic s he mal com o , ha is, when he he mal p e e ence o he use s is neu al,
sligh ly coole o sligh ly wa me (TPV
∈
[
−
1,1]). In o he wo ds, he sys em o ecas s
com o and i co esponds o he occupan ’s he mal sensa ion.
False Com o (FC): This alue is iden i ied when he com o law inco ec ly p edic s
he mal com o , ha is, he he mal p e e ence o use s is signi ican ly ou side he com o
ange (TPV
/∈
[
−
1,1]). In his case, he sys em does no accu a ely e lec he occupan ’s
he mal expe ience.
T ue Discom o (TD): This alue is iden i ied when discom o is p edic ed and he
he mal p e e ence o e o he occupan is ou side he com o ange (TPV
/∈
[
−
1,1]). In
his case, he sys em aligns wi h he he mal s a e expe ienced by he occupan .
False Discom o (FD): This alue is iden i ied when discom o is inco ec ly p edic ed,
since he occupan ’s he mal p e e ence is wi hin he com o ange (TPV
∈
[
−
1,1]). This
Appl. Sci. 2025,15, 1243 16 o 17
20.
Healy, S. Ai -Condi ioning and he ‘Homogeniza ion’ o People and Buil En i onmen s. Build. Res. In . 2008,36, 312–322.
[C ossRe ]
21.
Ca losena, L.; Ruiz-Pa do, Á.; Rod íguez-Ja a, E.Á.; San amou is, M. Wo ldwide Po en ial o Emissi e Ma e ials Based Radia i e
Cooling Technologies o Mi iga e U ban O e hea ing. Build. En i on. 2023,243, 110694. [C ossRe ]
22.
Rincón-Casado, A.; Rod íguez Ja a, E.Á.; Ruiz Pa do, A.; Salme ón Lissén, J.M.; Sánchez de la Flo , F.J. Reducing he Cooling
Ene gy Demand by Op imizing he Ai low Dis ibu ion in a Ven ila ed Roo : Nume ical S udy o an Exis ing Residen ial
Building and Applicabili y Map. Appl. Sci. 2024,14, 6596. [C ossRe ]
23. Fishman, D.S.; Pimbe , S.L. The The mal En i onmen in O ices. Ene gy Build. 1982,5, 109–116. [C ossRe ]
24. Foun ain, M.; B age , G.; de Dea , R. Expec a ions o Indoo Clima e Con ol. Ene gy Build. 1996,24, 179–182. [C ossRe ]
25.
Hellwig, R.T.; Teli, D.; Schweike , M.; Mo a, R.; Choi, J.-H.; Rawal, R.; Lee, M.C.J.; Wang, Z.; Al-A ash, F. Guidelines o Low
Ene gy Building Design Based on he Adap i e The mal Com o Concep —Technical Repo : IEA EBC Annex 69: S a egy and
P ac ice o Adap i e The mal Com o in Low Ene gy Buildings. Aalbo g Uni e si y: Aalbo g, Denma k, 2022.
26.
De Dea , R. Adap i e Com o Applica ions in Aus alia and Impac s on Building Ene gy Consump ion. In P oceedings o
he Six h In e na ional Con e ence on Indoo Ai Quali y, Ven ila ion and Ene gy Conse a ion in Buildings: Sus ainable Buil
En i onmen , Sendai, Japan, 28–31 Oc obe 2007; pp. 28–31.
27. Deuble, M.P.; de Dea , R.J. G een Occupan s o G een Buildings: The Missing Link? Build. En i on. 2012,56, 21–27. [C ossRe ]
28. Aguile a, J.J.; Boga u, D.-I.; Kazanci, O.B.; Angelopoulos, C.; Coakley, D.; Olesen, B.W. Com o -Based Con ol o Mixed-Mode
Buildings. Ene gy Build. 2021,252, 111465. [C ossRe ]
29.
Peng, Y.; Lei, Y.; Tekle , Z.D.; An anu i, N.; Lau, S.-K.; Chong, A. Hyb id Sys em Con ols o Na u al Ven ila ion and HVAC in
Mixed-Mode Buildings: A Comp ehensi e Re iew. Ene gy Build. 2022,276, 112509. [C ossRe ]
30.
Lambe i, G.; Con ada, F.; Kindinis, A. Explo ing Adap i e S a egies o Cope wi h Clima e Change: The Case S udy o Le
Co busie ’s Mode n A chi ec u e Re o i ing. Ene gy Build. 2024,302, 113756. [C ossRe ]
31.
Ko ek, M.; G iese , J.; Beck, C.; Rudol , B.; Rubel, F. Wo ld Map o he Köppen-Geige Clima e Classi ica ion Upda ed. Me eo ol.
Zei sch i 2006,15, 259–263. [C ossRe ]
32.
Salme ón, J.M.; Ál a ez, S.; Molina, J.L.; Ruiz, A.; Sánchez, F.J. Tigh ening he Ene gy Consump ions o Buildings Depending on
Thei Typology and on Clima e Se e i y Indexes. Ene gy Build. 2013,58, 372–377. [C ossRe ]
33. Mo gan, C.; de Dea , R. Wea he , Clo hing and The mal Adap a ion o Indoo Clima e. Clim. Res. 2003,24, 267–284. [C ossRe ]
34.
Ba badilla-Ma ín, E.; Salme ón Lissén, J.M.; Guadix Ma ín, J.; Apa icio-Ruiz, P.; B o as, L. Field S udy on Adap i e The mal
Com o in Mixed Mode O ice Buildings in Sou hwes e n A ea o Spain. Build. En i on. 2017,123, 163–175. [C ossRe ]
35.
Apa icio-Ruiz, P.; Ba badilla-Ma ín, E.; Salme ón-Lissén, J.M.; Guadix-Ma ín, J. Building Au oma ion Sys em wi h Adap i e
Com o in Mixed Mode Buildings. Sus ain. Ci ies Soc. 2018,43, 77–85. [C ossRe ]
36.
Kim, J.; de Dea , R. Is Mixed-Mode Ven ila ion a Com o able Low-Ene gy Solu ion? A Li e a u e Re iew. Build. En i on. 2021,
205, 108215. [C ossRe ]
37.
Deuble, M.P.; de Dea , R.J. Mixed-Mode Buildings: A Double S anda d in Occupan s’ Com o Expec a ions. Build. En i on. 2012,
54, 53–60. [C ossRe ]
38.
Luo, M.; Cao, B.; Ji, W.; Ouyang, Q.; Lin, B.; Zhu, Y. The Unde lying Linkage be ween Pe sonal Con ol and The mal Com o :
Psychological o Physical E ec s? Ene gy Build. 2016,111, 56–63. [C ossRe ]
39.
Mo eno-Teje a, S.; Sil a-Pé ez, M.A.; Lillo-B a o, I.; Ramí ez-San igosa, L. Sola Resou ce Assessmen in Se ille, Spain. S a is ical
Cha ac e isa ion o Sola Radia ion a Di e en Time Resolu ions. Sol. Ene gy 2016,132, 430–441. [C ossRe ]
40.
Ca lucci, S.; Bai, L.; de Dea , R.; Yang, L. Re iew o Adap i e The mal Com o Models in Buil En i onmen al Regula o y
Documen s. Build. En i on. 2018,137, 73–89. [C ossRe ]
41.
ISO-10551; E gonomics o he Physical En i onmen - Subjec i e Judgemen Scales o Assessing Physical En i onmen s. ISO:
Gene a, Swi ze land, 2019.
42.
Jazizadeh, F.; Ma in, F.M.; Bece ik-Ge be , B. A The mal P e e ence Scale o Pe sonalized Com o P o ile Iden i ica ion ia
Pa icipa o y Sensing. Build. En i on. 2013,68, 140–149. [C ossRe ]
43.
Kim, J.; Schia on, S.; B age , G. Pe sonal Com o Models – A New Pa adigm in The mal Com o o Occupan -Cen ic
En i onmen al Con ol. Build. En i on. 2018,132, 114–124. [C ossRe ]
44.
Ba badilla-Ma ín, E.; Ragel-Bonilla, J.C.; Apa icio-Ruiz, P.; Guadix, J. Analysing he Indoo Tempe a u e Ranges o an E icien
Con ol o HVAC Sys ems. Ene gy Build. 2023,301, 113681. [C ossRe ]
45.
Zhang, S.; Yao, R.; To um, J.; Essah, E.; Li, B. Machine Lea ning-Based App oach o P edic The mal Com o in Mixed-Mode
Buildings: Inco po a ing Adap i e Beha io s. J. Build. Eng. 2024,87, 108877. [C ossRe ]
46.
Nicol, J.F.; Raja, I.A.; Allaudin, A.; Jamy, G.N. Clima ic Va ia ions in Com o able Tempe a u es: The Pakis an P ojec s. Ene gy
Build. 1999,30, 261–279. [C ossRe ]
47.
Rijal, H.B.; Yoshida, H.; Umemiya, N. Seasonal and Regional Di e ences in Neu al Tempe a u es in Nepalese T adi ional
Ve nacula Houses. Build. En i on. 2010,45, 2743–2753. [C ossRe ]
Appl. Sci. 2025,15, 1243 17 o 17
48. B age , G.; Paliaga, G.; De Dea , R. Ope able Windows, Pe sonal Con ol and Occupan Com o . ASHRAE T ans. 2004,110.
49.
Humph eys, M.A.; Rijal, H.B.; Nicol, J.F. Upda ing he Adap i e Rela ion be ween Clima e and Com o Indoo s; New Insigh s
and an Ex ended Da abase. Build. En i on. 2013,63, 40–55. [C ossRe ]
50.
Li, D.; Menassa, C.C.; Kama , V.R. Pe sonalized Human Com o in Indoo Building En i onmen s unde Di e se Condi ioning
Modes. Build. En i on. 2017,126, 304–317. [C ossRe ]
Disclaime /Publishe ’s No e: The s a emen s, opinions and da a con ained in all publica ions a e solely hose o he indi idual
au ho (s) and con ibu o (s) and no o MDPI and/o he edi o (s). MDPI and/o he edi o (s) disclaim esponsibili y o any inju y o
people o p ope y esul ing om any ideas, me hods, ins uc ions o p oduc s e e ed o in he con en .
